Dec 8, 2003 - epilepsy surgery and were subsequently seizure free. ...... tients with emergency department (ED) visits with the diagnosis of NC ..... CT/MRI scans revealed organic brain damage in 64 patients with pre- ... Conclusions: Our results indicate that one third of the newly diag- ...... and to make power spectra.
4
AES PROCEEDINGS
after infancy, and how these targets can best be exploited for therapeutic purposes. (Supported by NINDS, NICHD, Primary Children’s Medical Center Foundation, Epilepsy Foundation of America.)
IW.11 THE ROLE OF TONIC GABA INHIBITION IN CONTROL OF BRAIN EXCITABILITY 1 George B. Richerson, 2 Robert L. MacDonald, 3 Istvan Mody, and 4 Dimitri M. Kullmann (1 Neurology, Yale University, New Haven, CT; 2 Neurology, Vanderbilt University, Nashville, TN; 3 Neurology, UCLA, Los Angeles, CA; and 4 Clinical & Experimental Epilepsy, Institute of Neurology, University College London, London, United Kingdom) γ -Aminobutyric acid (GABA)ergic inhibition plays a prominent role in control of brain excitability and is the target of many anticonvulsants. It has recently been recognized that there are two forms of GABAergic inhibition. Phasic inhibition is the rapid response to synaptic release of GABA, whereas tonic inhibition is the continuous inhibition of neurons by the finite level of ambient GABA that normally exists in the extracellular space. These two forms of GABAergic inhibition can be separated based on differences in their pharmacology. Tonic inhibition is mediated, at least in part, by specific αβδ GABAA receptors that are extrasynaptic, have high affinity for GABA, and have no fast desensitization. Phasic inhibition is mediated primarily by specific αβγ GABAA receptors that are subsynaptic, have low affinity for GABA, and have prominent desensitization. There are differences in the relative importance of tonic versus phasic inhibition in different parts of the brain, and within the hippocampus, there are even differences between different cell types, with interneurons normally receiving greater tonic inhibition than pyramidal cells. Tonic and phasic inhibition are also modulated differently by a variety of pharmacologic agents, including neurosteroids, benzodiazepines, barbiturates, GABA-uptake blockers, and vigabatrin. The GABA transporter plays an important role in controlling the level of tonic inhibition, and transporter reversal can be a source of GABA release that increases ambient [GABA]. This workshop will focus on the role of tonic inhibition in normal brain function and in epilepsy, and will discuss how the two forms of inhibition may play complementary roles in control of brain excitability. Defining the role of tonic inhibition may lead to a greater understanding of the mechanisms of seizures and new avenues for treatment of epilepsy.
IW.12 NONCONVULSIVE STATUS EPILEPTICUS AND ICU EEG MONITORING 1 Lawrence J. Hirsch, 2 David M. Treiman, and 3 Frank W. Drislane (1 Neurological Institute, Columbia University, New York, NY; 2 Barrow Neurological Institute, Phoenix, AZ; and 3 Beth Israel Deaconess Medical Center, Harvard University, Boston, MA) Due to improvements in technology, continuous EEG monitoring (cEEG) in critically ill patients is becoming an increasingly used procedure, especially in patients with acute brain disorders. Nonconvulsive seizures and nonconvulsive status epilepticus are being recognized more frequently than expected with this technique. In addition, cEEG can detect cerebral ischemia, monitor level of sedation/anesthesia, assess reactivity and variability, and provide a continuous evaluation of brain activity in patients in whom the neurologic examination is limited. In this clinical investigators’ workshop, we will discuss the indications for cEEG, findings in patients with a variety of critical illnesses, methods, pitfalls, data reduction and quantitative EEG analysis, and the potential for additional clinical applications of this technology. A major focus of this symposium will involve the interpretation of controversial EEG patterns such as periodic or stimulus-induced rhythmic discharges: Which patterns are contributing to impaired mental status? Which are causing neuronal injury and need to be treated? Which are simply markers of preexisting injury? How can we answer these questions? Dr. Treiman will tackle these issues and will present relevant animal and human data. Dr. Drislane will then provide an update on the treatment of refractory status epilepticus, including use of novel agents. He will address the issue of the goal of treatment (EEG suppression or cessation of seizures), how long to treat, and which agent to use. Finally, he will lead a discussion on Epilepsia, Vol. 44, Suppl. 9, 2003
the feasibility of conducting a multicenter trial to answer some of these questions more definitively.
December 8, 2003 Presidential Symposium—The Developing Epileptic Brain 8:30 a.m.–11:00 a.m. PS.01 PRESIDENTIAL SYMPOSIUM: THE DEVELOPING EPILEPTIC BRAIN Jeffrey L. Noebels, Christopher A. Walsh, Pat R. Levitt, and Ben A. Barres (Baylor College of Medicine, Houston, TX; Beth Israel Hospital, Harvard Medical School, Boston, MA; Vanderbilt University, Nashville, TN; and Stanford University School of Medicine, Palo Alto, CA) Epilepsy reflects the biology of the brain before birth, when a single molecular event can either protect the nervous system from damage, or set the stage for a lifetime of seizures. This symposium will focus on extraordinary advances in understanding the genes and molecular signals that control neuronal proliferation, migration, and synaptogenesis in the developing epileptic brain. Speakers will point to human case examples and experimental models to address three major questions. How do single genes disrupt the assembly of the neocortex to produce epilepsy? Where do interneurons come from? What is a glial cell for? In the case of developing neocortex, diverse molecules with entirely different intracellular functions are being isolated from families with cortical malformations and epilepsy, helping to decipher the importance of brain size, shape, lamination, and gyral patterns. Inhibition appears to be one of their major targets, and there is new information on the origin of cortical interneurons deep in the brain, when they proliferate, and their tangential migratory pathway to the neocortex. Mouse models reveal how the arrival of interneurons interacts with the structure and function of cortical microcircuits. Finally, astrocytes constitute nearly half of all cells in the brain, yet one of their more mysterious functions is only now becoming clear. Astrocytes cause a sevenfold increase in the number of mature, functional synapses and are required for synaptic stabilization. Most synapses form concurrent with the development of glia, and those that develop in their absence are functionally immature. The signals now being discovered that regulate the formation and stability of various populations of cortical neurons and synapses may prove to be therapeutic targets of the future.
December 8, 2003 Poster Session 1 11:00 a.m.–5:00 p.m. Clinical Epilepsy—Adult 1.001 UNILATERAL GIANT SOMATOSENSORY EVOKED FIELDS AS A LATERALIZING SIGN IN FRONTAL LOBE EPILEPSY 1 Valerie A. Carr, 1 Susanne Knake, 1 Hideaki Shiraishi, 1 Deirdre M. Foxe, 1 Keiko Hara, 1 P. Ellen Grant, 2 Donald L. Schomer, 2 Blaise F. Bourgeois, 1 Eric Halgren, and1Steven M. Stufflebeam (1 Martinos Center for Biomedical Imaging, MGH/Harvard Medical School, Charlestown, and 2 Clinical Neurophysiology, Beth Israel Deaconess Medical Center, and 3 Epilepsy & Clinical Neurophysiology, Children’s Hospital, Boston, MA) Rationale: In patients (pts) with intractable focal epilepsies who are surgical candidates, magnetoencephalography (MEG) may be used to lateralize and localize the irritative zone and eloquent cortex. Somatosensory evoked fields (SEFs) are often used as landmarks to localize the central sulcus and precentral gyrus. Here we report the use of SEFs to lateralize the epileptogenic region in two pts with intractable frontal lobe epilepsy.
AES PROCEEDINGS Methods: Pts were evaluated by using 306-channel MEG and 64channel EEG. Spontaneous data and median nerve SEFs were obtained, and tibial nerve SEFs were collected for one pt. Equivalent current dipoles (ECDs) were calculated for interictal spikes (IISs) and SEFs (N20m, P35m). Results: Case 1: 37-year-old, right-handed (rh) woman with two seizure types: (a) loss of vision → inability to speak and move, and (b) right shoulder twitching → loss of consciousness. Routine EEG, positron emission tomography (PET), single-photon emission computed tomography (SPECT), and neurologic examination were normal; EEG showed right temporal and bicentral spikes; magnetic resonance imaging (MRI) showed enlarged ventricles. ECDs calculated for IISs localized to the right somatosensory cortex. Left median nerve stimulation showed a giant SEF response (218 fT/cm, 35 ms), and was 3 times larger than that for the right median nerve (74 fT/cm, 35 ms). ECDs calculated from the giant SEF localized to the right central sulcus in the same vicinity as the IISs. Case 2: 13-year-old, rh girl with simple motor seizures of unknown etiology and the following semiology: numbness right foot, hand, or head → clonic movements right arm and leg. EEG showed right centroparietal and left parasagittal spikes; MRI was normal. ECDs of IISs clustered in the left hand and foot areas of the precentral gyrus. Right median nerve stimulation showed a giant SEF 8 times larger in amplitude than that for the left median nerve (see Fig. 1). Right tibial nerve SEF (161 fT/cm, 42 ms) was remarkably larger than that for the left (10 fT/cm, 40 ms). Dipoles fitted to the giant right median and tibial nerve SEFs localized to the left hand and foot areas, respectively.
5
(TBI). PTE is often refractory to medical treatment and commonly presents diagnostic and therapeutic challenges. This study was undertaken to assess the clinical utility of long-term video-EEG monitoring evaluation of patients with medically refractory PTE. Methods: The records of 1,634 patients consecutively referred for video-EEG monitoring from January 1998 to December 2002 were reviewed. Patients and family members were questioned as to possible risk factors for epilepsy. Patients were included if they sustained moderate to severe TBI before the onset of epilepsy. Moderate to severe TBI was defined as injury resulting in prolonged (>30 min) loss of consciousness or amnesia, hospitalization, or neuroradiologic evidence of TBI. All patients underwent surface ictal and interictal video-EEG recordings, and data were reviewed by board-certified clinical neurophysiologists. All patients also underwent magnetic resonance imaging (MRI) with a technique sensitive for detecting sclerosis of the mesial temporal structures. Results: One hundred thirty-one patients (8% of patients evaluated over 4 years) met inclusion criteria. All had frequent, disabling spells, which had been diagnosed as epileptic by the referring physicians. All but two were being treated with antiepileptic drugs (AEDs), in most cases with combinations of these. Video-EEG evaluation was effective in establishing a diagnosis in 108 (82%). The average duration of monitoring was 4.6 days (SD, 2.4; median, 4) and was not different for the nondiagnostic cases. Of the diagnosed cases, 68 (63%) had epileptic seizures recorded, which were temporal in 35 (32%), frontal in 21 (19%), parietal in three (3%), and occipital in two (2%). Of the temporal lobe epilepsy cases, 18 (51%) had mesial temporal sclerosis. An additional seven (6%) patients had generalized epileptic disorders, of whom four had the clinical features of idiopathic generalized epilepsy. Thirty-nine (29.7% of the total cohort) had psychogenic nonepileptic seizures (NESs) as the exclusive seizure type recorded. Conclusions: The diagnostic yield of a reasonably short evaluation (average, 1 year. LTLE was defined when discrete lesion or ictal-onset zone in an invasive study was located outside collateral sulcus, and subgrouped as anterior and posterior TLE (ATLE and PTLE) with reference to the line across the cerebral peduncle. Total of 96 seizures, 11 ATLE, six PTLE, and 21 MTLE patients were reviewed. Results: Epigastric aura and psychic aura were less common in PTLE than in ATLE and MTLE, although there was no statistical significance. Oroalimentary automatism (OAA) was not observed initially in PTLE. Initial hypomotor symptoms were observed frequently in PTLE (p = 0.003). Dystonia and focal motor signs including facial twitching and clonic movement of limbs were not observed in PTLE. Generalized tonic–clonic seizures (GTCSs) occurred significantly earlier in PTLE than in ATLE and MTLE. Ictal scalp EEG is not helpful to differentiate between ATLE and PTLE. Conclusions: There are a number of clinical differences between ATLE and PTLE, although these are not sufficient to differentiate them. 1.008 EMOTION ASSOCIATED WITH DREAMS MAY ASSIST DIFFERENTIATION BETWEEN EPILEPTIC AND NONEPILEPTIC EVENTS Fumisuke Matsuo (Neurology Department, University of Utah School of Medicine, Salt Lake City, UT) Rationale: Mimicry between complex partial epileptic seizures (CPEs) and psychogenic dissociative attacks (PDAs) often leads to prolonged empiric antiepileptic drug (AED) treatment and perpetuates the unsatisfactory diagnosis of “pseudoseizures.” Whereas the relation between PDAs and a range of psychiatric disorders is yet to be defined, many patients with posttraumatic stress disorder (PTSD) exhibit PDAs. The objective was to compare sleep history between patients with CPE and PDAs, provisionally diagnosed as the first clinical encounter. Methods: The investigator arrived at a single clinical diagnosis, based on detailed description of clinical events provided by the patient and, when available, the witness, the chronology of their recurrences, precipitating factors, the results of diagnostic investigations, empiric AED treatment outcome, the psychosocial background, the history of psychosexual abuse, and a review of episodic psychic symptoms. Clinical sleep symptoms were identified according to American Sleep Disorders Association (ASDA) criteria. Dream experiences were recorded as a narrative summary of most emotionally charged dreams, and rated by an analog scale (no recall, nondistressing dreams, distressing dreams, and nightmare). The patient was not included, when CPEs and PDAs could not be differentiated, and dual diagnoses were not allowed. Results: Sleep history was available in 97 CPEs and 119 PDA cases between 1997 and 2002. Insomnia, associated in some with daytime hypersomnia, and sleep talking were twice often seen with PDAs. Contrast between the two groups was most apparent in the dream scale (Table 1), yet 36 (30%) of PDA patients either denied dream recall or were unable to describe dreams.
7
TABLE 1. Correlation between dream scale and diagnoses Dream scale CPE cases (%) PDA cases (%)
Indifferent Nondistressing Distressing Nightmare 77 30
14 8
5 24
4 38
CPE, complex partial epilepsy; PDA, psychogenic dissociative attack.
Conclusions: The history of psychosexual abuse was emphasized in differential diagnosis between CPEs and PDAs, even though the prevalence of a history in patients with CPE has not been defined. Although sleep history is more readily accessible, standardization of data collection is indicated to assure its reliability, when evaluation involves multiple interviews and multiple investigators. Further, dream scale is expected to change, reflecting the impact of disabling paroxysmal symptoms. This preliminary investigation was designed to show that the emotion associated could be assessed reliably, even though analysis of dream content may not assist the clinician. Multimodality diagnostic criteria of PDA can be expected to improve the clinician’s confidence, when assisted by clinical neurophysiologic investigations. Extended clinical monitoring with or without therapeutic interventions has led to dual diagnoses in some cases.
1.009 JUVENILE MYOCLONIC EPILEPSY AND IDIOPATHIC PHOTOSENSITIVE OCCIPITAL LOBE EPILEPSY: IS THERE OVERLAP 1 Isabella Taylor, 1 Carla Marini, 1,2 Samuel F. Berkovic, and 1,2,3 Ingrid E. Scheffer (1 Epilepsy Research Institute, Austin Health, University of Melbourne, Heidelberg, 2 Neurology, Royal Children’s Hospital, Parkville, and 3 Neurosciences, Monash Medical Centre, Clayton, Victoria, Australia) Rationale: Seizures triggered by photic stimulation are a common manifestation of both occipital epilepsies and idiopathic generalized epilepsies, especially juvenile myoclonic epilepsy (JME). Idiopathic photosensitive occipital epilepsy (IPOE) was described as a focal epilepsy with colorful elementary visual auras, often with prominent conscious tonic head and eye version; myoclonus is not a feature. All seizures are induced by photic stimuli. JME is characterized by myoclonic jerks (MJs) and generalized tonic–clonic seizures (GTCSs), with photosensitivity occurring in 30% of patients; visual auras are not recognized. JME is associated with generalized spike-and-wave (GSW) on EEG, whereas IPOE has occipital spikes with or without GSW. Intermittent photic stimulation (IPS) may produce GSW in both; in IPOE, occipital spikes are seen. We describe four families with overlap between JME and IPOE. Methods: Four families were identified in which two or more affected individuals had visual auras and electroclinical features of an idiopathic epilepsy. Available living affected family members underwent detailed electroclinical assessment. A 21-channel EEG including IPS and hyperventilation was performed. An additional Oz electrode was used in some cases. International classification was used to classify seizure types. Results: There were 12 affected individuals in four families, of whom 11 were female. Clinical onset was 8–21 years (mean, 11 years). Ten patients had visual auras including colored circles or bright flashes of light, six with conscious head version. Five also experienced MJ. Of the remaining individuals, one had MJ and occipital spikes; the other had GTCSs without visual aura or MJ. Of the 10 individuals with visual auras, seven had nonphotic-induced GTCSs. Of 10 patients with EEG studies, eight had GSW and six had occipital spikes. All had photosensitivity with GSW, and four had photic-induced occipital spikes. Conclusions: There is overlap between the clusters of clinical features used to diagnose IPOE and JME. Half of the affected individuals with visual aura had MJ; the former is characteristic of IPOE, and the latter of JME. Importantly, visual aura is not regarded as part of JME, nor is myoclonus part of IPOE, but our data emphasize that these symptoms may occur in both disorders. Moreover, two thirds of individuals with visual aura had spontaneous GTCSs; the latter feature is not described in IPOE. Understanding the genetic basis of these disorders must account for the striking female predominance, the variable phenotypes Epilepsia, Vol. 44, Suppl. 9, 2003
8
AES PROCEEDINGS
associated with photosensitivity, and the overlap of clinical features classically regarded as distinguishing focal and generalized syndromes. JME with photosensitivity and IPOE are clinically useful entities, but these family data suggest more clinical overlap than previously appreciated. This is probably due to shared genetic determinants in their complex genetic architecture. (Supported by National Health and Medical Research Council.) 1.010 SEIZURE SEMIOLOGY OF PARIETOOCCIPITAL LOBE SEIZURES Fawziah A. Bamogddam, Dileep Nair, Silvia Neme, and Prakash Kotagal (Neurology, The Cleveland Clinic Foundation, Cleveland, OH) Rationale: Parietooccipital lobe epilepsy is less frequently encountered than is temporal or frontal lobe epilepsy. The seizure semiologic features from parietooccipital lobe epilepsy may depend on whether patients have central, basal, lateral, or mesial parietooccipital lobe seizure onsets. The aim of this study was to analyze the seizure types seen in various forms of parietooccipital lobe epilepsy in patients who underwent epilepsy surgery and were subsequently seizure free. Methods: We included patients who had undergone video-EEG monitoring at our institution, and who were seizure free for ≥12 months after surgery. The videos of the patient’s seizures were reviewed separately by two of the authors, P.K. and F.B. Seizure symptoms were classified based on the Cleveland Clinic seizure classification. We classified our patients based on areas of resection in the parietooccipital lobe into central (n = 3), lateral (n = 3), mesial (n = 5), and basal (n = 1). Results: We identified 12 patients (four children and eight adults) with medically intractable parietooccipital lobe epilepsy, between 1994 and 2002, who were 12 months seizure free after surgery. Fifty-five seizures were analyzed. Auras were seen in seven patients, including somatosensory (n = 2), visual (n = 2), vertiginous (n = 2), epigastric (n = 1), autonomic (n = 1), and undefined (n = 2). Alimentary automatisms were seen in six patients, behavioral arrest (n = 5), bilateral eye blinking (n = 3), bilateral tonic (n = 3), unilateral clonic (n = 3), and head and/or eye version (n = 4). Duration of seizures ranged from 10 to 169 s (mean, 42 s). Somatosensory auras were seen in the central and lateral groups; visual auras, in the mesial group; and vertiginous auras, in the lateral group. Eye blinking was seen more in the lateral>mesial group, automatisms and behavioral arrest in the lateral and mesial groups, bilateral tonic in the central, lateral, and mesial groups, unilateral clonic in the mesial and lateral groups, and versive seizures in the mesial and lateral groups. Conclusions: The semiology of parietooccipital lobe seizures is influenced greatly by the location of the seizure focus and spread patterns. This can be helpful in localizing seizure onset within the parietooccipital lobe in candidates being evaluated for epilepsy surgery. 1.011 A NEW VIDEO-PROCESSING METHOD FOR QUANTITATIVE EVALUATION OF SEIZURE SEMIOLOGY 1 Jo˜ ao Paulo S. Cunha, 2 Christian Vollmar, 1 Zhanjian Li, 2 Berend Feddersen, and 2 Soheyl Noachtar (1 Electronics and Telecomm./IEETA, University of Aveiro, Aveiro, Portugal; and 2 Neurology, Klinikum Grosshadern, University of Munich, Munich, Germany) Rationale: The analysis of seizure semiology is well established in the evaluation of patients considered for epilepsy surgery. The clinical significance of several lateralizing ictal phenomena, such as head deviation, is still controversial. Visual inspection of ictal movements is usually performed by review of videocassettes in a qualitative way, which is prone to bias. Our objective was to evaluate the reliability of 2D quantitative video analysis by using standard video equipment. Methods: Videos were acquired by using the routine setting of an epilepsy-monitoring unit. They were recorded in analog S-VHS videocassettes with full PAL resolution. Selected video sequences were digitized to a PC computer with 720 × 525 pixels resolution, compressed into MPEG4 format, leading to a pixelsize of 5.4 × 3.8 mm. A reference system of reflective markers was designed to provide multiple reference segments distributed in the video image. Another set of reflective markers is placed in several patients’ anatomic positions. Based on these
Epilepsia, Vol. 44, Suppl. 9, 2003
markers and the geometry model of the scene, video-analysis software is used to perform several measurements of distances, linear velocity, angles, angular velocity, and movement frequency during epileptic seizures captured on digital video. Results: We paid particular attention to the upper part of the body, where most of the “movements of interest” are located. This approach allows quantitative measurements like angle, angular velocity, and head extension from standard 2D videos with an error 0.05). No patient had a history of coronary artery disease or arrhythmia. Obesity was not particularly more prevalent in any group (p > 0.05). Baseline HR was similar (p > 0.05) in psychogenic (74.6 ± 1.4) and TLE (72.7 ± 2.5) but was increased in aura events: 86.1 ± 2.6 beats/min (p < 0.05). During or after each psychogenic event, HR did not increase significantly: 77.6 ± 1.9 and 76 ± 1.6 beats/min, respectively (p > 0.05). HR also remained unchanged during or after each aura event: 91.2 ± 2.6 and 88.4 ± 3 beats/min (p > 0.05). However, during each TLE event, HR was significantly increased to 109.4 ± 3.2 beats/min (p < 0.05), returning to baseline in the postictal phase: 81.9 ± 2.4 beats/min. No events with bradycardia were observed. Fifty-five percent of the TLE events started on the right and 45% on the left, with 100% contralateral hemisphere spreading. Conclusions: HR increases significantly in dialeptic TLE and not in nonconvulsive (dialeptic-like) nonepileptic events. A baseline state of enhanced HR may precede the self-perception of dialeptic epileptic auras in TLE. HR analysis can differentiate dialeptic TLE from nonepileptic events and epileptic auras. 1.025 CLINICAL ASPECTS OF NONELECTROGENIC EPILEPTIFORM DISORDERS (NEEDS) OF PSYCHOGENIC ORIGIN Kalarickal J. Oommen and Mustafa Gursoy (Neurology, University of Oklahoma Health Sciences Center, Oklahoma City, OK) Rationale: Hysteroid epilepsy (Gowers), psychogenic seizures (ancient Egyptians and Greeks), pseudo seizures, and nonepileptic seizures are but a few of the terms used to describe epileptiform attacks without an abnormal, excessive electrical discharge (Jackson) from the brain. Today, NEEDS represent a group of physiologic and psychologic conditions that mimic epilepsy, and constitute 10 to 20% of children and ≤ 40% of adults admitted to epilepsy-monitoring units. Advances in diagnosis and management of epilepsy in the last quarter century have made the diagnosis easier, but have not dispelled the myths remaining in the literature regarding the clinical profile of this condition. Our objective was to study the demographic profile and analyze the clinical features of the psychogenic variety of NEEDS by video-EEG characteristics in a comprehensive epilepsy program setting. Methods: (a) Medical records of patients who underwent video-EEG monitoring at the Comprehensive Oklahoma Program for Epilepsy for a 6-year period, from the beginning in March 1995 to February 2000, were reviewed for demographic data, medical and psychiatric history and comorbid conditions; (b) the video-EEGs for the same period were
AES PROCEEDINGS reviewed to characterize the clinical features of the recorded events during monitoring; and (c) the seizure type, stereotypy, and time to record the first event were tabulated. Associated features such as pelvic thrusting, crying, body position at onset were scrutinized. Results: Of 844 patients monitored, 178 (21.1%) patients had NEEDS without a physiologic basis. The male (48) to female (130) ratio was 1:3.7. The age range was 35.9 ± 12 years, and duration of the disorder was 9.1 ± 12.2 years. 141 (79.2%) were receiving one through four antiepileptic drugs (AEDs), although only nine (5%) had documented epileptic seizures. Psychiatric history was present in 58 (32.6%), and family history of seizures, in 35 (19.6%). 103 (57.2%) had their first seizure within 24 h of admission (mean + SD = 24.8 ± 24.6 h). The most common clinical manifestation was generalized motor activity in 144 (80.1%), and the least common (4%) was loss of tone (atonic seizure). Other features included tremors (34.3%), partial motor activity (28.7%), sensory symptoms (25.8%), and staring and unresponsiveness (25.3%). Stereotypy was noted in 114 (64%), pelvic thrusting occurred in 38 (21.3%), associated crying in 20 (11.2%), and seizures started while in the prone position in 12 (6.7%) of patients. Interictal EEG abnormalities were noted in 45 (25.3%). Conclusions: (a) Psychogenic NEEDS occur more commonly in women (M/F ratio of 1:3.7); (b) generalized motor activity is the most (80.1%) and loss of tone (4%) the least common clinical presentation; (c) the vast majority (79.2%) were receiving one to four medications for an average of 9 years at a high cost to the patient; and (d) video-EEG monitoring yielded the diagnosis within 2 days of admission in most patients and can help patients save the cost of medications and avoid their long-term toxic side effects. 1.026 BODY HABITUS AND NONEPILEPTIC SEIZURES: IS THERE A LINK? Anna Vinter Marquez, Michelle Apperson, Sarah T. Farias, Suzanne Koopmans, Anthony R. Lima III, Alan Shatzel, and Taoufik Alsaadi (Department of Neurology, University of California, Davis Medical System, Sacramento, CA) Rationale: We observed anecdotally that patients admitted for videoEEG (vEEG) monitoring who we diagnosed with nonepileptic seizures (NESs) seemed to have a larger body habitus than similar patients with epilepsy. NESs are often considered physical manifestations of psychological distress. There is some limited evidence that weight problems are more common in people with psychiatric illness, and we postulated that there might be a higher prevalence of overweight and obesity in patients with NESs. The goal of this study was to test our hypothesis that there was a measurable difference in body mass index (BMI) in patients with NESs compared with their epileptic counterparts. Methods: Study subjects and controls were chosen from the list of all patients admitted to our inpatient video-EEG (vEEG) monitoring unit between January 2000 and March 2003. The diagnosis of NESs was made only when the vEEG-recorded events were identified by the patient and/or family as typical, and those events were not accompanied by any EEG abnormalities. All patients older than 18 years who we diagnosed with NESs were included in the study. We selected age- and gender-matched controls whose diagnosis of epilepsy was confirmed by vEEG. Patients who had both NESs and epileptic seizures were excluded. The height and weight of each patient was collected from the patient’s intake history and physical, or from the hospital pharmacy computer system. These data were entered into an Excel spreadsheet and the BMI was calculated. Data were analyzed by using the SPSS for Windows statistics program. Results: The mean BMI of NES patients was 29.9 (SD, 9.1), and the mean BMI of patients with epilepsy was 26.4 (SD, 8.7). There was no statistically significant difference between the two groups (p = 0.10; t test). Categoric data analysis also showed no statistically significant difference between the groups: 22 of 36 (61%) NES patients were overweight or obese (defined as a BMI ≥25), compared with 15 of 35 (43%) epilepsy patients (p = 0.16, Fisher’s exact test). Conclusions: We found no significant difference in BMI between patients with NESs and their age- and gender-matched epilepsy counterparts. These results emphasize the importance of avoiding bias and stereotyping when we evaluate and treat patients with unexplained or
13
refractory seizures. One factor we did not control for was the use of antiepileptic drugs, which could potentially have affected the results. Of interest, the percentage of overweight and obese NES patients in our study (61%) is almost identical to that of the general adult American population–64% as measured by the National Center for Health Statistics NHANES survey of 1999–2000.
1.027 MOST PATIENTS WITH PSYCHOGENIC NONEPILEPTIC SEIZURES DO NOT EXERT VALID EFFORT ON NEUROCOGNITIVE TESTING 1,3 David J. Williamson, 2 Daniel L. Drane, 2 Elizabeth S. Stroup, 2 John W. Miller, and 2 Mark D. Holmes (1 Clinical Communications, OrthoMcNeil Pharmaceutical, Inc., Clearwater, and 2 School of Medicine, University of Washington, Seattle, WA; and 3 Department of Clinical & Health Psychology, University of Florida, Gainesville, FL) Rationale: Patients with objective evidence of epileptic seizures (ESs) are often difficult to separate from those with psychogenic nonepileptic seizures (PNESs) on the basis of neurocognitive testing. This difficulty has been attributed to subtle neurologic damage or diminished ability to handle emotional stress (Wilkus, Dodrill, & Thompson, 1984; Binder et al., 1998). Green et al. (2001) recently demonstrated that effort, as quantified by modern symptom validity testing, often explains more variance in neurocognitive test performance than does the presence of the neurocognitive disorder itself. To date, only one group has studied the effects of effort on testing in patients with ESs and PNESs (Binder et al., 1998), but these studies were done with tests that may be less sensitive than more recently validated tests. This study is meant to reexamine the role that objectively quantified effort plays in mediating performance of these groups on neurocognitive testing. Methods: Fifty-seven patients consecutively referred for continuous video-EEG monitoring for evaluation of apparent uncontrolled seizures at the University of Washington were administered the Word Memory Test as part of a comprehensive epilepsy neuropsychological evaluation. Patients were classified on the basis of their ictal EEG recordings and behavioral presentation as experiencing ESs (n = 32; evidence of definite EEG abnormalities), PNESs (n = 14; episodes of unresponsiveness or behavioral abnormality in the absence of EEG changes), nonepileptic spells of other origin (NESO; n = 5; spells related to other causes, such as syncope), or indeterminate spells (IS; n = 6; no spells during monitoring or subjective feelings only, in the absence of EEG abnormality and without unresponsiveness or behavioral abnormality). Results: The majority (64%) of PNES patients “failed” symptom validity testing. In marked contrast, the ESs and indeterminate groups failed such tests at rates of 13% for ESs, 17% for ISs, and 0% for NESOs. These differences are far greater than those expected by chance (p < 0.001) and far exceed the rates of symptom validity failures reported in other studies using older techniques. Conclusions: For the majority of PNES patients in this sample, behavioral, or emotional difficulties are more likely to explain neurocognitive test performance than actual neurocognitive deficits. These results strongly suggest that newer tests of symptom validity should be used in neuropsychological evaluations of patients with apparent uncontrolled seizures. Further exploration of these results may assist in the psychometric differentiation of patients with PNESs from patients with ESs. Moreover, we should gain a better understanding of the extent to which the difficulty differentiating these groups may be a function of effort rather than the hypothesized occult neuropathology reported in the histories of patients with PNES.
1.028 TREATMENT OF NONEPILEPTIC SEIZURES BY USING A PSYCHOEDUCATIONAL APPROACH IN A GROUP SETTING Charles M. Zaroff, Lorna Myers, and Daniel Luciano (Department of Neurology, NYU Comprehensive Epilepsy Center, New York, NY) Rationale: Nonepileptic seizures (NESs) refer to seizure-like events not associated with epilepsy. There is currently no standard of care for the treatment of NESs. Presentation of the diagnosis in an informative and nonjudgmental manner may in itself reduce or even resolve NESs
Epilepsia, Vol. 44, Suppl. 9, 2003
14
AES PROCEEDINGS
whereas inadequate/unclear diagnostic presentation can lead to avoidance of necessary follow-up treatment. Traditionally, psychopharmacology in the form of psychotropic medication, and/or psychotherapy have been used with NES patients. The need for proper support and treatment contained in such approaches is not in doubt. However, outcome measures are lacking, as a reduction in NESs may be associated with symptom substitution. The current study aimed to derive a specific treatment modality for NES in which illness-related symptoms were assessed and monitored after treatment completion. Methods: After diagnosis of NESs based on seizures captured during video-EEG monitoring, patients were interviewed by the staff’s clinical psychologist. Psychiatric interview was conducted, and patients were referred for an outpatient psychotherapy treatment. Before the initial session, patients completed standardized questionnaires to assess quality of life (Quality of Life in Epilepsy-31; QOLIE-31), coping skills (Coping Inventory for Stressful Situations; CISS), traumatic symptoms (Davidson Trauma Scale; DTS), and dissociative tendencies (Curious Experiences Survey; CES). A time-limited psychotherapy group was initiated by using a psychoeducational approach. Group leaders provided instructional handouts at the onset of each of the first nine sessions describing NES, anger expression and management, NES comobordities, trauma and abuse, and stress-coping mechanisms, among others. A final group session was provided for review and was followed by completion of questionnaires administered pretreatment. Results: Results of standardized questionnaires revealed a significant decline in posttraumatic symptoms on the DTS [t(3) = 3.922, p = 0.029]. Nonsignificant trends were observed indicating decreased dissociative symptoms on the CES and increased coping skills and quality of life on the CISS and QOLIE-31, respectively. Conclusions: Group therapy with a psychoeducational approach is warranted for individuals with NESs. Individuals report significant declines in post-traumatic symptoms. Reduction in dissociative symptoms occurred along with improved quality of life and coping strategies. Psychoeducation served not only to provide information on NES diagnosis but also helped to elucidate the mechanism by which NESs may develop, thus empowering individuals and contributing to seizure control or remission. Group therapy aided in providing support and proved to be a powerful means of identification.
1.029 RESPONSE TO TREATMENT IN NEWLY DIAGNOSED EPILEPSY Martin J. Brodie and Rajiv Mohanraj (Epilepsy Unit, Divison of Cardiovascular and Medical Sciences, Western Infirmary, Glasgow, Scotland, United Kingdom) Rationale: Prediction of the response to antiepileptic drug (AED) treatment in newly diagnosed epilepsy could help to refine pharmacologic decision making and provide a rationale for early surgical intervention. Methods: Outcomes in unselected patients in whom a diagnosis of epilepsy was made and first AED begun at the Epilepsy Unit in Glasgow since 1982 will be presented. Results: Data are available from 780 previously untreated patients (males, 52%; females, 48%; median age, 29 years, range, 9–93 years; median follow-up, 79 months, range, 2–22 years). Overall 504 (65%) patients reported ≥1 years’ seizure freedom, 42 (5%) of whom subsequently relapsed after a median 25 months (range, 12–97 months) and never again become seizure free. Adequate seizure control was never achieved in the remaining 276 (35%) patients; 92% of the 462 (60%) patients achieving remission did so within 3 years of starting treatment. Of these, 53% never had another seizure after taking the first dose of AED treatment. Of the patients in remission, 92% received monotherapy with the first (n = 359) or second (n = 53) AED. Only 37 (8%) patients went into remission with duotherapy, whereas just one patient each remained controlled with three or four drugs, respectively. Patients with idiopathic epilepsy (n = 222; 66% remission) did better than those with cryptogenic (n = 314; 57% remission, p < 0.05) or symptomatic (n = 244; 56% remission, p < 0.05) epilepsy. Outcomes in patients with cerebrovascular disease (n = 63; 70% remission, p < 0.01) and cerebral atrophy (n = 42; 71% remission; p = 0.028) did better, whereas those with posttraumatic epilepsy (n = 65; 35% remission; p < 0.001) did
Epilepsia, Vol. 44, Suppl. 9, 2003
worse than the remainder of the symptomatic group. Remission rates in patients with underlying cortical dysplasia (n = 15; 60% remission), hippocampal sclerosis (n = 14; 50% remission), and brain tumours (n = 25; 57% remission) were not significantly different from other patients with symptomatic epilepsy. Conclusions: This analysis supports our previous observation (1) of two populations of epilepsy patients (i.e., those that respond to AED therapy usually with the first- or second-choice treatment and those with refractory epilepsy de novo in whom a different therapeutic strategy may be required). Outcomes in patients with newly diagnosed epilepsy on the basis of underlying cortical dysplasia and hippocampal sclerosis were no worse than those with other causes of symptomatic epilepsy. A patient who does not attain seizure control with the first two to three AED regimens within 2–3 years of starting treatment is unlikely ever to have a useful period of remission and can be said to have refractory epilepsy. REFERENCE 1. Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med 2000; 342:314–9.
1.030 ANTICONVULSANT PROPHYLAXIS IN NEUROSURGERY: A NATIONWIDE RETROSPECTIVE MULTICENTRIC SURVEY 1 Bertrand C. Devaux, 2 Jean-Pierre Castel, 3 Herve Vespignani, 1 Elisabeth Landre, 4 Florian Von Raison, 4 Aurelie Grun-Overdyking, 1 Francois-Xavier Roux, and Epinec Investigator Group (1 Department of Neurosurgery, Sainte-Anne Hospital, Paris, 2 Department of Neurosurgery, Pellegrin Hospital, Bordeaux, 3 Department of Neurology, SaintJulien Hospital, Nancy, and 4 Pfizer Global Pharmaceutics, Paris, France) Rationale: Anticonvulsant prophylaxis (AP) is recommended in various neurosurgic diseases, but no definite agreement is obtained regarding indication and choice of an antiepileptic drug (AED), doses, and treatment duration. Methods: A multicentric retrospective nonrandomized survey was conducted to (a) assess current practice regarding AP in most French neurosurgical centers; (b) document the incidence of seizures in a given disease; and (c) provide further recommendations for seizure prevention and treatment in neurosurgical practice. The survey enrolled 67 academic, general hospitals, and private centers of neurosurgery. One hundred consecutive patients admitted in each participating center from March 1, 1998, for the treatment of intracranial supratentorial disease were entered in the study. The SAS software was used for descriptive statistics, and a stepwise logistic regression model assessed the weight of different risk factors for seizures. Results: Included in the main descriptive analysis were 5,852 patients. Mean age was 50.8 years; 63.1% of them had a surgical procedure; 45.5% had a craniotomy. Main diseases were tumoral (35.4%), traumatic (34.7%), and vascular (20.5%). Mean observation period was 301 days; 57.1% of patients received an AED at any moment of the observation period (main AEDs were valproate, carbamazepine, and phenobarbital). Adverse effects were observed in 3.7% of patients. Among 5,037 patients evaluable for epileptic seizures, a seizure occurred in 21.6% of patients at any time of the observation period, whereas 78.4% had no seizure. AED was prescribed in prevention in 44.4% of all cases, and after a first seizure, in 9%. Success rates (percentage of patients receiving prophylactic AED and having no seizure) were 90.6% for traumatic, 81.6% for vascular, 70.4% for tumoral, and 66.6% for infectious lesion. Failure rates (percentage of patients receiving prophylactic AED and having at least a seizure) were 9.4%, 18.4%, 29.6%, and 33.3%, respectively. Factors positively correlated with seizure occurrence were tumoral lesion, long-term antiepileptic treatment, craniotomy, and cortical/subcortical lesion location. Factors negatively correlated with seizure occurrence were traumatic lesion, hydrocephalus, age younger than 16 or older than 60 years. Odds ratios were estimated at 3 for long-term AED treatment and corticosubcortical lesion location, 2 for chronic alcoholism, 1.8 for craniotomy, and 1.4 for tumoral lesions. Conclusions: The present study demonstrated that relative risks for seizure occurrence are different in various neurosurgical diseases.
AES PROCEEDINGS Therapeutic management should take into account that patients having (a) corticosubcortical lesion, (b) chronic alcoholism, (c) craniotomy, and (4) cerebral tumor are more at risk of seizure. (Supported by Pfizer Global Pharmaceutics France.) 1.031 THE IMPACT OF SEIZURE CLUSTERING ON ADULTS WITH EPILEPSY Sheryl Haut, Shayna Aster, and Shlomo Shinnar (Comprehensive Epilepsy Management Center, Montefiore Medical Center and the Albert Einstein College of Medicine, Bronx, NY) Rationale: Patients with epilepsy often experience seizures in clusters. We performed preliminary analysis of a prospective study of seizure clustering, to examine risk factors for seizure clustering by history, and the effect of clustering on anxiety, depression, and quality-of-life measures. Methods: Detailed intake questionnaires were administered to all subjects. Seizure clustering was defined as three or more seizures in a 24-h period. Subjects with seizure clustering were subdivided into two groups: those whose typical seizure pattern was clustering, and those with a history of ever having experienced a seizure cluster. Magnetic resonance imaging (MRI) and EEG data were obtained by chart review. Lesional epilepsy was defined as epilepsy associated with significant head trauma, neoplasm, stroke, or vascular malformation, excluding mesial temporal sclerosis (MTS). All subjects completed Beck Anxiety Inventories (BAIs), Beck Depression Inventories (BDIs), and Quality of Life in Epilepsy (QOLIE) testing. Results: Of the initial 114 subjects recruited, complete data were available for 103 subjects. Overall, 41 (39%) had a history of clustering, including 28 (27%) with a clustering pattern. Lesional epilepsy was significantly associated with clustering (p = 0.019) and with clustering pattern (p = 0.017); however, specific MRI findings or EEG abnormalities were not associated with clustering. Poor seizure control was significantly associated with a history of clustering pattern (p = 0.025), but not with overall clustering. In this selected population, there was no significant association between lesional epilepsy and poor control. Mean BAI scores (15.1, cluster group; 15.0, cluster pattern group; 11.7, nonclustered group) and mean BDI scores (17.2, cluster group; 17.3, cluster pattern group; 11.4, nonclustered group) were higher for any history of seizure clustering, but only the mean BDI score in subjects who had ever experienced seizure clustering reached statistical significance (p = 0.015). Similarly, mean QOLIE scores (55.6, cluster group; 62.1, nonclustered group) were lower for any history of seizure clustering, but were of borderline statistical significance (p = 0.086). Conclusions: These preliminary results demonstrate the adverse impact of seizure clustering on prognosis of epilepsy and on quality of life. Those who experienced seizure clustering as the typical seizure pattern are most affected. Further analysis is planned with a larger sample and longer follow-up to evaluate the specific predictors of poor and favorable outcome as well as identify potential areas where intervention may be helpful. [Supported by K23 NS02192 (P.I., S.H.).] 1.032 SEIZURES IN NEUROCYSTICERCOSIS AT LAC+USC MEDICAL CENTER David Y. Ko, Abbas Medhdi, Iceland Huston, and Soma Sahai (Neurology, University of Southern California, Los Angeles, CA) Rationale: Seizures are one of the most common symptoms of neurocysticercosis (NC). This study looked at the prevalence of seziures in patients with the diagnosis of NC at LAC+USC Medical Center in the mid 1990s. A comparison of this study is made with the results of the prior studies of NC at the same medical center from the 1970s and 1980s. Methods: This was a retrospective chart review of inpatients or patients with emergency department (ED) visits with the diagnosis of NC at LAC+USC Medical Center, a large public hospital in the city of Los Angeles where NC is a common disorder. The period of 1995–1998 was sampled. The patients’ diagnosis of NC was mainly based on neuroimaging, as well as clinical criteria. Results: For the 4-year period in the 1990s, 469 patients were evaluted in the ED or admitted at LAC+USC Medical Center with the diagnosis
15
of NC. Nearly all the patients were hispanic, with 56% being male and 44% female, with a mean age of 35 years. The most common clinical presentation was seizures in 192 (40%) patients, followed by signs of increaesed intracranial pressure in 26.8% and headache in 13.8%. The study of cases during 1970–1980 had 127 cases of NC, of whom 55.1% had seizures (1). The study of cases during 1981–1986 had 238 cases, of whom 56.3% had seizure disorder (2). The other alarming fact is the number of cases of NC, which increased from 12.7/per year in the 1970s to 117.25/per year in the 1990s in the same institution, a 10-fold increase over a 30-year period. In the period 1918–1965, there were 14 with diagnosis of NC among 82,700 autopsy cases at this instituiton. This study and the two prior studies were mainly of inpatients, but there were a signifciant number of outpatient cases, thus underestimating the prevalance of NC in the LA County health care system. Because of the large public medical health care system at LA County with multiple outlying medical clinics, it was difficult to get outpatient charts of these patients to determine what percentage of these patients with seizures had epilepsy. Nevertheless, cysticercosis is one of the most common causes of symptomatic epilepsy seen at LAC+USC. Of interest is that a number of intractable epilepsy cases are seen at the presurgical conference; there are a number of patients with NC who also have mesial temporal lobe sclerosis, thus having dual pathology. Conclusions: The incidence of seizures due to NC is slightly less than that in two prior studies at the same institution but is still high at 40%. With the fact the overall number of cases of NC is increasing—almost 10 times the number of cases now compared with 30 years ago—still makes NC one of the most common causes of seizures and epilepsy in Southern California. REFERENCES 1. McCormick G, Zee C, et al. Cysticercosis cerebri: review of 127 cases. Arch Neurol 1982;39:534–9. 2. Scharf David: Neurocysticercosis: two hundred thirty eight cases from a California hospital. Arch Neurol 1988;45:777–80. 1.033 PREDICTING OUTCOMES IN NEWLY DIAGNOSED EPILEPSY Rajiv Mohanraj and Martin J. Brodie (Epilepsy Unit, Division of Cardiovascular and Medical Sciences, Western Infirmary, Glasgow, Scotland, United Kingdom) Rationale: The majority of patients diagnosed with epilepsy gain complete control of seizures with antiepileptic drug (AED) therapy. Identifying those who are unlikely to enter remission early in the course of the disorder can help target specialist intervention including epilepsy surgery. Methods: The 890 patients diagnosed with epilepsy and prescribed their first AED between August 1982 and May 2001 were evaluated longitudinally for treatment outcomes. Factors of prognostic importance were collected for each patient by review of research case notes. Results: Treatment outcomes were known for 780 (88%) patients newly diagnosed with epilepsy (405 male, 375 female; median age, 29 years); 222 patients had idiopathic generalised epilepsy, and 558 had localisation-related epilepsy. Longitudinal evaluation revealed three main outcome groups: patients who had been seizure free for >12 months and were in that state at the end of follow-up (remission), those who never had a 12-month period of seizure freedom (uncontrolled); and those who were seizure free for ≥12 months before becoming uncontrolled (relapse). At the end of a median period of 79 months (range, 24–240 months) of follow-up, 462 (59%) patients were in remission, 276 (35%) were uncontrolled, and 42 (6%) had relapsed. Patients who entered remission did so within the first year of treatment in 83% of cases. Patients with idiopathic generalised epilepsy were more likely to enter remission than were those with localisation-related epilepsy [odds ratio (OR), 1.47; p = 0.019]. Analysis by age groups showed that elderly patients (older than 65 years at diagnosis, n = 90, 85% remission) were significantly more likely to enter remission than were other adults (age 20–64, n = 520, 53% remission) or adolescents (age younger than 20, n = 170, 65% remission). History of head injury (n = 86; OR, 3.11; p < 0.001), febrile convulsions (n = 35; OR, 2.69; p = 0.02), family history
Epilepsia, Vol. 44, Suppl. 9, 2003
16
AES PROCEEDINGS
of epilepsy (n = 91; OR, 2.02; p = 0.02), partial seizures with or without secondary generalisation (n = 357; OR, 1.54; p = 0.002), presence of psychiatric comorbidity (OR, 2.13; p < 0.001), younger mean age at onset (p = 0.001), and large number of pretreatment seizures (p < 0.001) were associated with the probability of not achieving remission. Duration of epilepsy before starting treatment, seizure clustering, previous status epilepticus, results of brain imaging, surface electroencephalographic findings, and presence of learning disability or neurologic deficit were found not to have a significant correlation with outcome. Multivariate analysis of all prognostic factors will be presented. Conclusions: Prediction of outcomes in individual patients with epilepsy is possible by using information available early in the course of the disorder. A clinical prediction model for each syndrome, appropriately validated, will help identify patients requiring early specialist intervention.
1.034 FREQUENCY OF SEIZURES IN MULTIPLE SCLEROSIS PATIENTS TREATED WITH INTRATHECAL BACLOFEN 1 Stephan U. Schuele, 1 Christoph Kellinghaus, 1 Francois A. Bethoux, 2 Nicholas Boulis, and 1 Tobias Loddenkemper (1 Neurology and 2 Neurosurgery, Cleveland Clinic Foundation, Cleveland, OH) Rationale: Baclofen has been associated with the occurrence of epileptic seizures and status epilepticus in a small number of cases. An increased frequency of epileptic events was seen in patients with posttraumatic brain injury taking intrathecal baclofen (ITB), predominantly after initiation of therapy or dose adjustements. The goal of our study was to investigate the incidence of epileptic seizures in a series of 99 patients with multiple sclerosis (MS) treated with ITB. Methods: Data were obtained from the Mellen Center ITB Therapy Registry. The registry contains records of all patients since 1990 with a definite diagnosis of MS who had an ITB pump implanted. Results: Ninety-nine patients, 31 male and 68 female, underwent implantation of a baclofen pump between 1990 and 2002. Mean age at pump implantation was 46.3 years; mean disease duration, 16 years. Mean baclofen rate at the most recent visit was 327.7 µg/day; 89% of patients were nonambulatory with an EDSS ≥7.0. A total of 297 cumulative pump years was reviewed. Epileptic seizures were noted in 11 patients, two male and nine female. Mean age at pump implantation in this group was 47.1 years; mean disease duration, 17 years. The mean ITB rate at onset of seizures was 382.5 µg/day and 372.3 at last visit. Patients were followed up on average 4 years after pump implantation. Four patients had preexisting epilepsy, but no increase in seizure frequency was detected during the follow-up period. Seven patients had new onset of epileptic seizures on average 2 years and 3 months after initiation of ITB therapy. Five patients had single seizures, all associated with additional aggravating factors: two patients had a febrile illness; baclofen overdose, perioperative state, and low serum sodium were seen in one patient, respectively. Two patients developed recurrent seizures, >3 times per year. Three patients developed status epilepticus. Conclusions: In our registry of 99 patients, we observed the new onset of epileptic seizures in 7% of patients within an average follow-up period of 4 years, often associated with additional factors increasing their seizure susceptibility. None of the four patients with preexisting epilepsy had worsening of their seizure frequency. In epidemiologic studies, patients with MS show a threefold increase in risk for developing epilepsy compared with the general population. The cumulative risk for developing epilepsy after onset of MS symptoms is estimated at 3.1% by 15 years’ disease duration. ITB therapy seems to bear a definite risk for epileptic seizures in MS patients. [Supported by Innovative Medizinische Forschung, WWU Muenster (FoeKz. LO 610101, Dr. Loddenkemper; FoeKz. KE 620201, Dr. Kellinghaus).]
1.035 IS IT DEMENTIA OR EPILEPSY? Ahmed Al-Sadat, Jagdish Shah, Darren Fuerst, Dorothy Muriel, and Craig Watson (Department of Neurology, Wayne State University School of Medicine, Detroit, MI)
Epilepsia, Vol. 44, Suppl. 9, 2003
Rationale: Dementia is a common neuropsychiatric condition and occurs in 5–11% of the population by age 65 and in ≤50% beyond age 85 years. Thus a careful search for reversible causes of cognitive impairment is mandatory. One such example is newly diagnosed or poorly controlled epilepsy. Amnesia of the event is a common feature of complex partial seizures (CPSs). Transient epileptic amnesia and epilepsy presenting as dementia have been occasionally reported. We report three patients masquerading as dementia, who had with memory and cognitive impairment caused by frequent CPSs and/or temporal lobe electrographic seizures (EEG Sz). Methods: Of 600 consecutive patients admitted to the EEG/video monitoring unit, we identified three patients who had with memory disturbance and cognitive decline. All patients underwent scalp-sphenoidal EEG/video monitoring, fluorodeoxyglucose–positron emission tomography (FDG-PET) scanning, magnetic resonance imaging (MRI), and neuropsychological evaluation. Results: On EEG/video monitoring, all three patients had frequent CPSs or EEG Sz. Seizure frequency ranged from one to seven seizures per hour. No obvious clinical manifestations were seen on video recording. Subsequently, all three patients were loaded with intravenous antiepileptic medication, after which the EEG showed no further electrographic seizures. Family members noted significant improvement in memory and cognition. MRI results: Normal in one (patient 1), left temporal lobe encephalomalacia with left hippocampal sclerosis (HS) (patient 2), and left HS with increased signal in left temporal lobe white matter (patient 3). FDG-PET results: Left temporal hypometabolism (patient 1), left temporoparietal hypometabolism (patient 2), and left temporoparietal hypometabolism with left thalamic hypometabolism (patient 3). Conclusions: Frequent CPSs or temporal lobe EEG Sz may present with memory and cognitive impairement. The usual clinical manifestations of CPSs may be absent in these patients, and thus they may pose a diagnostic challange. A high degree of suspicion and awareness of this phenomenon may lead to proper diagnosis and managment. EEG/video monitoring is an invaluable tool in this setting.
1.036 TOLERABILITY OF LEVETIRACETAM IN ELDERLY PATIENTS WITH COGNITIVE DISORDERS, ANXIETY, AND EPILEPSY 1 Joyce A. Cramer, 2 Katrien De Rue, 3 Ilo E. Leppik, and 4 Guenter Kraemer (1 Psychiatry, Yale University School of Medicine, West Haven, CT; 2 UCB Pharma, Brussels, Belgium; 3 MINCEP Epilepsy Care, Minneapolis, MN; and 4 Swiss Epilepsy Center, Zurich, Switzerland) Rationale: The purpose of this analysis was to compare treatmentemergent adverse events (TEAE) related to use of levetiracetam (LEV) reported by young and elderly patients with cognitive disorders, anxiety, and epilepsy. Methods: The LEV database includes reports of TEAE from placebocontrolled clinical trials from which patients were grouped as young (younger than 65 years) or elderly (older than 65 years). Patients were diagnosed with a cognitive disorder (419 young; 319 elderly), an anxiety disorder (1,440 young; 169 elderly), or localization-related epilepsy (1,023, all young) who participated in trials lasting ≤16 weeks. Results: The TEAEs occurring most frequently in the LEV-treated groups were abdominal pain, asthenia, headache, anorexia, weight loss, dizziness, insomnia, somnolence, and tremor. The only significant differences in TEAEs were seen between young and elderly groups with anxiety disorders (>3% higher for LEV- than for placebo-treated patients) in headache (difference between LEV and placebo: 5.2%, elderly; −0.9%, young; p = 0.041), and tremor (5.2%, −0.5%, respectively; p = 0.022) and between young anxiety patients and young epilepsy patients for somnolence (−0.7%, 5.4%, respectively; p = 0.036). For the other TEAEs, there was no evidence for consistent differences between young and elderly patients and between patients with different CNS disorders. The preponderance of adverse events was reported during the first 4 weeks of treatment for all groups. The proportion of LEV- and placebo-treated patients discontinuing was low, exceeding 2% higher in LEV than placebo groups only for somnolence (all young LEV-treated groups), dizziness (both elderly LEV-treated groups), tremor (elderly LEV-treated anxiety patients only), and asthenia (elderly LEV-treated cognition group). Overall, elderly anxiety and young epilepsy patients
AES PROCEEDINGS had the highest incidences of TEAE, whereas those with cognitive disorders and young anxiety patients had the lowest incidences of TEAE. However, all of the epilepsy patients were receiving at least one other antiepileptic drug that may have contributed to this finding. Conclusions: Overall, LEV was well tolerated by all patient groups. The favorable adverse event profile suggests that LEV might be suitable for use by elderly patients. (Supported by UCB Pharma.) 1.037 USE OF EEG AND DIAGNOSIS OF EPILEPTIC SEIZURES IN ELDERLY IN A GERIATRIC CENTER SETTING: A DESCRIPTIVE PILOT MULTICENTER STUDY 1 Bernard Gueguen, 1 Christine Soufflet, 2 Florian Von Raison, and 3 Jean Marie Vetel (1 Dept. of Clinical Neurophysiology, Centre Hospitalier Sainte-Anne, and 2 Scientific and Medical Department, Pfizer PGP, Paris, and 3 Dept of Gerontology, Centre Hospitalier du Mans, Le Mans, France) Rationale: Clinical diagnosis of epilepsy remains difficult in elderly. The aim of the present study was to obtain prospective data to justify the use of the electroencephalogram (EEG) in the emergency room when facing with symptoms suggesting epileptic seizure in elderly people. Methods: This prospective observational study was carried out in line with inpatients in five geriatric centres. Full consent from patients/family was obtained before entering in the study. To be included, patients (older than 70 years) had to have clinical signs suggesting epilepsy and no previous epilepsy diagnosis. There were no inclusion/exclusion criteria regarding concomitant diagnoses or treatment. The EEG was to be performed 3 h); 25 patients were then included for further analysis (23 female; two male; mean age, 87 years; range, 72–103). According to the first EEG reading, 96% of analyzed patients showed at least one electrical abnormality: spikes (41%), spike-and-wave complexes typical (23%) or atypical (41%), all of these abnormalities being focal (41%) or diffuse (64%), intermittent (27%), or in bursts (9%). Patients could have more than one electrical abnormality. The first expert analysis of EEG confirmed the presence of typical electrical abnormalities (76%) and atypical abnormalities (16%). The a posteriori review of all EEGs by the expert showed typical epileptiform abnormalities (20%), atypical epileptiform activities (44%), and nonepileptiform abnormalities (32%). One patient had a noninterpretable EEG, but typical epileptic symptoms. Clinical manifestations of seizure were isolated mental confusion (36%), tonic–clonic seizures (36%), loss of consciousness with fall (32%), and loss of contact (20%). Postictal confusion was found in 20% of patients, and incontinence, in 16%. Hyper- or hypotonia, tongue biting, acute psychic signs, focal deficit syndrome, autonomic disturbance, motor automatism (wandering), and chewing were reported in one patient each. Conclusions: Clinical symptoms and EEGs are often atypical in elderly people. EEG performed within the first 3 h after the onset of a clinical event is a valuable tool to confirm its epileptic nature and may help to initiate or delay an antiepileptic treatment. EEG has to be more easily accessible in emergency units and elderly care structures. (Supported by Pfizer.) 1.038 ZONISAMIDE ADJUNCT THERAPY FOR ELDERLY INTRACTABLE SEIZURE IN VA OUTPATIENT CLINIC Hisanori Hasegawa (Neurology Service, Aleda E. Lutz Saginaw VA Medical Center, Saginaw, MI) Rationale: Zonisamide (ZNS) was introduced in Japan in 1991 and has been acknowledged as a safe and efficient anticonvulsant (AED). There was a suggestion that ZNS was safe for and better tolerated in elderly patients in a postmarket multicenter safety study in Japan. Its longer elimination half-life allows a reliable once-a-day drug regimen.
17
Side effects of ZNS included gastrointestinal upset (GI), dizziness, skin rash, and acute psychosis in Japanese articles, but the incidence of them is believed less in older patients for unknown reasons. There has been no report for utilization of ZNS exclusively in the elderly epilepsy population. Methods: This is a retrospective outcome review in the outpatient neurology clinic in Saginaw VA Medical Center. ZNS was used in those patients whose seizures was not under control (more than one episode per month) with other AEDs previously. Twenty consecutive patients with complex partial seizures, whose seizure control was resistant to at least three preexisting AEDs, age range from 49 to 89 years (average, 64.1), were identified. ZNS was added to the concurrent AEDs in all cases, expecting adjunctive benefit. The starting dose was 100 mg p.o. qd. The patients were followed up for 2 months, and the dose was increased to 100 mg b.i.d. or 200 mg qd, if patients could tolerate ZNS, but clinical improvement was not sufficient. The dose was not increased if patients could attain significant seizure-frequency reduction >50%. Follow-up period varied from 4 to 12 months. Results: Among the identified 20 patients, four patients had postanoxic etiology, three had posttrauma, three had poststroke etiology, and 10 had unknown causes. Three patients became seizure free; 16 (80%) patients of 20 experienced >50% reduction in seizure frequency with the low adjunctive ZNS with no ≤200 mg/day. Clinical improvement did not correlate with etiology. Three patients discontinued ZNS. One experienced skin rash, one experienced visual hallucination, and the other had excessive drowsiness. None experienced paradoxic aggravation of seizure frequency. Among the 16 patients whose seizures improved, five patients experienced mild drowsiness and GI upset, but they continued ZNS. No patients experienced kidney stones or acute psychosis. Once-a-day regimen was allowed, and no single patient had a compliance problem. Conclusions: ZNS adjunctive therapy was well tolerated and successful in elderly patients with intractable complex partial seizure disorder in a neurology outpatient clinic. Seizure-responder rate, defined as >50% seizure-frequency reduction, was unexpectedly good in elderly epileptsy patients with minimal side effects. Once-a-day regimen is easy for elderly patients and contributes to better compliance. The fndings of this preliminary retrospective study suggest that ZNS adjunction is effective in low dose with fewer adverse effects in elderly epilepsy patients.
1.039 ETIOLOGY AND TYPE OF THE FIRST-EVER EPILEPTIC SEIZURES IN THE ELDERLY Krystyna Niedzielska, Aleksandra Wierzbicka, Maria BaranskaGieruszczak, Wanda Lojkowska, Iwona Kurkowska, Waldemar Lechowicz, Lidia Wolkow, and Danuta Ryglewicz (Department of Clinical Neurophysiology, Institute of Psychiatry and Neurology, Warsaw, Poland) Rationale: In last decades, rapid growth of the elderly population and progressively higher incidence of epilepsy in this age group has been reported. The diagnosis of epileptic seizures at admission to hospital brings a variety of specific problems. We are often faced with difficulties in obtaining the reliable reports on seizure course, and the interictal discharges in EEG are found less frequently than in the younger epilepsy population. A correct recognition of nonconvulsive status epilepticus (SE) may cause a special diagnostic problem, because of a noncharacteristic clinical picture dominated by disturbances of consciousness. Therefore, the data on occurrence of epileptic seizures in the elderly have been different in particular reports. The aim of our prospective study was to assess the number of admissions of elderly patients to the neurologic ward with the first epileptic seizures and to evaluate the etiology and clinical features of the seizures. Methods: During the period of 2000–2002, 2,986 patients aged 60 or older were admitted to our Neurological Departments; among them there were 137 (4.6%) patients with the first-ever epileptic seizures. There were 66 males and 71 females in this group, aged from 60 to 98 years. In all patients, computed tomography (CT) or magnetic resonance imaging (MRI) and repeated EEGs or video/EEG monitoring were performed. Ninety-one (66.4%) patients were admitted because of single seizures, and in the remaining 46 (33.6%), SE was diagnosed. Results: In the group of 91 patients with single seizures, the etiology was established in 94.5%. The dominating cause (63.7%) was vasogenic
Epilepsia, Vol. 44, Suppl. 9, 2003
18
AES PROCEEDINGS
brain lesion (acute stroke, 8.8%; history of stroke, 54.9%). Generalized tonic–clonic seizures were the most often observed (74.4%), less frequent were partial seizures (20.8%) and myoclonic seizures (4.4%). In EEGs, localized slow waves were the most frequent finding (41%), whereas interictal epileptiform discharges were rarely observed (17.6%). CT/MRI scans revealed organic brain damage in 64 patients with predominance of multifocal lesions. In the group of 46 patients with de novo SE, the symptomatic etiology was established in 42 (91.3%); in the majority of cases, these were vascular lesions: acute stroke (19.6%) and poststroke changes (56.5%). CT/MRI examinations showed focal brain lesions in 76% of cases; multiple lesions were the most common findings. On the basis of EEG in 23 patients, convulsive SE and in 23 nonconvulsive SE (NCSE): partial in 21, generalized in two subjects were identified. Conclusions: Our results indicate that one third of the newly diagnosed seizures in the elderly are manifested as SE. De novo SE occurs in half of the cases in form of NCSE, mainly as a partial complex NCSE. The majority of newly identified seizures in the elderly were of vascular origin, with predominance of multifocal vascular brain lesions. 1.040 OBSERVATIONS ON THE DELAY IN DIAGNOSIS OF SEIZURES IN THE ELDERLY: UPDATE 3 1 Mark C. Spitz, 1 Jacquelyn L. Bainbridge, 1 K. Vanessa Winzenburg, 2 R. Eugene Ramsay, 2 Flavia M. Pryor, 3 Basim M. Uthman, 3 Brenda Smith, 1 Sharon Edwards, 2 Juanita Johnson, 3 Anita Wilson, 3 Ashlee Stocco, and DVA CSP 428 Study Group (1 Neurology, UCHSC, Denver, CO; and 2 Neurology, Miami VA Medical Center, Miami, and 3 Neurology, University of Florida, Gainesville, FL) Rationale: The onset of epilepsy often occurs in the elderly population. From our observations, the time to correct diagnosis is often delayed. Our interest is to investigate this phenomenon. Methods: We reviewed the charts of 159 patients enrolled in the Veterans Administration Cooperative Study of new-onset epilepsy in the elderly. No patients were profoundly demented or had known fatal illnesses. Concomitant medical diseases were permitted. Results: One hundred fifty-one men and two women aged 59–96 years (mean, 71.3 years) had a mean time to correct diagnosis of seizures of 1.7 years (median, 2.5 months). Forty-nine of these were diagnosed immediately. Fifty-eight patients had generalized tonic–clonic (GTC) seizures, and half (50%) were immediately diagnosed. Eighty-six patients had complex partial seizures (CPSs), and only 20.9% had an immediate correct diagnosis. This was dependent primarily on delays by the patient for GTCs, and on both the patient and the health care providers for CPSs. If a history of cardiovascular disease was known, the correct diagnosis was significantly delayed in patients with GTCs, CPSs, and simple partial seizures (SPSs). More severe concomitant medical disorders, such as cancer, caused greater delays in the time to diagnosis. Patients with only SPSs (n = 29) waited a median of 2 months to seek medical attention. They were diagnosed with SPSs 1month later (median, 3 months from onset to diagnosis). Twenty-three percent of patients with SPSs were incorrectly diagnosed originally as having transient ischemic attacks (TIAs). Patients with SPSs were mostly self-referred (p < 0.01), whereas those with CPSs or GTCs were mostly referred by family members or health care providers. Preliminary data suggest that patients older than 80 are more likely to be diagnosed with GTCs, and less likely to be diagnosed with SPSs, than are patients between the ages of 60 and 80. Conclusions: We observed significant delays in the correct diagnosis of seizures in this population. Two reasons we believe contribute to this problem are the lack of awareness of partial seizures by the public and health care providers, and an attempt to attribute all of a patient’s symptoms to a single diagnosis.
Basic Science 1.041 ALTERED MGLUR-MEDIATED INHIBITION IN THE PERFORANT PATH AFTER PILOCARPINE TREATMENT IN MICE Kristopher J. Bough and Raymond J. Dingledine (Pharmacology, Emory University, Atlanta, GA) Epilepsia, Vol. 44, Suppl. 9, 2003
Rationale: Previous reports have shown that mGluR-mediated inhibition within the dentate is altered after pilocarpine (PILO)-induced status epilepticus. Whether these changes contribute to the development of the disease (i.e., occur during the latent period), however, remains an important and unresolved issue. Methods: Field excitatory postsynaptic potentials (fEPSPs) were recorded in the lateral (LPP) and medial perforant path (MPP) simultaneously in adult mouse hippocampal slices. Paired stimuli (0.1-ms duration, 150-ms interpulse interval) were delivered to help differentiate MPP (paired-pulse depression) from LPP (paired-pulse facilitation) responses and to assess drug-induced increases in the paired-pulse ratio that are consistent with a presynaptic site of action. fEPSP slopes (10– 90%) were measured in control and PILO-treated mice in response to selective group II and group III agonists. Results: Application of the selective group II mGluR agonist DCG-IV (300 nM) reversibly depressed the fEPSP slopes in both the LPP (−30 ± 2%; n = 7) and the MPP (−45 ± 3%; n = 7) in control mice; by comparison, PILO-treated animals showed similar levels of inhibition within the LPP (−25 ± 2%; n = 10), but inhibition within the MPP was significantly enhanced (−61 ± 4%; n = 10; p < 0.01, ANOVA). Treatment with a selective group III agonist (L-AP4, 600 µM) reversibly depressed fEPSP slopes in both the LPP (−40 ± 5%; n = 7) and the MPP (−14 ± 5%; n = 7) in controls; in PILO-treated animals, L-AP4 produced the same level of inhibition as was observed in controls (−43 ± 2%; n = 10); however, inhibition within the MPP was markedly diminished (−4 ± 2%; n = 10; p < 0.05, ANOVA). The specific mGluR8 receptor agonist S-3,4-DCPG (3 µM) reversibly inhibited the LPP (−31 ± 2%; n = 11) but not MPP in control animals; there were no differences for PILO-treated mice (LPP only, −28 ± 2%; n = 9). Conclusions: In the MPP, mGluR7-mediated inhibition was reduced 3–9 days after PILO, whereas mGluR2/3-mediated inhibition was enhanced. In the LPP, there were no significant changes in mGluR-mediated inhibition after PILO treatment. Consistent with previous work, these data indicate that mGluR-mediated inhibition within the dentate is altered before the onset of spontaneous recurrent seizures. Future studies are required to decipher whether these changes are compensatory or contribute to epileptogenesis. [Supported by The Charlie Foundation (K.J.B.), NINDS grant NS043032 (K.J.B.), and the NINDS (R.J.D.).]
1.042 ALTERATIONS IN NMDA-RECEPTOR SUBUNIT EXPRESSION AFTER PERINATAL HYPOXIA-INDUCED SEIZURES Weimin Dai, Rachel Levada, and Frances E. Jensen (Department of Neurology, Children’s Hospital and Harvard Medical School, Boston, MA) Rationale: Perinatal hypoxia increases hippocampal network excitability and long-term seizure susceptibility (Jensen et al., 1992). We previously showed that maturational differences in AMPA subtypes of glutamate receptors play a role hypoxia-induced epileptogenesis (Sanchez et al., 2000). In addition, hypoxia increases calcineurin (CaN) activity and expression within 24 h (Dai et al., 2002). N-methyl-Daspartate (NMDA) subtypes play a major role in epileptogenesis and are developmentally regulated via CaN, with CaN inducing a switch from NR2B to NR2A (Shi et al., 2000). We used a rodent model of perinatal hypoxia to determine whether early-life seizures induced alterations in NMDA-receptor subunit expression and function. Methods: Rats at postnatal day 10–11 were placed in 4–7% O2 for 15 min to induce seizures; 24 h after seizures, hippocampal slices were prepared and incubated at room temperature for 1 h before recording. Patch-clamp recordings in whole-cell configuration and bath application of the NMDA-receptor antagonists Ro25-6981 and D-AP5 were used to compare the seizure-induced alterations in NMDA currents in pyramidal cells of hippocampal slices from hypoxic and normal rats. Membrane protein from hippocampal homogenates was prepared 24 h after seizures, and NMDAR subunit expression was analyzed by Western blotting and immunodetected with NR2A, NR2B, or NR1 antibodies. Results: NMDA (200 µM) evoked large responses in hippocampal pyramidal neurons from control rats (−138.5 ± 24 pA; n = 25) and in rats at 24 h after hypoxic seizures (−115.8 ± 19.6 pA; n = 16; p > 0.05). Ro25-6981, a highly selective and potent NR2B-receptor antagonist (Fischer et al., 1997) markedly inhibited NMDA responses from control rats by 62% (37.8 ± 4.4% of the control current; n = 9;
AES PROCEEDINGS p < 0.0001). In contrast, Ro25-6981 had little effect on NMDA responses in slices from rats 24 h after hypoxia (88 ± 4.6% of the control current; n = 5; p > 0.05). D-AP5 (50 µM) completely blocked the residues of the Ro25-6981–resistant component of the NMDA responses in these slices from rats 24 h after hypoxia (n = 4). The lack of effect of Ro25-6981 indicates that NR2B subunits contribute minimally to NMDA currents in rats after hypoxia. Western blot analysis showed a significant decrease (32 ± 7%; n = 8) in NR2B subunit protein in samples from rats 24 h after hypoxic seizure compared with control rats (n = 7; p < 0.05). There were no changes in NR2A and NR1 expression 10 ± 14% (n = 4) and 1 ± 12% (n = 4), respectively compared with control rats (n = 4; p > 0.05). Conclusions: NR2B-receptor function and expression is downregulated in hippocampus at 24 h after hypoxia-induced seizures, coincident with the time window in which we observed increased CaN activity and function. These data suggest seizures may modify NR2 expression via CaN. Future studies will be required to determine whether these changes are proepileptogenic or compensatory in the immature brain. [Supported by NINDS RO1 NS31718 (F.E.J.).]
1.043 INCREASED ADENOSINE RELEASE IN HIPPOCAMPAL AREA CA1 DUE TO ELEVATED CARBON DIOXIDE LEVELS 1 Chris G. Dulla, 2 Susan A. Masino, and 1,2,3 Kevin J. Staley (1 Neuroscience Program, 2 Department of Pharmacology, and 3 Department of Pediatrics, University of Colorado HSC, Denver, CO) Rationale: Adenosine is an inhibitory neuromodulator recently shown to decrease seizure activity in a kindling model of epilepsy. In the hippocampus, adenosine A1 -receptor activation results in decreased neurotransmitter release and hyperpolarization of neurons. Although the mechanisms that control the release of adenosine are unknown, a decrease in intracellular pH is common to many adenosine-releasing stimuli. Increases in brain pCO2 decrease intracellular pH and may result in adenosine release. We investigated the effects of increased CO2 on adenosine release, neurotransmission, and intracellular pH changes in area CA1 of rat hippocampal slices. Methods: Hippocampal slices were prepared from 6- to 8-week-old Sprague–Dawley rats: 400-µm slices were cut in ice-cold artificial cerebrospinal fluid (aCSF) bubbled with 95% oxygen/5% carbon dioxide. Field excitatory postsynaptic potentials (fEPSPs) were evoked by stimulation of Schaffer collateral axons with a bipolar stimulating electrode, and the fEPSP slope was measured with a recording electrode placed in the stratum radiatum of area CA1. Extracellular adenosine levels were measured by using a double-barreled subtractive adenosine sensor (Dale, J Physiol, 1998). The adenosine sensor was placed across the s. pyramidale in CA1. Intracellular pH measurements were made by using BCECF-AM loaded slices imaged on a Zeiss 510 2-photon microscope. Results: When hippocampal slices were exposed to aCSF acidified with 10% CO2 and 26 mM bicarbonate, there was a significant increase in the extracellular concentration of adenosine. This increase in extracellular adenosine concentration was accompanied by a decrease in fEPSP amplitude and a decrease in BCECF fluorescence. The depression of fEPSP amplitude was attenuated by application of theophylline, an adenosine-receptor antagonist. When hippocampal slices were exposed to 10% CO2 and 52 mM bicarbonate (pH similar to control), extracellular adenosine levels appeared to decrease. Application of 10% CO2 and 52 mM bicarbonate did not alter fEPSP or change BCECF fluorescence. Conclusions: Based on these studies, it can be concluded that exposure of hippocampal slices to increased CO2 levels, as would occur during seizure or hypoxia, causes an increase in the release of adenosine into the extracellular milieu of the slice. Increased adenosine release due to elevated CO2 is mediated by the acidification that accompanies increased CO2 levels. Adenosine levels decrease on addition of 10% CO2 and 52 mM bicarbonate. This may be due to additional buffering capacity, which can reduce physiologic pH transients and thus reduce adenosine release, or to inhibition of high-voltage–activated calcium currents by bicarbonate. This study also suggests that the sedative and anticonvulsive effects of hypercapnia, and the proconvulsive effects of hypocapnia, may be due to changes in extracellular adenosine levels. (Supported by NIH grant 29173.)
19
1.044 TOPIRAMATE, AN AMINOSULFONYL-CONTAINING ANTICONVULSANT, PROLONGS AMPA RECEPTOR SYNAPTIC CURRENTS IN BASOLATERAL AMYGDALA NEURONS: IMPLICATIONS FOR SELECTIVE INHIBITION OF GLUR5 KAINATE RECEPTOR RESPONSES Divina S. Gryder and Michael A. Rogawski (Epilepsy Research Section, NINDS-NIH, Bethesda, MD) Rationale: We previously reported that topiramate (TPM) inhibits both the AMPA and kainate components of the excitatory synaptic response elicited by stimulation of the amygdalar capsule, but is substantially more potent on the component of the response mediated by GluR5 kainate receptors than by AMPA receptors. We proposed that selective blockade of kainate receptors is key factor in the therapeutic activity of TPM. Recognizing that TPM contains an aminosulfonyl functional group, as do cyclothiazide and other cyclothiazide-like modulators of AMPA-receptor desensitization, we sought to determine whether TPM alters the time course of AMPA-receptor–mediated synaptic responses. Methods: Whole-cell patch recordings were made from visually identified BLA principal neurons in 450-µm-thick coronal slices from the rat brain. With a bipolar tungsten electrode placed on the amygdalar capsule, synaptic responses were evoked with 100-µs monophasic stimuli. AMPA receptors were isolated by perfusing a cocktail of antagonist including 100 µM D-AP5, 10 nM LY 293558, 10 µM bicuculline methiodide, and 10 µM SCH 50911 to block NMDA, kainate, γ -aminobutyric acid (GABA)A and GABAB receptors, respectively. Results: The mean peak AMPA-receptor current amplitude was 1,594 ± 398 pA; the mean 10–90% increase time was 7.7 ± 4.8 ms; and the mean 90–10% decay time was 23.2 ± 4.6 ms (seven slices); the mean decay time constant value (τ c ) from single exponential fits was 11.6 ± 1.4 ms. TPM (0.1–10 mM) caused modest reduction in the peak amplitude of the AMPA-receptor–mediated synaptic response associated with a significant slowing of the decay time constant by 189 ± 150% for 1 µM TPM and 156 ± 151% for 10 µM TPM (p < 0.004), without a significant change in the rise time. TPM did not significantly affect the 10–90% rise time or the 90–10% decay time of GluR5 kainate-receptor synaptic currents, which were, respectively, 6.1 ± 1.1 ms and 39.2 ± 6.7 ms under control conditions. Conclusions: TPM slowed the time course of AMPA but not kainatereceptor–mediated synaptic currents, an action that could be due to effects on AMPA-receptor desensitization similar to that of cyclothiazide. Cyclothiazide-like desensitization modulators not only slow AMPAreceptor desensitization, but also increase the peak amplitude of AMPAreceptor currents. It is interesting to speculate that, whereas TPM may block (reduce the amplitude) of both AMPA and kainate currents, its relatively weaker activity on AMPA currents could be due to an effect on AMPA-receptor desensitization that would functionally oppose (partially reverse) the block. The overall weak activity of TPM on AMPA receptors likely accounts for its lack of dramatic neurobehavioral side effects in comparison with more potent CNS AMPA-receptor antagonists. The extent to which prolongation of AMPA-receptor current decay contributes to the various clinical actions of TPM remains to be determined. (Supported by NINDS.) 1.045 INDUCTION OF EPILEPTIFORM ACTVITY BY GROUP I METABOTROPIC GLUTAMATE RECEPTORS IN THE HIPPOCAMPUS 1 Linda Karr and 1,2 Paul A. Rutecki (1 Neurology, Wm. S. Middleton VA Medical Center, and 2 Neurology and Neurosurgery, University of Wisconsin, Madison, WI) Rationale: During conditions of increased glutamate release such as seizures, group I metabotropic glutamate receptors (mGluR 1 and 5) activate many second-messenger pathways and may contribute to epileptogenesis. In hippocampal slices, the group I mGluR agonist 3,5dihyroxyphenylglycine (DHPG) results in the induction of epileptiform discharges that persist for hours after DHPG is removed. We investigated the sensitivity of persistent epileptiform activity to mGluR 1 and 5 antagonists, and the need for phospholipase C (PLC), a second messenger activated by DHPG, in producing long-term changes in excitability. Methods: Hippocampal slices were prepared from Sprague–Dawley rats (100–200 g), and incubated either in the presence of DHPG (100 µM) Epilepsia, Vol. 44, Suppl. 9, 2003
20
AES PROCEEDINGS
alone or in an artificial cerebrospinal fluid (aCSF) with a specific antagonist followed by DHPG for 90–120 min. Slices were transferred to an interface chamber and bathed in aCSF ([K+ ]o = 5 mM) for 1 h and then monitored extracellularly in the CA3 region for spontaneously occurring epileptiform activity that consisted of either brief interictal (2 s) ictal-like discharges. After evaluation, slices were exposed to bicuculline (BMI, 10 µM) and assessed for epileptiform activity. Results: The combined blockade of mGluR1 with LY367385 (30 µM) and the mGluR5 with MPEP (50 µM) prevented the induction of ictal activity by DHPG exposure in all 30 slices, with six slices having interictal discharges. Control slices prepared from the same animals demonstrated ictal discharges in 46% of slices and interictal activity in 38% (n = 26; p < 0.05, χ 2 ). In slices exposed to group I mGluR blockade, BMI produced interictal activity in all slices, with three displaying ictal discharges. Exposure to LY367385 alone prevented DHPG-induction of ictal discharges in aCSF, but 68% of slices had interictal activity (n = 19). MPEP alone also blocked the DHPG-induced generation of ictal discharges, with only 4% of slices displaying ictal patterns and 36% interictal patterns (n = 28), compared with control slices that had 21% ictal and 61% interictal patterns (p < 0.05). Blockade of PLC with U73122 (1 µM) also prevented ictal discharges that result from DHPG exposure, although 50% of 16 slices demonstrated interictal activity in ACSF compared with control slices that displayed ictal discharges in 30% and interictal discharges in 65% (n = 20; p < 0.05). In U73122-exposed slices, BMI application resulted in interictal activity in 94% of slices, and 6% had ictal activity compared with the controls slices that had 45% interictal and 55% ictal activity (n = 16; p < 0.05). Conclusions: The induction of persistent ictal epileptiform activity by DHPG was dependent on activation of mGluR1, mGluR5, and PLC. The induction of interictal discharges was less sensitive to blockade by group I mGluR or PLC antagonists. These results point to the cooperative effect of group I mGluRs and the necessity of PLC second-messenger activation in the epileptogenic effects of DHPG. (Supported by VA Research.)
1.046 MINIATURE EXCITATORY POSTSYNAPTIC CURRENTS IN HILAR MOSSY CELLS Charles Livsey and Anne Williamson (Department of Neurosurgery, Yale University School of Medicine, New Haven, CT) Rationale: Hilar mossy cells (MCs) are large glutamatergic neurons found within the dentate hilus and are known to receive at least two main inputs, mossy fiber inputs from the dentate granule cells and recurrent collateral and commissural inputs from the CA3 pyramidal cells and probably other mossy cells. Although many mossy cells are lost during the development of temporal lobe epilepsy, there is evidence that at least a portion remain. Therefore, we predict that there are two populations of mossy cells in the normal hilus. Methods: Miniature excitatory postsynaptic currents (mEPSCs) were recorded from mossy cells in the hippocampal hilus from 250-µm brain slices from young to adult rats by using whole-cell patch-clamp recording techniques. Mossy cells were identified during the recordings by their large size and characteristic, slowly decaying EPSCs. In addition, Lucifer Yellow fills were used further to identify these cells by their large spines and thorny excressences. Recordings were made at room temperature in artificial cerebrospinal fluid (aCSF) containing tetrodotoxin (TTX; 1 µM) and bicuculline (30 µM). Data were acquired at 70 kHz after being filtered at 10 kHz with a 4 Pole Bessel lowpass filter. Data were further digitally filtered at 2–5 kHz for analysis. Results: We recorded mEPSCs from MCs and found two distinct populations of events based on their kinetics and quantal size. The most obvious distinguishing feature of these events is a distinct quantal size (∼9 pA for mossy fibers and 6 pA for CC inputs). Interestingly, mEPSCs comprise not a single quanta as would be expected, but may be made up of one (9 pA), two (18 pA), three (27 pA), or many multiples of the quantal size, as would be expected for evoked EPSCs. CC inputs, conversely, were 6 pA, 12 pA, 18 pA, etc. The events attributed to the mossy fibers also have slower kinetics (∼8–12 ms) than the more distally occurring collateral commissural events (4–8 ms). The two populations of events can be separated by blocking the mossy fiber events with DCGIV (0.1–1 µM). In the presence of this group II selective mGluR agonist, Epilepsia, Vol. 44, Suppl. 9, 2003
discrete quantal peaks are more obvious. The proportion of mossy fiber and CC inputs varied between cells from ∼30 to 70%. Conclusions: MCs receive two distinct inputs in normal hippocampal slices. These inputs are readily distinguished by quantal size, and by rise and decay time constant kinetics. We hypothesize that our observation of mEPSCs of varying quantal size indicates that there is a varying probability of release or multiple receptor clusters resulting in the multiple quanta at individual synapses. If so, this is a novel finding that alters our understanding of synaptic transmission. Furthermore, the proportion of these two inputs varies from cell to cell and could underlie the variability in susceptibility to cell death between MCs. (Supported by NIH.)
1.047 DECREASED SENSITIVITY TO IFENPRODIL OF AN AUGMENTED NMDA-RECEPTOR–MEDIATED INPUT TO KINDLED DENTATE GYRUS GRANULE CELLS Nils O. Dalby and Istvan Mody (Neurology, UCLA, Los Angeles, CA) Rationale: N-methyl-D-aspartate (NMDA) receptor–mediated synaptic transmission is increased in the hippocampi of epileptic rats and in humans. However, the precise level at which this increase is mediated remains unknown because of the difficulties inherent to recordings of NMDA-receptor–mediated spontaneous synaptic events. Here we describe the basic properties of pharmacologically isolated spontaneous and evoked NMDA-receptor–mediated synaptic events in dentate gyrus granule cells of control and kindled rats. Methods: Male Wistar rats were surgically implanted with bipolar stimulating electrodes in the hippocampal commissures and were stimulated once daily, 5 days a week until the occurrence of ≥10 consecutive grade 5 seizures. Horizontal brain slices were prepared 24–48 h after the last seizure for whole-cell “blind” voltage-clamp recordings of dentate gyrus granule cells. Recordings were performed at 34◦ C in modified aCSF containing 5 µM Mg, 30 µM picrotoxin, 10 µM DNQX, and 10 µM D-serine. A small stimulus (40 µs) was given every 30 s in the perforant path to compare spontaneous and evoked event characteristics. Ifenprodil (10 µM), or D-APV (50 µM) or D-CPP (25 µM) was perfused after 10-min recording of basal activity. Alternatively, D-APV or DCPP was perfused 10 min after ifenprodil. Recordings were digitized at 8 kHz, filtered at 2 kHz, and analyzed by using the EVAN Labview-based software. Results: Average peaks and 10–90% rise times (RTs) of spontaneous NMDA-receptor–mediated events were not significantly different between control (n = 17) and kindled cells (n = 9). The decay time constants of the events were significantly increased in kindled cells (60.1 ± 4.9 ms; mean ± SEM) compared with 41.1 ± 2.9 ms in controls. Kindling significantly increased the baseline variance (10.8 ± 1.5 vs. 7.5 ± 1.0 pA2 in controls) and a tonic NMDA-receptor–mediated conductance, measured as the decrease in holding current on perfusion of D-APV or D-CPP (46.5 ± 7.3 vs. 18.8 ± 3.0 pA in controls). In control cells, the NR2B antagonist ifenprodil (10 µM) had no effect on the kinetic properties of spontaneous events but significantly reduced their frequency. It also reduced the peak of the stimulus-evoked events. In sharp contrast, ifenprodil had no significant effects on any of the properties of the spontaneous or stimulus-evoked events in the kindled rats. Conclusions: Granule cells of kindled rats display an increase in the charge carried by spontaneous NMDA-receptor–mediated synaptic events, mainly because of an increase in the decay time constant. Furthermore, the much increased baseline variance and effect of D-APV in kindled rats suggest that the activity of most likely extrasynaptically localted NMDA receptors is also increased. Finally, the lack of ifenprodil sensitivity in kindled rats points to an altered NMDA-receptor composition in the granule cells themselves or in the neurons of the circuitry providing excitatory input to the dentate gyrus. (Supported by NIH/NINDS grant NS 02808 and the Coelho Endowment to I.M.)
1.048 PROTONS RELIEVE ZINC INHIBITION OF KAINATE RECEPTORS David D. Mott and Raymond J. Dingledine (Pharmacology, Emory University School of Medicine, Atlanta, GA)
AES PROCEEDINGS Rationale: Kainate receptors have been implicated in the pathogenesis of epilepsy and contribute to seizures in hippocampal area CA3. The epileptogenic effect of these receptors may result from their ability to both reduce γ -aminobutyric acid (GABA)ergic inhibition and directly excite principal cells. We have previously shown that synaptic kainate receptors are inhibited by physiologically released zinc in hippocampal mossy fiber synapses. In addition, with recombinant kainate receptors, we have found that zinc inhibition of these receptors is subunit dependent. We have also found that kainate receptors are inhibited by protons. Proton concentrations increase markedly during seizures, with pH levels declining to ≤6.5. The better to understand the impact of seizures on kainate receptor–mediated neurotransmission, we have now examined the effect of protons on zinc inhibition of kainate receptors. Methods: Two-electrode voltage clamp was used to record currents from recombinant receptors expressed in Xenopus oocytes. Results: Protons shifted the zinc dose–response curve to the right, indicating that when proton concentrations increased, a given concentration of zinc produced less inhibition. The effect of protons on zinc inhibition of kainate receptors was subunit dependent. Lowering pH from 7.5 to 6.0 had a substantially greater effect on the median inhibitory concentration (IC50 ) for zinc inhibition of GluR6R/KA2 receptors (150fold increase) than on homomeric GluR6R receptors (sixfold increase). Examination of the proton inhibition curve for GluR6/KA2 receptors revealed that in the presence of zinc, protons potentiated the current over a physiologic range of pH values with greater potentiation at more acidic pH levels. Proton potentiation was greater as the concentration of zinc was increased from 1 to 1,000 µM, indicating that the potentiation reflects relief of zinc inhibition. Conclusions: These findings indicate an interaction between proton and zinc inhibition of kainate receptors. During repetitive synaptic activity, such as seizures, released zinc inhibits kainate receptors. Acidification of the synaptic cleft during the seizure would relieve this inhibition, thereby increasing kainate neurotransmission. Thus a major effect of a transient decrease in pH during a seizure could be an unmasking of kainate receptor–mediated synaptic transmission. [Supported by NINDS (D.M., R.D.) and NARSAD (D.M.).]
1.049 CONTRIBUTION OF SHORT-TERM FREQUENCY-DEPENDENT PLASTICITY TO SEIZURE GENERATION AND SPREAD 1 Stephan Rueegg, 2 Michael Kaplan, and 2 Marc A. Dichter (1 Neurology, University of Basel, Basel, Switzerland; and 2 Neurology, University of Pennsylvania, Philadelphia, PA) Rationale: Neurons in the cortex fire with short, high-frequency bursts or trains during interictal epileptiform discharges and with longer trains during ictal activity. The nature of the responses in neurons that are synaptically coupled to these neurons, however, is not well characterized. Our laboratory has been analyzing short-term synaptic plasticity in both excitatory and inhibitory circuits in response to repetitive firing of the presynaptic neuron to understand the mechanisms by which seizures may develop and spread throughout the brain. Methods: Isolated pairs of synaptically connected hippocampal neurons in cell culture were studied with patch-clamp electrodes, in the WCVC mode, and in current clamp to more closely mimic conditions that exist physiologically. Extracellular medium was similar to rat CSF (145 mM NaCl, 3 mM KCl, 2.0 mM CaCl2 , 1 mM MgCl2 , 10 mM glucose, 10 mM Hepes, pH 7.3). Action potentials (one, two, or trains at 5, 20, and 100 Hz) were evoked in the presynaptic neurons. Excitatory postsynaptic potentials (or currents) were recorded in first the absence and then presence of blockers of γ -aminobutyric acid (GABA)A and N-methyl-D-aspartate (NMDA) receptors. Results: The majority of excitatory synapses (137 of 154, 84%) showed paired-pulse depression (PPD) in the WCVC mode at interstimulus intervals of 200 ms. Only 8% showed facilitation. By contrast, repetitive activation of 58 monosynaptic excitatory connections, in the WCVC mode and with NMDA receptors blocked, demonstrated three kinds of short-term plasticity (STP): (a) 32 pairs (55%) showed tetanic depression, with the average EPSC declining to 200% of control value; and (c) nine pairs (16%) showed first depression and then facilitation or
21
a variable response within the train. Each of these patterns was a result of changes in presynaptic release of neurotransmitter. When the neurons were studied in a more physiologic condition, repetitive activation at 5, 20, and 100 Hz produced marked progressive neuronal depolarization, usually reaching threshold for activation of the postsynaptic neuron and often well outlasting the stimulus train, especially at 20 and 100 Hz. At 100 Hz, simply stimulating one presynaptic neuron with 10 APs frequently induced repetitive “seizure-like” discharges in the follower cell that lasted for seconds at least. Blocking NMDA receptors reduced but did not eliminate the large underlying postsynaptic depolarizations. Conclusions: Despite reduction in neurotransmitter release with repetitive activation of excitatory synapses, small excitatory circuits demonstrate dramatic facilitation of synaptic transmission. This is mediated by both AMPA- and NMDA-receptor–mediated currents. Such powerful short-term synaptic facilitation can clearly play an important role in the development and spread of seizures. [Supported by NS 24260 (M.A.D.).]
1.050 SHORT-TERM BLOCKADE OF ADENOSINE A1 RECEPTORS INDUCES LONG-TERM DECREASES IN INTERICTAL ACTIVITY 1 Dan A. Sdrulla and 1,2 Kevin J. Staley (1 MSTP/Neuroscience Program, and 2 Neurology and Pediatrics, University of Colorado HSC, Denver, CO) Rationale: Adenosine is an inhibitory synaptic modulator in the central nervous system (CNS). The main adenosine receptor in the CNS is A1R, and its activation inhibits neurotransmitter release and hyperpolarizes neurons. Because adenosine release appears to be activity dependent, it may play an important role in regulating synaptic activity during periods of heightened neuronal firing such as occur during epileptiform activity. We investigated the effects of temporary blockade of adenosine receptors on neuronal firing in an in vitro model of interictal activity. Methods: We tested the effects of the adenosine A1 receptor (A1R) antagonists theophylline (250 µM) or DPCPX (100 nM), on the CA3 burst probability by using extracellular and intracellular recordings. Hippocampal coronal slices were prepared from 4- to 6-week-old Sprague– Dawley rats. Spontaneous bursting of the CA3 network was induced by blockade of γ -aminobutyric acid (GABA)A and B conductances with 100 mM picrotoxin and 1 mM CGP55845A, respectively. Intracellular recordings were performed in the presence of picrotoxin (100 µM), CGP55845A (1 µM), and dAP5 (50 µM). Results: Repetitive 10-min applications of adenosine A1R blockers resulted in progressive decreases in extracellular CA3 burst frequency on washout. This effect was independent of NMDARs, mGluR receptors, L-type calcium channels, or PKC activation. Activation of PKA during A1R blockade resulted in cessation of CA3 network bursting during the wash of A1R blockade. This decrease in CA3 burst frequency could be observed during subsequent adenosine A1R blockades, suggesting that A1R activation is not necessary for its expression. Preliminary voltageclamp intracellular recordings revealed a decrease in the frequency of spontaneous excitatory postsynaptic currents (sEPSCs) without a change in the average sEPSC amplitude, suggesting a presynaptic locus of this effect. Conclusions: These data suggest that temporary blockade of A1R results in lasting decrease in CA3 network burst probability. This decrease in network activity together with the decrease in sEPSCs frequency at CA3 excitatory synapses suggest that transient A1R blockade results in a long-term decrease in glutamate-release probability. Together, these data suggest that therapeutic manipulations of adenosine activity during interictal or ictal epileptiform activity may induce long-lasting decreases in the excitation of epileptic networks. (Supported by NIH.)
1.051 POLYAMINES FACILITATE SYNAPTIC RESPONSES MEDIATED BY GLUR2-DEFICIENT AMPA RECEPTORS IN IMMATURE LAYER V PYRAMIDAL NEURONS Jieun Shin and John R. Huguenard (Department of Neurology and Neurological Sciences, Stanford University School of Medicine, Stanford, CA) Epilepsia, Vol. 44, Suppl. 9, 2003
22
AES PROCEEDINGS
Rationale: AMPA receptors mediate the majority of fast excitation in the CNS and play key roles in generalization of seizure activity. The presence or absence of GluR2 in synaptic AMPA receptors will influence postsynaptic mechanisms such as Ca permeability and short-term plasticity. Receptors lacking GluR2 should have enhanced paired-pulse facilitation because of polyamine-dependent unblocking (Rozov and Burnashev. Nature 1999;40:594–8). We tested whether such unblocking occurs in immature neocortical pyramidal cells, which have recently been shown to express relatively low levels of functional GluR2 at synapses (Kumar et al. J Neurosci 2002;22:3005–15). Methods: Whole-cell recordings were used to obtain synaptic responses in layer V pyramidal neurons of P12–14 or P16–18 rats, and spermine (50 mM) was either included or excluded from the pipette solution. We recorded the responses to paired (50-Hz) stimuli and measured the rectification index (RI) and paired-pulse ratio (PPR) at the two developmental stages. Results: At the immature stage (3 months) changes of inhibition in dentate granule cells after kainate-induced status epilepticus. Previous data with this model revealed that these rats develop spontaneous seizures and have hilar neuron loss with robust mossy fiber sprouting. Methods: Hippocampal slices were prepared from Sprague–Dawley rats 4–7 days or >3 months after kainate or saline treatment. Whole-cell patch-clamp techniques were used to record spontaneous and miniature inhibitory postsynaptic currents (sIPSCs and mIPSCs) of dentate granule cells. The sIPSCs were obtained with CsCl-containing pipets, and the holding potential at −70 mV in DNQX (50 µM) and AP-5 (50 µM). The mIPSCs were obtained by adding TTX (2 µM) to the medium. Bicuculline (30 µM) was routinely added at the end of each experiment to verify that the recorded currents were mediated by GABAA receptors. Most animals were coded, so the experimenter was blind to the treatment. Results: Most recorded neurons (n = 73) showed relatively highfrequency and large-amplitude sIPSCs. Application of TTX eliminated all large-amplitude IPSCs and revealed mIPSCs. In both short- and longterm groups, the mean sIPSC frequency of the saline- and kainate-treated rats was not significantly different. However, mIPSC frequency from the
Epilepsia, Vol. 44, Suppl. 9, 2003
kainate-treated rats of the short-term group was ∼30% lower than that of saline controls (p < 0.05). The long-term epileptic animals also had significantly fewer mIPSCs than did control animals (∼30%; p < 0.05) (Dudek and Shao, 2002). In contrast, the mean amplitude of the sIPSC or mIPSC was not reduced in kainate-treated rats from both short- and long-term groups. Conclusions: Our results are consistent with a recent study using the pilocarpine model (Kobayashi and Buckmaster. J Neurosci 2003;23:2240–52). Our data on sIPSCs are also consistent with other experiments, particularly in vivo recordings, which have shown robust or even enhanced inhibition in the dentate gyrus in rats with chronic recurrent seizures after status epilepticus. The reduction of mIPSCs frequency (i.e., ∼30%) shortly after kainate treatment suggests that the status epilepticus caused a partial loss of the inhibitory interneuron population that innervates granule cells. A similar reduction (∼30%) of mIPSCs in the long-term kainate group further suggests that the interneuron damage remained unchanged during epileptogenesis. (Supported by NS 16683.) 1.065 VARIABILITY OF GABAERGIC SYNAPTIC INPUTS REGULATES THE EXCITABILITY OF HIPPOCAMPAL PYRAMIDAL CELLS: COMBINED DYNAMIC CLAMP AND MODELING STUDY Ildiko Aradi, Vijayalakshmi Santhakumar, and Ivan Soltesz (Anatomy and Neurobiology, University of California, Irvine, Irvine, CA) Rationale: γ -aminobutyric acid (GABA)ergic synaptic inputs to principal cells are heterogeneous in terms of their anatomic, molecular, and physiologic properties. Whether diversity in GABAergic synaptic inputs affects the efficacy of GABAergic inhibition is not understood. Methods: Dynamic clamp recordings were carried out from CA1 and CA3 pyramidal cells in hippocampal slices to study if alterations in the variance of inhibitory postsynaptic current (IPSC) populations can modulate the excitability of hippocampal pyramidal cells. The effects of changes in the IPSC variability were also studied in biophysical models of CA1 and CA3 pyramidal cells. Results: In response to increases in the variability of the inhibitory synaptic peak conductances, CA1 neurons increased their tonic firing frequency, whereas CA3 cells produced burst firing. In addition, both computational models and dynamic clamp recordings of CA1 cells showed that increased variance in the populations of IPSCs affected the CA1 pyramidal cells’ firing rate during theta–gamma oscillations. Conclusions: These results indicate the functional importance of the diversity of interneurons in cortical and hippocampal circuits, and suggest that GABAergic synaptic variability is a key parameter of neural networks that can regulate neuronal excitability under both normal and pathologic conditions. (Supported by the NIH, NS35915, to I.S.)
1.066 VALPROIC ACID ANTICONVULSANT EFFECTS REQUIRE INTACT INHIBITORY SYNAPTIC TRANSMISSION IN CA3 1 Audrey S. Yee and 1,2 Kevin J. Staley (1 Neurology and 2 Pediatrics, University of Colorado Health Sciences Center, Denver, CO) Rationale: The periodic discharge of the in vitro hippocampal CA3 pyramidal cell network is a well-studied model of pathological network synchronization. Many experimental manipulations that induce seizures in vivo also produce episodic depolarizations and bursts of action potentials of CA3. This synchronized bursting closely resembles interictal epileptiform activity recorded in vivo that underlies interictal spikes on the human electroencephalogram (EEG). During conditions of high epileptiform activty in humans, such as periodic lateralized epileptiform discharges (PLEDS) or status epilepticus (SE), anticonvulsants that conventionally abort seizures may be ineffective. To study this phenomenon, we produced bursting in CA3 under conditions of “high burst probability,” in which the burst frequency mimics PLEDs. Valproic acid (VPA) is used to treat SE. However, the underlying mechanisms of VPA remain unclear. We evaluated the anticonvulsant effects of VPA on CA3 under conditions of high burst probability. We measured changes in the interburst interval, burst duration, and thresholds for burst initiation and
AES PROCEEDINGS termination, before and after blockade of γ -aminobutyric acid (GABA)A receptors. Methods: In adult rats, 400-µm slices were prepared by using a modified aCSF with high sucrose/glucose (75/25 mM), high Mg2+ (7 mM), and low Ca2+ (0.5 mM). We used extracellular field recordings to examine spontaneous bursting induced under conditions of high burst probability ([K+ ]o 8.5 mM) with and without addition of 100 µM picrotoxin to block GABAA synaptic transmission. Anticonvulsant effects of VPA (200 µM, 1 mM, 5 mM) on CA3 were evaluated. Results: Our preliminary data show that in high burst probability when GABAA activity was blocked, addition of VPA (200 µM to 5 mM) did not significantly alter the interburst interval and burst duration. Similarly, in high burst probability, when GABAA activity was intact, application of 200 µM and 1 mM VPA did not significantly alter interburst interval or burst duration. However, 5 mM VPA increased the interburst interval from control by 306% and decreased the burst duration by 15%. This corresponds to an increase in the burst start threshold of 44% and increase in the burst end threshold of 42%. Conclusions: These observations suggest that the anticonvulsant effects of VPA are mediated via the GABAergic inhibitory system. We plan future studies to examine VPA effects in low burst probability (3.3 mM K+ ) with and without blockade of the GABAergic system. (Supported by NIH.) 1.067 GABA(A)-RECEPTOR AGONISTS REGULATE THE SEXUAL DIFFERENTIATION OF INFANTILE RAT SUBSTANTIA NIGRA RETICULATA NEURONS 1 Aristea S. Galanopoulou and 1,2,3 Solomon L. Mosh´ e (1 Neurology, 2 Neuroscience, and 3 Pediatrics, Albert Einstein College of Medicine, Bronx, NY) Rationale: The substantia nigra pars reticulata (SNR) is a sexually dimorphic structure involved in movement and seizure control. We have previously shown that at postnatal day 15 (PN15), the γ -aminobutyric acid [GABA(A)]-receptor agonist muscimol increases intracellular calcium concentration in male but not in female SNR neurons. We hypothesized that muscimol may thus have sex-specific effects on the transcription of calcium-regulated genes in infantile SNR. We studied, therefore, in male and female PN15 rat SNR, the in vivo effects of muscimol on the expression of the active phosphorylated form of the transcription factor cyclic adenosine monophosphate (cAMP)-responsive element binding protein (phosphoCREB), which regulates calcium-dependent gene transcription. Moreover, we tested the muscimol effects on the messenger RNA (mRNA) expression of KCC2, a calcium-regulated, neuronspecific potassium chloride cotransporter, which is involved in the functional maturation of GABA(A) receptors. Methods: Male and female PN15 rats were injected with a single subcutaneous injection of muscimol (4 µg) or its vehicle. A group of rats was killed 1 h after injection, transcardially perfused with paraformaldehyde, and processed with a phosphoCREB-specific immunochemistry (Upstate Biotechnology). A second group of rats was killed 4 h after injection, the brains were snap frozen in dry ice/isomethylbutane, and were processed with a KCC2-specific, digoxigenin-labeled in situ hybridization. Densitometry of stained cells was performed. Results: One hour after muscimol injection, phosphoCREB immunoreactivity was increased in male but not in female PN15 SNR neurons. Four hours after muscimol injection, KCC2 mRNA expression increased in male but decreased in female SNR neurons. Conclusions: Drugs that modulate the activity of GABA(A) receptors can activate calcium-regulated signaling processes and gene transcription in male but not in female PN15 SNR neurons. The distinct translational effects of GABA(A)-receptor agonists promote thus the sexual differentiation of the SNR. Through their divergent regulatory effects on KCC2 expression, such drugs may promote the functional maturation of GABA-sensitive, SNR-mediated seizure control in male but inhibit it in female infantile rats. [Supported by NIH grants NS45243 (A.S.G.) and NS20253 (S.L.M.), and the Heffer Family Foundation.] 1.068 REGIONAL SPECIFICITY OF SEIZURE-INDUCED INTERNEURON LOSS AND REDUCED INHIBITION IN THE EPILEPTIC RAT HIPPOCAMPUS
27
1 Joshua W. Hagen, 1,3 Umit Sayin, 2 Sue Osting, 1 Paul Rutecki, and 1,2 Thomas P. Sutula (1 Department of Neurology, and 2 Department of Anatomy, University of Wisconsin, Madison, WI; and 3 Institute of
Forensic Medicine, Istanbul University, Istanbul, Turkey) Rationale: Recent anatomic and physiological studies have demonstrated that after pilocarpine-induced status epilepticus, there is reduction of parvalbumin and somatostatin interneurons and the frequency of spontaneous inhibitory postsynaptic currents (IPSCs) in granule cells of the dentate gyrus preceding the development of spontaneous seizures (Kobayashi, Buckmaster. J Neurosci 2003;23:2440–52). After ∼100 evoked class V kindled seizures, there is reduction of cholecystokinin (CCK) and GAT-1 interneurons providing axosomatic and axoaxonic inhibition to granule cells in the dentate gyrus, which is accompanied by reduced amplitude and duration of evoked IPSCs and the emergence of spontaneous seizures (Sayin et al. J Neurosci 2003;23:2759–68). As these observations suggest that seizure-induced loss of specific interneuron subclasses in the dentate gyrus may contribute to emergence of spontaneous seizures, it was of interest to determine if spontaneous seizures are also associated with loss of specific interneurons and inhibition in CA3. Methods: Paired-pulse inhibition and GAT-1 and CCK interneuron populations in CA3 were assessed in kindled rats after a range of three afterdischarges (Ads) to >100 class V seizures and were compared with age-matched controls. Paired-pulse inhibition was measured in hippocampal slices by recording field potentials in the stratum pyramidale in response to stimulation of the granule cell layer of the dentate gyrus at interstimulus intervals of 15–300 ms. CCK and GAT-1 interneuron populations were examined by immunohistochemical and stereologic methods in kindled rats that experienced three, 60–80, or >100 class V seizures. Results: After 90–100 evoked class V seizures, there was a loss of paired-pulse inhibition at 15- to 25-ms interpulse intervals in CA3 (n = 20), which was not observed in age-matched controls (n = 58) or kindled rats that experienced three ADs or three to 35 class V seizures (n = 21; p < 0.001, ANOVA). The loss of paired-pulse inhibition in CA3 was consistent with loss of γ -aminobutyric acid (GABA)A inhibition and was temporally associated with emergence of spontaneous seizures and loss of inhibition in the dentate gyrus. In contrast to the loss of both CCK and GAT-1 interneurons in the dentate gyrus at this kindling stage, GAT-1 immunoreactivity was reduced in CA3, but no loss of CCK interneurons was detected by stereologic counting methods (n = 4). Conclusions: The emergence of spontaneous seizures is associated with loss of paired-pulse inhibition in both dentate gyrus and CA3. In the dentate gyrus, the loss of inhibition may be caused by seizure-induced loss of CCK and GAT-1 interneurons providing axosomatic and axoaxonic inhibition. In CA3, the lack of evidence for CCK interneuron loss indicates that seizures induce regionally specific interneuron loss in hippocampal circuitry, and that specific seizure-induced cellular alterations are likely to contribute to the loss of inhibition in different regions of hippocampal circuitry. (Supported by NINDS 25020, VA Resesarch.)
1.069 PARVALBUMIN-POSITIVE NEURONAL DENSITY IN THE KINDLED AND IRRADIATED RAT HIPPOCAMPUS 1 Kenneth A. Jenrow, 2 Alexander E. Ratkewicz, 1 Lemke W. Lemke, 3 Madhuri Kadiyala, 3 David Burdette, and 1 Kost V. Elisevich (1 Neurosurgery, 2 Radiation Oncology, and 3 Neurology, Henry Ford Hospital, Detriot, MI) Rationale: Several clinical and experimental reports have suggested that irradiation of epileptic foci at subnecrotic doses can reduce the incidence of spontaneously recurring seizures. We hypothesized that these effects may be mediated, in part, by a selective depletion of parvalbuminpositive γ -aminobutyric acid (GABA)ergic interneurons within the hippocampus. Networks of these perisomatic inhibitory neurons, which are connected via both chemical and electrical synapses, are critically involved in the emergence of large-scale synchrony among principal neurons within the hippocampus, and might therefore contribute to the initiation or elaboration of limbic seizures within this structure. Methods: Male Wistar rats were kindled by daily (afterdischarge threshold) electrical stimulation of the left amygdala, until a minimum of 10 stage 5 seizures were observed. The left amygdala was irradiated via Epilepsia, Vol. 44, Suppl. 9, 2003
28
AES PROCEEDINGS
dorsoventral exposure to a beam-collimated 18-MV x-ray source, with a 2.5-mm radius (90% isodose), at isocenter doses of either 18 or 25 Gy. Kindled control animals received sham irradiation. Animals were transcardially perfused 160 days after irradiation with fixative containing 4% paraformaldehyde and 0.08% gluteraldehyde, postfixed overnight, paraffin embedded, microtome sectioned (7 µm), and processed for parvalbumin immunohistochemistry. Parvalbumin and total cell counts were performed bilaterally by using the method of Mouritzen Dam in all hippocampal subfields. Volume calculations for each of these subfields were performed by using Scion (NIH) Image software. Results: There were no effects of irradiation at either 18 or 25 Gy on seizure threshold or seizure duration during the entire postirradiation interval before perfusion. Similarly, there were no significant differences in parvalbumin-positive neuronal densities associated with exposure to ionizing radiation at either 18 or 25 Gy. Parvalbumin-positive neuronal density was significantly increased (p < 0.003) in the pyramidal layer of CA1 in the hemisphere in which kindling was initiated, relative to the contralateral hemisphere. No consistently significant differences in total (principal) cell density within the hippocampus were observed among our groups as a function of either kindling or irradiation. Conclusions: We conclude that parvalbumin-positive neuronal density within the hippocampus is not affected by irradiation at the doses used in these investigations. Further investigations are required to determine whether analogous radiation exposures might affect chemical and/or electrical synaptic function among these neurons. The apparent increase in parvalbumin-positive neuronal density within the hippocampus as a function of kindling also warrants further investigation. This may reflect kindling-induced increases in parvalbumin expression with CA1.
1.070 PERSISTENT CHANGES IN SEIZURE SUSCEPTIBILITY AND GABAERGIC NEUROTRANSMISSION AFTER RECURRENT EARLY-LIFE SEIZURES 1 Doo-Kwun Kim, 2 Richard A. Hrachovy, and 1 John W. Swann (1 The Cain Foundation Laboratories, Department of Pediatrics, and 2 The Peter Kellaway Section of Neurophysiology, Department of Neurology, Baylor College of Medicine, Houston, TX) Rationale: The effects repeated seizures have on the developing brain have long been debated. Studies reported here examined the long-term effects of seizures in infant rats by a unilateral intrahippocampal injection of tetanus toxin on postnatal day 10. Methods: Video-EEG recordings were conducted between postnatal day 50 and 60. Alterations in seizure susceptibility were assayed on day 22 or 50 by using the flurothyl method. Changes in selected pre- and postsynaptic markers for γ -aminobutyric acid (GABA)ergic synapses were also determined by quantitative immunoblotting on day 50. Results: Behavioral seizures were first observed on day 12. Seizures gradually declined in frequency over the next 5 days and were not observed after day 17. On average, rats had 8.4 ± 2.7 seizures during 10 daily 1-h behavioral monitoring sessions. As adults (days 50–60), all rats displayed interictal spikes in the hippocampus and/or overlying cortex. Brief (2–5 s) electrographic seizures without behavioral accompaniments were recorded in only one of five animals. Because rats did not display spontaneous behavioral seizures, flurothyl was used to assay changes in seizure susceptibility. When rats were tested on day 50, they had a small (15%) but significant decrease in the time to onset of seizures when compared with littermate controls. However, when another group of rats was tested on day 22, a marked increase in seizure susceptibility was observed. Thus rats appear to progress from a period of overt spontaneous seizures (day 12–17), through a period of marked seizure susceptibility (day 22), to one of lessened seizure susceptibility (day 50). The hypothesis was put forward that alterations in seizure susceptibility reflect a gradual recovery of the brain to a normal state. To address this, pre- and postsynaptic markers for GABAergic neurotransmission were assayed. Contrary to our hypothesis, results showed that the expression levels of GAD67 and the α1 subunit of the GABAA receptor were actually increased. These changes were observed on both the CA3 and CA1 subfields but not in the dentate gyrus. The expression levels of the β2-3 and γ 2 subunits of the GABAA receptor were unchanged. Conclusions: The hippocampus appears to compensate for earlylife seizures and possibly persistent interictal activity by increasing the Epilepsia, Vol. 44, Suppl. 9, 2003
expression of GAD67, the GABA synthetic enzyme, and at least one GABAA -receptor subunit. Such compensatory changes could contribute to the progressive diminution in seizure susceptibility observed. (Supported by NIH, NINDS grants NS18309 and NS37171.) 1.071 ALTERED GABAA -RECEPTOR SUBUNIT EXPRESSION IN MATURE BUT NOT IMMATURE DENTATE GRANULE NEURONS AFTER STATUS EPILEPTICUS Brenda E. Porter, Xiaonan Cui, and Amy R. Brooks-Kayal (Child Neurology, The Children’s Hospital of Philadelphia, Philadelphia, PA) Rationale: Changes in hippocampal dentate physiology after status epilepticus (SE) may contribute to the development of spontaneous seizures. There is both altered expression of γ -aminobutyric acid (GABA)A receptor (GABAR) subunits, and a dramatic increase in the birth of newborn dentate granule neurons (DGNs) after SE. Here we determine whether the newborn, mature, or both populations of DGNs contribute to the changes in GABA-receptor expression after SE during postnatal development. Methods: Postnatal day 20 (P20) rats received either a dose of lithium and saline (control) or lithium-pilocarpine (pilo) to induce SE. They were killed and perfused on P34, and 10-µm frozen sections were stained for PSA-NCAM and NeuN. Immature, PSA-NCAM positive (+) DGNs or mature NeuN-positive (+) DGNs were individually dissected and underwent single-cell RNA amplification; the aRNA from each cell was then hybridized against a reverse Northern blot for the relative expression of the 16 GABAR subunits. Results: We find that the mature NeuN+DGNs from pilo-treated animals have increased relative expression of the GABA α1,β1,β2,γ 2 receptor subunits compared with mature DGNs from control animals. The most dramatic increase is an approximate fivefold increase in the α1 subunit. In contrast, immature DGNs do not differ in the relative concentrations of these GABAR subunits between control and pilo-treated animals. There are, however, differences in GABAR expression between mature and immature DGNs within both the control and pilo-treated groups. In control animals, the immature DGNs express higher levels of β1,γ 2 than do mature DGNs. However, in pilo-treated animals, the immature and mature DGNs express similar levels of β1,γ 2 because of an increased expression of these subunits in mature DGNs after SE. The mature DGNs from the pilo-treated animals now express increased α1 and β2 compared with the immature DGNs because of the increase in α1 and β2 in the mature DGNs after SE. Conclusions: Immature and mature DGNs in both control and pilotreated animals have different patterns of GABAR subunit expression, suggesting that within the dentate as DGNs mature, GABAR genes are developmentally regulated. After SE at P20, there is increased expression of α1, β1, β2, γ 2 in mature, but not immature DGNs, implying that only DGNs present at the time of SE undergo increases in GABAR subunit expression. The pattern of GABAR expression after SE is different, depending on the age of the animal at the time of SE, as adult animals after SE have a downregulation of α1, opposite to our current findings. [Supported by NINDS (A.B.K.); Child Neurology Foundation (A.B.K., B.E.P.).] 1.072 QUANTITATION OF GABAA -RECEPTOR SUBUNIT GENE EXPRESSION IN THE HIPPOCAMPUS WITH REAL-TIME QUANTITATIVE TAQMAN RT-PCR ASSAY Doodipala S. Reddy (Department of Molecular Biomedical Sciences, North Carolina State University College of Veterinary Medicine, Raleigh, NC) Rationale: The γ -aminobutyric acid (GABA)A receptor plays a critical role in epilepsy. The pentameric chloride channel is formed from seven different classes of subunits with multiple variants, which define its physiological and pharmacologic properties. Quantitation of different subunits in various brain regions is important to understand and develop novel drugs for epilepsy. However, conventional RNA assays are tedious and less sensitive. The aim of this study was to develop and validate a novel assay of GABAA -receptor subunit gene expression by “real-time” reverse transcriptase and polymerase chain reaction (RT-PCR) and to explore GABAA -receptor subunit gene expression in the hippocampus.
AES PROCEEDINGS Methods: Total RNA was extracted from hippocampus and cDNA prepared by reverse transcription. Real-time RT-PCR was applied to quantitate the expression of the target GABAA -receptor genes. For each subunit gene, a set of primers and TaqMan fluorogenic probe was designed specifically to amplify the target cDNA. TaqMan reaction conditions were optimized and quantitated by the threshold cycle (CT , the point during PCR amplification at which fluorescence significantly increases above background) by using the iCycler iQ detection system (Bio-Rad). Data were normalized to the housekeeping gene GAPDH. Results: Real-time TaqMan RT-PCR assay yielded superior sensitivity and accurate expression when compared with traditional RNA assays. The combination of primer mix (600 nM) and TaqMan probe (300 nM) that showed the highest fluorescence was used for all further real-time PCR assays. The method detected very low levels of gene expression (∼1 pg cDNA). We validated this technology for the quantification of GABAA receptor α1–4, β2, γ 2, and δ subunits in the mouse hippocampus. A standard curve for GAPDH and one of the target genes was constructed (3–3,000 pg cDNA). The slopes for GAPDH and target genes were around −3.4 (r2 = 0.998), reflecting similar optimum PCR efficiencies. The ratio of the CT value of the target gene to GAPDH provided the relative expression in the hippocampus. We used this assay to quantify progesterone-induced increase (approximately threefold) in transcription of GABAA -receptor α1 subunit in the hippocampus. Conclusions: The study provided a simple and sensitive method for quantification of GABAA -receptor subunits in brain tissues. This realtime assay allows rapid, high-throughput transcriptional profiling of the complete GABAA -receptor superfamily, and thus provides a new powerful tool for investigations of the mechanisms involved in antiepileptic drugs and GABAergic neurotransmission in seizure susceptibility. (Supported by NCSU-CVM.)
1.073 KAINATE-INDUCED STATUS EPILEPTICUS ALTERS GABAMEDIATED NEUROTRANSMISSION IN THE IMMATURE MOUSE HIPPOCAMPUS Moushumi Sehgal and John W. Swann (The Cain Foundation Laboratories, Department of Pediatrics, Baylor College of Medicine, Houston, TX) Rationale: Neuronal and synaptic reorganization of inhibitory circuits within the adult rodent hippocampus after kainate-induced status epilepticus has been previously described. Alterations of γ -aminobutyric acid (GABA)ergic transmission in the immature mouse brain could contribute to the known resistance of developing hippocampal neurons to seizure-induced damage and a resulting chronic epilepsy. Methods: Intraperitoneal injection of kainic acid in mouse pups on postnatal day (PND) 10 was used to induce pronounced limbic convulsions, progressing to status epilepticus. At two different time points after seizure induction (PNDs 14 and 28), we investigated GABAergic interneurons through immunohistochemical markers (parvalbumin and calbindin D28k) and also analyzed GAD-67 and the major GABA (A)-receptor subunits expressed in the hippocampus by Western blot. Results: On PND 14, a partial loss of parvalbumin-positive interneurons was noted, suggesting either acute degeneration of these cells or a downregulation in the expression of this calcium-binding protein. This was mainly seen in the CA1 region of the hippocampus. Calbindin staining of interneurons was spared throughout the hippocampus. The decreased immunoreactivity of the parvalbumin interneuron population was transient and appeared to have fully recovered by PND 28. This supports the conclusion that changes are more likely due to decreased expression rather than to cell loss. In keeping with this conclusion are results from GAD-67 immunoblotting, which did not show a change in expression between experimental and sham animals at either of the two investigational time points. Thus a significant loss of GABAergic neurons appears unlikely. At a postsynaptic level, on PND 14, the expression of the GABA (A)-receptor subunits α-1 and γ -2 was significantly increased, although the β-subunit expression remained unaltered. These early changes in GABA (A)-receptor subunit composition also subsided completely by PND 28. Conclusions: It can be postulated that subunit upregulation produces changes in the GABA(A)-receptor soon after status epilepticus, which, in turn, transiently decreases CNS excitability and mediates a neuroprotec-
29
tive homeostatic process. This may help to explain why status epilepticus does not produce chronic epilepsy in the immature animal. (Supported by NIH grants NS18309 and NS37171.)
1.074 LONG-TERM CHANGES OF γ -AMINOBUTYRIC ACID (GABA)A -RECEPTOR SUBUNIT IN HIPPOCAMPAL DENTATE GRANULE CELLS AFTER PROLONGED SEIZURES IN IMMATURE RATS 1 Guojun Zhang, 1 Yogendra Sinh H. Raol, and 1,2 Amy R. Brooks-Kayal (1 Neurology, Children’s Hospital of Philadelphia, and 2 Neurology and Pediatrics, University of Pennsylvania, Philadelphia, PA) Rationale: Previous studies from our laboratory suggest that 70% of rats that had experienced status epilepticus (SE) at postnatal day (P20) go on to develop spontaneous seizures, whereas 30% do not. The morphologic examination of the hippocampus of these rats does not identify any specific changes, which could contribute to the development of spontaneous seizures. Long-term alterations in postsynaptic GABAA receptor (GABAR) function and gene expression has been observed in various animal models after SE. For example, a decrease in α1-mRNA levels was found in the rats that had experienced SE in adulthood, whereas adult rats that had experienced SE at P10 had higher α1 levels. In the present study we determined the effect of SE at P20 on GABAR expression and function. Methods: Rat pups of 20 days postnatal age (P20) were subjected to lithium-pilocarpine–induced SE. Occurrence of SE was confirmed by EEG and behavioral monitoring. Rats were video-EEG monitored in adulthood (>P90) to identify animals with spontaneous seizures. Wholecell patch-clamp recording and single-cell antisense RNA amplification (aRNA) techniques were used to examine the expression of different GABAR subunit mRNAs and associated changes in the pharmacologic properties in single hippocampal dentate granule cells (DGCs). Results: Relative expression of α1 subunit mRNA of GABAR (compared with total β-actin) were increased (twofold) significantly in all rats that experienced seizures at P20 compared with lithium control rats (p < 0.01). Notably, expression of α1 subunit mRNA (compared with total GABA subunit) was significantly elevated (35%) in rats without spontaneous seizures compared with rats with spontaneous seizures (p < 0.01). The potentiation of zolpidem on GABA currents was increased in rats that had experienced seizures in P20 compared with lithium control rats. Conclusions: Prolonged seizures at postnatal day (P20) lead to longterm alterations in GABAR subunit mRNA expression and function in DGCs, and these changes occur differentially in animals that become epileptic and those that do not. These findings suggest that levels of α1 subunit mRNA may play a determining role in development of epilepsy after early seizures. (Supported by NIH NS38595 and Epilepsy Foundation of America to A.B.K.)
1.075 ENDOGENOUS NEUROPEPTIDE Y PREVENTS RECURRENCE OF EXPERIMENTAL FEBRILE SEIZURES 1 Celine M. Dube, 1 Kristen L. Brunson, 2 Mariam Eghbal-Ahmadi, 1 Rebeca Gonzales-Vega, and 1,2 Tallie Z. Baram (1 Anatomy/Neurobiology and 2 Pediatrics, University of California at Irvine, Irvine, CA) Rationale: Febrile seizures typically occur at the onset of a fever episode and do not recur within the same episode despite persistent or increased hyperthermia. Indeed, seizure recurrence within hours in a febrile episode are considered an atypical, complex feature that alters prognosis. Here we used an appropriate-age, characterized immature rat model to test the hypothesis that a first febrile seizure influences the threshold temperature for subsequent ones, and to determine the underlying mechanisms. Methods: Experimental febrile seizures were induced 2 or 3 times at 4-h intervals. To determine whether neuropeptide Y (NPY) was involved in increased thresholds for subsequent seizures, the NPY Y2-receptor antagonist (BIIE0246, 100 nm/kg), was infused i.c.v. before induction of the second seizure, and release-provoked upregulation of hippocampal NPY mRNA expression was determined 4 and 24 h after seizures.
Epilepsia, Vol. 44, Suppl. 9, 2003
30
AES PROCEEDINGS
Results: Experimental febrile seizures (involving the hippocampal formation) emerged at temperatures similar to those of febrile infants. Threshold temperatures for the onset of a second and third seizure were significantly and progressively higher than those required for the first. NPY Y2-receptor antagonist abolished the increased threshold for a second compared with a first seizure. NPY mRNA expression was increased 4 h [in dentate gyrus (DG)] and 24 h (DG and CA1) after febrile seizures. Conclusions: The basis for the single-seizure-per-febrile-episode phenomenon occurring in most children involves inhibitory actions of endogenous NPY, suggesting that the signaling cascade of this peptide may provide therapeutic targets. [Supported by NIH NS28912, NS35439 (T.Z.B.); Epilepsy Foundation of America Postdoctoral Research Training Fellowship (C.D.).] 1.076 β-HYDROXYBUTYRATE CAN SUBSTITUTE FOR GLUCOSE UNDER HYPOGLYCEMIC CONDITIONS IN THE RAT HIPPPOCAMPUS 1 Brenda Wu and 2 Anne Williamson (1 Neurology and 2 Neurosurgery, Yale University School of Medicine, New Haven, CT) Rationale: The mechanisms underlying the efficacy of the ketogenic diet are unclear, but it has been hypothesized that the metabolism of ketones, including β-hydroxybutyrate, is a critical aspect of the therapy. BHB can replace glucose to support synaptic functioning in young rats, but not in adults. Here we examined the effects of BHB on synaptic transmission in mildly hypoglycemic conditions to test the hypothesis that BHB can partially restore synaptic transmission in adult animals during periods of metabolic demand. Methods: Hippocampal slices were prepared from adult Sprague– Dawley rats and maintained in an interface recording chamber. Fieldpotential recordings were made in the cell body layer of both the CA1 and the dentate gyrus. Input–output curves and the responses to 5- and 10Hz repetitive orthodromic stimulation (10-s trains) were obtained under three different recording conditions: control aCSF (10 mM glucose); aCSF containing 2 mM glucose made isoosmotic with sucrose; and aCSF containing 2 mM glucose and 8 mM BHB. The data were normalized to the peak response recorded in control medium. Results: Bath application of low glucose (2 from 10 mM) was associated with a slight decrease in the amplitude of the population spikes (PSs) and in the PS slope in both the CA1 and dentate gyrus (DG) (n = 6). The slope, but not the amplitude of the CA1 PS, partially recovered after bath application of BHB in CA1, but not in the DG. When we examined the responses elicited by repetitive stimulation in low glucose, we found that there was a depression of the synaptic response (PS amplitude) during the train in 2 mM glucose as compared with 10 mM in both CA1 and the DG. When BHB was present, this depression was reduced and approached the levels seen in control aCSF. This effect of BHB was seen in both CA1 and in the DG, but was more significant for the CA1 region. In contrast, this effect was not seen when acetate, a substrate for the glial monocarboxylic acid transporter, was added to the aCSF (n = 2). Conclusions: In hypoglycemic conditions, we found that BHB, but not acetate, is able to substitute for glucose in both the CA1 and DG under specific conditions. This effect is stronger in the CA1 than in the DG, suggesting that there are regional differences in the ability of neurons in these areas either to take up BHB through the neuronal monocarboxylic acid transporter or in the ability of these cells to metabolize BHB. The BHB effect is seen primarily when the tissue is metabolically stressed, suggesting that BHB can be metabolized by adult neurons, but that glucose metabolism must be impaired. Taken together, these data indicate that improving neuronal, as compared with glial, metabolism in normal tissue exposed to mild hypoglycemia is critical to supporting synaptic function. These data may be useful for understanding the efficacy of the ketogenic diet. (Supported by NIH grant PO1NS39092.) 1.077 LONG-TERM PLASTICITY CHANGES IN THE EXPRESSION OF THE CANNABINOID RECEPTOR ASSOCIATED WITH EPILEPTOGENESIS AND REGULATION OF SEIZURE FREQUENCY AND DURATION IN THE PILOCARPINE MODEL OF EPILEPSY
Epilepsia, Vol. 44, Suppl. 9, 2003
1 Robert Blair, 2 Lisa Wallice, 2 Katherine Falenski, 2 Billy Martin, and 1,2,3 Robert DeLorenzo (1 Neurology, 2 Pharmacology and Toxicology, and 3 Molecular Biophysics and Biochemistry, Virginia Commonwealth
University, Richmond, VA) Rationale: Cannabinoids are known to have anticonvulsant properties, and recent studies have demonstrated that the anticonvulsant effects of cannabinoids and endocannabinoids are mediated through the activation of the cannabinoid 1 (CB1) receptor, the most highly expressed G protein–coupled receptor in the brain (Eur J Pharmacol 2001;428:52; Eur J Pharmacol 2002;452:295). However, little is known concerning the role of the CB1 receptor in regulating seizure discharge in epilepsy. This study was initiated to investigate the endogenous role of the CB1 receptor in the pilocarpine model of epilepsy. Methods: Epilepsy was caused in rats by using the pilocarpine model (Brain Res 2001;903:1). The effects of cannabinoid agonists and antagonists on seizure frequency and duration were determined by using video-EEG monitoring of the epileptic animals. Cannabinoid agonists and antagonists were administered by intraperitoneal injection. Western blot and immunohistochemical analyses were used to evaluate hippocampal CB1 protein expression from control and epileptic tissue. The anatomic distribution of the CB1 receptor in control and epileptic brains was studied by using immunohistochemical staining of the CB1-receptor protein on coronal sections. In addition, acute seizures were shown to increase levels of endogenous cannabinoids, demonstrating a possible physiologic role for these agents in modulating seizure expression. Results: The marijuana extract, d-9-tetrahydrocannabinol (THC) and other cannabinoid agonists completely abolished spontaneous seizure activity in the epileptic animals. Administration of the CB1-receptor antagonist, SR141716A, significantly increased both seizure duration and frequency. Moreover, CB1-receptor antagonism in some animals induced prolonged EEG and behavioral seizures that met the definition of status epilepticus. Western blot and immunohistochemical studies on control and epileptic tissue showed that expression of CB1-receptor protein was significantly increased in the epileptic hippocampus. Conclusions: By elucidating a role for the endogenous cannabinoid system in regulating seizure activity, the results from this study illustrate a function for the endogenous cannabinoid system in modulating seizure frequency and duration in epileptic animals and suggest that inhibition of the CB1 receptor may play a role in the development of status epilepticus. The long-term plasticity changes in expression of the CB1 receptor in association with epileptogenesis further suggest a major role for the endogenous cannabinoid system in epileptogenesis. (Supported by NIHNINDS grant R01NS23350 and P50NS25630 to R.D. and NIDA grant P50DA05274 to R.D. and B.M.)
1.078 ELECTROGRAPHIC CORRELATES OF STATUS EPILEPTICUS–INDUCED EPILEPTOGENESIS 1,3 Anatol Bragin, 1,3 Charles L. Wilson, 1 Simone Kastropil Benassi, 1 Vishal Jivan, 1 Navid Redjal, and 1,2,3 Jerome Engel (1 Departments of Neurology, 2 Neurobiology; and 3 Brain Research Institute, David Geffen School of Medicine at UCLA, Los Angeles, CA) Rationale: The mechanisms by which trauma or status epilepticus trigger epileptogenesis in some people but not in others are not clear. The development of markers to identify the changes associated with the process of epileptogenesis may provide new means for initiating earlier treatment to aid in prevention of recurrent spontaneous seizures. We performed a comparative analysis of changes in electrical activity recorded in multiple brain areas after status epilepticus in rats that develop or do not develop recurrent spontaneous seizures. Methods: Under anesthesia, recording microelectrodes (tungsten, 50 µm) were implanted in six different areas of hippocampal–entorhinal circuitry and piriform cortex. Stimulating electrodes were implanted in the right perforant path (PP), and a cannula was implanted in the right posterior hippocampus. After 1 week for recovery, baseline wide-band electrical activity (0.1 Hz–5 kHz) and evoked responses to PP stimulation in the right dentate gyrus (DG) were recorded. Recordings were subjected to power spectral analysis (PSA) of frequency bands ranging from 0.1 to 200 Hz, and brain excitability was measured during stimulation with input/output (I/O) curves. After obtaining baseline data, kainic acid (KA;
AES PROCEEDINGS 0.2 µl/0.2 µg) was injected though the cannula to produce acute status epilepticus, and the same tests were carried out at intervals during daily recordings over a period of 40 days after KA injection. Results: Of the 16 rats injected, nine (56%) developed recurrent spontaneous seizures. In comparing rats with and without seizures, we found no significant relation between severity of status epilepticus and development of epilepsy. Also, there was no consistent change in power of any frequency band of EEG activity between rats that became epileptic and those that did not. Rats that became epileptic showed a significantly steeper I/O curve of the DG response evoked by PP stimulation during the latent period after status. Interictal EEG spikes were observed bilaterally in both groups of rats. Fast ripple (FR) oscillations were observed only in the posterior hippocampus and right entorhinal cortex unilateral to the KA lesion in the group of epileptic rats with rare seizures (three to five per month; n = 7). In rats with frequent seizures (>10 per month; n = 2), FRs were recorded bilaterally in the hippocampus and entorhinal cortex. Conclusions: Our results suggest that a predisposition to develop epilepsy after brain damage may exist for different rats. In this study, global characteristics of EEG measured with PSA did not reflect ongoing development of epileptogenesis. The increased I/O slope of the PP-evoked DG response was the only significant marker of epileptogenesis in animals that developed seizures in comparison to those rats that did not. Finally, the more sites recording FRs, the greater the frequency of seizures, suggesting that FRs are not only localizing markers but also reflect the severity of epilepsy. (Supported by NIH grants NS-33310 and NS-02808.)
1.079 BDNF EXPRESSION IN DENTATE GRANULE CELL GIANT MOSSY FIBER BOUTONS 1 Steve C. Danzer, 1 Guoping Feng, and 1,2,3 James O. McNamara (1 Neurobiology, 2 Medicine (Neurology), and 3 Pharmacology and Molecular Cancer Biology, Duke University, Durham, NC) Rationale: Genetic and pharmacologic manipulations indicate that BDNF plays an important role in epileptogenesis. Furthermore, epileptogenesis is associated with increased BDNF expression in the mossy fiber pathway of the hippocampus. This pathway contains both the presynaptic terminals of dentate granule cells (giant mossy fiber boutons), and the postsynaptic terminals of CA3 pyramidal cell proximal dendrites (thorny excrescences). To identify which, if any, of these processes express BDNF, and to determine the likely targets of increases in BDNF protein, BDNF immunohistochemistry was examined in control and pilocarpine-treated transgenic mice expressing green fluorescent protein (GFP) in dentate granule cells and CA3 pyramidal cells. Colocalization of GFP in mossy fiber boutons or thorny excrescences with BDNF immunoreactivity permits establishing the cellular localization of BDNF. Methods: Two- to three-month-old GFP-expressing mice from the M line (Feng et al., 2000) were treated either with scopolamine (1 mg/kg) followed 15 min later by pilocarpine (340 mg/kg) or sterile saline. All procedures conformed to NIH and institutional guidelines for the care and use of animals. After 3 h of status epilepticus (or control treatment) mice were treated with diazepam (10 mg/kg) and allowed to recover for 48 h before killing. BDNF immunohistochemistry was conducted by using a polyclonal rabbit anti-BDNF antibody (1 µg/ml; Amgen) followed by Alexa Fluor 546 secondary antibodies (1:500; Molecular Probes). Neurons were imaged by using a Leica confocal microscope equipped with a ×63 objective. Results: In control animals, ∼36% of GFP-expressing mossy fiber boutons were also positive for BDNF (116 boutons from five animals). In contrast, no BDNF-positive CA3 pyramidal cell thorny excrescences were found (37 from four animals). Interestingly, BDNF-negative thorny excrescences were often observed adjacent to BDNF-positive mossy fiber boutons. Forty-eight hours after pilocarpine-induced status epilepticus, ∼80% of mossy fiber boutons were BDNF immunopositive. Insufficient numbers of thorny excrescences were available for analysis. Conclusions: The present study establishes the localization of BDNF protein in dentate granule cell giant mossy fiber boutons in normal animals and demonstrates that the seizure-induced increase in BDNF is also, at least in part, in mossy fiber boutons. Because CA3 pyramidal cells
31
are the postsynaptic targets of granule cell boutons in stratum lucidum, these findings indicate that seizures increase presynaptic BDNF at the granule cell/CA3 pyramidal cell synapse, and suggest that increased release of BDNF onto pyramidal cells may participate in epileptogenesis. Autocrine activation of TrkB receptors on mossy fiber boutons may occur as well. [Supported by a postdoctoral fellowship from the American Epilepsy Society (SCD) and by NIH grants NS07370 and NS32334 and NINDS grant NS17771.] 1.080 DIFFERENTIAL REGULATION OF MYOCYTE ENHANCER FACTOR 2 (MEF2) CELL FATE DETERMINATION MOLECULES IN DENTATE GRANULE CELLS AFTER PILOCARPINE-INDUCED STATUS EPILEPTICUS 1 Robert C. Elliott, 2 Fulvio A. Scorza, 1 Brian Kruegel, and 3 Daniel H. Lowenstein (1 Neurology, Beth Israel Deaconess Medical Center, Boston, MA; 2 Clinical Neurology and Neurosurgery, Universidade Federal de S˜ao Paulo/Escola Paulista de Medicina, Sao Paulo, Brazil; and 3 Neurology, University of California San Francisco, San Francisco, CA) Rationale: Previous studies in our laboratory have implicated the basic helix-loop-helix (bHLH) family of cell fate–determination molecules as potentially involved in normal granule cell development as well as epilepsy-associated reorganization. bHLH members have been shown to cooperate with members of another family of cell-fate determinants, the myocyte enhancer factor 2 (MEF2) family of transcription factors, in directing the development of specific cell phenotypes in muscle myogenesis. The objective of this study was to investigate the possible involvement of the MEF2 family in aspects of neurogenesis and axon outgrowth associated with the onset of certain forms of epilepsy. To do so, we analyzed the patterns of MEF2 family mRNA expression in the rat dentate gyrus after pilocarpine-induced status epilepticus (SE). Methods: Adult, male Sprague–Dawley rats (180–200 g) were given i.p. atropine methylbromide followed 20 min later by pilocarpine hydrochloride to induce SE. Seizure activity was monitored behaviorally and terminated with diazepam after 2 h of convulsive SE. Control rats received saline instead of pilocarpine. Animals were killed 3, 7, 14, 28, or 60 days later, and perfusion-fixed brains were processed for nonradioactive in situ hybridization analysis of MEF2 family mRNAs. Digoxygenin-labeled in situ probes were transcribed from DNA templates generated by polymerase chain reaction (PCR) from neonatal or adult rat hippocampal cDNA libraries. Results: With a pilocarpine-induced rat model of human temporal lobe epilepsy (hTLE), mRNA levels for each MEF2 family member (MEF2A, 2B, 2C, and 2D) were evaluated by in situ hybridization over a broad time course of epileptogenesis. Each MEF2 member was detected throughout the granule cell layer of normal, untreated adult rats. During epileptogenesis, MEF2C and 2D were markedly increased within 7– 14 days after SE, and remained elevated above control levels for ≤60 days. In contrast, MEF2A and 2B demonstrated no apparent transcriptional regulation at any time after SE. Conclusions: The differential regulation in the dentate gyrus of various members of the MEF2 family suggests a potential role for these molecules, particularly MEF2C and 2D, in epilepsy-associated reorganization of granule cell neurons. In addition, the coincident regulation of members of the MEF2 family, as shown here, and members of the bHLH family, as shown previously, may be indicative of a cooperative role of these two families of cell fate–determination molecules in granule cell alteration after SE. [Supported by NIHRO1NS39950 (D.H.L.) and FAPESP (Brazil) (F.A.S).] 1.081 MOSSY CELL DELETION DOES NOT CAUSE HYPEREXCITABILITY IN THE DENTATE GYRUS: IMPLICATIONS FOR EPILEPSY Anna d.H. Ratzliff, Allyson L. Howard, and Ivan Soltesz (Anatomy & Neurobiology, University of California, Irvine, CA) Rationale: Loss of mossy cells and interneurons from the hilus of the dentate gyrus is a major histologic hallmark of human temporal lobe epilepsy. Mossy cells can influence the activity of granule cells both directly through excitatory synapses and indirectly by exciting
Epilepsia, Vol. 44, Suppl. 9, 2003
32
AES PROCEEDINGS
interneurons. However, the net effect of mossy cell activity in normal circuits and the consequences of mossy cell loss in epileptogenesis remain controversial. Methods: Mossy cells were identified and differentiated from hilar interneurons in living slices by using a fluorescent retrograde tracer injected in vivo into the contralateral dentate gyrus. Acute ablation of either prelabeled mossy cells or hilar-granule cell border interneurons was performed by visually directed, controlled, rapid single-cell aspiration by using an enlarged patch pipette. Perforant path evoked granule cell field responses were recorded before and after ablation. Results: Ablation of mossy cells caused a significant decrease in granule cell layer excitatory postsynaptic potential (fEPSP) peak amplitude (84.6 ± 1.1% of preablation) and population spike amplitudes (60.2 ± 5.7% of preablation). In contrast, control ablation of interneurons caused a significant increase in fEPSP amplitude (111.8 ± 1.2% of preablation). To assess the effect of mossy cell activity along the septotemporal axis of the hippocampus, mossy cells were ablated in slices cut along the longitudinal axis. Even at a distance >2 mm away from the site of ablation, a significant decrease in the granule cell fEPSP (87.3 ± 6.1% of preablation) was recorded. Conclusions: Using novel techniques to study mossy cells in the normal dentate circuit, we demonstrate that the acute ablation of mossy cells in both the horizontal and axial slice preparations causes hypoexcitability in granule cells. These data demonstrate that the net effect of mossy cell activity is to excite granule cells and is in agreement with the “irritable mossy cell” hypothesis (Santhakumar et al. J Physiol 2000; Ratzliff et al. TINS 2002) of epileptogenesis. [Supported by: the NIH (NS35915) to I.S. and UCI-MSTP to A.R. and A.H.]
1.082 DISAPPEARANCE OF ICTAL ACTIVITY: A DEVELOPMENTAL MODEL OF THE RESOLUTION PHASE OF CHILDHOOD EPILEPSY 1 Charles E. Niesen and 2 Shundi Ge (1 Division of Pediatric Neurology, Cedars-Sinai Medical Center, and 2 Division of Neurology, Children’s Hospital Los Angeles, Los Angeles, CA) Rationale: Given appropriate medical treatment, most children “outgrow” their seizure disorders, and many will do so after seizure suppression for only 1–2 years. Clearly, for children, epilepsy is not a permanent or irreversible state. Understanding the molecular events that underlie these changes may provide new treatment options. In our tetrodotoxin (TTX) rat hippocampal brain slice model of chronic epilepsy, we have demonstrated and studied just such a resolving period of seizure-like activity. Methods: Hippocampi were prepared from 7-day-old Wistar rats according to the method of Niesen and Ge (1999). TTX at 0.5 µM was addded to the culture media for 7 days and then removed. Whole-cell patch recordings were performed on CA1 neurons after TTX removal, and the microelectrode contained (in mM): 140 KGluc, 10 HEPES, 0.1 CaCl2 , 1.1 BAPTA, 2 Mg-ATP, and 0.1 GTP. GABAA -mediated potentials were studied in the presence of AMPA and NMDA-receptor blockers CNQX (50 µM) and APV (30 µM). The day of TTX removal was designated E0, and the days thereafter were numbered accordingly. Results: On days E0–2, CA1 neurons displayed robust seizure-like activity, including spontaneous and evoked depolarizing bursts (200– 500 ms) with multiple superimposed action potentials and multiple population spikes on evoked field potentials. By E3–5, the duration of the synaptic depolarizing potentials (100–300 ms) and number of action potentials had decreased. Spontaneous bursts were infrequent. By E6–7, the duration of evoked synaptic potentials decreased to 0.01) were significantly higher for patients compared with controls. There were no significant group differences on measures of sleep efficiency (88 vs. 90%), mean stage 2 sleep%TST (46.5 vs. 49), mean slow-wave sleep%TST (17.9 vs. 17.1), or respiratory disturbance index (2.98 vs. 2.8). One patient had three electrographic seizures, but no patients had clinical seizures during the 2 nights of the study. Overall attentional performance was significantly worse in patients compared with controls (10.5 ± 7.1 vs. 4.56 ± 3.7; p > 0.01). There was a significant correlation between increased REM latency, increased WASO, and abnormal daytime attention function in patients (r = 0.70; r = 0.60, respectively). Conclusions: This controlled pilot study, involving a modest sample size, demonstrates disturbed sleep architecture among children with idiopathic generalized epilepsy compared with healthy controls. Disturbed sleep architecture appears to be of clinical consequence given its relation to significantly poorer daytime attentional function. (Supported by Clinical Research Fellowship Grant: Epilepsy Foundation; UW GCRC Grant: 2M 01 RR03186-18.)
1.155 MULTIPLE BAND FREQUENCY ANALYSIS FOR NEONATAL ELECTROENCEPHALOGRAPHIC SEIZURES Michihiro Matsumoto, Hiroshi Otsubo, Ayako Ochi, Koji Iida, Shahid Parvez, Bill Chiu, Masaomi Kitayama, and O. Carter Snead III (Division of Neurology, The Hospital for Sick Children, Toronto, ON, Canada) Rationale: The clinical outcome of neonatal seizures has been studied with some success. Combinations of EEG, clinical findings, and neuroimaging studies are useful predictors of developmental outcome, but do not correlate well with postneonatal seizures (PNSs). We applied new frequency-analysis method, Multiple Band Frequency Analysis (MBFA), for neonatal seizures. MBFA provides both high frequency and temporary resolution to analyze frequency components and time course of neonatal EEG seizures. We compared the MBFA findings between the group with PNSs at age 18 months (Group A) and the other group without PNSs (Group B). Methods: We reviewed medical records of neonates admitted to the Hospital for Sick Children (HSC) from 1996 to 1998, which documented electrographic seizures. We sifted the records that had clinical follow-up ≥18 months at HSC. We selected ictal EEG that lasts >10 s. We used digital EEG system with 16 channels bipolar neonatal EEG and extracerebral monitoring channels, including EOG, ECG, EMG, and thoracic (respiratory) movements. Digital EEGs are recorded ≥30 min at sampling rate of 200 Hz. We used MBFA to analyze EEGs and to make power spectra. MBFA applied multiband-pass filter to
AES PROCEEDINGS decompose EEG signals to particular frequency components. MBFA measured amplitudes of decomposed frequencies in every assigned time window. MBFA provides power spectra with obtained amplitudes of every assigned frequency band. MBFA can trace the highest power frequency in every assigned time window. In this study, we analyzed the neonatal EEGs that range from 1 to 20 Hz with 0.2-Hz frequency resolution in every 100 mS. We collected duration, spectral power, and frequency of the sustained dominant seizure activities that calculated by MBFA. We compared those data between Group A and Group B with Mann–Whitney’s U test. Results: Twenty-eight patients had neonatal seizures. We excluded 13 patients as they lacked postneonatal follow-up or qualified EEG recordings. Group A had 11 neonates, and Group B had four neonates. We obtained 52 electrographic seizures in Group A (two to seven seizures per patient; average, 5 seizures) and 16 seizures in Group B (two to seven seizures per patient; average, four seizures). Group A had the longer duration of the sustained peak frequency (mean ± SD; 9.5 ± 5.4 s) than that of Group B (4.8 ± 3.0 s) with a statistical significance (p < 0.01). There were not any significant differences in the peak frequency (2.0 ± 1.5 vs. 1.8 ± 0.9 Hz) and the power of the peak frequency (26.1 ± 18.1 vs. 30.7 ± 13.2 dB) between two groups. Conclusions: The neonates who had longer duration of sustained peak frequency developed PNS. MBFA is a useful tool to analyze frequency components and time course of neonatal EEG seizures for prediction of developing epilepsy in postneonatal period.
1.156 ORAL AND MANUAL DYSPRAXIA IN BENIGN CHILDHOOD EPILEPSY WITH CENTROTEMPORAL SPIKES Carmen Silvia Miziara and Maria Luiza Manreza (Neurology, Hospital das Cl´ınicas from Medicine School of Sao Paulo University, Sao Paulo, SP, Brazil) Rationale: To evaluate the oral and manual praxia in benign childhood epilepsy with centrotemporal spikes (BECTS) compared with control group. Methods: We studied 40 right-handed children (16 girls, 24 boys) with BECTS. They were 7–13 years old and were compared with a control group composed of children as close as possible in age at testing, sex, and level in school. The control groups were composed of 154 children for assessment of oral praxia and other 82 for assessment of manual praxia. The oral praxia screening included volitional tongue and lip, cheek, and jaw movements, isolation and sequence. The manual praxia was tested with single sequence of three movements and a different sequence of two and three movements. Results: In the total subtests, the group of BECTS revealed abnormalities in 166 subtests (4.15 failure for child) and 191 subtests in the control group (1.13 failure for child). There was no difference of age at testing, sex, level in school, laterality of EEG discharges, Illinois Test Psycholinguistic Ability, and dichotic listening test. All subtests that examined oral and movement in sequence (seven of 78 subtests) revealed abnormalities in BECTS when compared with control group (Mann–Whitney U test, p < 0.001). In relation to manual praxia, we found statistically significant difference of p < 0.05 between BECTS group and control group in a sequence of three different movements (Mann–Whitney U test, p = 0.027) and single sequence of three movements (Mann–Whitney U test, p = 0.007). Conclusions: Children with BECTS showed oral and manual dyspraxia in comparision with the control group.
1.157 MAGNETOENCEPHALOGRAPHY UNDER GENERAL ANESTHESIA FOR CHILDREN WITH EPILEPSY 1 Ayako Ochi, 1 Hiroshi Otsubo, 1 Elizabeth Pang, 1 Shelly K. Weiss, 1 Amrita Hunjan, 1 Irene Elliott, 2 Sheelagh M. Kemp, 3 Sylverster Chuang, and 1 O. Carter Snead III (1 Neurology,. 2 Anaesthesia, and 3 Diagnostic Imaging, The Hospital for Sick Children, Toronto, ON, Canada) Rationale: Magnetoencephalography (MEG) has been increasingly used as a diagnostic tool for patients with epilepsy. The accuracy of the MEG source localization onto magnetic resonance imaging (MRI) is
57
dependent on the ability of the patient to keep still in the dewar. Thus children with epilepsy who are uncooperative require general anesthesia (GA) for the MEG and MRI studies. GA has a tendency to eliminate the epileptic discharges. We evaluated the MEG studies under GA compared with prolonged video-EEG (VEEG) results in children with epilepsy. Methods: A total of 185 pediatric patients with epilepsy underwent MEG study with simultaneous 19-channel scalp-EEG at The Hospital for Sick Children in Toronto between August 2000 and March 2003. These patients also had VEEG for 16–96 h with 19 scalp electrodes. Twenty-two of 185 children underwent GA for MEG and MRI studies. Propofol and remifentanil were used for GA. Whole-head 151-channel gradiometers were used for MEG recording (CTF, Port Coquitlam, BC, Canada). We performed MEG dipole analysis using single moving dipole modeling. We compared the interictal discharges among MEG, simultaneous EEG, and VEEG recording. Fewer than six MEG dipoles were defined as being inconclusive MEG results. Results: Twenty-two children consisted of 10 girls and 12 boys. Age at time of MEG study ranged between 7 months and 9 years (mean ± SD, 4.6 ± 2.6 years). Age at seizure onset ranged from 1 day of life to 3.5 years (mean ± SD, 0.9 ± 1.0 years). Nineteen had MRI abnormalities. There were no interictal MEG spikes in four patients and fewer than six MEG dipoles in two. In these six patients with fewer than six MEG dipoles, VEEG captured intermittent interictal spike–waves. The remaining 16 patients showed six or more MEG dipoles. In these 16 patients, VEEG showed continuous interictal spike–waves in six, intermittent spike–waves in eight, and no spikes in two. Sensitivity of MEG under GA was 73% (16 of 22 patients). The efficacy of MEG results under GA, which provided valuable information for epilepsy management, was 63% (14 of 22 patients). Conclusions: GA influenced interictal epileptiform discharges in short-term MEG recording. Continuous spike–waves can exist under GA and be localized by MEG, although sensitivity appears to be decreased by GA. Short-term MEG study under GA has a certain disadvantage in collecting interictal spikes compared with the prolonged video-EEG monitoring in uncooperative children with epilepsy. 1.158 DIFFERENT EEG RESULTS IN INPATIENT AND OUTPATIENT, ICU, AND OTHER CASES: 2,584 CONSECUTIVE EEGS S. Robert Snodgrass [Neurology (Pediatrics), Childrens Hospital Los Angeles, Keck/USC School of Medicine, Los Angeles, CA] Rationale: Nonspecialists often assume that abnormal EEGs have the same meaning in all patients. We find that abnormalities are more frequent in inpatients and that the kinds of abnormalities differ between inpatients and outpatients, sick, ventilated, and healthy patients. Methods: Results of all EEGs except videotelemetry studies performed between July 2001 and January 2003 were reviewed. An electrographic seizure is defined as a continuous discharge lasting ≥5 s. Results: Most EEGs were outpatient EEGs. Outpatients comprised 73.2% of all studies, 1,899 of 2,584 during this 18-month period. Fiftyeight percent of all outpatient studies included sleep. Twenty-three were sedated or drug-induced sleep (1.2% of outpatient EEGs). The percentage of normal studies was similar between waking only and waking and sleep (WS) studies: 61.3 and 62.3%, but WS studies included significantly more seizures and epileptiform discharges: 18.7% (W) and 30.3% (WS). Recorded seizures were also more frequent in WS studies: 97 outpatient EEGs included seizures, of which 61% were primary generalized type. Only 14% of 61 inpatient EEGs with seizures had primary generalized seizures. Inpatients on ventilators were 2.5 times more likely to have abnormal EEGs than those not on ventilators. Only two of 23 sedated sleep studies showed epileptiform discharges, and only three of 28 STAT EEGs (excluding brain-death studies) included seizures or major background abnormalities. If we exclude NICU patients, we were more likely to obtain sleep in outpatients (58%) than inpatients (39%) (Fig. 1). Conclusions: Sick inpatients more often have abnormal EEGs than do outpatients, but less often have epileptiform EEGs. Six percent of EEGs included seizures; seizure types were different in inpatients and outpatients. Outpatient EEGs including natural sleep had a higher incidence of epileptiform abnormalities than sedated sleep studies or studies limited to waking. STAT EEGs rarely revealed dramatic abnormalities. (Supported by Division funds.) Epilepsia, Vol. 44, Suppl. 9, 2003
58
AES PROCEEDINGS 1 Katherine M. Wambera, 1 Mary B. Connolly, and 2 Peter K.H. Wong (1 Pediatric Neurology and 2 Clinical Neurophysiology, British Columbia
Children’s Hospital, Vancouver, BC, Canada)
1.159 THE ACUTE EFFECT OF MUSIC PERIODICITY ON ROLANDIC SPIKES: A RANDOMIZED, SINGLE-BLINDED, CROSSOVER, PILOT CLINICAL TRIAL OF THE ACUTE EFFECT OF MUSIC WITH LONG-TERM PERIODICITY AND REPEATED MELODIC LINE (M-LTP/RML) ON INTERICTAL SPIKE DISCHARGES (ISD) IN CHILDREN WITH BENIGN CHILDHOOD EPILEPSY WITH CENTROTEMPORAL SPIKES (BCECTS) Robert P. Turner (Neurology, Pediatrics, and Neurological Surgery, Medical University of South Carolina, Charleston, SC) Rationale: To investigate the effect of exposure to music with longterm periodicity and repeated melodic line (M-LTP/RML) on frequency of interictal spike discharges (ISDs) in children with benign childhood epilepsy with centrotemporal spikes (BCECTS). The goal of this pilot study was to demonstrate decreased ISDs due to exposure to MLTP/RML. Exposure to M-LTP/RML (Mozart’s Sonata for Two Pianos, K. 448) has been shown to enhance spatial–temporal functioning. Both antiepileptiform and antiseizure properties of M-LTP/RLM have been demonstrated by the seminal work of John Hughes, without validation by clinical trial (Gates JR. Letter to the editor: the Mozart effect. Epilepsy Behav 2002;3:483; Hughes JR. Review: the Mozart effect. Epilepsy Behav 2002;2:396–417). A mechanism of this effect, not due to relaxation or enjoyment of the music, has been proposed based on the trion model of Mountcastle’s columnar organization of the neocortex (Rauscher FH, Shaw GL. Key components of the Mozart effect. Percept Mot Skills 1998;86:835–41; Shaw GL, Bodner M. Music enhances spatial-temporal reasoning: towards a neurophysiological basis using EEG. Clin Electroencephalogr 1999;30:151–5). Methods: Four subjects with BCECTS, aged 5–9 years, were recruited for this prospective, randomized, single-blinded, crossover, pilot clinical study. ISD frequency/minute was averaged over each of three periods/hour, over 4 h of continuous EEG monitoring: (a) Silence, 15 min; (b) Exposure, M-LTP/RML or control (placebo) music (Beethoven’s F¨ur Elise) (18 min); and (c) Wash-out period, 27 min. Mean ISD count/epoch, standard deviations, variance, and correlation data were calculated. Results: A significant (>30%) decrease in mean ISDs was demonstrated comparing baseline with exposure to M-LTP/RML, but not to control music, in two subjects demonstrating sufficient ISDs for data collection and statistical analysis. Conclusions: Demonstration of decreased ISDs from exposure to MLTP/RML indicates an effect on mechanisms of spike generation. If reproducible in a sufficiently powered prospective, randomized clinical trial, this effect would contribute to understanding epileptogenesis and new treatment strategies for aborting and preventing seizures. A larger clinical trial is proposed to study this effect on spikes as well as seizures, with subsequent studies of mechanisms indicated. [Supported by Outpatient General Clinical Research Center (GCRC), Medical University of South Carolina.]
1.160 PERIODIC LATERALIZED EPILEPTIFORM DISCHARGES (PLEDS) IN PEDIATRIC EEG RECORDINGS
Epilepsia, Vol. 44, Suppl. 9, 2003
Rationale: To describe the clinical, neuroimaging, EEG features, and neurologic outcome in pediatric patients, outside of the neonatal period, with periodic lateralized epileptiform discharges (PLEDs) on EEG. Methods: Children were identified from the EEG database in the Department of Clinical Neurophysiology at British Columbia’s Children’s Hospital. Clinical records, neuroimaging, EEG data, and outcome were reviewed. Results: From July 1, 1993, to December 31, 2002, 24,535 EEGs were performed. PLEDs were seen in 132 EEGs (0.5%). Of these recordings, 113 were performed in 74 children, age 1 month or older (n = 42 male, n = 32 female). At the time of detection of PLEDs, the mean age of the patients was 4.3 years; range, 6 weeks–19 years. Thirty-one patients had preexisting developmental delay or learning problems; 23 patients had a preexisting epilepsy; congenital brain abnormality, or neurocutaneous syndrome (n = 12), remote infarction (n = 4), metabolic disorders (n = 3), chromosomal disorders (n = 2), and Dravet syndrome (n = 2). Of patients, 91% presented with seizures, 80% had altered level of consciousness, and 53% required ventilation. On examination, 30% were hemiparetic, and 10% had brainstem signs. Acute neuroimaging was obtained in 52 patients [computed tomography (CT) only, n = 28; CT + magnetic resonance imaging (MRI), n = 20; MRI only, n = 4]. Neuroimaging demonstrated focal or diffuse cerebral edema (n = 58), and congenital brain abnormalities (n = 9).Underlying etiology was infectious or postinfectious encephalopathy (n = 29), a known seizure disorder with or without an acute exacerbation (n = 23), hypoxic ischemic encephalopathy (n = 9), cerebral infarction (n = 6), and tumor (n = 3). PLEDs were focal in 70 and multifocal in 43 EEG recordings. Focal neuroimaging changes were concordant with location of PLEDs in 14 (87%) of 16 and discordant in two (13%) of 16. Clinical and/or electrographic seizures (n = 37) were also seen during the EEGs. Eighteen patients had PLEDs detected on serial EEGs. The location of the PLEDs remained constant in each recording in 10 patients and varied in eight. Outcome data are available for 55 patients. Death occurred in 16 patients (21%), and three survivors were lost to follow-up. In 55 survivors, where follow-up is available, microcephaly was seen in six, moderate to severe mental retardation in 21, cerebral palsy in 30, and epilepsy in 41. Conclusions: PLEDs are rare in pediatric EEGs. They are associated with abnormal neuroimaging. The prognosis is poor, with significant mortality and moderate to severe mental retardation, cerebral palsy, and epilepsy in many survivors.
1.161 NOVELTY EVENT-RELATED POTENTIAL ABNORMALITIES IN CHILDREN WITH INFANTILE SPASMS 1,2 Klaus G. Werner, 2 Torsten Baldeweg, 1 Stewart G. Boyd, 1,3 Rod C. Scott, and 1 Brian G. Neville (1 Neurosciences Unit, 2 Developmental Cognitive Neuroscience Unit, and 3 Radiology and Physics Unit, Institute of Child Health, London, United Kingdom) Rationale: Infantile spasms (ISs) are associated with acute cognitive regression, long-term learning disability, and autistic spectrum disorder. Although there may be a structural basis to the epilepsy, seizure activity itself is implicated in the pathogenesis of these disabilities. There are strong indications of temporal lobe dysfunction in children with this and related epileptic regressions; from the site of lesions in tuberous sclerosis and early-onset developmental tumours, and from the EEG localisation in the Landau–Kleffner syndrome. We therefore tested if the temporal lobe is functionally abnormal in children with ISs by recording eventrelated potentials (ERPs) to novel environmental sounds known to be generated in temporal lobe regions. Methods: Twenty-six term infants (age range, 2–10 months) and 24 infants with ISs (range, 2–10 months) were recruited. EEG was recorded continuously from 19 electrodes (10-20 system). An oddball paradigm was used with frequent sounds (p = 80%, 1 kHz), deviants (p = 10%, 1.5 kHz), and brief novel sounds (p = 10%), delivered binaurally (interstimulus interval, 700 ms) via speakers at a distance of 30 cm. Infants were either asleep (stage II) or awake/drowsy. Three blocks were recorded, with 240 deviants and 240 novels.
AES PROCEEDINGS Results: Robust and reproducible ERPs to novel sounds were detected in sleep in all normal infants showing a negativity (N2) at 250 ms and positivity (P3) at 500–900 ms over inferior temporal electrodes. Polarity inversion at central electrodes is consistent with a generator site within the temporal lobe. Control infants showed decreasing P3 latencies with age [F(2, 17) = 22.3, p < 0.0001]. Despite the presence of severe EEG abnormalities, ERPs to novel sounds could be recorded in 75% of patients with ISs. Patients with ISs had significantly prolonged P3 [F(1, 40) = 13.6, p = 0.001], but normal N2 latencies compared with controls, with age as a covariate. Conclusions: 1. Novelty ERPs can be recorded reliably from age 2 months in control infants and infants with ISs. Their topography suggests a bilateral generator in the posterior superior temporal lobe. 2. ERPs show markedly decreasing latencies of the P3 in the first year of life. 3. Patients with ISs show increased P3 latencies during the first year of life, consistent with the hypothesis of abnormal temporal lobe development. (Supported by Hospital Savings Association, The Wellcome Trust.)
1.162 STUDIES ON CEREBROSPINAL FLUID IONIZED CALCIUM AND MAGNESIUM CONCENTRATIONS IN CONVULSIVE CHILDREN Hitoshi Yamamoto, Hiroshi Murakami, Noriko Kamiyama, Yusaku Miyamoto, and Miho Fukuda (Pediatrics, St. Marianna University School of Medicine, Kawasaki, Kanagawa, Japan) Rationale: The levels of ionized calcium (iCa) and magnesium (iMg) were measured in the cerebrospinal fluid (CSF) of convulsive and nonconvulsive children to investigate the relation between seizure manifestation and CSF iCa and iMg levels. Standard levels of CSF iCa and iMg were also established. Methods: CSF samples from 23 patients, ages 0–15 years, with various forms of seizures and 26 age-matched controls were collected by lumbar puncture. CSF was obtained anaerobically, and the concentrations of CSF iCa and iMg were measured with an electrolyte analyzer (Stat Profile Ultra, NOVA, U.S.A.) immediately. Results: The levels of CSF iCa were significantly higher in control children younger than 11 months compared with children older than 12 months. The levels of CSF iMg in control children did not differ significantly with aging. The levels of CSF iCa in convulsive children did not differ significantly from control levels. The levels of CSF iMg in convulsive children were significantly lower than in controls. Conclusions: These results suggest that seizure manifestation may be related to age-dependent changes in iCa and decreased iMg in the brain.
1.163 OROFACIAL AUTOMATISMS INDUCED BY ACUTE WITHDRAWAL FROM HIGH-DOSE MIDAZOLAM MIMICKING NONCONVULSIVE STATUS EPILEPTICUS IN A CHILD 1 David K. Epstein and 2 Marc P. DiFazio (1 Neurology and 2 Child and Adolescent Neurology Service, Walter Reed Army Medical Center, Washington, D.C.) Rationale: Nonconvulsive status epilepticus (NCSE) is an uncommon occurrence in children. It is often difficult to identify without an electrographic correlate, especially in children who are critically ill or receiving medications that may alter level of consciousness. Additionally, NCSE has been induced in patients after introduction or acute withdrawal of medications such as benzodiazepines (BZDs). We report a 4-year-old child with the acute onset of mental status changes in association with orofacial automatisms clinically consistent with complex partial seizures after acute withdrawal of high-dose midazolam (MDL). Methods: The patient was admitted to the ICU for treatment of epiglottitis. He required high-dose MDL for agitation and airway stabilization. He received this up to a rate of 270 µg/kg/h. Infusion was stopped 24 h before extubation. After extubation, the patient was alert, awake, with normal speech and ability to identify parents. Six hours later, he had acute onset of mental status changes, manifest as a “twi-
59
light state”: eyes open with orofacial automatisms and low-amplitude myoclonic jerks. He was occasionally able to comply with simple commands. No change was noted over the next 8 h. On transfer, concerns regarding NCSE arose. An EEG demonstrated diffuse slowing with no evidence of electrographic seizures. Although the EEG was reassuring, the clinical manifestations prompted continued suspicion of NCSE. A trial-dose of intravenous lorazepam (LZP) was given without clinical or electrographic change. Orofacial automatisms continued with waxing and waning level of consciousness over the next several days. Magnetic resonance imaging (MRI) of brain, spinal tap, and laboratory values were normal. Forty-eight hours after symptom onset, the patient began to manifest unusual stereotypies, visual hallucinations, and inappropriate laughter. He returned slowly to baseline over the next 72 h without residual symptoms or deficits. Results: This patient’s initial clinical manifestation, occurring after prolonged administration of high-dose MDL, closely mimicked NCSE. Reports of NCSE after BZD withdrawal prompted concerns regarding this syndrome. However, in children, withdrawal from high-dose MDL has been reported to cause a syndrome distinct from NCSE, consisting of acute-onset dyskinesias, chorea, ataxia, and mental status changes. Evaluation for an epileptic substrate in several of these patients was negative. This syndrome typically resolved over days to weeks, without residua in the majority of children. Conclusions: Acute withdrawal from high-dose MDL is associated with a distinct clinical syndrome closely mimicking NCSE in children. Children who manifest orofacial dyskinesias/automatisms in association with mental status changes after such withdrawal should be urgently evaluated for NCSE. If diagnostic studies are negative, a movement disorder induced by MDL withdrawal should be considered. This disorder is typically short-lived, and full recovery can be expected.
1.164 MANAGEMENT OF REFRACTORY STATUS EPILEPTICUS IN CHILDREN Syed A. Hosain and Novette S. Green (Pediatric Epilepsy Center, New York Presbyterian/Weill Cornell Medical College, New York, NY) Rationale: Only a handful of studies report on the efficacy of the treatment of refractory status epilepticus (RSE). In this report we present our experience on the efficacy of barbiturates and benzodiazapine (BZD) coma in the treatment of RSE. Methods: In a retrospective chart review of patients admitted to the Pediatric Intensive Care Unit (PICU) from 1997 through 2002, we identified 17 patients with SE. Seven of these patients had RSE. Patients with RSE were treated with continuous midazolam (MDL) and/or pentobarbital (PTB) infusion. Continuous electroencephalography (EEG) monitoring with 16-channel EEG was used to guide management. We collected information on demographics, pharmacologic, treatment and EEG. Efficacy of treatment was the primary outcome measure. EEG changes during treatment and treatment-emergent side effects were secondary measures. Results: Mean age of the population was 4.5 years (range, 1.5 weeks–5 years). The mean dose of MDL was 0.5 mg/kg/h (range, 0.2– 1.4 mg/kg/h). The mean dose of PTB was 3 mg/kg/h (range, 0.9–5 mg/kg/h). Five patients initially treated with MDL were then treated with PTB. One patient received MDL only. One patient received PTB only. Treatment outcome: 1. Five of six patients treated with MDL did not respond. 2. All five nonresponders to MDL responded to PTB. 3. All six patients who received PTB responded. 4. Only one patient receiving MDL responded. EEG findings: 1. No patient receiving MDL (n = 6) achieved suppression burst pattern (S-B). 2. Six of seven patients receiving PTB showed S-B pattern. Adverse effects: 1. Three patients with PTB developed hemodynamic cardiovascular instability. 2. Only one patient with MDL developed cardiovascular instability.
Epilepsia, Vol. 44, Suppl. 9, 2003
60
AES PROCEEDINGS
Conclusions: In this small cohort of patients, PTB appeared more potent for treatment of RSE. EEG changes are more predictive with use of PTB. MDL is less likely to cause hemodynamic instability when compared with PTB. More studies are needed to determine the most effective and safe treatment modality for RSE.
1.165 NEW-ONSET SEIZURES PRESENTING AS STATUS EPILEPTICUS IN A POPULATION OF CHILDREN Tena Rosser, Taeun Chang, Adeline Vandever, Christine Kennedy, and William D. Gaillard (Neurology, Children’s National Medical Center, Washington, D.C.) Rationale: Status epilepticus (SE) is characterized by continuous seizure activity and is a medical emergency. The purpose of this study was to characterize prospectively all cases of new-onset seizures presenting as SE in a large population of children. Methods: Over 27 months, all children presenting to our institution with a new-onset seizure were prospectively evaluated as part of a clinical care pathway. Results: Of 473 children with new-onset seizures, 65 (14%) presented with SE (seizures lasting >20 min). Twenty-two seized for >60 min. Patient ages ranged from 0.2 to 15.6 years (mean, 3.8 years; median, 2.9 years). Multiple seizure types were noted at presentation including complex partial (48%), generalized (34%), focal with secondary generalization (15%), and simple partial (3%). The average age was 2.2 years for onset with generalized seizures, 5.1 years for focal seizures, and 4.2 years for focal seizures with generalization. Etiologies included idiopathic (32%), remote symptomatic (24%), illness related (26%), CNS infection (9%), electrolyte abnormalities (5%), trauma (3%), and hemorrhage (1%). None of the cases of SE was secondary to tumor or stroke. Twenty-eight patients (43%) had normal EEGs at presentation. Focal spikes or slowing were identified in 30 patients, whereas 14 patients had generalized features on EEG. Two children had PLEDs. Head computed tomography (CT) was performed on all patients and was normal in 69%. Abnormal findings included nonspecific changes, edema, trauma, cortical dysplasia, metabolic disease, periventricular leukomalacia, and acute disseminated encephalomyelitis. Conclusions: In conclusion, 14% of children had SE as an initial presentation of a new-onset seizure. Complex partial SE was the most common seizure type at presentation. The average age of SE onset was at a relatively young average age of 3.8 years. Toddlers tended to present with generalized seizures, whereas older children were more likely to present with focal seizures with or without generalization. A wide range of etiologies was identified, with a third of cases idiopathic. Almost half of the population had a normal EEG at presentation, but among abnormal EEGs, focal findings were most common. The majority of patients had a normal head CT at presentation, but diagnostically relevant abnormalities were identified in a minority of patients. These findings may be helpful in the diagnosis, workup, and management of children presenting with new-onset SE.
1.166 OXCARBAZEPINE-INDUCED STATUS EPILEPTICUS DURING TREATMENT OF INTRACTABLE COMPLEX PARTIAL SEIZURES IN CHILDREN WITH INTELLECTUAL DISABILITY Inna I. Vaisleib, Shelley D. Williams, Robyn A. Neft, and Jamie L. Bengel (Child Neurology, Children’s Hospital of Pittsburgh, Pittsburgh, PA) Rationale: Oxcarbazepine (OXC) was approved as adjunctive therapy in the treatment of partial seizures in children aged 4–16 years. Unwanted provocation of status epilepticus (SE) was previously reported in two mentally retarded patients from Finland. We report three patients with static encephalopathy, mental retardation, and OXC-induced SE. The purpose of this analysis is to identify the patients at risk for development of this serious complication while being treated with OXC. Methods: Retrospective chart review was performed on patients seen at Children’s Hospital of Pittsburgh with childhood epilepsy treated with
Epilepsia, Vol. 44, Suppl. 9, 2003
OXC. Data regarding demographics, diagnosis, treatment duration, medication doses, and EEG findings were collected and analyzed. Results: Twenty children with partial seizures with or without cognitive problems were treated with OXC. Four experienced worsening of seizures when OXC was added (four male; mean age, 11.25 years; range, 9–13; average age at seizure onset, 3 years; on average, three other anticonvulsants used). Two children had Lennox–Gastaut syndrome (one microcephaly, one Dandy Walker syndrome), two mild learning problem. Average OXC dose was 18.5 mg/kg/day; one patient was rechallenged twice. Three patients developed convulsive SE; one had significant increase in seizures after an average 0.75 per month. Previous SE was documented in one. Discontinuation of OXC and treatment with lorazepam (Ativan) and phenytoin resulted in clinical and electrographic resolution of SE. Conclusions: Children with intellectual disability may be at the greater risk of developing SE while treated with OXC. Perhaps OXC exacerbated primary generalized activity in those patients. More data are needed to evaluate this hypothesis. Of an additional interest: in three patients, prior treatment with carbamazepine did not result in this complication. 1.167 THE RISK FACTORS OF STATUS EPILEPTICUS IN CHINESE CHILDREN Yi Wang, Daokai Sun, Jin Zhang, Yiyun Shi, and Hui Xu (Neurology Department, Children’s Hospital of Fudan University, Shanghai, China) Rationale: Status epilepticus (SE) is one of the most important neurological emergencies. Between 15% to 27% of patients with epilepsy will experience at least one episode of SE. The goal of this study is to identify the risk factors associated with SE in the population of children with epilepsy in China. Methods: We designed a case–control study for patients with SE attending neurology clinic of Children’s Hospital of Fudan University in Shanghai from January 2000 to December 2002. SE was defined as a clinical seizure lasting ≥30 min, or repeated seizures without recovery. Patients who had at least one SE were enrolled in the study. Control subjects were selected, matched for age and sex, from a group of non-SE patients with epilepsy in the same hospital and during the same period. Information on etiology, family history, febrile convulsion history, age at onset, seizure types, EEG findings, computed tomography/magnetic resonance imaging (CT/MRI) data, intellectual disability, initiation of treatment, anticonvulsants (AEDs) administered, number of episodes of SE were ascertained from personal interview and review of the medical records and were analyzed with conditional logistic regression analysis. Results: Seventy-seven patients with SE and 154 control patients with epilepsy but non-SE were enrolled in this study. The mean age was 5.1 ± 3.8 years old. Twenty-four variables were analyzed. On univariate analysis, a significant positive association was found with certain etiology, seizure as SE before, retractable seizure, complex partial or partial secondary to the generalized seizure, abnormal neurologic image, abnormal background or focal spike in EEG, and poor compliance with AEDs. On multiple factor analysis, symptomatic epilepsy (OR, 5.62; 95%CI, 3.68–8.14); seizure as SE before (OR, 4.50; 95%CI, 3.46– 5.26); complex partial seizure (OR, 2.96; 95%CI, 2.13–4.20); partial secondary to the generalized seizure (OR, 2.40; 95%CI, 2.16–2.89); abnormal EEG background (OR, 2.06; 95%CI, 1.08–4.10); poor compliance with AEDs (OR, 1.72; 95%CI, 1.05–2.33) were found to be risk factors for SE. Conclusions: Our study suggests that these six factors increase a subject’s risk of developing SE as compared with age- and sexmatched controls. (Supported by Children’s Hospital of Fudan University)
Genetic Models and Human Genetics
1.168 MICROARRAY-BASED EXPRESSION PROFILING OF GENES REGULATED BY EARLY-LIFE SEIZURES
AES PROCEEDINGS 1 Sookyong Koh, 2 Robert C. Elliott, 1 Hyokwon Chung, and 3 Keri E. Pearson (1 Pediatrics, Children’s Memorial Hospital/Northwestern University, Chicago, IL; 2 Neurology, Beth Israel Deaconess Medical Cen-
ter, Boston, MA; and 3 Neurogenomics Program, The Translational Genomics Research Institute, Phoenix, AZ) Rationale: Early childhood convulsions have been correlated with memory impairment, and hippocampal neuron loss in patients, with intractable temporal lobe epilepsy. Animals subjected to kainate (KA) seizures at postnatal day (P)15 sustain more extensive hippocampal neuronal injury after second KA in adulthood. To study the molecular basis of this priming effect of early-life seizures, we probed >8,000 genes by using high-density oligonucleotide arrays and identified gene-expression patterns in the hippocampus at various times after KA seizures at P15. Methods: At 1, 24, and 240 h after KA (3 mg/kg, i.p.) or PBS injection at P15, animals were anesthetized with isoflurane, decapitated, and the hippocampi were dissected. Total RNA was isolated from each animal individually, and equal amounts of RNA from four animals were pooled for each Affymetrix Genechip. Three independent hybridizations were performed per condition per time point (total of 18 U34A profiles). Microarray Suite was used for the probe-level and GeneSpring, for genelevel data analysis. KA samples were normalized against controls, and genes with expression levels close to the background were removed. We selected genes showing at least twofold changes, and used Welch’s approximate t test, one-way ANOVA, and K-means clustering. Results: Among 40 known genes differentially regulated at 1 h, 32 were upregulated in the rats treated with KA. More than 80% were IEGs/transcription factors including ania2,4,6a,7, egr1,2,4, nr4a1, 2,3,cfos, junB, homer1a, arc, narp, btg2, C/EBP, and CREM. Microtubuleassociated-protein2 and ubiquitin-conjugating enzyme were among eight downregulated at 1 h. All but eight genes returned to baseline by 24 h, and a new set of 152 genes (2/3 up, 1/3 down) were identified at 24 h. Noteworthy were induction of inflammation-related genes (19) such as TNFα, MHC classII proteins, interferon-induced protein, CD9, c1q, and cytokine A (Sigje). Downregulated were receptors (nine) and enzymes [19, including kinases (CaMKII, CaMKK, IP3K, PKC) and phosphatases]; a notable exception was glutathione peroxidase (up threefold). At 240 h, only 19 genes were differentially regulated. A K-means clustering (specifying nine groups) of the 129 transcripts significantly changed over time revealed the following prominent five patterns: (a) increase at 24 h; (b) increase at 1 h; decrease at 24 h; (c) increase at 24 h; decrease at 240 h; (d) decrease at 24 h; and (e) increase at 1 h; normalize by 24 h. Conclusions: Complex and dynamic patterns of hippocampal gene expression were induced by early-life seizures. Conspicuously absent were the genes involved in the cell-death pathway, consistent with absence of cell death after KA at P15. Instead, we found a change in expression of survival-promoting and inflammation-related genes. In stark contrast to adults, returning of the majority of genes in our study to control levels by 240 h suggests plastic, homeostatic response of immature brain to excessive, dysregulating stimulation. (Supported by K08NS02068 and the NINDS/NIMH Microarray Consortium.)
1.169 COMORBID BEHAVIOR OF RATS SELECTIVELY BRED TO BE SEIZURE PRONE VERSUS RESISTANT BASED ON AMYGDALA KINDLING: AN ANIMAL MODEL OF ADHD Dan C. McIntyre, Krista L. Gilby, and Hymie Anisman (Psychology, Neuroscience Institute, Carleton University, Ottawa, ON, Canada) Rationale: Based on amygdala-kindling rates, we selectively bred rats to be seizure-prone or seizure-resistant. The strain differences in excitability are comorbid with important behavior/personality attributes. One of these reflects learning/attention problems with heightened activity in Fast rats, behaviors reminescent of attention-deficit hyperactivity disorder (ADHD) in humans. The present experiments explore the learning impairments in Fast compared with Slow rats by measuring their controlling factors, distractability, and tendency for hyperfocusing. Methods: Naive male rats of the Fast and Slow strain were tested for escape speed and accuracy to a submerged platform in the Morris water maze (MWM) by using protocols that measured spatial versus cued learning. In two versions of the spatial procedure, reference memory
61
was measured in one, and working memory in the other, with or without pretraining procedures to familarize the rats with the task concepts. In other experiments, cued learning was assessed, where the location of the submerged platform was indicated by an overhanging cue. In one procedure, the predictive, overhanging cue was accompanied by a second, distracting cue that did not signal the platform. The distracting cue was presented either at the beginning or the end of the learning procedure. Last, activity and habituation were monitored in the open field over 2 days, in the presence or absence of a low dose of amphetamine (0.5 mg/kg), which was introduced only on day 1. Results: The Fast rats performed inferior to the Slow rats in acquisition of both the reference and working-memory versions of the spatial MWM task. The disability in Fast rats was partially ameloriated by pretraining procedures, familarizing them with the task concepts. The Fast rats were also impaired in acquisition of the cued task, particularly when a distracting cue was available early in learning procedure. By contrast, when the distracting cue was available at the end of learning, Fast rats were not distracted by it, and remained completely focused on it, even when it was no longer predictive (hyperfocusing). Activity measurements indicated higher scores in Fast compared with Slow rats. However, on day 2, Slow rats showed normal habitation, whereas Fast rats did not. In rats treated with amphetamine, the drug increased activity in Slow but not Fast rats, whereas on day 2, without drug, both groups registered normal habituation. Conclusions: The Fast rats showed spatial learning problems related to concept formation, which could be corrected with pretraining. In cued versions of the task, learning problems in Fast rats were accentuated by early distractability, but later cue fixation with no distractability. These results are analogous to those in humans with ADHD, who show distraction during initial learning, but hyperfocus in familar circumstances. Epilepy is ≤20 times more prevalent in those with ADHD than in others. Thus seizure-prone Fast rats provide an interesting new animal model of ADHD. (Supported by CIHR grants to D.C.M. and H.A.)
1.170 INCREASED GABA(A) RECEPTOR α4 PROMOTER ACTIVITY AFTER PILOCARPINE-INDUCED STATUS-EPILEPTICUS IN ADULT RATS 1 Yogendra Sinh H. Raol, 1 Margaret A. Maronski, 1 John H. Wolfe, 2 Shelly J. Russek, and 3 Amy R. Brooks-Kayal (1 Neurology, Children’s Hospital of Philadelphia, Philadelphia, PA; 2 Pharmacology, Boston University School of Medicine, Boston, MA; and 3 Neurology and Pediatrics, Children’s Hospital of Philadelphia, University of Pennsylvania, Philadelphia, PA) Rationale: The adenoassociated virus (AAV) vector has been preferentially used in the central nervous system because of its predominantly neuronal transduction, higher stability, and lack of any toxicity. Earlier studies have shown increased expression of α4 subunit and decreased α1 subunit in dentate granule cells after pilocarpine-induced status epilepticus (SE) in adult rats. In the present study we explore the possibility of using AAV 2 vector as a genetic tool to examine the regulation of γ -aminobutyric acid (GABA)A receptor subunits in dentate gyrus after SE in adult rats. Methods: Adult male Sprague–Dawley rats were injected with 2 µl of AAV 2 virus (5.2 × 1012 genomic particles/ml) containing an α4 promoter driving the expression of eYFP reporter gene into upper and lower blades of dentate gyrus. Two weeks after viral injections, rats were injected with pilocarpine to induce SE. The control rats were injected with 1/10 of the convulsive dose. One and two weeks after SE, rats were perfused with 4% paraformaldehyde. Nonradioactive in situ hybridization was used to detect the expression of eYFP reporter mRNA. Results: An increase in α4 promoter driving reporter gene expression was found both 1 and 2 weeks after pilocarpine-induced SE compared with control animals. Conclusions: We show increased expression of eYFP reporter gene driven by α4 promoter after SE, suggesting that previously reported increases in α4 subunit after SE are occurring at a transcriptional level. This system may be useful for modifying gene expression in dentate granule cells after SE. (Supported by NIH NS42363-01 to A.B.K. and S.J.R.)
Epilepsia, Vol. 44, Suppl. 9, 2003
62
AES PROCEEDINGS
1.171 EPIGENETIC CONTROL OF GENE EXPRESSION IN EPILEPSY Avtar S. Roopra, Timothy Daley, Romena Qazi, Thomas Sutula, and Barry Scheonike (Neurology, University of Wisconsin, Madison, WI) Rationale: Neural circuits undergo a diverse variety of long-term structural and functional alterations in response to pathological insults and repeated seizures. Insights into how injury and seizures induce alterations in neural circuits are of interest for understanding epileptogenesis and the consequences of seizures, and may provide opportunities for therapeutic intervention in these processes. Long-term alterations in neural circuits induced during epileptogenesis and seizures include neurogenesis, neuronal death, gliosis, axon sprouting, and modifications of membrane receptors and signaling pathways influencing cellular functions. These long-term alterations require long-term changes in gene-expression patterns. Despite intensive research into seizureinduced immediate-early gene (IEG) expression, very little is known about how changes in IEG expression initiate and regulate cascades of gene-expression changes underlying long-term structural and functional reorganization of epileptic neural circuits. It is now apparent that in addition to the direct effects of IEGs on promoters, gene expression is powerfully controlled by modifications of chromatin structure through posttranslational acetylation and methylation of histones. Whereas these mechanisms for regulation of gene expression are increasingly appreciated as basic aspects of molecular and cellular biology, the potential role of histone modification in seizure-induced gene expression has not been addressed Methods: Immunoprecipitations were performed on extracts from cell lines, control and kainate-treated adult rat brain extracts. Chromatin precipitations were performed on the same protein samples. Results: We have demonstrated that a seizure-induced transcription factor, neuron restrictive silencing factor (NRSF), interacts with a chromatin-modifying methylase enzyme. This interaction potentially establishes long-term, stable alterations in gene-expression patterns that could regulate long-term gene expression underlying neural circuit alterations induced by seizures through the epigenetic mechanism of methylation of chromatin. We provide evidence for a direct interaction between the IEG NRSF and histone methylases. We also show that neuronal genes (e.g., BDNF, the sodium channel NaV1.2 and muscarinic receptor type 4) are associated with methylated histones in vivo and discuss stable alterations that occur in the methylation status of chromatin after seizure. Conclusions: We propose that transient increases in NRSF levels after seizure activate permanent posttranslational modifications of chromatin through histone methylation to influence subsequent seizure-induced gene-expression patterns and contribute to important functional consequences of repeated seizures through the epigenetic mechanism of chromatin modification (Supported by Department of Neurology, University of Wisconsin-Madison, Madison, WI.)
1.172 CHANGES IN HIPPOCAMPAL GENE EXPRESSION CHARACTERIZED BY DNA MICROARRAY ANALYSIS ONE YEAR AFTER THE INDUCTION OF EPILEPSY IN THE PILOCARPINE MODEL 1 Mohiuddin Taher, 2 Sajida Hassan, 2 Lisa Wallace, 1 Robert Blair, 1,2 Michael Miles, and 1,2,3 Robert DeLorenzo (1 Neurology, 2 Pharmacology and Toxicology, and 3 Molecular Biophysics and Biochemistry, Virginia Commonwealth University, Richmond, VA) Rationale: Acquired epilepsy is a condition associated with the development of persistant seizures in previously normal brain tissue, epileptogenesis. Long-term plasticity changes have been observed in association with epileptogenesis, and it has been suggested that long-term gene changes play a role in the development of the epileptic condition (Neuroscientist 1999;5:886). This study evaluates whether long-term changes in gene expression occurred for up to a year after the induction of epilepsy in the pilocarpine model using the powerful technique of DNA microarray analysis to characterize the expression of large numbers of genes. Methods: Epilepsy was induced in rats using the pilocarpine model (Brain Res 2001;903:1). Animals were monitored to confirm the induc-
Epilepsia, Vol. 44, Suppl. 9, 2003
tion of the epileptic condition. Animals with epilepsy for >1 year and the appropriate controls were killed, and hippocampal tissue was rapidly isolated and processed for RNA extraction and characterization (J Neurosci 2003;23:2218). Labeled cRNA samples were analyzed on Affymetrix Rat Genome array Set A, which represents nearly 7,000 named genes and ∼1,000 EST sequences. Sample fragmentation, hybridization, and array scanning were performed by using standard conditions. The microarray data was analyzed by using MAS 5 (Affymetrix) and the S-score algorithm developed in our laboratory (J Mol Biol 2002;317:225). Results: Microarray studies were performed on at least four individual animals from control (C), pilocarpine treated (P) without ongoing epilepsy, or epileptic (E) animal groups. S-score analysis followed by statistical filtering using a permutation method, and k-means clustering identified three groups of genes: altered uniquely in E animals, altered in common between E and P groups, and only altered in P animals. Strikingly, groups of genes recently identified by our array studies at 2 weeks after pilocarpine (J Neurosci 2003;23:2218) were also seen in these animals at 1 year after pilocarpine. A substantial group of calcium signaling genes were uniquely downregulated in the E animals, perhaps reflecting compensation to long-term alterations in calcium levels with epilepsy. Several genes upregulated in the P group may play a role in preventing development of epilepsy in those animals. Conclusions: The results of this study demonstrate that changes in gene expression were observed for as long as 1 year after the induction of epileptogenesis. Using the advancement of DNA microarray technologies, it was possible to evaluate the hypothesis that long-term gene expression contributes to the plasticity changes associated with epilepsy. The results demonstrate expression patterns associated with epilepsy, but also patterns that may protect against epileptogenesis. (Supported by grant R01AA13678 to M.F.M. and grants R01NS23350 and P50NS25630 to R.J.D.)
1.173 ABNORMAL GIANT CELLS AND TUMOR FORMATION IN A RODENT MODEL OF TUBEROUS SCLEROSIS D. Koji Takahashi and Scott C. Baraban (Neurological Surgery, UCSF– University of California, San Francisco, San Francisco, CA) Rationale: Tuberous sclerosis complex (TSC) is an autosomal dominant human genetic disease caused by mutations of either TSC2 (tuberin) or TSC1 (hamartin) genes. In the CNS, TSC is characterized by cortical tubers, subependymal nodules, and abnormal giant cells. Clinical manifestations invariably include seizures and mental retardation. Eker rats carry a spontaneous germline mutation of TSC2 (TSC2+/− ). Although renal cell carcinomas, subependymal and subcortical hamartomas have been found in these animals, little evidence for cortical tubers or abnormal giant cells exists. In the present study, we attempted to induce cortical tubers or giant cells using a prenatal “second hit” approach. Hydroquinone, a nongenotoxic carcinogen, was chosen for these studies because of its ability to increase renal cell carcinomas in the Eker rat (Yoon et al., 2001). Detailed histologic and immunohistochemical studies were performed on brain sections from “second hit” litters as well as aged Eker rats. Methods: Pregnant TSC2+/− females were injected once a day with hydroquinone (HQ) according to the following protocols: (a) intraperitoneal (i.p.) injections of 50, 100, 150 mg/kg of HQ between embryonic days 15 and 20 (n = 5); or (b) i.p. injections of 50 or 100 mg/kg HQ between E8 and E19 (n = 9). Offspring were killed at p14 or p28, and tissue sections throughout the CNS were prepared and stained for cresyl violet and tuberin. In separate studies, brains of older Eker rats (∼18 months old) that died of natural causes were sectioned for anatomic analysis. Sagittal and coronal sections every 40 µm were stained for NeuN (neuron-specific antibody), GFAP (glial-specific antibody), vimentin, nestin, tuberin, hematoxylin and eosin (H&E), and cresyl violet. Results: Tissue sections stained with cresyl violet did not reveal any difference between HQ-treated Eker (TSC2+/− ) rats and siblings (TSC2+/+ ) at p14 or p28. None of the hallmark traits of human TSC including cortical tubers, hamartomas, or giant cell astrogliomas were found. However, abnormal giant cells in deep cortical layers were observed in brain sections from untreated aged Eker rats (∼18 months). Abnormal giant cells observed with cresyl violet or H&E, were GFAP
AES PROCEEDINGS and tuberin positive, but did not express NeuN. In addition, large tumors connected to the brainstem and cerebellum were observed in three of four aged Eker rats. Within tumors, large ectopic GFAP-positive cells with abnormal morphology could also be found, along with increased angiogenesis. Conclusions: Our data suggest that cortical tuber formation in Eker rats is a rare event (Mizuguchi et al., 2000) and that prenatal exposure to a nongenotoxic carcinogen such as hydroquinone is not sufficient to induce tuber formation. However, with advanced age, there may be an increased likelihood of tumor formation, and the emergence of giant cells in the Eker rat brain; each of these abnormalities could contribute to neurologic problems associated with TSC. As such, further analysis of this rodent model may be warranted. (Supported by Tuberous Sclerosis Alliance.)
1.174 INCREASED SEIZURE SUSCEPTIBILITY IN NR1 KNOCKOUT MICE OF THE NMDA RECEPTOR Victor E. Campos, Nicole R. Collins, and Yuqing Li (Molecular and Integrative Physiology, University of Illinois at Urbana Champaign, Urbana, IL) Rationale: Epilepsy is caused by excessive excitation or insufficient inhibition. We have generated a cortex-specific NR1 knockout of the excitatory N-methyl-D-aspartate (NMDA) receptor. Despite a decrease in neuronal excitability, these mice undergo spontaneous seizures. This unexpected phenotype indicates an insufficient understanding concerning the role of the NMDA receptor in seizures. Methods: We have characterized temporal seizure susceptibility by the administration of the chemoconvulsant pentylenetetrazol (PTZ). We intraperitoneally injected knockout and control animals and recorded seizure behavior for 1 h. Results: We found that NMDA-receptor knockout mice have significantly greater seizure susceptibility. We also found that this seizure susceptibility is age dependent. Conclusions: NMDA-receptor knockout mice have increased seizure susceptibility. Furthermore, this seizure susceptibility is differentially expressed during development (Fig. 1). (Supported by National Institute of Neuronal Diseases and Stroke.)
63
an outbred, transgenic mouse strain, originally designed to cause a tissuespecific increase in polyamine levels, was observed having occasional spontaneous seizures. Tests were conducted to determine if the observed seizures correlated with the insertion of the transgene. Methods: Male and female mice, age 10 weeks, were used. The mice were generated from a colony that had been started with the F1 generation of a C57B1/CBA crossing. Of the 30 mice tested in this pilot study, 17 carried a transgene that linked the β-4 galactosyltransferase-I promoter with the antizyme inhibitor gene. The other 13 mice evaluated were littermates that tested negative for the transgene insertion. To determine if a difference in seizure thresholds existed between the two groups, a heparinized metrazol solution, (0.5%) was infused at a constant rate of 0.34 ml/min into a lateral tail vein of an unrestrained mouse. Infusion was by means of a Sage syringe pump (model 341A) and a 10-ml B-D plastic syringe connected to a length of no. 20 PE tubing. A 27-gauge stainless steel needle with the hub removed was connected to the tubing and inserted into a vein and secured to the tail by a piece of adhesive tape. At the start of the infusion, a hemostat clamped to the tubing to prevent backflow was removed, the infusion started, and two stopwatches started. The time in seconds from the start of the infusion to the appearance of the “first twitch” and the onset of sustained clonus were recorded. The times to each end point were converted to mg/kg of metrazol for each mouse. A two-sample test of means, using Smith’s Statistical Package for the Macintosh, was used to determine statistical significance between the two groups. Results: Table 1 demonstrates the calculated differences between the control group of mice, and those carrying the transgene.
TABLE 1. mg/kg Metrazol First twitch Clonus
Control group (n = 13)
Test group (n = 17)
p Values
26.8 ± 1.7 35.6 ± 2.3
19.6 ± 1.38 25.4 ± 1.6
0.002 0.001
Values represented as mean ± standard error.
The dose of metrazol necessary to induce both the first twitch and clonus was reduced by nearly 30% in the carrier group. Conclusions: These data show a significant decrease in the seizure threshold of the transgene carrier mice in our colony. Further investigations are currently under way to determine if this difference in response is a result of an increased level of polyamines in sensitive regions of the brain, or the inadvertent disruption of another gene. Regardless of the cause, the mice show potential as a new model for the study of epilepsy. (Supported by internal sources.)
1.176 LONG-TERM EFFECTS OF CHRONIC–INTERMITTENT DEEP BRAIN STIMULATION OF THE NCL. SUBTHALAMICUS (STN) ON SEIZURE ACTIVITY IN A RAT MODEL OF ABSENCE EPILEPSY 1 Brita Fritsch, 1 Hristos Karakizlis, 1 Hajo M. Hamer, 2 Astrid Dempfle, 1 Wolfgang H. Oertel, and 1 Felix Rosenow ( 1 Epilepsy Center, Department of Neurology, and 2 Institute of Biostatistics, Philipps-University Marburg, Marburg, Hessen, Germany) 1.175 DEVELOPMENT OF A NEW TRANSGENIC SEIZURE MODEL IN MICE 1 Greg L. Christensen, 1 Douglas T. Carrell, 2 Ivaylo P. Ivanov, 2 John F. Atkins, 3 Jose H. Woodhead, and 3 Steve White (1 Andrology, 2 Human Genetics, and 3 ADD Program, Pharmacology and Toxicology, University of Utah, Salt Lake City, UT) Rationale: Many new genetic models of epilepsy have been developed in mice over the past several years. In many cases, targeted disruption, modification, or overexpression of a gene has resulted in an epileptic phenotype that was not anticipated. The current study was initiated when
Rationale: An anticonvulsive effect by inhibiting the substantia nigra pars reticulata (SNpr), the major output nucleus of the basal ganglia, via consecutive disinhibition of the collicullus superior, has been reported. Experimental data and case reports of patients with epilepsy treated with STN stimulation (HFS) suggest an anticonvulsive effect, probably mediated by an inhibition of the STN, which interferes with the excitatory subthalamiconigral pathway. Methods: In the present study the influence of a bilateral, chronic– intermittent HFS of the STN on seizure activity was investigated in rats of the WAG/Rij strain (n = 12), with well-defined absence seizures. Bipolar, concentric deep brain electrodes and screw-scull electrodes were implanted by stereotaxy. A stimulation frequency of 130 Hz with an
Epilepsia, Vol. 44, Suppl. 9, 2003
64
AES PROCEEDINGS
single-impulse duration of 60 µs, applied as a train of 30 s every 5 min for 1 h, was used. The procedure was repeated after 1 week. Effects were evaluated for both experimental days by cumulated seizure duration (CSD) and seizure frequency (SF) in EEG during a baseline period (B1 and B2 for 1 h), stimulation (S1 and S2), and directly after stimulation (P1 and P2 for 1 h). For statistical analysis, the non parametric Friedman test followed by Wilcoxon test was performed. Results: The chronic–intermittent bilateral HFS of the STN resulted after the first stimulation in a decrease of cumulated seizure duration (B1, 119 s; S1, 61 s; P1, 38 s) and in further reduction even during baseline measurement at the second experimental day 1 week later (B2, 20 s; S2, 13 s; P2, 3 s). This reduction in CSD and SF was statistically significant (CSD, B1 vs. B2, p = 0.006; B1 vs. P1, p = 0.008; SF, B1 vs. B2, p = 0.006; B1 vs. P1, p = 0.013). Conclusions: This study provides evidence for anticonvulsive properties of chronic–intermittent, bilateral HFS of the STN in absence epilepsy. Ineffectiveness of chronic–tonic stimulation seen in previous studies suggests dependency of anticonvulsive effects on stimulation paradigms. In addition, a long-term anticonvulsive effect lasting for ≥1 week was assessed and was boosted by the second stimulation. (Supported by the Urlan-Professorship for Neurology/Epileptology, F.R.)
1.177 DIVERGENT EXPRESSION PATTERNS OF GENES INVOLVED IN LIPID METABOLISM DURING EARLY DEVELOPMENT MAY UNDERLIE DIFFERENCES IN ADULT BRAIN MORPHOLOGY IN SEIZURE-PRONE AND SEIZURE-RESISTANT RATS 1 Krista L. Gilby, 2 Peter Crino, and 1 Dan C. McIntyre ( 1 Psychology, Neuroscience Institute, Carleton University, Ottawa, ON, Canada; 2 Neurology, University of Pennsylvania Medical Center, Philadelphia, PA; and 3 Psychology, Neuroscience Institute, Carleton University, Ottawa, ON, Canada) Rationale: Rat strains that were selectively bred to be seizure prone or seizure resistant have been shown to exhibit differences in adult brain morphology. In seizure-prone animals, the ventricles and piriform cortex are significantly larger than in seizure-resistant animals, whereas the corpus callosum and dorsal hippocampus are significantly smaller. The mechanism through which these differences in brain morphology arise remains unknown. Methods: We used a number of gene-screening technologies to investigate putative differences in gene expression that arise during ontogenesis of the strains, which may, in turn, implicate underlying genetic control mechanisms that operate differently during critical periods of development. These techniques included superarrays and differential display. Results: Of nearly 2,500 genes sampled, the results of this study revealed only three differences in gene expression between the strains during embryogenesis. When cDNA from the head of embryonic day 21 (E21) rats from each strain was probed by using superarray technology, mRNA levels for apolipoprotein E (ApoE) and the β2 subunit of the sodium channel were shown to be significantly reduced in seizure-prone rats. In addition, when differential display was used to examine differences in gene expression between the strains at varying embryonic time points (E11, E13, E15, E19, and E21), α 2 -macroglobulin was shown to be expressed at E11 in seizure-prone rats but was not expressed until E13 in seizure-resistant rats. Conclusions: Clearly, divergent expression patterns occur for a very limited number of developmentally important genes during early embryogenesis of the strains. Interestingly, both ApoE and α 2 macroglobulin are intricately involved in lipid metabolism and reelinbased control of brain formation. These findings suggest that alterations in the timing and/or degree of expression for ApoE or α 2 -macroglobulin, such as those shown here in the two strains, may be critical to initiate developmental cascades that give rise to the different brain morphologies, and possibly the differing seizure vulnerabilities characterized by the seizure-prone and the seizure-resistant phenotypes. (Supported by CIHR.)
Epilepsia, Vol. 44, Suppl. 9, 2003
1.178 EARLY INTERVENTION STUDY USING GABA-A RECEPTOR β3 SUBUNIT KNOCKOUT MOUSE MODEL FOR ANGELMAN SYNDROME 1 Krista L. Gilby, 2 Peter Crino, and 1 Dan C. McIntyre ( 1 Brain Research Institute, UCLA School of Medicine, Los Angeles, CA; and 2 Department of Anesthesiology, University of Pittsburgh, Pittsburgh, PA) Rationale: Some of the more catastrophic epilepsies are associated with childhood and arise through abnormal development. One example is the genetically based disorder Angelman syndrome, which is associated with abnormal craniofacial development, severe mental retardation, atypical absence and myoclonic epilepsy, sleep deficits, and hyperactivity. One mouse model for Angelman syndrome is the knockout (KO) mouse for the γ -aminobutyric acid (GABA)-A receptor β3 subunit, based on close phenotypic similarities of high neonatal mortality due primarily to cleft palate, deficiencies in learning and memory, hyperactivity and circling, and myoclonic and aptypical absence epilepsy. Diazepam (DZP; Valium) is a benzodiazepine with anxiolytic and sedative properties that acts on the GABA-A receptor to enhance channel activity, and is used to treat seizure disorders. We are using the β3 KO mouse to study the effects of early intervention with DZP. Methods: β3(−/−) mice are recognizeable at birth by their pink eyes, although only 10% survive the first 24 h. β3(−/−) and littermates (C) were injected s.c. with either physiologic saline (SAL) or DZP, 1.5 mg/kg, on a daily basis for either the first (PN WK1) or second (PN WK2) postnatal week. At 11 weeks, injected mice were implanted with two bipolar cortical electrodes, and EEGs recorded periodically. Behavior was examined by using open-field, rotarod, and elevated plus maze assays. Results: β3(−/−) DZP WK1: overall EEG baseline calmer and less disorganized than β3(−/−) SAL but exhibit severe ongoing myoclonic twitches, rare ictal spikes and electrographic seizures, and convulsions with facial distortions, protruding tongue, and forelimb clonus; show no tendency to circling behavior and are much calmer than β3(−/−) SAL, resembling C mice in the open-field test and rotarod; show slightly decreased anxiety on elevated plus maze. β3(−/−) DZP WK2: overall calming effect on EEG, with no seizures and much suppression of baseline irregularities; only moderate tendency toward the frantic circular running of β3(−/−) SAL, and locomotor activity intermediate between β3(−/−) SAL and C; on the elevated plus maze, coordination and anxiety comparable to C but increased novelty-seeking (jumping off the platform). Both DZP WK1 and DZP WK2 have irregular edges of the iris, indicating abnormal eye development. Conclusions: Early intervention with DZP in this epileptic mouse model has some therapeutic advantages, but may produce severe side effects in PN WK1. This is likely due to the developmental change of GABA from excitatory to inhibitory at the end of PN WK1; enhanced excitatory action of GABA induced by DZP in PN WK1 would produce an increase in intracellular calcium through activation of the NMDA receptor, resulting in downstream modifications. These results indicate narrow stage-specific effects of DZP on fetal brain and eye development and suggest increased caution on the part of expectant mothers taking DZP. (Supported by NIH grant NS35985.)
1.179 METABOLIC CONTROL OF EPILEPSY IN ADULT EL MICE WITH THE KETOGENIC DIET AND CALORIC RESTRICTION John G. Mantis, Nicole Centeno, Mariana T. Torodova, Richard McGowan, and Thomas N. Seyfried (Biology Department, Boston College, Chestnut Hill, MA) Rationale: Glucose uptake into the brain is greater during epileptic seizures than during most other brain activities, suggesting a key role for glucose in the initiation and spread of seizure activity. Under conditions of fasting or caloric restriction (CR), however, brain cells can also derive energy from ketone bodies (acetoacetate and β-hydroxybutyrate). Diet therapies for epilepsy are as old as the disease itself, and the high-fat, low-carbohydrate ketogenic diet (KD) was developed as an alternative to fasting for seizure management. Although the mechanisms by which KD inhibits seizures remain speculative, ketone bodies and alterations
AES PROCEEDINGS in brain energy metabolism are likely involved. We previously showed that CR inhibits seizure susceptibility by reducing blood glucose in the epileptic EL mouse, a model of multifactorial idiopathic epilepsy. CR is a natural dietary therapy that involves the reduction of total dietary energy intake while maintaining adequate levels of essential vitamins and minerals. In this study, we compared the antiepileptic efficacy of the KD with that of CR in adult EL mice. Methods: Adult EL mice (∼12 months old) that experienced ≥15 recurrent complex partial seizures were fed either a standard (chow) diet unrestricted (SD-UR) or restricted (SD-R), and either a KD unrestricted (KD-UR) or restricted (KD-R). Seizure susceptibility, body weights, and the levels of plasma glucose levels were measured once a week over an 8-week treatment period in each diet group. Results: Body weights and blood glucose levels remained stable over the 8-week testing period in the SD-UR and the KD-UR groups, but were significantly (p < 0.001) reduced by ∼25% in the SD-R and KD-R groups. Results from ketone body measurements will also be presented. Seizure susceptibility remained high in both UR diet groups, but decreased significantly (p < 0.001) after 2 weeks in both R diet groups. Seizures were completely managed in both R groups after 8 weeks of diet therapy. Conclusions: The results indicate that seizure susceptibility is dependent on plasma glucose levels in EL mice fed either an SD or a KD diet and that seizure management depends more on the amount than on the origin of dietary calories. A reduction in plasma glucose levels coupled with an increase in plasma ketone levels is predicted to control the epileptic seizures in EL mice. (Supported by NIH grant HD39722, and the Boston College Research Expense Fund.)
1.180 MICE WITH A KCNQ2 C-TERMINAL DELETION EXHIBIT INCREASED SEIZURE SUSCEPTIBILITY AND ALTERED SENSITIVITY TO COMPOUNDS ACTING AT THE M-CHANNEL 1 James F. Otto, 2 Yan Yang, 2 Wayne N. Frankel, 1 Karen S. Wilcox, and 1 H. Steve White ( 1 Pharmacology and Toxicology, University of Utah, Salt Lake City, UT; and 2 Jackson Labs, Bar Harbor, ME) Rationale: Benign familial neonatal convulsions (BFNCs) are caused by reduction-of-function mutations in the genes that encode the KCNQ2 or KCNQ3 subunits of the M-type K+ channel, which underlies the M current. We used the B6-Szt1/+ (Szt1) mouse, which contains a spontaneous C-terminal deletion in the KCNQ2 subunit (Yang et al. Hum Mol Genet 2003;12:975–84), as a model of genetically altered M-current function. Using Szt1 and wild-type B6 mice, we (a) examined differences in seizure threshold, and (b) evaluated the ability of drugs that act at the M channel to modify seizure susceptibility. Methods: Transcorneal stimulation was used to evoke three types of seizures: partial psychomotor, minimal clonic, and minimal tonic hindlimb extension (THE). Convulsive current (CC) curves for each seizure type were constructed using B6 and Szt1 mice. At the calculated CC10 , CC90 , and CC50 values for each strain, mice were injected with either linopirdine (LPD, an M-channel blocker), retigabine (RGB, an enhancer), or LPD + RGB, respectively, and tested for seizure incidence. All drug injections were i.p., 10 mg/kg, and methyl cellulose (MeCell) was used as a vehicle control. RGB dose–response curves were also constructed in the partial psychomotor model. Results: In all seizure types tested, we observed significant decreases in the thresholds of Szt1 mice relative to B6 mice. In both strains, we also observed significant gender-dependent differences in seizure threshold. The RGB dose–response data show that in the psychomotor model, RGB is less potent in Szt1 mice than in B6 mice. Moreover, several straindependent differences in seizure incidence arose in response to treatment with LPD and RGB (see Table 1). Conclusions: Our data suggest that mice with presumably compromised M-current function have reduced seizure thresholds, and that this increase in net neuroexcitability can also confer increased sensitivity to M-channel–modulating drugs. Our results also indicate that these altered drug senstivities are dependent on the seizure type elicited. Finally, these data shed light on the implications of genetically altered M channels and their role in altered pharmacoresponsiveness and seizure susceptibility. [Supported by NIH 5-RO1-NS-40246 (W.N.F., H.S.W.), Primary Children’s Medical Center Foundation 51001142 (K.S.W.).]
65 TABLE 1. Strain-dependent differences in pharmacoresponsiveness Minimal clonic seizure
Stimulation/treatment CC90 MeCell RGB CC10 MeCell LPD CC50 MeCell RGB+LPD
Partial psychomotor seizure
B6
Szt1
B6
Szt1
9/10 4/10a
8/10 3/10a
9/11 2/11a
7/9 7/9
2/15 6/10a
1/10 10/10a,b
1/10 6/10a
1/10 7/10a
7/15 13/15a,b
6/11 2/11
6/9 7/9b
12/22 12/25
CC, convulsive current; RGB, retigabine; LPD, linopirdine. Data represented as number seizing/total tested. Table data from female mice only. a p < 0.05, Fisher Exact for within-strain drug effect. b p < 0.05, Mantel–Haenszel χ 2 for between-strains drug effect.
1.181 ALTERED INHIBITION AND INTERNEURON DISTRIBUTION IN THE DENTATE OF P35 KNOCKOUT MICE 1 Leena S. Patel, 2 H. Jurgen Wenzel, and 2 Philip A. Schwartzkroin (1 Physiology & Biophysics, University of Washington, Seattle, WA; and 2 Neurological Surgery, University of California at Davis, Davis, CA) Rationale: Previous work has identified the p35 (neuronal specific activator of cyclin-dependent kinase 5) knockout mouse as a useful model in which to study the relation between dysplasia and spontaneous behavioral and electrographic seizures. One feature identified in the knockout animal that may contribute to seizure activity is recurrent excitation of the dentate granule cells via aberrant mossy fiber collaterals and basal dendrites. However, little is known about the local inhibitory circuit and its contribution to the suppression (or facilitation) of granule cell excitability. We hypothesize that p35-dependent migrational abnormalities affect dentate interneuron (IN) connectivity and function, and that these abnormal circuitry features contribute to seizure susceptibility and influence the net output of granule cells. Methods: Acute hippocampal slice preparations from both wild-type and knockout animals were used for intracellular recordings from dentate granule neurons. Sharp electrodes were used to measure the electrophysiologic responses of granule cells to orthodromic (perforant path) stimulation. Electrodes backfilled with 2% biocytin were also used to label intracellularly electrophysiologically identified interneurons. After filling, slices were fixed and processed for light and ultrastructural analyses. Immunocytochemistry (ICC) of γ -aminobutyric acid (GABA) and parvalbumin (Parv) was carried out in parallel studies. Results: Of 24 electrophysiologically identified granules cells in p35 knockout animals, 13 showed a reduction in the early GABAA -mediated inhibitory postsynaptic potential (IPSP), and 21 demonstrated a reduction in the late GABAB -mediated IPSP after stimulation of the perforant path (compared with wild-type mice; n = 7). ICC revealed a small reduction in Parv-positive interneurons in −/− mice, with a significant decrease in immunoreactivity in the axonal plexus around the GCs. GABA and Parv ICC from p35 −/− tissue also revealed displaced (heterotopic) GABA/Parv-positive IN distribution across the granule cell and molecular layers (GCL and ML). Biocytin-labeled GABAergic basket cells were displaced within the GCL/MLs. These cells exhibited a normalappearing axonal arbor within the GCL; however, extensive axonal arborization was also observed in regions where GCs were dispersed (i.e., hilus and ML). Whereas interneurons in wild-type mice show a characteristic laminar distribution of their axonal arbors, in p35 knockout mice, these INs expand their axonal territories across the dentate layers into the hilus and the IML, to “follow” the dispersed GCs. Conclusions: These studies identify a feature of the p35 neuronal migration disorder that may contribute to seizure activity: an abnormal inhibitory circuit. Electrophysiologic recordings from granule cells
Epilepsia, Vol. 44, Suppl. 9, 2003
66
AES PROCEEDINGS
demonstrate a reduction in inhibitory input. ICC and intracellular labeling of individual interneurons from p35 knockout mice confirm abnormalities in soma and axon distribution of inhibitory interneurons. (Supported by NIH grant NS18895.)
1.182 THE INFLUENCE OF STRAIN AND HOUSING CONDITION ON THE DEVELOPMENT AND EXPRESSION OF TWO TYPES OF SPIKE–WAVE DISCHARGES IN RATS Ulrich Schridde and Gilles van Luijtelaar (Department of Biological Psychology, NICI, University of Nijmegen, Nijmegen, Netherlands) Rationale: In the electroencephalogram (EEG) of WAG/Rij rats, a genetic model of absence epilepsy, two types of spike–wave discharges (SWDs) can be recorded: type 1 SWDs, bilateral generalized, and type 2 SWDs, more local. Earlier it was established that all adult WAG/Rij rats show type 1 SWDs and that the pattern of inheritance obeys mendelian rules. In the present experiment, we studied the influences of strain and housing condition on the development and expression of type 1 and type 2 SWDs. The strains were WAG/Rij and ACI rats; the latter strain has very few type 1 SWDs but a substantial number of type 2 SWDs. The environment was manipulated by housing the rats in an impoverished (IC) or enriched (EC) environment. Methods: Seventy-five WAG/Rij and 80 ACI male rats were housed in an IC or EC environment from weaning until age 6 months. At age 3 months, housing condition changed for half of the rats from IC to EC or vice versa, whereas for the other half, it stayed the same. EC housing consisted of a group of eight to 10 rats housed in a complex environment, and IC housing implied single housing in Makrolon type III cages. The EEG was recorded at age 3 and again at 6 months. We analyzed whether strain or housing condition before age 3 months or after age 3 months had an influence on the number and the mean duration of type 1 and type 2 SWDs. Results: We found that the number of type 1 SWDs was higher for WAG/Rij than for ACI rats. EC housing led to an increase in mean duration of type 1 SWDs in WAG/Rij rats only. Regarding type 2 SWDs, we found that rats that were recently EC housed had a higher number of discharges than IC-housed rats. Next, at age 6 months, WAG/Rij rats that were EC housed after age 3 months had the highest number of type 2 SWDs of all groups. Finally, ACI rats had longer type 2 SWDs than WAG/Rij rats. Conclusions: We demonstrated that the development and expression of type 1 and type 2 SWDs are differently influenced by strain and housing condition. The number of type 1 SWDs was influenced by strain, whereas the number of type 2 SWDs was influenced by housing condition. Whereas the mean duration of type 1 SWDs was influenced by housing condition, the mean duration of type 2 SWDs was strain dependent. Our data also revealed a genotype–environment interaction and a sensitive period for influences of housing condition on type 2 SWDs, inferred from the observation that only when WAG/Rij rats were EC housed after age 3 months, they had more type 2 SWDs than rats from all other groups. In all, our data underline the different properties of type 1 and type 2 SWDs, whereas a comparison with the literature suggests opposite effects of EC housing on absence and convulsive epilepsy. [Supported by Dutch Organization for Scientific Research (NWO), grant 425-20-401.]
1.183 SEIZURES IN NQO1 KNOCKOUT MICE Janet L. Stringer, Delwin J. Long II, LaToia M. Marks, and Anil K. Jaiswal (Pharmacology, Baylor College of Medicine, Houston, TX) Rationale: NAD(P)H:quinone oxidoreductase 1 (NQO1) is a widely expressed enzyme thought to protect cells from the effects of reactive quinones and their derivatives. NQO1 is part of an oxidative stress– induced cellular defense mechanism that includes induction of gene expression. NQO1 −/− mice have alterations in intracellular redox status and altered glucose and fatty acid metabolism and have been observed to have spontaneous seizures. We compared the responsiveness of the NQO1 −/− mice to kainic acid-, pentylenetetrazol (PTZ)-, and insulininduced seizures to the background strain to try to determine whether
Epilepsia, Vol. 44, Suppl. 9, 2003
there is a change in seizure expression or initiation in the absence of this protein. Methods: Wild-type (C57Bl/6) and NQO1 −/− mice were injected intraperitoneally with either 15 mg/kg kainic acid (n = 9, −/−; n = 8, +/+) or 30 mg/kg PTZ (n = 8, +/+; n = 4, −/−). Seizure threshold was tested using PTZ infusion into the tail vein in 10 +/+ mice and nine −/− mice. Insulin-induced seizures were studied by intraperitoneal injections of 14-IU/kg insulin (n = 4, +/+; n = 3, −/−). Seizures were monitored until normal baseline behavior returned. In addition, brains from the −/− animals were examined histologically after perfusion with 4% paraformaldehyde. Sections, 35 µm, were Nissl stained and stained for glial fibrillary acidic protein (GFAP) immunoreactivity. Results: The spontaneous seizures in the NQO1 −/− mice occurred more commonly in older animals, but were still quite rare (once every 3–4 days). The seizures were characterized by tonic extension of the front legs, tension of the tail, salivation, and sometimes loss of righting. Some of the seizures included generalized clonic activity. The seizure threshold determined by the PTZ infusion was not different between the −/− group and the +/+ group. The wild-type animals took 0.29 ± 0.025 mg/kg PTZ to the first muscle twitch, whereas the −/− mice took 0.32 ± 0.02 mg/kg. The seizures induced by intraperitoneal injections of PTZ and kainic acid were not different between the two groups of mice. After injection of insulin, both groups became lethargic by 45 min, and the first seizure activity occurred between 50 and 65 min after the insulin injection. There was no difference in the morphology of the brains from the −/− mice with either the Nissl stain or GFAP immunoreactivity. Conclusions: The NQO1 −/− mouse is an example of a knockout mouse with spontaneous seizures, but no obvious morphologic changes in the brain and no changes in the responses to standard convulsive agents. Additional work has shown that NQO1 in the brain is localized primarily to the oligodendrocytes, and a role in myelin formation has been hypothesized. These data, plus the role of NQO1 in metabolism, suggest that the spontaneous seizures in these mice may be due to metabolic abnormalities due to the altered biochemical status of the animals. Further experiments are needed to confirm this hypothesis and to determine whether this animal may be useful in understanding seizures initiated by changes in systemic metabolism. (Supported by NS39941 to J.L.S. and ES07943 to A.K.J.) 1.184 PATHOLOGICAL CELL FEATURES IN CEREBRAL LESIONS IN THE EKER RAT MODEL OF THE TUBEROUS SCLEROSIS COMPLEX: AN IMMUNOCYTOCHEMICAL AND ELECTRON MICROSCOPIC STUDY 1 H. Jurgen Wenzel, 2 Peter B. Crino, 1 Mareike E. Wenzel, and 1 Philip A. Schwartzkroin ( 1 Neurological Surgery, University of California at Davis, Davis, CA; and 2 Neurology & Epilepsy Center, University of Pennsylvania, Philadelphia, PA) Rationale: Tuberous sclerosis (TSC) is an autosomal dominant disorder caused by mutations of either TSC1 or TSC2 genes. In the brain, the CNS lesions most closely associated with epilepsy and autism are regions of cortical dysplasia called tubers. The Eker rat, which carries a spontaneous germline mutation of the TSC2 gene, has been studied as an animal model of TSC. Recently, we showed that early postnatal irradiation, a “second hit” affecting Eker TSC2 +/− cells, caused abnormal cell features that resemble some of the TSC pathologies seen in human patients. We extend our analysis in the irradiated Eker rat brain, to characterize further the neuronal and glial phenotypes of abnormal cells that might be linked to seizures. Methods: Brain sections from irradiated and nonirradiated Eker (TSC+/−) rats were used for immunocytochemistry (ICC) of neuronal markers (NeuN, neurofilament, MAP2, GAD67, GABA) and markers for glial and/or immature cells (GFAP, vimentin, nestin) using doubleimmunofluorescence (IF) techniques. In addition, to characterize abnormal cells that may lack functional tuberin, a phosphoribosomal S6protein antibody was used as an ICC marker. Individual cells located in the neocortex or in the hamartomas displaying pathological features were examined by means of light and electron microscopy (EM). Results: Our previous study showed that irradiation of Eker (+/−) rats generated such neuropathologic features as dysmorphic (cytomegalic) neurons, giant astrocytes, and hamartoma-like lesions. Double-IF analysis combining neuronal and glial markers reveals that dysmorphic
AES PROCEEDINGS cells with neuron-like morphology also display a neuronal ICC (GFAPnegative) phenotype. At the EM level, these cells have excitatory and inhibitory synapses on soma and dendritic processes, suggesting that they are involved in neuronal circuitry. Giant astrocytes (immunonegative for neuronal markers) are closely associated with blood vessels. Interestingly, many of these large aberrant cells demonstrate robust immunoreactivity for phospho-S6 protein expression, a marker for cells lacking tuberin. Further ICC analyses of cells in hamartomas (subcortical white matter heterotopia) reveal a mixed glioneuronal phenotype, which expresses both GFAP and GAD67/MAP2 (but not NeuN). At the EM level, phospho-S6-protein–positive cells exhibit different intensities of GFAP and intermediate filament expression. These results suggest cell populations with varying degree of maturity. Conclusions: Early postnatal irradiation of Eker (TSC +/−) rat pups generates some of the pathologic features observed in brain lesions of TSC patients. Future studies focusing on modifying this “second hit” approach, and on tuberin gene pathways, will provide a better understanding of how brain lesions in TSC are formed. [Supported by NIH NS 18895 (P.A.S.), and NS 04502 and Tuberous Sclerosis Alliance (P.B.C.).]
1.185 COMPARISON OF THE SEIZURE SUSCEPTIBILITY OF DIFFERENT STRAINS OF MICE IN ELECTRICAL KINDLING AND CHEMICALLY INDUCED SEIZURE MODELS Lijuan Yang and Harlan E. Shannon (Neuroscience Division, Lilly Research Laboratories, Indianapolis, IN) Rationale: Previous studies have shown that seizure thresholds vary in different strains of mice. However, there has been little direct comparison across strains in terms of their susceptibility to electrical kindling and chemically induced seizure thresholds. In the present studies, we compared the seizure susceptibility of four strains of mice: CF-1, a Swiss albino strain commonly used in anticonvulsant studies; CD1, a non-Swiss albino strain commonly used in pharmacology studies; C57/BL6, an inbred strain and common parental line for transgenic mice; and 129S6/SvEv, a common source for embryonic stem cells and parental line for transgenic mice. We compared the differences among these four strains of mice in the amygdala kindling model and in chemically induced acute seizure models. In addition, we determined the anticonvulsant effects of carbamazepine (CBZ) and levetiracetam (LEV) in kindled mice. Methods: A bipolar electrode was stereotaxically implanted into the right amygdala of CF-1, CD-1, C57/BL6, and 129S6/SvEv mice. Mice were given daily stimulations (500 mA, 60 Hz, 1 s) until ≥10 stage 5 (Racine scale) seizures occurred. Behavioral seizure scores (Racine scale) and EEG afterdischarge (AD) durations were recorded throughout. Postkindling seizure thresholds and AD durations were also determined. After thresholds stabilized, the anticonvulsant effects of CBZ (30 mg/kg, i.p.) and LEV (50 mg/kg, i.p.) were determined. In chemically induced seizure models, kainate (20 mg/kg, i.p.) or pilocarpine (100 mg/kg, i.p.) were injected every 20 min until the first limbic seizure. The number of doses and minutes to onset of limbic seizures were recorded. Results: During kindling development, the C57/BL6 mice had higher seizure scores but comparable AD durations compared with the other three strains of mice. After kindling, seizure thresholds were not significantly different among the four strains; however, the 129S6/SvEv strain had longer AD durations. CBZ and LEV increased the seizure thresholds of all strains. In addition, C57/BL6 mice required significantly fewer doses and less time to the onset of limbic seizures, and had more deaths, than the other three strains after kainate, whereas 129S6/SvEv mice were relatively more resistant to pilocarpine in that they required higher doses to induce seizures. Conclusions: C57/BL6 mice were more seizure prone than CF-1, CD1, or 129S6/SvEv mice in amygdala kindling and kainate seizure models, whereas 129S6/SvEv mice were less susceptible to pilocarpine-induced seizures. The present results demonstrate that differences among mouse strains exist in seizure susceptibility to amygdala kindling as well as kainate- and pilocarpine-induced seizures. In particular, C57/BL6 mice are relatively more seizure prone, a finding that should be taken into consideration in investigations of transgenic mice with the C57BL/6 strain as one of the parental lines. (Supported by Lilly Research Laboratories.)
67
1.186 SZT2: A NEW SEIZURE THRESHOLD MOUSE MUTANT OBTAINED BY MUTAGENESIS AND ECT Yan Yang, Barbara J. Beyer, Verity A. Letts, and Wayne N. Frankel (Neurogenetics Group, The Jackson Laboratory, Bar Harbor, ME) Rationale: The genome sequence availability coupled with efficient chemical mutagenesis and phenotypic screening increases the potential for assigning function to mammalian genes. We recently published Szt1 (Yang et al. Hum Mol Genet 2003;12:975–84), a mutation detected by the electroconvulsive threshold (ECT) test, hemizygous for two human epilepsy susceptibility genes. This result emboldened us to seek new seizure threshold mutants where seizure phenotype is primary and can be used to improve therapy. Methods: We screened progeny of ENU-treated C57BL/6J mice at the 3% response level for minimal clonic ECT seizures using transcorneal electrodes. Mice were further characterized genetically using DNA markers, as well as for other seizure threshold paradigms including maximal ECT (hindlimb extension seizures) and PTZ. Results: Szt2 (seizure threshold-2) was one of three mutants isolated in pilot screens. It has a low seizure threshold in several paradigms, including minimal and maximal ECT, and PTZ. Szt2 is semidominant and was mapped to distal Chr 4 in backcrosses with the BALB/cByJ strain. For higher-resolution mapping, we progeny-tested mice carrying informative recombinations, restricting Szt2 to an 11-Mb interval between D4Mit332 and D4Mit124. Candidate genes that might be involved in idiopathic seizure phenotype are being evaluated as the critical interval is being reduced further. Conclusions: Szt2 is a single locus mouse mutation with a low threshold in several seizure-induction paradigms, including ECT and PTZ. We are close to identifying the genetic defect. Seizure-threshold models with discrete genetic defects have the potential to provide practical and meaningful contributions to developing better antiepileptic drug therapies. [Supported by grants from the National Institutes of Health (NS40246 and NS31348 to W.N.F.) and a TJL Postdoctoral Fellowship to Y.Y.] 1.187 NO EVIDENCE FOR AN INCREASED MALIGNANCY RISK IN EPILEPSY CAUSED BY LGI1 MUTATION 1 Eylert Brodtkorb, 2 Karl O. Nakken, and 3 Steinlein K. Ortrud (1 Department of Neurology, Trondheim University Hospital, Trondheim, and 2 The National Centre for Epilepsy, Sandvika, Norway; and 3 Institute of Human Genetics, University Hospital Bonn, Bonn, Germany) Rationale: The LGI1 gene (leucine-rich glioma-inactivated-1) is supposed to be a tumor-suppression gene involved in the formation and progression of glial tumors. It has recently been shown that LGI1 is mutated in families with autosomal dominant temporal lobe epilepsy (ADLTE). An urgent question emerged: are LGI1 mutation carriers at a higher risk for brain tumors and other malignancies? Methods: We have studied the comorbidity in a large, five-generation Norwegian family with ADLTE caused by the heterozygous LGI1 mutation C46R. Results: In only one patient had a malignancy been diagnosed. This patient was treated with surgery 18 years ago and is now in good health. The obligate carriers in generation II and the likely carrier in generation I all reached above-average age (90–95 years). Neuroimaging in 11 of the subjects with clinical ADLTE revealed no signs of intracranial neoplams. Sudden death had occurred in two individuals with epilepsy, in one young man at age 27, and in a 64-year-old woman who had multiple chronic disorsers. Conclusions: Our clinical findings do not support the hypothesis that the present mutation increases the number of malignancies or lower the age at onset. 1.188 NO ASSOCIATION BETWEEN VARIATION IN THE PRODYNORPHIN GENE AND RISK FOR TEMPORAL LOBE EPILEPSY IN A U.S. COHORT 1 Russell J. Buono, 2 Micahel R. Sperling, 1 Theresa M. Scattergood, 3 Dennis J. Dlugos, 1 Stephanie O. Kratzer, 1 Danielle M. Press, 1 Wade H.
Epilepsia, Vol. 44, Suppl. 9, 2003
68
AES PROCEEDINGS
Berrettini, and 1 Thomas N. Ferraro ( 1 Psychiatry, University of Pennsylvania, 2 Neurology, Thomas Jefferson University Hospital, and 3 Neurology, Children’s Hospital of Philadelphia, Philadelphia, PA) Rationale: A previous report in a Austrian population identified positive association between a variation in the promoter of the prodynorphin gene (PDYN) and risk for acquiring temporal lobe epilepsy (TLE). This association was seen in TLE patients that had a family history of epilepsy (at least one first- or second-degree relative with epilepsy) but not in TLE patients without a family history. The variation affects promoter activity in an in vitro assay and is a plausible candidate for an epilepsy susceptibility gene. However, a subsequent study in a separate German population found no association between the PDYN polymorphism and TLE with familial history. We replicated these studies with twice the number of patient samples compared with the previous reports. Methods: All patients were recruited in Philadelphia, PA, U.S.A., and provided signed informed consents for these studies. DNA was obtained after extraction from a peripheral blood sample. The variation is a 68-bp repeat element that occurs in human populations in one, two, three, or four copies of the repeat per allele. PCR was used to amplify the region surrounding the repeat element, and genotypes were visualized on 5% nondenaturing polyacrylamide gels with ethidium bromide staining. Gels were photographed and scored by two independent readers. Genotypes and allele frequencies were analyzed by χ 2 contingency analysis. Results: Genotypes from a total of 158 TLE patients of European ancestry (74 with a positive family history) and 143 controls were analyzed. Alleles with one or two copies of the repeat were designated “low,” whereas alleles with three or four copies of the repeat were designated “high.” All genotypes were in Hardy–Weinberg equilibrium. We found no significant difference in genotype distribution or in allele frequency between the TLE patients and controls. Frequency of the high and low alleles in controls was 67% and 33%, respectively; in TLE patients, it was 66% and 34%, respectively. There was no significant difference in allele frequency between those TLE patients with and without a positive family history. High and low alleles frequencies were 68% and 32% in TLE without family history of epilepsy (n = 84) and 65% and 35% in TLE patients with a family history of epilepsy (n = 74). Conclusions: Our results suggest that the PDYN promoter variation under study is not associated with the phenotype of TLE in this patient population. It is possible that the discrepancy between our data and those published previously may be explained, in part, by increased genetic heterogeneity of the U.S. population compared with those studied in Europe. (Supported by NIH RO1NS-40396 to R.J.B. and The University of Pennsylvania Center for Neurobiology and Behavior.)
1.189 ASSOCIATION OF THE NEURONAL NICOTINIC ACETYLCHOLINE RECEPTOR SUBUNIT α4 Polymorphisms with Febrile Seizures 1 I.-Ching Chou, 2 Cheng-Chun Lee, 3 Chao-Ching Huang, 1 Chang-Hai Tsai, and 1, Fuu-Jen Tsai ( 1 Departments of Pediatrics and 2 Neurology, China Medical College Hospital, Taichung, 3 Department of Pediatrics, National Cheng Kung University Hospital, Tainan, and 4 Departments of Medical Genetics, China Medical College Hospital, Taichung, Taiwan) Rationale: The α4 subunit gene of the neuronal nicotinic acetylcholine receptor (CHRNA4) has been identified as the first gene underlying an idiopathic partial epilepsy syndrome in human, autosomaldominant nocturnal frontal lobe epilepsy. Studies provided evidences that the protein coded by CHRNA4 is one of the most abundant subunits of the neuronal nicotinic acetylcholine receptors in mammalian brain and mutations of CHRNA4 seem to cause neuronal excitation. The CHRNA4 gene may have a role in the development of febrile seizures (FSs), the majority of childhood seizures. The present study assessed the distribution of genotypes of CHRNA4 in patients with FSs. Methods: A total of 102 children with FSs and 80 normal control subjects were included in the study. Polymerase chain reaction was used to identify the C/T polymorphism of the CHRNA4 gene. Genotypes and allelic frequencies for the CHRNA4 gene polymorphisms in both groups were compared. Results: The number of individuals with heterozygous CHRNA4 (Ser543Ser)-C/T genotype was significantly greater (60.8 vs. 32.5%; p = 0.001), and the CHRNA4 (Ser543Ser)-T allele frequency was signifiEpilepsia, Vol. 44, Suppl. 9, 2003
cantly higher (p = 0.001) in patients with FSs compared with healthy controls. The odds ratio for developing FSs in individuals with the CHRNA4 (Ser543Ser)-CT genotype was 3.77 compared with individuals with two copies of the CHRNA4 (Ser543Ser)-C allele. Conclusions: This study demonstrated an association between the CHRNA4 gene and FSs. Individuals with the T allele had a higher incidence of FSs. These data suggest that the CHRNA4 gene or a closely linked gene might be one of the susceptibility factors for FSs.
1.190 CORTICAL TUBERS IN TUBEROUS SCLEROSIS COMPLEX ARE FORMED BY A SOMATIC INACTIVATING TSC GENE MUTATION 1 Jia Yu, 1 Marianna Baybis, 1 Allana Lee, 2 Raymond Yeung, 3 David Gutmann, 3 Erik Uhlmann, 4 Guy McKhann II, 5 Howard Weiner, 6 Katherine Nathanson, and 1 Peter Crino ( 1 Neurology, PENN Epilepsy Center, University of Pennsylvania, Philadelphia, PA; 2 Surgery, University of Washington, Seattle, WA; 3 Neurology, Washington University, St. Louis, MO; 4 Neurological Surgery, Columbia University, and 5 Pediatric Neurosurgery, New York University, New York, NY; and 6 Medical Genetics, University of Pennsylvania, Philadelphia, PA) Rationale: Tuberous sclerosis complex (TSC) is an autosomal dominant disorder resulting from mutations in the TSC1 or TSC2 genes encoding the proteins hamartin or tuberin. Tubers are developmental malformations of the cerebral cortex associated with epilepsy, characterized histologically by dysmorphic giant cells (GCs) that exhibit cytomegaly. A pivotal unanswered question is whether tubers arise as a consequence of a second-hit somatic mutation within the TSC1 or TSC2 genes. We identify a marker for GCs in human tubers, phosphorylated ribosomal S6 protein (phospho-RS6) and show that GCs are formed as a direct consequence of an inactivating somatic second-hit TSC gene mutation. Methods: Tubers were obtained from 13 patients with clinically diagnosed TSC (mean age, 7.25 years). One tuber specimen, frontal cortex, and cerebellum were obtained postmortem from one TSC patient (age 3 years). The Eker rat model results from a truncating Tsc2 mutation and exhibits single subependymal nodules in the lateral ventricles. Tsc1 astrocyte-specific conditional knockout (Tsc1 cKO) mice result in loss of hamartin expression in astrocytes. Brain sections from human specimens, Eker rats, and Tsc1 cKO mice were probed with phospho-RS6 (1:500 dilution) or tuberin antibodies (C-20, 1:200). Genomic DNA was extracted from the tuber, frontal cortex, and cerebellum obtained postmortem. Exon-specific flanking polymerase chain reaction (PCR) primers were generated, and amplicons from all 41 exons of TSC2 were sequenced. Single phospho-RS6 immunoreactive GCs as well as nonimmunoreactive pyramidal neurons in cortex and Purkinje cells in cerebellum were microdissected, and TSC2 mRNA was amplified by RT-PCR with overlapping cDNA primers, and sequenced. Studies were approved by the University of Pennsylvania Institutional Review Board. Results: Phospho-RS6 is a marker for cells such as astrocytes lacking hamartin in the Tsc1 cKO mouse and GCs lacking tuberin in the Eker rat model Phospho-RS6 is highly expressed in human GCs but absent in control neurons. DNA sequencing of TSC2 in a tuber, normal cortex, and the cerebellum from one patient revealed a nonsense mutation (1371 C>T, 458 R>X) in exon 13. RT-PCR amplification and sequencing of TSC2 mRNA from single microdissected, phospho-RS6 immunoreactive GCs revealed a multiexon–inactivating deletion resulting in loss of tuberin immunoreactivity in the GCs. This mutation was not identified in adjacent, phospho-RS6 nonimmunoreactive neurons. Conclusions: Expression of phospho-RS6 and single-cell mutational analysis provides direct evidence that GCs are generated as a consequence of loss of hamartin–tuberin pathway resulting from a somatic inactivating TSC gene mutation. [Supported (P.B.C.) by the Tuberous Sclerosis Alliance, NINDS NS045021, and the Esther A. and Joseph Klingenstein Fund.]
1.191 HYPOTHALAMIC HAMARTOMA DUE TO A SOMATIC MUTATION IN TRANSCRIPTION FACTOR GENE GLI3 1,5 Jeremy L. Freeman, 2,3 Robyn H. Wallace, 3 Paul A. Izzillo, 3 John C. Mulley, 1,4,5 A. Simon Harvey, and 4,5 Samuel F. Berkovic ( 1 Department of Neurology, Royal Children’s Hospital, Parkville, Victoria, Australia;
AES PROCEEDINGS 2 Department of Anatomy & Neurobiology, University of Tennessee, Memphis, TN; 3 Department of Cytogenetics & Molecular Genetics, Adelaide Women’s & Children’s Hospital, Thebarton, South Australia, Australia; and 4 Department of Neurology and 5 Epilepsy Research Institute, Austin & Repatriation Medical Centre, West Heidelberg, Victoria, Australia)
Rationale: The GLI3 gene (MIM 165240) encodes a DNAbinding transcription factor involved in downstream regulation of Sonic Hedgehog–mediated CNS development. Truncating (frameshift and nonsense) mutations located 3-prime of the zinc finger–encoding regions of GLI3 cause familial Pallister–Hall syndrome (PHS; MIM 146510), a rare entity characterized by hypothalamic hamartoma, central polydactyly, and other malformations. Hypothalamic hamartoma also occurs as an isolated CNS malformation in patients with an increasingly recognized clinical presentation incorporating early-onset gelastic seizures and often leading to secondarily generalized epilepsy. We aimed to investigate the role of GLI3 mutation in the etiology of nonsyndromal (sporadic) hypothalamic hamartoma associated with epilepsy. Methods: Twenty-four patients with hypothalamic hamartoma and refractory epilepsy were studied; none met clinical criteria for PHS. Whole-blood samples were obtained and DNA extracted using the QIAGEN Blood DNA Kit. In 12 patients, DNA from hypothalamic hamartoma tissue collected during transcallosal resection was extracted by using DNAzol (Invitrogen). Single-stranded conformation polymorphism screening (SSCP) of samples was performed: HEX-labeled primers were designed to amplify all 14 exons of GLI3; a 30-ng sample of patient DNA was amplified in a total volume of 10 µl; products were separated on nondenaturing 4% polyacrylamide gels containing 2% glycerol (GelScan 2000; Corbett Research). PCR products showing conformational changes were reamplified, by using unlabeled primers, from 100 ng of genomic DNA, and then sequenced (ABI PRISM BigDye Terminator v3.0 Ready Reaction Cycle Sequencing Kit; Perkin Elmer). Results: No GLI3 mutations were found in peripheral blood samples. In hypothalamic hamartoma tissue of one patient, a truncating mutation (Gln926Stop) was found in the CBP-binding domain of GLI3 that was absent in the blood sample. Although several gene polymorphisms were detected, disease-causing mutations were not identified in the other patients by SSCP, but may be revealed by sequencing experiments, currently under way. Conclusions: The location and type of the GLI3 mutation found in the hamartoma of our patient is similar to that described in some families with PHS, strongly suggesting that sporadic hypothalamic hamartomas can be due to somatic mutations in the Sonic Hedgehog pathway. This raises the possibility that other “sporadic” malformations of cortical development may be due to somatic mutations. 1.192 CONTRIBUTION OF GENETIC FACTORS TO SEIZURE RISK: A POPULATION-BASED STUDY OF 47,614 U.S., NORWEGIAN, AND DANISH TWIN PAIRS 1,2 Marianne J. Kjeldsen, 3 Linda A. Corey, 1 Mogens L. Friis, 4 Marit H. Solaas, 5 Jennifer R. Harris, 2 Kirsten O. Kyvik, 2 Kaare Christensen, and 6 Jack M. Pellock ( 1 Department of Neurology, Odense University Hospital, and 2 The Danish Twin Registry, Institute of Public Health, University of Southern Denmark, Odense, Denmark; 3 Department of Human Genetics, Virginia Commonwealth University, Richmond, VA; 4 Institute of Medical Genetics, University of Oslo, and 5 Department of Epidemiology, National Institute of Public Health, Oslo, Norway; and 6 Department of Neurology, Virginia Commonwealth University, Richmond, VA) Rationale: This study deals with the establishment of a sample of twins ascertained from population-based twin registries in three countries to study the importance of genetic factors for seizure liability. Methods: A sample of twins with seizures was established from twin registries in the U.S., Norway, and Denmark. Twins were screened by questionnaire/telephone interview for a history of seizures (epilepsy, febrile seizures, other seizures, and staring spells). Seizure frequencies were calculated both within and across twin samples. Classic twin analyses were performed, using probandwise concordance rates, to compare the similarity between monozygotic (MZ) and dizygotic (DZ) twins within and across twin samples for each of the seizure categories.
69
Results: Among the 47,614 twin pairs who provided health-history information included in the three twin samples, members of 6,234 pairs responded positively to at least one of the questions concerning seizures. Among those with a history of seizures, 1.6% reported epilepsy, 2.1% reported febrile seizures, 3.4% reported other seizures (without epilepsy and febrile seizures), and 1.2% reported a history of staring spells (without reporting epilepsy, febrile seizures, or other seizures). Table 1 provides number of affected pairs and concordance rates for MZ and DZ twins by seizure category for the combined twin sample.
TABLE 1. Number of affected twin pairs and concordance rates in the combined twin sample Epilepsy Febrile seizures Other seizures Staring spells MZ pairs DZ pairs CR-MZ CR-DZ
300 758 0.28 0.07
440 838 0.33 0.11
527 1,285 0.26 0.16
225 502 0.26 0.08
CR–MZ, concordance rate, MZ twins; CR–DZ, concordance rate, DZ twins. Conclusions: By merging data collected from three twin populations, we have created the largest unselected, population-based sample of twins with seizures to be developed to date. Our finding of MZ-DZ concordance rates that are consistent with the presence of genetic effects on the expression of epilepsy, febrile seizures, other seizures, and staring spells confirms the importance of genes in the occurrence of seizures. All twins who reported a history of seizures are currently being clinically evaluated by neurologists. Seizures and epileptic syndromes are being classified according to the ILAE. This sample is an important resource for studying the genetics of epilepsy subtypes and febrile seizures. [Supported in part by the following grants: NIH NINDS grant (NS-31564); The Clinical Research Institute, University of Southern Denmark; The grant committee of Consultancy Counsel, Odense University Hospital; The Danish Epilepsy Society Research Foundation; The Hede Nielsen Foundation; The K.A. Rohdes Foundation; The Novo Nordic Foundation, and the Norwegian Research Council.]
1.193 REGIONAL AND DEVELOPMENTAL LOCALIZATION OF ION-CHANNEL GENES COEXPRESSED IN HEART AND BRAIN 1 Alica M. Goldman, 1 Daniel L. Burgess, 2 Dawna D. Armstrong, 1 Mayra Mori, and 1 Jeffrey L. Noebels ( 1 Neurology and 2 Pathology, Baylor College of Medicine, Houston, TX) Rationale: Mutations of ion-channel genes are an important cause of inherited epilepsy and cardiac arrhythmia syndromes. It is our hypothesis that a defined group of ion channelopathies may underlie a specific phenotype of seizures combined with cardiac arrhythmias. The objective of this study is to verify the cerebral expression of two potassium ion genes linked to inherited cardiac rhythm disorders that are candidates for this syndrome. Methods: We performed regional mapping of KvLQT1 and KCNE1 genes, both linked to inherited LQT syndromes, in the developing and adult mouse brain and in dissected adult human brain regions obtained from postmortem tissue specimens by using the reverse transcription– polymerase chain reaction (RT-PCR) method. RNA was obtained from the whole brain of P0, P10 mice and from subregions of adult mouse brain (female C57BL/6). Spinal cord was studied in P10 and adult mice. Human RNA expression profiles were assessed in samples from neocortex, hippocampus, cerebellum, and spinal cord. RNA samples from mouse and human heart were used as internal positive controls. Results: KvLQT1 mutations account for ∼50% of all LQT cases, yet the expression of this gene in brain has never been unequivocally documented. We confirmed the expression of two functional KvLQT1 isoforms in human brain. Isoform 1, described in heart and several other peripheral tissues, was present in all brain regions studied. Isoform 2,
Epilepsia, Vol. 44, Suppl. 9, 2003
70
AES PROCEEDINGS
believed to be heart specific, was found in hippocampus, frontal, parietal, and occipital neocortex and at a lower levels in temporal cortex and spinal cord. Additional splice variants of uncertain functional significance were also noted. In mouse, we found two distinct isoforms expressed in heart, of which only one was clearly present in all CNS samples studied. Definite KvLQT1 expression in CNS was obvious in P0 animals and at older ages. Mutations in the Iks channel β subunit (MinK) are a less frequent source of human LQT mutations; however, when present, their clinical impact is profound. We found the gene to be expressed in all human CNS regions, as well as in mouse brain samples through early development. MinK RNA expression was weak in the P10 and adult mouse spinal cords. Conclusions: Mutations in LQT genes alter membrane excitability and cause cardiac arrhythmias. Of seven LQT genes identified to date (SCN5A, KvLQT1, HERG, MinkA, MiRP1, Kir2.1, AnkB), all had confirmed brain expression except KVLQT1 and MinK. We determined the unequivocal presence of these two LQT genes in mouse and human brain. Moreover, their developmental and regional expression patterns support our hypothesis that LQT genes coexpressed in heart and brain are likely candidate genes for the clinical phenotype of seizures combined with cardiac arrhythmias. [Supported by American Epilepsy Society/Epilepsy Foundation (A.M.G.), Blue Bird Circle, Charles A. Dana Foundation (J.L.N.).] 1.194 MUTATION ANALYSIS AND PHENOTYPIC CORRELATION OF TSC1 AND TSC2 IN KOREAN TUBEROUS SCLEROSIS PATIENTS Hee Hwang, Ji Eun Choi, Jong Hee Chae, Ki Joong Kim, and Yong Seung Hwang (Pediatrics, Seoul National University Children’s Hospital, Seoul, Republic of Korea) Rationale: Tuberous sclerosis (TS) is an autosomal dominant neurocutaneous syndrome, characterized by multisystemic hamartomatous lesions. More than 450 causable aberrations in TSC1 and TSC2, encoding hamartin and tuberin, were reported, but their phenotypic correlation is not clarified. We analyzed mutations of TSC1 and TSC2 in clinically diagnosed TS patients and investigated genotype/phenotype correlation. Methods: Forty-four patients (22 males, 22 females) met the criteria of definitive TSC by Tuberous Sclerosis Consensus Conference. Nine patients (20%) were familial cases. Genomic DNA was extracted from patients’ blood samples, and exons of TSC1 and TSC2 were PCR amplified. Mutation screening was performed by denaturing high-performance liquid chromatography (DHPLC), and confirmed by direct sequencing. Genotype–phenotype correlation was analyzed. Results: Among 44 patients, 41 patients (93%) had seizures. Hypopigmented macules were noted in 35 patients (80%), facial angiofibromatosis in 25 patients (57%), and Shagreen’s patch in 14 patients (32%). On brain MRI, subependymal nodules were revealed in 40 patients (91%), cortical tubers in 34 patients (83%), and giant cell astrocytoma in three patients (7%). In 13 patients (30%), renal cysts were identified. Renal and hepatic angiomyolipoma were noted in three patients (7%) and two patients (5%), respectively. No significant difference was proved between familial and sporadic cases. Mutations were identified in 13 of 44 cases (30%), composed of two mutations in TSC1 (5%) and 11 mutations in TSC2 (25%). Mutation types were as follows: five deletions (39%), four nonsense mutations (31%), three missense mutations (23%), and one insertion (8%). Eleven were novel mutations. Mutations in Cterminal half of TSC2 (amino acid 1100-1805) had significant correlation with cardiac rhabdomyoma greater than those in N-terminal half (amino acid 1-1099). Other clinical features did not show genotype–phenotype correlation. Conclusions: These data provide the first representative features of distribution and spectrum of TSC1 and TSC2 mutations in Korean patients. In our study, TSC1 and TSC2 mutation rates were relatively lower than the mutation rate reported in western countries. Eleven novel mutations were identified in our study.
1.195 THE CHANGE OF HOMOCYSTEINE LEVEL BY POLYMORPHISM OF 5,10-METHYLENETETRAHYDROFOLATE REDUCTASE AND METHIONINE SYNTHASE GENE IN EPILEPSY Epilepsia, Vol. 44, Suppl. 9, 2003
Ok Joon Kim (Department of Neurology, Pochon CHA University, Sungnam, Kyunggi-do, Republic of Korea) Rationale: Hyperhomocysteinemia was observed in epilepsy patients receiving anticonvulsants, especially homozygotes for mtehylenetetrahydrofolate reductase (MTHFR) gene C677T mutation. Recently it was reported that polymorphisms in methionine synthase (MS A2756G) also affect homocysteine level. Therefore, we examined whether the type of MS as well as MTHFR gene affects the level of homocysteine in epilepsy patients. Methods: We investigated folate, vitamin B12 , homocysteine, MTHFR, and MS gene in 181 patients with epilepsy. We analyzed the change of folate, vitamin B12 , and homocysteine level according to the type of genes. Results: The levels of folate and vitamin B12 were not affected by MTHFR and MS genes, but the level of homocysteine was affected by each other or combination of MTHFR and MS genes. When MTHFR/MS gene type was CC/GG, CT/GG, TT/AA, and TT/AG, the level of homocysteine was elevated compared with other type. Conclusions: We recommend higher vitamin supplementation with antiepileptic drugs in epilepsy patients according to the gene type after examining MTHFR and MS gene.
1.196 EPILEPTIC SEIZURES AND SYNDROMES IN TWINS: THE IMPORTANCE OF GENETIC FACTORS 1,3 Marianne J. Kjeldsen, 2 Linda A. Corey, 3 Kaare Christensen, and 1 Mogens L. Friis ( 1 Department of Neurology, Odense University Hospital, Odense, Denmark; 2 Department of Human Genetics, Virginia Commonwealth University, Richmond, VA; and 3 The Danish Twin Registry, Institute of Public Health, University of Southern Denmark, Odense, Denmark) Rationale: This study examined the contribution of genetic factors to risk for seizures and epilepsy overall, as well as to risk for the major epilepsy syndromes by using a twin study design. Methods: The study population was recruited from the nationwide, population-based Danish Twin Registry located at the University of Southern Denmark, Odense, Denmark. A total of 34,076 twins were screened for epilepsy. Cases were confirmed and classified by two neurologists according to the classification systems of the International League Against Epilepsy (ILAE). Probandwise concordance rates were used to assess the similarity of monozygotic (MZ) and DZ twins for seizures, epilepsy overall, and each of the major epilepsy categories. Biometric analysis was used to estimate heritability for epilepsy. Results: A total of 214 twin pairs with epileptic seizures and 190 pairs with epilepsy were ascertained. Significantly higher concordance rates were found for MZ compared with DZ twins for both epileptic seizures (0.56 for MZ and 0.21 for DZ pairs; p < 0.001) and for epilepsy (0.49 for MZ and 0.16 for DZ pairs; p < 0.001). Concordance rates were higher for generalized epilepsy (0.65 for MZ and 0.12 for DZ) than for localizationrelated epilepsy (0.30 for MZ and 0.10 for DZ). In twin pairs where both members had seizures, 83% of MZ and 65% of DZ pairs had the same major epilepsy syndrome. Genetic factors were found to account for two thirds of the liability to epileptic seizures and for more than one half of the liability for epilepsy. Conclusions: Analysis of this neurologist-verified epilepsy twin data set has confirmed that genetic factors have a substantial impact on the etiology of epileptic seizures as well as on the occurrence of both generalized and partial epilepsies. (Supported in part by the following grants: NIH NINDS grant (NS-31564); The Clinical Research Institute, University of Southern Denmark; The grant committee of Consultancy Counsel, Odense University Hospital; The Danish Epilepsy Society Research Foundation, and The Hede Nielsen Foundation)
1.197 A NEW MISSENSE MUTATION IN A PATIENT WITH CLASSIC LISSENCEPHALY 1 Fabio R. Torres, 1 Antonia Paula M. Faria, 2 Maria A. Montenegro, 2 Marilisa M. Guerreiro, 2 Fernando Cendes, and 1 Iscia Lopes-Cendes (1 Department of Medical Genetics and 2 Department of Neurology, Campinas University State, Campinas, Sao Paulo, Brazil)
AES PROCEEDINGS Rationale: Migration of postmitotic neurons from the ventricular zone to form the cortical plate comprises one of the most critical stages in brain development. Mutations in two genes, LIS1 and DCX, are responsible for neuronal migration disorders including classic lissencephaly (LIS) and subcortical band heterotopia (SBH). More than 100 deletions and 41 intragenic mutations have been described in the LIS1 gene in patients with LIS, but only five missense mutations were reported until now. The objective of this study is to report a new missense mutation in a patient with lissencephaly. Methods: All patients included in the study had high-resolution magnetic resonance imaging (MRI) scans performed in a 2-T scanner with T1- and T2-weighted images in three orthogonal planes or computed tomography (CT) scans. We searched for mutations in the coding exons of LIS1 and DCX genes in all patients. In addition, a total of 50 unrelated normal control subjects were also genotyped. Polymerase chain reaction (PCR) reactions were performed with primers designed to amplify the entire coding region and intron/exon boundaries of the LIS1 and DCX genes. PCR samples were analyzed by the single-strand conformation polymorphism (SSCP) method. The nucleotide sequence of all fragments showing an altered mobility in the SSCP experiment were subsequently determined using the Big Dye Terminator Sequencing kit for megaBACE 1000. Results: A total of 15 patients was analyzed; five of them have SBH, and 10 have LIS. SSCP analysis revealed one patient of the LIS group with altered band shift in exon 8 of the LIS1 gene and seven patients (four of the SBH and three of the LIS group) with abnormal migration patterns for exon 11 of the LIS1 gene. SSCP analysis of the DCX gene did not show any abnormal band shift. Sequencing of the abnormal PCR fragment amplified from exon 8 of the LIS1 gene showed a A → C transversion, which predicts a changes of a histidine for a proline in position 277 of the protein. MRI of this patient showed abnormalities more severe in the posterior regions, with agyria in the parietal and occipital lobes and pachygyria in temporal and frontal lobes. The abnormality found in exon 11 of the LIS1 gene was a C →T substitution in the beginning of the 3 -untranslated region, which was also present in nine control subjects. Conclusions: The malformation pattern of this patient is more severe in the posterior regions, a typical characteristic of individuals with LIS1 mutations. Otherwise this patient seems to have a LIS grade (3a), probably more severe than expected for individuals with missense mutations in LIS1.The new missense mutation we describe changes a basic histidine for a nonpolar proline. It is localized in the fifth WD domain of the lis1 protein, which is very conserved in different species, and is probably involved in interactions with b spectrin, a cytoskeletal protein. (Supported by CAPES and FAPESP.)
1.198 AUTOSOMAL RECESSIVE FORM OF PERIVENTRICULAR NODULAR HETEROTOPIA IN THREE BROTHERS WITH WEST SYNDROME 2 Marcelo M. Rodrigues, 1 Luis Otavio S.F. Cabloco, 1 Eliana Garzon, 3 Henrique C. Junior, 4 Fabio R. Torres, 4 Iara L. Brandao Almeida, 4 Iscia Lopes-Cendes, 2 Luis Celso P. Vilanova, 1 Elza Marcia T. Yacubian, and 1 Americo C. Sakamoto ( 1 Unit of Epilepsy Treatment and Research, 2 Department of Neurology, and 3 Department of Radiology, Sao Paulo Federal University EPM, Sao Paulo, and 4 Department of Medical Genetics, Campinas State University, Campinas, Sao Paulo, Brazil) Rationale: Periventricular heterotopia (PH) is characterized by the presence of neuronal nodules along the lateral ventricle walls. Familial PH is usually associated with prenatal lethality in males and mutations, especially in the first seven coding exons, of the X-linked gene filamin A (FLNA). Recently patients with PH associated with chromosome 5p15.1 duplication and 5p15.33 trisomy were reported, as well as an autosomal recessive form of PH in four unrelated families. We describe clinical and genetic characteristics of three male siblings with West syndrome secondary to PH. Methods: Members of the family had video-EEG recordings, MRI, and genetic studies performed. Genomic DNA was obtained by direct extraction from lymphocytes of fresh blood using standard manual technique. Polymerase chain reactions (PCRs) were carried out using the published primer sequences for the first seven exons of the FLNA gene.
71
PCR products were analyzed by single-strand conformation polymorphism (SSCP). Results: This consanguineous family has three affected male siblings, two monozygotic twins (9 months old), and a 4-year-old boy. Onset of flexor spasms, typical of West syndrome, was at ∼4 months of age in all of them. MRI revealed bilateral periventricular nodules of cerebral gray matter intensity typical of PH in the three children, whereas both parents had normal MRIs. Preliminary SSCP analysis in the first seven exons of the FLNA gene did not show any abnormalities. Conclusions: This is the first description of West syndrome in boys with PH. In addition, it is very likely that PH is segregating as an autosomal recessive trait in this family, making less likely the presence of FLNA gene mutations. (Supported by CAPES and FAPESP.)
1.199 DO GENETIC EFFECTS INFLUENCE THE FACTORS THAT PRECIPITATE SEIZURES? 1 Marit H. Solaas, 2 Marianne J. Kjeldsen, 2 Mogens L. Friis, 3 Karl O. Nakken, 3 Grethe B. Refsland, 4 Jack M. Pellock, and 5 Linda A. Corey ( 1 Institute of Medical Genetics, University of Oslo, Oslo, Norway; 2 Department of Neurology, Odense University Hospital, Odense, Denmark; 3 The National Centre for Epilepsy, Sandvika, Norway; and 4 Department of Neurology and 5 Department of Human Genetics, Virginia Commonwealth University, Richmond, VA) Rationale: In this study we examined the factors that provoke seizures in a sample of twins who were concordant for epilepsy to determine whether genetic effects have an impact on seizure-precipitating factors. Methods: Study participants were originally ascertained through twins who responded positively to queries regarding a history of seizures in a survey questionnaire sent to twins included in twin registries in the United States, Denmark, and Norway. Twins who indicated a history of seizures have been evaluated and their seizure and epilepsy types classified according to the ILAE. Twin zygosity was determined by DNA markers. The twins were asked about the presence of 36 factors known to provoke seizures. A twin pair was considered concordant for epilepsy regardless of seizure and epilepsy subtypes, and concordant for provoking factors if pair members shared at least one of these factors. Classic probandwise concordance rates were computed to compare similarity between monozygotic (MZ) and dizygotic (DZ) twins within and across twin samples for seizure-precipitating factors. Results: To date, members of 1,019 twin pairs from the three population-based twin samples have been tested for zygosity and classified according to the ILAE. Among these twin pairs, 87 MZ pairs and 40 DZ pairs were concordant for seizures, although some of the pairs were discordant for seizure and epilepsy subtypes. After excluding pairs concordant for febrile seizures, 45 MZ and 28 DZ pairs were found to be concordant for epilepsy. Among these twin pairs, the concordance rates for provoking factors were 0.80 for MZ pairs and 0.33 for DZ pairs. Conclusions: The results of this study suggest that genetic factors may contribute to the causes that provoke seizures. The extent to which the similarity found in the factors that provoke seizures in twin pairs concordant for epilepsy reflects concordance for specific epilepsy syndromes is not clear. We have initiated studies to examine further the degree to which provoking factors are characteristic for specific epilepsy syndromes. (Supported by NIH NINDS grant NS-31564.)
1.200 SEQUENCE VARIANTS OF GABRB3, GABRA5, AND GABRB3 ON CHROMOSOME 15Q 11.2-12 IN REMITTING CHILDHOOD ABSENCE EPILEPSY 1,7 Miyabi Tanaka, 2 Ignacio P. Castroviejo, 1 Jesus Machado-Salas, 1 Gregorio S. Pineda, 3 Berge Minnasian, 4 Marco T. Medina, 5 Chung Y. Fong, 1,6 Julia Bailey, 1 Dongsheng Bai, 7 Richard W. Olsen, and 1 Antonio V. Delgado-Escueta ( 1 Comprehensive Epilepsy Center, University of California in Los Angeles, Los Angeles, CA; 2 The Hospital of La Paz, Madrid, Spain; 3 Neurology, The Hospital for Sick Children, Toronto, Canada; 4 Neurology, National Autonomous University, Tegucigalpa, Honduras; 5 Neurology, Queen Mary Hospital, Hong Kong, Hong Kong; and 6 Neuropsychiatrics and 7 Pharmacology, University of California at Los Angeles, Los Angeles, CA)
Epilepsia, Vol. 44, Suppl. 9, 2003
72
AES PROCEEDINGS
Rationale: Because possible genetic linkage has been observed between A55CA1 (HLOD 1.8 with theta = 0.37, AR model and NPL-2.5 with p = 0.0004) [Robinson et al. Epilepsia 2002;43(suppl. 7):277] and GABRB3, 85CA, GABRA5 (all with LOD 2.5 with theta = 0) [Tanaka et al. Epilepsia 2000;41(suppl. 7):250) and remitting childhood absence epilepsy (r-CAE) in two separate and independent cohorts from Europe/London and Los Angeles/Mexico, we investigated the possible association of sequence variants in GABRB3, GABRA5, and GABRG3 on chromosome 15q 11.2-12 and r-CAE probands. Methods: We studied r-CAE probands of 10 families (three Canadian, two Spanish, one European American, four European/Amerind), which had originally contributed to a summed LOD of 2.5 (theta = 0). Each primer was designed to contain the full length of each exon. For GABRB3 and GABRA5, we extended the region to ∼200 bp forward from exon 1a. All r-CAE probands were initially screened with controls by denaturing high-performance liquid chromatography. Suspected variations were then sequenced. Results: Six single nucleotide polymorphisms (SNPs) were found. C30G (NCBI L04311) 10bp telomeric from exon 1a of GABRB3 and C170T (NCBI L04311) in exon 1a of GABRB3 are present in six of 10 r-CAE probands, in 10 of 41 European controls, and 32 of 39 European/Amerinds. In GABRA5, C231T in intron 5 (NCBI AF228447) and C268T, (NCBI AF228447) in exon 6 are each present in one of six r-CAE white probands, but in none of 39 white controls. A1473G (NCBI NM 000810) in exon 11 of GABRA5 and T413C, (NCBI AF269142) in exon 8 of GABRG3 did not show any differences with controls. Conclusions: Although our results are preliminary, the observed differences in SNP C231T in intron 5 and C268T in exon 6 of GABRA5 between r-CAE probands and controls are promising. Further recruitment of 100 new CAE probands is in progress, and results of their SNP screens will be presented.
1.201 SEARCHING FOR MUTATIONS IN THE EMX2 GENE IN A LARGE SAMPLE OF PATIENTS WITH SCHIZENCEPHALY 1 Fabio R. Torres, 2 Maria C.S. Rodrigues, 3 Maria M. Montenegro, 3 Ana Maria S.G. Piovesana, 3 Marilisa M. Guerreiro, 3 Fernando Cendes, 2 Juan C. Llerena, and 3 Iscia Lopes-Cendes ( 1 Department of Medical Genetics, Campinas University State–UNICAMP, Campinas, Sao Paulo, 2 Department of Medical Genetics, IFF/Fiocruz, Rio de Janeiro, Rio de Janeiro, and 3 Department of Neurology, Campinas University State, Campinas, Sao Paulo, Brazil) Rationale: Schizencephaly is a rare congenital defect of the cerebral cortex characterized by clefts of the cerebral mantle, extending from the pial surface to the lateral ventricles, and lined by heterotopic gray matter. The description of familial cases of schizencephaly raised the hypothesis that genetic factors could play a relevant role in the pathogenesis of this brain malformation. This evidence is supported by the presence, in a few patients, of germline mutations of the homeobox gene EMX2. The objective of this study was to search for mutations in the EMX2 gene in patients with schizencephaly Methods: All patients included in this study had high-resolution magnetic resonance imaging (MRI) scans performed in a 2-T scanner with T1 - and T2 -weighted images in three orthogonal planes or computed tomography (CT) scans. We searched for mutations in the three coding exons of EMX2 gene in all patients. In addition we genotyped 50 unrelated normal subjects. Polymerase chain reaction (PCR) reactions were performed with primers designed to amplify the entire coding region and intron/exon boundaries of the three exons of the EMX2 gene. PCR samples were analyzed by the single-strand conformation polymorphism (SSCP) method. The nucleotide sequence of all fragments showing an altered mobility in the SSCP experiment were subsequently determined using the Big Dye Terminator Sequencing kit for megaBACE 1000. Results: A total of 34 patients were analyzed in this study. Only one of our patients had a familiar history of epilepsy, but none of them had a family history of brain abnormalities or mental impairment. Ten patients had bilateral open-lip clefts, four patients had both types of clefts, four patients had unilateral closed-lip cleft, nine patients had unilateral openlip cleft, and seven patients had clefts not yet classified. SSCP analysis identified band shifts in exon-2 in four patients. Automatic sequencing Epilepsia, Vol. 44, Suppl. 9, 2003
has identified a C→ A substitution that does not change the wild-type amino acid arginine in position 156 of the protein. In addition, the same sequencing change was found in four individuals of the control group. Conclusions: Although the C→A substitution has been previously reported as disease related in schizencephaly patients, our results clearly show that this mutation is not involved in the etiology of this disorder in our group of patients, because it was present in 8% of our control group. In addition, it is likely that abnormalities in EMX2 are not a major cause of brain malformation because only a few reports confirmed the presence of these mutations in patients with schizencephaly. (Supported by CAPES and FAPESP.) 1.202 WILLIAM LENNOX’S TWIN STUDY COMPARED WITH RECENT DATA: LESSONS IN CLASSIFICATION AND GENETICS 1 Lata Vadlamudi, 2 Eva Andermann, 3 Cesare T. Lombroso, 4 Steven C. Schachter, 5 Roger L. Milne, 5 John L. Hopper, 2 Frederick Andermann, and 1 Samuel F. Berkovic (1 Epilepsy Research Institute, University of Melbourne, Melbourne, Victoria, Australia; 2 Montreal Neurological Institute, Montreal, Quebec, Canada; 3 Harvard Medical School and 4 Beth Israel Deaconess Medical Center, Boston, MA; and 5 Genetic Epidemiology Unit, University of Melbourne, Melbourne, Victoria, Australia) Rationale: William Lennox recognized the value of twin studies in analyzing the genetics of epilepsy and collected a series of twins between 1934 and 1958. His files were archived, which allowed us the opportunity to reevaluate this unique resource. We analyzed the data according to modern classifications of epileptic seizures and syndromes. We also compared concordances for epilepsy syndromes with recent Australian twin data. Methods: There were 143 twin pairs reviewed from Lennox’s original files. Each twin individual was classified for seizures and syndromes. The twin pairs were divided into 71 monozygous (MZ) and 72 dizygous (DZ) twin pairs for estimation of casewise concordances for epilepsy syndromes. Results: Seizure analysis demonstrated strong links between the contemporary seizure classifications and Lennox’s terminology. In 32 cases diagnosed with absence seizures, the term “petit mal” was used by Lennox in 29 cases. Lennox used the term “psychomotor seizures” in 47 cases, of which we regarded 45 as either simple or complex partial seizures. There was striking similarity between the two data sets for casewise concordances for all generalized epilepsies (Lennox:Australian, MZ 0.80:0.82; DZ, 0.27:0.26). In the partial epilepsies, higher MZ concordance was seen in the Australian twin data (MZ, 0.09:0.36; DZ, 0.06:0.05). In febrile seizures, both data sets demonstrated high MZ concordances (MZ, 1.0:0.58; DZ, 0.67:0.14). Conclusions: This is the first study to demonstrate that historical seizure terminology, as used by Lennox, corresponds closely to contemporary terms showing that his data remain valuable. With the exception of partial epilepsies, where the Lennox data did not show an inherited contribution (perhaps due to issues of power or errors in retrospective classification), the similarity with contemporary Australian twin data was remarkable. In particular, the major genetic contribution in generalized epilepsies and febrile seizures was highlighted, despite two populations differing in era and continent. This now allows the potential to combine these large data sets, for more detailed analyses of subsyndromes. 1.203 HUMAN PRION PROTEIN GENE VARIANT ALLELE IS ASSOCIATED WITH MESIAL TEMPORAL LOBE EPILEPSY AND PREDICTS ITS SURGICAL OUTCOME 1 Roger Walz, 2 Rosa M.R.P.S. Castro, 1 Tonicarlo R. Velasco, 1 Veriano Alexandre, Jr., 2 Marilene H. Lopes, 1 Jo˜ao P. Leite, 1 Antˆonio C. Santos, 1 Jo˜ ao A. Assirati, Jr., 1 Lauro Wichert-Ana, 1 Vera C. Terra-Bustamante, 1 Marino M. Bianchin, 2 Paulo C. Maciag, 2 Karina B. Ribeiro, 1 Ricardo Guarnieri, 1 David Ara´ujo, 1 Ot´avia Cabalero, 1 Ricardo Moura, 2 Anna C.M. Salin, 2 Kathleen Kindlmann, 2 Michele C. Landemberger, 1 Wilson Marques, Jr., 1 Regina M.F. Fernandes, 1 Luciano N. Serafini, 1 Helio R. Machado, 1 Carlos G. Carlotti, Jr., 2 Ricardo R. Brentani, 1 Am´erico C. Sakamoto, and 2 Vilma R. Martins 2 [1 CIREP, Centro de Cirurgia de Epilepsia, Hospital das Cl´ınicas, Faculdade de Medicina de Ribeir˜ao Preto, Universidade de S˜ao Paulo (FMRP-USP), Ribeirao Preto, and 2 Ludwig Institute, S˜ ao Paulo Branch, S˜ao Paulo, Brazil]
AES PROCEEDINGS Rationale: Mesial temporal lobe epilepsy related to hippocampal sclerosis (MTLE-HS) is the most common surgically remediable epileptic syndrome. Ablation of the cellular prion protein (PrPc ) gene (Prnp) enhances neuronal excitability of the hippocampus in vitro and sensitivity to seizure in vivo, indicating that PrPc might be related to epilepsy. We investigated the association among Prnp variant alleles and the surgical outcome of MTLE-HS patients. Methods: Prnp coding sequence of DNA from leukocytes of 100 consecutive patients with surgically treated MTLE-HS was compared with that from a group of healthy controls that were similar according to sex, age, and ethnicity (n = 180). The presence of Prnp variant alleles was correlated with sex, ethnicity, age at surgery, age at epilepsy onset (recurrent seizures), the duration of epilepsy, history of initial precipitating insult, epilepsy duration until surgery, family history of seizures in the first-degree offspring, monthly complex partial seizure frequency impairing awareness in the year before surgery, distribution of interictal spikes, neuroimaging, side of surgery, and surgical outcome. Surgical outcome was analyzed by Kaplan–Meier curve followed by Cox proportional hazards model to identify independent risk factors for having seizures after surgery. Results: A variant allele at position 171 (Asn→Ser), absent in controls, was found in heterozygosis (Asn171Ser) in 23% of patients (p < 0.0001). The Prnp genotypes were not correlated with any clinical and presurgical investigated data. However, patients carrying Asn171Ser variant had 5 times greater chance to remain with seizures after temporal lobectomy (95%CI, 1.65–17.33; p = 0.005) than those carrying the normal allele. At 18 months after surgery, 91.8% of patients with the normal allele at codon 171 were seizure free; Cox proportional hazards model was applied to identify independent risk factors for having seizures after surgery, in comparison to 68.2% of those carrying Asn171Ser (p = 0.005). Conclusions: Prnp variant allele Asn171Ser is highly prevalent in patients with medically untreatable MTLE-HS and independently predicts their surgical outcome. The results suggests that Prnp variant allele at codon 171 (Asn171Ser) is associated with epileptogenesis in MTLEHS. [Supported by grants from FAPESP (99/07124-8, 98/143352) and FAEPA-HC.]
1.204 INDEPENDENT CONFIRMATION OF DISTINCT GENETIC EFFECTS ON MYOCLONIC AND ABSENCE SEIZURES 1,2 Melodie R. Winawer, 4 Carla Marini, 6 Daniel Rabinowitz, 4 Samuel F. Berkovic, 4,5 Ingrid E. Scheffer, and 2,3,7 Ruth Ottman (1 Department of Neurology, 2 G.H. Sergievsky Center, Columbia University, and 3 Department of Epidemiology, Columbia University/Mailman School of Public Health, New York, NY; 4 Department of Neurology, Epilepsy Research Institute, University of Melbourne, Austin and Repatriation Medical Center and 5 Department of Neurology, Royal Children’s Hospital, Monash Medical Centre, Melbourne, Victoria, Australia; and 6 Department of Statistics, Columbia University, and 7 Epidemiology of Brain Disorders Research Department, New York State Psychiatric Institute, New York, NY) Rationale: Using a novel approach, we have recently shown that some of the genetic effects on myoclonic and absence seizures are distinct, whereas the shared versus distinct genetic effects on existing idiopathic generalized epilepsy syndromes (IGEs) are less clear. Here we apply this novel method to examine genetic contributions to myoclonic and absence seizures in an independent set of families to confirm our original findings. Methods: We examined evidence for distinct genetic contributions to myoclonic and absence seizures in families containing two or more individuals with IGE. For this purpose we used a permutation test to evaluate whether different seizure types co-occur within families to a degree greater than expected by chance. We performed a similar analysis to examine evidence for distinct genetic contributions to juvenile myoclonic epilepsy (JME), juvenile absence epilepsy (JAE), and childhood absence epilepsy (CAE). Results: The analysis included 41 families containing 103 individuals with either myoclonic seizures alone, absence seizures alone, or both seizure types. Overall, 30 of the families (73%) were concordant for seizure type. The observed number of concordant families was significantly greater than that expected (30 vs. 18.02; Z = 4.90; p < 0.0001). Thirty-six of the families had two or more individuals with CAE, JAE,
73
or JME. The observed number of families concordant for syndrome was significantly greater than expected when JME was compared with absence epilepsies (JAE+CAE; 27 vs. 16.36; Z = 4.34; p < 0.0001), but not when JAE was compared with CAE (13 vs. 11.40; Z = 0.85; p = 0.3953). Conclusions: These results confirm previous evidence for distinct genetic effects on absence and myoclonic seizures in an independent set of families. They also suggest that genetic differences among IGE syndromes may be driven by the seizure types that define them. The approach used here can direct the selection of subgroups of families more likely to be genetically homogenous for linkage analysis. In future linkage studies of IGEs, a focus on seizure type in addition to syndrome type may be useful. (Supported by NIH grants R01 NS20656, R01 NS43472, R01 NS36319, K23 NS02211; NIH/NIGMS R01 GM055978-06; National Health and Medical Research Council of Australia; Bionomics Ltd.) 1.205 ADVANCING EPILEPSY TISSUE BANKING BY USING IMAGE FUSION AND INTRAOPERATIVE NEURONAVIGATION: A RESOURCE FOR THE FUTURE 1 Edie E. Zusman, 2 Valerie C. Coon, 3 Jeff P. Gregg, 1 MichaelS.B. Edwards, 4 Lillian Lee, and 4 Nicholas A. Scott (1 Neurological Surgery, Sutter Neuroscience Institute, Sacramento, 2 Stanford University School of Medicine, Stanford, and 3 Pathology and 4 Neurological Surgery, University of California at Davis, Sacramento, CA) Rationale: Traditionally neurosurgeons have used neuronavigational technology and image fusion to improve the clinical safety and accuracy of surgical procedures. By coupling co-registered anatomic and functional images available intraoperatively with a precise tissue-banking protocol, we use neuronavigation to optimize molecular study designs for basic neuroscience research in epilepsy. Methods: After attainment of informed consent, patients with medically refractory focal epilepsy underwent craniotomy with neuronavigational guidance. Registrations were confirmed within 2-mm accuracy. Image-guided resection was performed based on EEG localization and imaging characteristics. Fused imaging data sets made anatomic and functional information available: MRI T1 ± Gad, T2 , EEG data from electrode localization, brain mapping, and magnetic resonance spectroscopy. For each tissue specimen obtained, the precise region of tissue collection was identified. This information was screen-saved and stored superimposed on the preoperative images. Each specimen was divided evenly, with one portion sent “fresh” to pathology for frozen section as appropriate and routine histopathologic studies. The other portion of each specimen was further subdivided into ∼6-mm portions and flashfrozen in N2 (l) within 60 s after tissue removal. Specimens remain stored in a −70◦ C freezer. Results: For each patient enrolled in the study, specimens were obtained from specifically identified brain regions based on MRI/MRS, functional, or microscopic characteristics. Intraoperative images confirming specimen localization were available for comparison with EEG and imaging data sets. Histopathologic and immunocytochemical analysis were available on most specimens. Gene expression profiling using GeneChip microarrays containing 22,000 genes was performed. Conclusions: Preliminary analysis is under way to evaluate alterations in gene expression based on electrocorticography, imaging, and histologic characteristics of tissue resected during surgery for epilepsy. This rigorous tissue-banking protocol reliably preserves RNA for geneexpression profiling and may lead to an improved understanding of the molecular genetic basis for epilepsy. Directions for future research: To identify molecular genetic profiles which correlate with seizure activity as well as those which may be unique to epileptic circuits. (Supported by Bronte Epilepsy Research Foundation.)
Imaging 1.206 VOLUMETRIC EVIDENCE OF CONTRALATERAL DAMAGE IN UNILATERAL MESIAL TEMPORAL LOBE EPILEPSY Epilepsia, Vol. 44, Suppl. 9, 2003
74
AES PROCEEDINGS
David Araujo, Tonicarlo Rodrigues Velasco, Lauro Wichert-Ana, Veriano Alexandre, Jr., Vera Cristina Terra Bustamante, Roger Walz, Antonio Gilberto Carlotti, Jr., Joao Antonio Assirati, Jr.Americo Ceiki Sakamoto, Antonio Carlos Santos, and Joao Pereira Leite (Department of Neurology, Department of Surgery, and Department of Radiology, Hospital de Cl´ınicas de Ribeir˜ao Preto, Ribeir˜ao Preto, S˜ao Paulo, Brazil) Rationale: There is increasing evidence that temporal lobe epilepsy (TLE) may be related to widespread changes involving extrahippocampal and contralateral structures. We investigated the presurgical contralateral mean volumes of hippocampus, amygdala, temporal pole, parahippocampal gyrus, and posterior segment of the temporal lobe in unilateral TLE patients. Methods: Sixty-nine consecutive patients and 23 normal controls were analyzed with magnetic resonance imaging (MRI) volumetry. All patients came from the Epilepsy Surgery of the Ribeir˜ao Preto Medical School, and were operated on by anterior temporal lobectomy with adequate resection of mesial and neocortical structures, after comprehensive presurgical evaluation. All patients had mesial temporal sclerosis (MTS) by pathological examination. Results: Our data show significant difference of the mean volumes of temporal pole, parahippocampal gyrus, and hippocampus (p < 0.05) between patients with surgery on the left or right side and controls. There was no difference between the mean volumes of amygdala and the posterior segment of the temporal lobe. Conclusions: MRI volumetric data show that damage in TLE due to MTS may be more widespread, even in patients with unilateral TLE by clinical and neurophysiological criteria.
1.207 THE VALUE OF ACCURATE ANATOMIC ASSESSMENT ON VOLUMETRIC ANALYSIS OF THE AMYGDALA Leonardo Bonilha, Eliane Kobayashi, Fernando Cendes, and Li Min Li (Laboratory of Neuroimaging, State University of Campinas, Campinas, Sao Paulo, Brazil) Rationale: There is a wide range of reported values from volumetric studies of the amygdala. The use of single-plane-thick magnetic resonance imaging (MRI) images may prevent the correct visualization of anatomic landmarks and yield imprecise results. We aimed to assess whether there is significant difference between volumetric analysis of the amygdala performed with single-plane MRI 3-mm slices and with multiplanar analysis of MRI 1-mm slices. Methods: We studied healthy subjects and patients with temporal lobe epilepsy. We performed manual delineation of the amygdala on T1 -weighted IR, 3-mm coronal slices, and manual delineation of the amygdala on three-dimensional volumetric T1 -weighted images with 1mm slice thickness. The data were compared using dependent t test. Statistic level of significance was set at p < 0.05. Results: Dependent t test showed difference between the volumes obtained by the coronal plane–based measurements and the volumes obtained by the three-dimensional analysis of the right and left amygdala among controls (right amygdala, t(14) = 6.832, p < 0.001; left amygdala, t(14) = 4.814, p < 0.001); of the right and left amygdala among patients with right TLE (right amygdala, t(5) = 2.668, p < 0.05; left amygdala, t(5) = 3.328, p < 0.05); and of the right and left amygdala among patients with left TLE (right amygdala, t(8) = 5.756, p < 0.001; left amygdala, t(8) = 2.956, p < 0.05). Conclusions: The incorrect estimative of the amygdala volume may preclude the correct analysis of the biologic effects of alterations in the amygdaloid volume. The three-dimensional analysis should be preferred because it is based on more extensive anatomic assessment and closer to postmortem studies. [Supported by FAPESP (00/04710-2).]
1.208 SUBCORTICAL ABNORMALITIES IN PARTIAL EPILEPSY: A STUDY USING T2 -Relaxometry At 3 Tesla 1,2 Regula S. Briellmann, 1 Gaby S. Pell, 1 Ari Syngeniotis, 1 Steve Fleming, 1 David F. Abbott, 1,3 L. Anne Mitchell, and 1,2 Graeme D. Jack-
Epilepsia, Vol. 44, Suppl. 9, 2003
son [1 Brain Research Institute, Austin and Repatriation Medical Centre, Heidelberg West, and 2 Department of Medicine (Neurology) and 3 Department of Radiology, University of Melbourne, Parkville, Victoria, Australia] Rationale: There is increasing evidence that brain abnormalities in focal epilepsy are not restricted to the seizure focus area. T2 relaxometry has been established as a reliable tool for the assessment of MR signal changes, particularly in patients with partial epilepsy associated with hippocampal sclerosis (HS). Here we use T2 -relaxometry to investigate abnormalities in the mesial temporal lobe and subcortical nuclei in partial epilepsy. Methods: T2 -maps were acquired in 45 patients and 50 controls. Nineteen patients had HS, 14 had malformations of cortical development (MCD), and 12 had no obvious abnormalities on anatomic MR (normal MR). Ten coronal slices with eight images per location (CPMG, TE, 28– 231ms; TR, 5,000 ms) were acquired on a 3-T GE LX Horizon scanner. T2 maps were generated using a proprietary GE software package (Functool). Nine bilateral regions of interest (ROIs) were measured (hippocampus, anterior temporal lobe, amygdala, frontal and parietal white matter, caudate, putamen, pallidum, thalamus). All findings were expressed as percentage of control values [100/(mean control T2 -measurement] × (individual patient T2 -measurement). Level of significance was set at 1%. Results: In controls, there was no difference between corresponding right- and left-hemispheric ROIs, and there was no gender difference. T2-relaxometry values were increased in the ipsilateral hippocampus in all three patient groups (HS, 121 ± 7%; MCD, 106 ± 8%; normal MR, 105 ± 6%). Only patients with normal MR had a contralateral hippocampal signal increase as well (103 ± 7%). HS patients also showed signal increase in the amygdala (104 ± 6%) and anterior temporal lobe (105 ± 6%), but in none of the other assessed regions. Patients with MCD also showed signal increase in the anterior temporal lobe (104 ± 5%), caudate (106 ± 10%), pallidum (112 ± 12%), and a trend for signal increase in further four ROIs (amygdala, parietal lobe white matter, thalamus, putamen). Patients with normal MR also showed signal increase in the putamen (106 ± 7%) and a trend for it in the caudate. The AI was different in patients with HS (−0.16 ± 0.04; p < 0.0001) compared with controls (0.00 ± 0.04), but in none of the other regions. Conclusions: Abnormalities in the mesial temporal structures were found not only in HS patients, but also in patients with MCD or normal MR. Therefore, refractory focal epilepsy is commonly associated with signal change in the mesial temporal structures. Subcortical nuclei showed signal increase in patients with MCD and to a lesser degree in patients normal MR, but not in HS. This implies that the nature of the associated MR abnormality influences the degree of involvement of subcortical structures. Patients with HS may have a more localized disease process, compared with patients with MCD or normal MR. (Supported by National Health and Medical Research Council, Neurosciences Victoria, and Brain Imaging Research Foundation, Australia.)
1.209 RELEVANCE OF TEMPORAL POLE ABNORMALITIES INVESTIGATED BY CORONAL FLAIR IMAGES IN MESIAL TEMPORAL SCLEROSIS Henrique Carrete, Jr., Ivanilson A. Oliveira, Fl´avia S. Miyashira, Eliana Garzon, Am´erico C. Sakamoto, and Elza M.T. Yacubian (Unidade de Pesquisa e Tratamento das Epilepsias, Escola Paulista de Medicina, Universidade Federal de S˜ao Paulo, S˜ao Paulo, S˜ao Paulo, Brazil) Rationale: Abnormalities of the anterior temporal lobe in patients with intractable temporal lobe epilepsy and mesial temporal sclerosis (MTS) have been discussed in the literature. The cause of these magnetic resonance imaging (MRI) abnormalities has not been established. The aim of this study is to evaluate the diagnostic relevance of loss of the gray–white matter differentiation (LGWD) in the temporal pole (TP) on coronal fluid-attenuated inversion recovery (FLAIR) imaging in patients with MTS and age-matched control subjects. Methods: Sixty-five patients with MTS visually detected by MRI and 20 normal control subjects were included in this study. Only the first four or five coronal FLAIR images of the TP were visually assessed by two experienced neuroradiologists. Observers were unaware of lateralizing
AES PROCEEDINGS electroclinical data, and the images did not include the hippocampus. Patients and control subjects were divided into two groups according to LGWD on FLAIR images. This abnormality consists of an indistinct gray–white matter demarcation and/or an increased signal intensity of the TP. The topography of the LGWD in the TP was also evaluated, defined as occurring in the anteromedial region, in the lateral region, or both. The associations between LGWD and MRI signs of hippocampal sclerosis, history of initial precipitant insults (IPIs), age at onset, seizure frequency, and duration of illness were surveyed. Results: The MRI feature of LGWD was observed in 48 (74%) of the 65 patients with MTS. There was no LGWD in the 20 normal subjects. Considering the unilateral hippocampal abnormalities, present in 59 patients (91%), they were found in the same temporal lobe in all patients with unilateral LGWD. Bilateral hippocampal sclerosis was detected in six patients (9%), and in five of them (83%) we found LGWD ipsilateral to the more affected hippocampus. The only case in which both hippocampi were affected equally, LGWD was found on the side pointed out by electroclinical data. All 48 patients with LGWD had the abnormalities in the medial region of TP, being purely medial in 35 (73%) and involving both medial and lateral regions in 13 (27%) of 48. Presence of IPIs, represented mainly by febrile seizures, was correlated with LGWD because it was found in 21of 48 patients (44%) with LGWD and in five of 17 (29%) of those without. LGWD failed to correlate with age at onset, seizure frequency, duration of illness, and region of involvement of the TP. Conclusions: The MRI feature of LGWD can improve the visual diagnosis of MTS, and it is a valuable additive sign for correct seizure lateralization, particularly using coronal FLAIR sequences that include the anterior temporal lobe in full. In this series, LGWD had a medial predominance and was related to the presence of IPI. (Sponsored by Funda¸ca˜ o de Amparo a` Pesquisa do Estado de S˜ao Paulo.)
1.210 SEMIAUTOMATED CORTICAL TUBER IDENTIFICATION IN TUBEROUS SCLEROSIS: HIGH-DEFINITION IMAGING AT 3 TESLA WITH SURFACE COILS 1 Colin P. Doherty, 2 P. Ellen Grant, 1 Bradford C. Dickerson, 2 Evalina Busa, and 1 Elizabeth A. Thiele (1 Pediatric Epilepsy Service, Department of Neurology, Massachusetts General Hospital and Harvard Medical School, Boston, and 2 MGH/MIT/HMS Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital NMR facility, Charlestown, MA) Rationale: Cortical tuber counting in tuberous sclerosis complex (TSC) has been suggested as a biomarker of disease burden. Most techniques use simple manual counting methods using high-resolution T2 weighted fast spin echo (FSE) or fluid attenuated inversion recovery (FLAIR) sequences on conventional 1.5 Tesla imaging machines. We chose to employ higher field strengths and whole head phased array surface coils for better image resolution and applied cortical reconstruction techniques in post-processing for semi-automated identification of cortical tubers. Methods: Eight adults with a diagnosis of TSC were imaged using a 3Tesla Siemen’s ‘Trio’ whole body magnetic resonance scanner together with locally designed phased array surface coils. Detailed T1 -weighted FSE scans and two 3D sagittal MPRAGE scans were collected along with coronal and axial FLAIR images. The two 3D acquisitions were averaged in the first postprocessing step to reduce further signal-to-noise ambiguities. Novel image-processing software (FREESURFER) was then used for brain inflation and cortical reconstruction of the MPRAGE scans, which resulted in cortical-thickness measures. Cortical tubers were automatically identified using a statistical thickness map, and this was compared with manual counting on the FLAIR images using the same rater in a blinded technique. Results: Detailed T1 -weighted images at 3 Tesla using the phased array coils visualized cortical tubers at a resolution unmatched by even pathological samples. Using a threshold significance level of 0.5 for cortex >0.4 cm in thickness there was a high correlation (r = 0.92) between the number of tubers identified using the semiautomated and manual techniques. Results further correlated with detailed clinical, neurophysiologic, molecular, and neuropsychologic data.
75
Conclusions: We suggest that such enhanced techniques for tuber identification will refine the current specificity of cortical tuber count for disease status, surveillance, and prognosis. The ability to so clearly distinguish dysplastic from normal tissue argues for wider use of 3T imaging and phased array coils for the study of the natural history of TSC. (Supported by Tuberous Sclerosis Alliance Innovative Science Research Award to Elizabeth A. Thiele M.D., Ph.D.)
1.211 THE ROLE OF THE ENTORHINAL CORTEX IN THE PATHOGENESIS OF TEMPORAL LOBE EPILEPSY: A QUANTITATIVE MRI STUDY Mary Fitzsimons, Kevin Murphy, Colin Doherty, and Norman Delanty (Epilepsy Programme, Beaumont Hospital, Dublin, Ireland) Rationale: Reciprocal glutamatergic pathways of the entorhinal cortex (EC) connect the hippocampus to the surrounding cortex. Damage to the EC is associated with increased risk of seizure generation. Neuronal loss of layer III of the EC has been reported in surgical specimens from patients with temporal lobe epilepsy (TLE). Similarly, its strategic position and anatomical connections may implicate it in the propagation of seizures from the hippocampus to extrahippocampal areas. Atrophic change may be determined from magnetic resonance imaging (MRI)based volume estimation. We studied a group of patients with TLE to establish the frequency of coexisting EC and hippocampal volume loss. Methods: High-resolution 3D brain image data was acquired from 20 normal control subjects and 20 patients with TLE using a SPGR MRI imaging sequence. Standard anatomic landmarks were used to identify boundaries of the hippocampus and entorhinal cortex on which in vivo volumes were estimated from the image data using a computer-based stereologic technique. Results: Normal control volumes were estimated at 3.30 (±0.38)cc (mean ± SD) on the right and 3.13 (±0.35) cc on the left for the hippocampus, 1.06 (±0.16) ccon the right and 1.04 (±0.15) cc on the left for the EC. 70% of the TLE patients had ipsilateral hippocampal volume loss. Of these, 60% had coexisting ipsilateral EC volume loss, whereas the remainder had normal EC volumes. Conclusions: Results support the existence of a spectrum of changes that range from isolated hippocampal atrophy, isolated EC atrophy, and coexisting hippocampal and EC atrophy. The position on the spectrum may influence seizure frequency and severity and may have implications for treatment. The pattern of atrophy demonstrated raises questions about the role of the EC in the pathogenesis of TLE. Further elucidation of these questions requires larger studies and clinical correlation.
1.212 HIGH-DIMENSIONAL MAPPING OF THE HIPPOCAMPUS IN MTS COMPARED WITH CONTROLS 1 R. Edward Hogan, 2 Lei Wang, 3 Richard D. Bucholz, 1 L. James Willmore, 1 Mary E. Bertrand, and 2 John G. Csernansky (1 Neurology, 2 Psychiatry, and 3 Division of Neurosurgery, Saint Louis University, St. Louis, MO) Rationale: Magnetic resonance imaging (MRI)-based high dimensional mapping (HDM) in patients with mesial temporal sclerosis (MTS) demonstrates regional hippocampal surface anatomy changes, which are likely influenced by normal right–left hippocampal asymmetries and the underlying histopathology of MTS. Using HDM, we compare MTS subjects with controls to investigate the influence of hemispheric asymmetries in the hippocampal surface anatomy changes of MTS, and compare surface anatomy changes with the histopathology of MTS. Methods: The epilepsy surgery series at Saint Louis University was consecutively reviewed for subjects with histopathologic confirmation of MTS, and MR-based hippocampal volume (HV) loss (with >10% asymmetry) concordant with the side of MTS. A group of control subjects, with no history of CNS disease, also had MR studies. All MRs were performed on the same MR scanner. MRs from epilepsy subjects were grouped according to the side of MTS. The right- and left-MTS groups (each considered independently) were matched for both average age and intracranial size with a group of 10 normal control subjects. Hippocampal deformation segmentations and comparisons were
Epilepsia, Vol. 44, Suppl. 9, 2003
76
AES PROCEEDINGS
performed as previously described (Wang et al. Neuroimage 2001;14:531–45), calculating right- and left-sided differences in MTS and control groups, then comparing asymmetries of groups. This enables discrimination of effects related to normal right/left asymmetries from those of MTS. For consistency in measuring eigenvector functions predictive of three-dimensional surface structure changes, right to left asymmetry was measured from both RMTS and LMTS groups. Final deformation results were projected on a representative hippocampal surface for each group. Results: There were 15 subjects in both RMTS and LMTS groups, and 10 subjects in each of the matching control groups. In analysis of HV and hippocampal shape (using the first 9 eigenvectors), using repeated measures ANOVA, results showed p < 0.0001 for both LMTS versus controls and RMTS versus controls. Analysis for best subset of eigenvectors to discriminate between groups, using logistic regression, yielded eigenvectors 1, 2 for LMTS vs. control, and 1, 3 for the RMTS vs. control, suggestive that there are subtle differences in shape changes in LMTS subjects as compared with RMTS subjects. Visual inspection of surface anatomy changes showed similar patterns of regional volume loss in both the LMTS and RMTS groups as compared with controls, with the most accentuated involvement of the region correlating with Sommer sector in the hippocampal body. Conclusions: HDM shows similar patterns of involvement in hippocampal surface anatomy in both RMTS and LMTS. Accentuated involvement in surface anatomy correlates with the underlying pathology of MTS (Table 1).
TABLE 1. Hippocampal volumes
R MTS R MTS controls L MTS L MTS controls
Left HV (mm3 )/SD
Right HV (mm3 )/SD
2,568/363 2,937/268 1,556/215 2,786/346
1,954/404 3,336/343 2,766/468 3,135/429 causes and consequences of cerebellar atrophy may be more revealing when segregated based on these patterns of atrophy (Figs. 1 and 2).
1.213 THREE MRI PATTERNS OF CEREBELLAR ATROPHY IN PEOPLE WITH EPILEPSY Michel J. Berg, Virginia P. Moreno, and Lynn C. Liu (Department of Neurology; Strong Epilepsy Center, University of Rochester, Rochester, NY) Rationale: Cerebellar atrophy occurs to various degrees in people with epilepsy. Usually it is attributed to chronic medication use, although this mechanism has been debated. Coordination dysfunction has been correlated to cerebellar atrophy. Cerebellar atrophy has been postulated to worsen seizure control and severity as the cerebellum is generally thought to be an inhibitory structure. Methods: We investigated 57 consecutive patients with available magnetic resonance imaging (MRIs) and video-EEG monitoring proven epilepsy. The degree of cerebellar atrophy was determined by making seven measurements on each of three coronal views (anterior, mid, and posterior) and two measurements on each of three sagittal views (mid left, midline, and mid right). The measurements were designed to define the widest distances between the folia and the degree of separation of the cerebellar margin from the dura. Values of zero were used as controls (i.e., no measurable distance between the folia or cerebellar margin should be present on the MRI.). Results: We observed three distinct MRI patterns of cerebellar atrophy in people with epilepsy: (a) folia atrophy (Fig. 1), (b) global atrophy without folia involvement (Fig. 2), and (c) a combination of both folia and global atrophy. Of the 57 patients, 10 (18%) had atrophy affecting only the folia, two (4%) had global atrophy alone, and eight (14%) had both folia and global atrophy. These different patterns of cerebellar atrophy would not have been apparent using total cerebellar volume measurements. Conclusions: Three patterns of cerebellar atrophy occur in people with epilepsy: (a) folia atrophy, (b) global atrophy without individual folia involvement, and (c) folia and global atrophy. Future analysis of the
Epilepsia, Vol. 44, Suppl. 9, 2003
1.214 UTILITY OF MRI IN ADULT FIRST SEIZURE PATIENTS: DEVELOPMENT OF A CLINICAL ALGORITHM 1 Adrian J. Lowe, 1 Christine J. Kilpatrick, 1 Zelko Matkovic, 2 Donald C. Campbell, and 1 Terence J. O’Brien (1 Department of Neurology and 2 Clinical Epidemiology and Health Service Evaluation Unit, Royal Melbourne Hospital, Melbourne, Victoria, Australia) Rationale: Magnetic resonance imaging (MRI) is better than computed tomography (CT) at identifying structural abnormalities that are relevant to epilepsy, but MRI is more expensive and not immediately available in many regions. A first epileptic seizure is a common presentation, but little research has investigated the utility of MRI in this population, and in particular its additional value to that of CT, and its ability to predict seizure recurrence. Methods: A prospective cohort study, based in a hospital-based First Seizure Clinic, was conducted. For 105 consecutive adults presenting with a first seizure, the frequency that MRI added value to CT was examined, and a logistic regression was performed to determine if clinical variables could predict when this would occur. The ability of neuroimaging to predict seizure recurrence was assessed using Kaplan–Meier curves. Results: MRI provided useful information additional to CT in 16.2% of patients, identifying six lesions missed by CT, and clarifying the nature of the CT lesion in 11, including three CT lesions that were shown to be artifactual. MRI provided additional information more often in patients who had a CT showing a lesion that required clarification (p = 0.001), a partial-onset seizure (p = 0.002), or a history of head-injury (p = 0.009), and significantly less often in patients with a family history of epilepsy (p = 0.046). If MRI was only offered to patients with a CT lesion that required clarification; or a partial-onset seizure; or a history of head injury, then only 28.5% of patients would have received a MRI and 53% (16 of 30) of these MRIs would have added value to the CT. Of the patients not offered a MRI, only 1.3% (one of 75) would have gained
AES PROCEEDINGS value from this test. Neuroimaging did not predict seizure recurrence (p = 0.78). Conclusions: The results of this study identify a subgroup of patients in whom MRI has a high yield for adding value to CT, and indicates that MRI is not necessary in all patients with a first seizure. (Supported by Department of Human Services, Victoria, Australia.) 1.215 FOCAL LESIONS IN THE SPLENIUM OF THE CORPUS CALLOSUM AFTER ANTIEPILEPTIC DRUG THERAPY MODIFICATION Darin T. Okuda, Steve S. Chung, Joseph E. Heiserman, John F. Kerrigan III, and David M. Treiman (Neurology, Barrow Neurological Institute, Phoenix, AZ; Neurology, Barrow Neurological Institute, Phoenix, AZ; Radiology, Barrow Neurological Institute, Phoenix, AZ; Child Neurology, Barrow Neurological Institute, Phoenix, AZ; Neurology, Barrow Neurological Institute, Phoenix, AZ) Rationale: Six cases involving transient focal lesions in the splenium of the corpus callosum (SCC) after the modification of antiepileptic drug therapy are reported. Discrete, noncontrast enhancing, wellcircumscribed, nonhemorrhagic lesions in the SCC are a rare finding occurring in patients with epilepsy. The etiology and pathogenesis of these lesions are currently unknown. Current hypotheses include transient interictal focal white matter edema associated with transhemispheric connection of seizure activity, antiepileptic drug toxicity, demyelination, and alteration of vasopressin concentrations in addition to vitamin deficiencies. However, a consistent relation between these factors and its occurrence has not been observed. Methods: We reviewed clinical material on six patients who were found to have focal lesions involving the SCC on magnetic resonance imaging (MRI) of the brain. Results: Six patients of white, Hispanic, and Native American ethnicities with ages ranging from 27 to 33 were studied. Four of the six patients were admitted to the epilepsy monitoring unit where carbamazepine, phenytoin, and lamotrigine were tapered or discontinued for seizure monitoring. One patient was admitted for neutropenia with oxcarbazepine, phenytoin, and lamotrigine discontinued and valproic acid started, the other for seizure activity related to poor medication compliance with levetiracetam. MRI of the brain was obtained from 1 to 10 days after the last seizure event in all patients. MRI of the brain demonstrated focal hypointensities on T1 - and hyperintensities on T2 -weighted images in the SCC. Contrast enhancement was not observed in all lesions and dissociative symptoms were not observed in all patients. Of those individuals undergoing repeated MRI, complete resolution of the lesions in the SCC were seen within 5 days to 1.5 months after initial identification. Conclusions: Lesions involving the SCC in patients with epilepsy require no invasive diagnostic procedures or treatment. Repeated MRI brain scans demonstrated spontaneous and complete resolution of lesions within days to months after initial identification. Reasons for these changes are currently not well understood; however, recent seizure activity in addition to acute changes in the antiepileptic drug regimen may be responsible for its occurrence. Determining why these lesions occur may provide a better understanding of the functional and anatomic significance of the corpus callosum in epilepsy. Additional studies assessing the occurrence of these lesions are needed. 1.216 MRI-BASED CORTICAL SURFACE AREA MEASUREMENTS USING EXHAUSTIVE VERTICAL SECTIONS 1 Lisa Ronan, 2 Norman Delanty, 2 Colin Doherty, and 1 Mary Fitzsimons (1 Epilepsy Programme and 2 Neurology, Beaumont Hospital, Dublin, Ireland) Rationale: The folding or gyrification of the cerebral cortex is a developmental process, allowing a large cortical surface area and interneuron connectivity without a disproportionate increase in cranial size. Cortical developmental malformations (CDMs) may give rise to disruptions in the folding process. The gyrification index (GI), a measure of the degree of cortical folding, is defined as the ratio of the total cortical surface area to superficially exposed surface area. With the advent of magnetic
77
resonance imaging (MRI), CDMs are increasingly recognised as a cause of epilepsy. An abnormal GI may be a surrogate for subtle CDMs. Direct visual inspection of two-dimensional images of patients with partial epilepsy show that ∼25% are normal, without the use of additional quantitative measurement techniques. Inclusion of in vivo MRI-based GI measurements may enhance diagnostic imaging techniques. The stereologic technique of exhaustive vertical sections allows mathematically unbiased estimates of a 3D structure’s surface area based on 2D image slices. The method involves systematic sampling of the surface area around the axis of a fixed vertical plane. We describe the method and application of the stereologic technique of exhaustive vertical sections to MR brain images to produce unbiased estimates of cortical surface area. This technique may ultimately be used for in vivo measurement of the gyrification index, which to date has not been reported for an epilepsy population. Methods: High-resolution 3D SPGR MRI image data of a water-filled phantom was used to calibrate and test the method of exhaustive vertical sections. The same MRI sequence was used to acquire brain images from five normal control subjects. Rules were established for identification of the pial layer on these images, on which cortical surface areas were based. Using the computerised stereologic method of vertical sections, experiments were carried out to establish appropriate sampling rules to achieve accurate, precise, and repeatable measurements. Results: From repeated measurements, the surface area estimated from the phantom image data measured 340.78 cm2 ± 10.3% (mean ± CE), which compared well (±4.97%) to actual surface area of 358.6 cm2 . Surface area measurements from brain data (1,886–2,179 cm2 ) were found to be accurate within experimental error, which ranged consistently between 10.3% and 10.8%. Conclusions: Robust rules have been established for stereologic surface area estimation based on the pial layer. Our experiments show that the technique of exhaustive vertical sections is reliable and accurate. All brain measurements were carried out under identical stereologic conditions of sampling slice separation and orientation. Further experiments are required to optimise this method in cortical surface area estimation. The ultimate application of the technique to the measurement of GI is epilepsy may be of clinical utility. (Supported by the Irish Brain Research Foundation.)
1.217 MAGNETIZATION TRANSFER IMAGING DETECTS ABNORMALITIES IN FOCAL EPILEPSY PATIENTS WITH NORMAL CONVENTIONAL BRAIN MRI 1 Tuuli M. Salmenpera, 1 Mark R. Symms, 1 Fergus J. Rugg-Gunn, 1 Philip A. Boulby, 2 Gareth J. Barker, 3 Tarek A. Yousry, and 1 John S. Duncan (1 MRI Unit, The National Society for Epilepsy, Chalfont St Peter, Buckinghamshire, and 2 Neuroimaging Research Group, Institute of Psychiatry, and 3 Lysholm Department of Neuroradiology, The National Hospital for Neurology and Neurosurgery, London, United Kingdom) Rationale: Thirty percent of all patients with epilepsy develop refractory seizures, and are therefore candidates for surgical treatment. However, no relevant structural pathology is identified in 20% of these patients with current conventional optimal magnetic resonance imaging (MRI) scans, comprising T1 - and T2 -weighted, proton density (PD), and FLAIR images. Magnetization transfer imaging (MTI) is a new imaging technique that visualises protons tightly bound to macromolecular structures. The exchange of magnetization between bound protons and free water is quantified by magnetization transfer ratio (MTR), which provides a measure of macromolecular structural integrity. Our objective was to examine whether MTI can identify areas of abnormal MTR in patients with intractable focal epilepsy and normal conventional MRI. Methods: Fifty-two patients with focal epilepsy and unremarkable conventional MRI scans and ictal videotelemetry recordings were scanned with MTI. Statistical parametric mapping was used to compare the cerebral structure of an individual patient with a template created from the MTI data of 30 control subjects. Results: MTR was significantly reduced in 22 of the 52 intractable focal epilepsy patients with no visual abnormality on the conventional MRI. Additionally, four of the 52 patients showed an MTR signal increase. Proton density was decreased in two of the 22 patients with
Epilepsia, Vol. 44, Suppl. 9, 2003
78
AES PROCEEDINGS
abnormally decreased MTR. In one of these patients, the decreased proton density colocalized with the decrease in MTR signal. Conclusions: Observed decreases in MTR may reflect either focal lesion in cerebral tissue or localized atrophic change. Our findings suggest that advanced MRI sequences such as MTI are promising techniques for identifying cerebral lesions that were previously not demonstrable with MRI in patients with intractable focal epilepsy. The successful identification of a focus aids the assessment of possible surgical treatment in these patients. (Supported by Academy of Finland, Action Research, Emil Aaltonen Foundation, Vaajasalo Foundation.)
1.218 TEMPORAL LOBE WHITE-MATTER SIGNAL CHANGES IN TEMPORAL LOBE EPILEPSY: A T2 Relaxometry Study 1 Tiffany N. Townsend, 1,2 Neda Bernasconi, 1,2 G. Bruce Pike, 1 Frederick Andermann, and 1,2 Andrea Bernasconi (1 Neurology & Neurosurgery, Montreal Neurological Institute, and 2 Brain Imaging Center, McGill University, Montreal, Quebec, Canada) Rationale: Visual inspection of magnetic resonance imaging (MRI) reveals increased T2 signal abnormalities in the temporopolar region ipsilateral to the seizure focus in 30–60% of temporal lobe epilepsy (TLE). T2 signal changes posterior to the temporal pole, as well as bilateral T2 signal changes, are difficult to appreciate visually and have not been explored quantitatively. In addition, the relation of these abnormalities to hippocampal changes found in TLE has not been adequately investigated. The purpose of this study was to assess temporal lobe white matter (WM) T2 -signal changes quantitatively using T2 relaxometry, and to examine the relation of these changes to hippocampal volume and hippocampal T2 signal abnormalities in patients with medically intractable TLE. Methods: We studied 56 consecutive TLE patients (mean age, 35 years) and 30 neurologically normal controls. For computation of T2 values, images were acquired using a conventional dual-echo spinecho pulse sequence (TE, 20, 120 ms; TR, 3,240 ms; slice thickness 5 mm with 0.5 mm interslice gap) on a 1.5-T scanner. Twenty-three contiguous oblique coronal slices covering the whole brain with a voxel size of 0.94 × 0.94 mm2 were acquired. Slices were oriented orthogonal to the axis of the hippocampal body. For each individual, single-exponential-decay equations were fit to the T2 imaging data obtained in each pixel, and then T2 relaxation times were calculated for each pixel. For each slice, a map was then constructed in which pixel-intensity corresponded to the calculated T2 relaxation time. Averages of six slices were used to calculate left and right hippocampal T2 relaxation times (Hippo-T2 ) and left and right temporal lobe WM T2 relaxation times (WM-T2 ). Mean T2 relaxation time was calculated in each ROI. Finally, a total mean WMT2 and Hippo-T2 was calculated by averaging the values of all ROIs in all slices. Volumetric MRI of the hippocampus showed unilateral hippocampal atrophy (HA) in 27 patients, whereas 29 patients had normal hippocampal volumes (NV). Results: WM-T2 was increased ipsilateral to the seizure focus in TLE patients with HA and those with NV (p < 0.001). Contralateral WMT2 was increased in left and right TLE with HA (p < 0.001) and in right TLE with NV (p = 0.002). There was no difference in the mean ipsilateral WM-T2 between patients with HA and those with NV. We found a positive correlation between Hippo-T2 and WM-T2 ipsilateral (p < 0.001) and contralateral (p = 0.01) to the seizure focus. There was no correlation between WM-T2 and duration of epilepsy. There was no difference in mean WM-T2 between patients with a history of febrile convulsions and those without, and patients with frequent or rare/no secondary generalized seizures. Conclusions: T2 relaxometry demonstrates a bilateral increase in T2 relaxation time throughout the temporal lobe WM in patients with intractable TLE. These abnormalities may occur regardless of the presence of HA.
1.219 THE ASSOCIATION OF MALFORMATIONS OF CORTICAL DEVELOPMENT OCCURRING AT DIFFERENT STAGES OF THE EMBRYONIC PROCESS 1 Kette Valente, 2 Marcelo Valente, 1 Alessandra Freitas, 3 Rosa Valerio, and 2 Claudia Leite (1 Psychiatry, 2 Radiology, and 3 Neurology, University of Sao Paulo, Sao Paulo, SP, Brazil) Epilepsia, Vol. 44, Suppl. 9, 2003
Rationale: Development of the human brain occurs in stages that encompass proliferation and differentiation, migration, and cortical organization. These organized stages take place at different moments of the embryonic development. Interruption of this process may result in different types of malformations of cortical development (MCDs). It is believed that the type of MCD is stage-related, being determined by the moment of disruption. We describe the neuroimaging findings of a series with MCD and document the association of more than one MCD occurring at different moments of embryogenesis. Methods: Magnetic resonance imaging (MRIs) were reviewed by two neuroradiologists, blinded to clinical details, to determine the imaging features and classify according to the main MCD according to Barkovich et al. MCD were classified as proliferative, migrational, and organizational in nature. Clinical information was prospectively collected with a standard questionnaire and by review of hospital records. Results: Our group consisted of 74 patients with epilepsy determined by MCD, mean age, 8 years; 33.8% had a malformation of proliferation or differentiation, 31.1% of migration, and 35.1% of organization. The main malformations were focal cortical dysplasia (29.7%), polymicrogyria (27%), nodular heterotopia (16.4%), schizencephaly/cleft (8.1%), transmantle dysplasia (5.4%), subcortical band heterotopia (5.4%), hemimegalencephaly (4.1%), and pachygyria/lissencephaly (4.1%). A total of 42 patients (56.8%) showed one abnormality (cortical or subcortical) other than the most important for classification. A total of 33 (44.6%) patients had other cerebral malformations including commissural dysgenesis (20.3%); posterior fossa malformation (18.9%, affecting the cerebellum in 12 patients); ventricular dilatation or dysmorphism (14.9%), and white matter alterations (6.8%), excluding the hypersignal observed in FCD. Five of these patients had associated hipocampal atrophy. Fifteen patients (20.3%) had more than one MCD consistent with defects occurring at more than one developmental stage. From this group, nodular heterotopia was the most frequently (50%) associated with abnormalities in the overlying cortex. Commissural abnormalities occurred in 38.5% of patients with polymicrogyria/schizencephaly and cerebellar dysgenesis in 16% of patients with nodular heterotopia. Conclusions: This study illustrates that the spectrum of MCD is not completely well defined and that many patients may have more than one malformation occurring at different stages of the embryonic process. Because it is highly unlikely that two insults have occurred at two different moments in this expressive number of patients, this process may not be restricted to one specific moment, and when the process goes wrong it may affect different stages. It remains to be determined whether a genetic predisposition may underlie a more complex response observed in some of these patients.
1.220 LOCALIZATION OF INVASIVE ELECTRODES IN RELATION TO CORTICAL SURFACE ANATOMY 1 Alexandre Bastos, 2 Blaise M. Wheatley, 3 Roch Comeau, and 4 Donald W. Gross (1 Department of Radiology and 2 Department of Surgery, University of Alberta, Edmonton, Alberta; 3 Rogue Research Inc., Montreal, Quebec; and 4 Department of Internal Medicine, University of Alberta, Edmonton, AB, Canada) Rationale: Subdural grid electrodes are commonly used for localization of the seizure generator and cortical mapping in epilepsy patients being considered for resective surgery. A potential drawback of this method, however, relates to the difficulty in identifying the position of electrodes in relation to brain surface anatomy. Previous studies using curvilinear multiplanar reformatting (CMPR) have successfully addressed this issue. Nonetheless, local fluid collection and brain displacement still may prevent satisfactory electrode demonstration. This study aims to further improve grid localization by using technical advances in image acquisition and processing. Methods: Two patients who were investigated with subdural grid electrodes had an magnetic resonance imaging (MRI) examination after electrode placement. A 3.0-mm-thick T2− weighted gradient-echo sequence was acquired along the axial plane, with no gap. The images were transferred to a workstation and resampled into 1-mm thickness. CMPR was then performed using a software implementation that provided minimal intensity projection of curved image slabs. In a subsequent step, the electrodes identified on the grid MRI dataset were registered into a preimplantation volumetric MRI.
AES PROCEEDINGS Results: In both patients, the images obtained allowed accurate localization of the electrode grid, which was confirmed during surgery for removal of epileptogenic tissue. Conclusions: The developed implementations helped to overcome common technical difficulties associated with demonstration of intracranial electrodes. Because of its simplicity and effectiveness, this method has become a routine procedure at our institution, allowing accurate localization of grid electrodes thus enhancing interpretation of neurophysiologic data.
1.221 FDG-PET HYPOMETABOLISM IS ASSOCIATED WITH GOOD SURGICAL OUTCOME IN TEMPORAL LOBE EPILEPSY WITH NORMAL MRI Ebru E. Altay, James A. Fessler, Victoria P. Vahle, Hrayr P. Attarian, and Frank G. Gilliam (Neurology, Washington University, St. Louis, MO) Rationale: Although mesial temporal sclerosis is a good predictor for surgical outcome in patients with temporal lobe epilepsy, there is no well-known predeterminant for a group of patients with normal magnetic resonance imaging (MRI) findings but mesial temporal lobe seizures. Methods: We studied 19 consecutive patients who had normal MRI but mesial temporal lobe seizures defined by clinical and EEG findings from scalp and/or intracranial video/EEG monitoring. Fluorodeoxyglucose–positron emission tomography (FDG-PET) scans were performed for each patient as a part of standard phase 1 evaluation for epilepsy surgery. The presence of interictal regional delta slowing was also determined from the scalp recordings. All patients had subsequent epilepsy surgery and have been followed up for ≥6 months. Data were stored in SPSS database and analysed by using the χ 2 statistics. Results: The presence or the severity of PET hypometabolism ipsilateral to the resection side was significantly correlated with good surgical outcome with absence of seizures for ≥6 months (p < 0.05). Twelve of 14 patients who were seizure free after surgery had ipsilateral PET hypometabolism. Of five patients with continued seizures, one patient had ipsilateral hypometabolism on the PET. However, the presence of interictal regional delta slowing ipsilateral to the surgery side did not reveal a significant correlation. Conclusions: The presence and the severity of PET hypometabolism is closely correlated with good surgical outcome in mesial TLE patients with normal MRI findings. This observation may indicate that PET hypometabolism can be an independent predictor for surgical outcome in patients with mesial temporal lobe epilepsy.
1.222 TAYLOR-TYPE FOCAL CORTICAL DYSPLASIAS WITH SUBTLE MRI ABNORMALITIES: USEFULNESS OF 18 FDG-PET 1,2,3 Francine Chassoux, 2 Frederic Beuvon, 3,4 Lucie Hertz-Pannier, 1 Bertrand C. Devaux, 1 Elisabeth Landre, 1 Baris Turak, 1 Michael Mann, 3 Elisa Lopez, and 3 Franck Semah (1 Department of Neurosurgery and 2 Department of Neuropathology, Sainte-Anne Hospital, Paris, 3 SHFJ, CEA, Orsay, and 4 Pediatric Radiology, Necker-Enfants Malades, Paris, France) Rationale: Detection of Taylor-type focal cortical dysplasias (TTFCDs) with 18 FDG-PET has been reported in early series, but its usefulness has not been evaluated in cases with subtle or even absent abnormalities at current optimal magnetic resonance imaging (MRI).The aim of the study was to evaluate the contribution of PET in cases in which MRI failed to clearly demonstrate the dysplastic lesion. Methods: We analyzed imaging data of 13 consecutive patients (eight M, five F, aged from 12 to 40 years; mean, 25) operated on for early-onset severe partial epilepsy associated with TTFCDs during the 3 last years. High-resolution MRI scans were performed at 1.5 T (Signa GE) with 3D SPGR acquisition (1.2-mm-thick contiguous slices), T2 -weighted, and FLAIR images.Twelve patients underwent an 18 FDG-PET scan using a high-resolution PET camera (ECAT 953/31B Siemens: five cases, HR+ CTI Siemens: seven cases). MRI and PET scans were superimposed with ANATOMIST software. All patients were operated on after stereo-EEG after the PET scans. All MRIs were reviewed by a trained radiologist to describe the MRI abnormalities, without knowledge of PET results and
79
localization of the lesion. Cases with typical MRI features of TTFCDs were excluded from the study (n = 5). Results: Eight patients had either subtle or nonspecific MRI abnormalities (n = 6) or normal MRI (n = 2). Subtle findings consisted of isolated minor gyral abnormalities (unusual gyral orientation or organization) in four patients, slight blurring of grey and white matter in one, and mild linear signal abnormality in white matter in one. Location of the MRI abnormalities was frontal in three, central in one, parietal in two, and concordant with electroclinical data. The two patients with normal MRI had frontotemporal and central focus. PET scans (performed in seven of eight patients) showed a focal and severe hypometabolism corresponding to a single gyrus or a part of gyrus, associated with a surrounding mild to moderate hypometabolism in the region of MRI abnormalities, when present. Metabolic data were concordant with stereo-EEG findings. A continuous rhythmic spike activity was recorded within the hypometabolic gyrus, which was also the site of spontaneous seizure onset. Histologic examination demonstrated the typical features of TTFCDs with presence of giant dysmorphic neurons and balloon cells clearly corresponding to the severely hypometabolic gyrus. All patients are seizure free after a limited resection (follow-up, 2 months to 3 years; mean, 22 months). Conclusions: In a significant number of cases (eight of 13 in our series), TTFCDs remain difficult to detect even with optimal MRI. However, high-resolution 18 FDG-PETcoregistred with MRI proved very sensitive and accurate in these patients operated on at the adult age but with epilepsy onset in childhood, contributing to improved surgical strategy and results.
1.223 RELEVANCE OF 18-FLUORODEOXYGLUCOSE POSITRON EMISSION TOMOGRAPHY IN PRESURGICAL EVALUATION OF PATIENTS WITH REFRACTORY EPILEPSY 1 Felix A. Chen, 1 Sarah Farias, 1 Antony Lima III, 2 James E. Boggan, 2 Edie Zusman, and 1 Taoufik M. Alsaadi (1 Department of Neurology and 2 Department of Neurosurgery, University of California Davis Medical Center, Sacramento, CA) Rationale: 18-Fluorodeoxyglucose positron emission tomography (18 FDG-PET) is an imaging modality commonly used to detect a seizure focus in patients with temporal lobe epilepsy in conjunction with other localizing modalities such as magnetic resonance imaging (MRI) and video-EEG telemetry. The localizing value of PET imaging in the presurgical evaluation for temporal lobe resection remains unclear. We wished to evaluate the value of PET imaging in relation to presurgical MRI findings. Methods: We investigated retrospectively the relation between brain PET and MRI in 27 refractory epilepsy patients evaluated at the UC Davis Medical Center for anterior temporal lobe resection. This cohort of patients was selected on the basis of having received a PET imaging study as part of their presurgical evaluation. To ensure accurate localization, only patients with ictal video-EEG recordings according to previously established criteria were included in the study. MRI studies were evaluated for evidence of mesial temporal sclerosis (MTS) according to previously established criteria. MRI study results were used to classify the patients into those with either MTS or non-MTS. Patients were also classified based on PET imaging reports showing lateralizing temporal lobe metabolic changes or nonlateralizing changes. Of the patients studied, nine patients proceeded to temporal lobotomy surgery. The surgical outcomes were classified into either seizure free (Engel class I) or not seizure free (Engel class II–IV). Results: Of the 27 patients studied, 16 patients had normal MRI studies. Of these 16 patients, seven had normal PET studies (44%). In the remaining patients with abnormal MRI studies, 10 had evidence of unilateral MTS with concordant PET findings in eight. Two of the unilateral MTS MRI patients had symmetric PET studies. In the seven patients with normal PET studies, all seven (100%) also demonstrated normal MRI studies. There were 17 patients with lateralizing PET study findings. Of these patients, seven demonstrated concordant MRI findings (41%), and eight had normal MRI scans. Review of the data suggests that patients with normal MRI are more likely to have normal PET scans compared with patients that have evidence of MTS on MRI scan (p < 0.001). Five of the nine surgical patients became seizure free, of whom four patients had MRI findings concordant to the resected lobe; four patients also had Epilepsia, Vol. 44, Suppl. 9, 2003
80
AES PROCEEDINGS
PET findings that were concordant to the surgical site. One seizure-free patient had a normal MRI, whereas another had symmetric PET findings. Of the four non–seizure-free patients, three had concordant MRI and PET findings. Conclusions: The preliminary results of this study suggest that a normal MRI is more strongly correlated with a normal PET study. With regard to surgical outcomes, it appears that there is no significant difference in seizure-free outcome between patients with lateralizing and nonlateralizing PET findings. 1.224 POSTOPERATIVE CHANGES OF REGIONAL CEREBRAL GLUCOSE METABOLISM IN PATIENTS WITH MESIAL TEMPORAL LOBE EPILEPSY 3 Hyun Jeong Han, 1 Woo Suk Tae, 1 Jee-Hyun Kim, 1 Ki-Young Jung, 1 Dae Won Seo, 2 Seung Chyul Hong, and 1 Seung Bong Hong (1 Neurology and 2 Neurosurgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, and 3 Neurology, Myongji Hospital, College of Medicine, Kwandong University, Koyang-City, Republic of Korea) Rationale: Interictal 18 F-fluorodeoxyglucose (FDG) positron emission tomography (PET) in refractory mesial temporal lobe epilepsy (TLE) usually shows unilateral diffuse regional hypometabolism of one mesiolateral temporal area, with or without ipsilateral extratemporal cortical hypometabolism or contralateral temporal hypometabolism. To investigate the postoperative changes of hypometabolism outside the epileptic focus, we performed SPM (statistical parametric mapping) analysis between pre- and postoperative FDG-PET. Methods: Twenty patients with mesial TLE who had the surgery and have been seizure free for >2 years postoperatively were included. There were eight right TLE and 12 left TLE. All showed hippocampal sclerosis on pathology. The differences of regional cerebral glucose metabolism between pre- and postoperative FDG-PET scans were tested by SPM in right TLE and left TLE separately. For the SPM analysis, all SPECT images were spatially normalized to PET template with linear 12 parameters affine transformation and then smoothed with 14-mm FWHM (full width at half maximum), and paired t test was used. The significance level was set to false discovery rate corrected p < 0.05, and spatial extent was set to k > 50. Results: After seizure-free state for >2 years since the anterior temporal lobectomy with amygdalohippocampectomy was performed, regional cerebral glucose metabolism increased significantly in insula, inferior frontoparietal gyrus, and cingulate gyrus of the ipsilateral hemisphere to the epileptic focus, and the cingulate gyrus, orbitofrontal gyrus, and basal ganglia of the contralateral hemisphere. However, cerebral glucose metabolism decreased in fusiform gyrus and caudate nucleus of ipsilateral hemisphere. Conclusions: The brain regions with postoperative recovery of cerebral glucose metabolism appear to be preferential pathways for the propagation of ictal and interictal discharges from mesial temporal region. This finding suggests that extratemporal hypometabolism of mesial TLE is related to epileptic discharges (either interictal or ictal) and is reversible. 1.225 METABOLIC PATTERN OF NEOCORTICAL EPILEPSY QUANTIFIED BY STATISTICAL PROBABILISTIC ANATOMIC MAPS 1 Kwang-Ki Kim, 1 Sang Kun Lee, 1 Chang Ho Yun, 2 Jin-Chul Paeng, 2 Dong-Soo Lee, and 3 Chun-Kee Chung (1 Neurology, 2 Nuclear Medicine, and 3 Neurosurgery, Seoul National University College of Medicine, Seoul, Korea) Rationale: This study evaluated the patterns of metabolism in neocortical epilepsy, diagnosed by fluorodeoxyglucose positron emission tomography (FDG-PET), and tried to provide the basis of understanding the epileptic circuit, using a probabilistic atlas of the human brain (statistical probabilistic anatomical maps: SPAM). Methods: Thirty-six frontal lobe epilepsy (FLE), 16 parietal lobe epilepsy (PLE), and 10 occipital lobe epilepsy (OLE) patients and 22 age-matched controls were spatially normalized to the average brain PET template of international consortium of brain mapping (ICBM). All neocortical epilepsy patients underwent invasive studies and focal
Epilepsia, Vol. 44, Suppl. 9, 2003
neocortical resection. Counts from normalized PET images were multiplied by the probability from 98 volumes of interest (VOIs) of SPAM. Asymmetry indexes (AIs) reflecting the severity of hypometabolism or hypermetabolism were calculated by counts of selected all VOIs from SPAM images. The degree of hypometabolism or hypermetabolism was considered to be abnormal when AI was outside 3 SD (p < 0.01) of those of normal controls. Ictal-onset zone confirmed by invasive study was regarded as the epileptic focus. Results: SPAM images showed abnormally lateralized hypometabolism including epileptic focus in 22 FLE, 11 PLE, and eight OLE patients. Localized hypometabolic VOIs consistent with epileptic focus were found in five FLE, one PLE, and one OLE patients. Ipsilateral subcortical hypermetabolic areas were found in 17 FLE, nine PLE, and five OLE patients. The frequent ipsilateral hypermetabolic structures were thalamus, subthalamic nucleus, putamen, and hippocampal formation. Variable other cortical VOIs were also ipsilaterally hypermetabolic. Conclusions: Our results demonstrated that SPAM images could have a role in the diagnosis of neocortical epileptic focus. Ipsilateral hypermetabolic subcortical structures and other cortical areas suggest the disinhibition or the role of these structures in epilepsy. SPAM images can be used in the analysis of metabolic patterns of various epileptic syndromes. 1.226 PREDICTING OUTCOME OF TEMPORAL LOBE RESECTION IN MEDICALLY REFRACTORY EPILEPSY: STATISTICAL MAPPING ANALYSIS OF FDG-PET 1 Cynthia L. Mayer, 2 John W. Miller, 1 Barbara Lewellen, 1 Donna J. Cross, and 1,3 Satoshi Minoshima (1 Radiology, 2 Neurology, and 3 Bioengineering, University of Washington, Seattle, WA) Rationale: This study investigates the value of preoperative [F18]fluorodeoxyglucose (FDG) brain PET imaging in predicting outcome with respect to seizure control in epilepsy patients who undergo temporal lobe surgery. Results obtained using visual image inspection are compared with those of statistical mapping. Methods: Eighteen epilepsy patients (age 20–59 years) with discordant semiology, EEG, and MR findings underwent standard FDGPET brain imaging preoperatively. For statistical mapping, individual PET images were compared with those of normal adult controls (n = 22l age, 37 ± 9 years) using pixelwise Z-score mapping (abnormal pixels with p < 0.05) after anatomic standardizaton and extraction of gray matter activity using three-dimensional stereotactic surface projections (3D-SSP). A metabolic pattern typical of temporal lobe epilepsy (TLE) was established separately from a group of patients who underwent temporal lobectomy with Engel class I outcome. Based on PET patterns of metabolic activity, individual cases in the study group were categorized as having (a) findings typical of TLE (metabolic reductions in the medial and lateral temporal cortices, inferior frontal cortex, and thalamus); and (b) atypical of TLE (abnormalities not limited to the above structures). Conventional transaxial PET images were interpreted visually by multiple observers and compared with results of statistical mapping. Postoperative outcomes after resection of temporal cortices suspected to have seizure foci were classified as seizure-free, worthwhile improvement (>75% seizure reduction), or no substantial improvement (5) atypical for TLE had no substantial improvement postoperatively. Visual reading identified these cortical abnormalities in only one patient. MRI findings showed no temporal abnormality in three of five patients with seizure-free outcome and seven of nine patients with no substantial improvement. Conclusions: Refractory epilepsy patients who demonstrated FDGPET findings atypical for TLE had significantly less postoperative improvement in seizure control as compared with those having PET patterns typical for TLE. Use of statistical mapping improves predictive outcome
AES PROCEEDINGS in patients with findings atypical for TLE as compared with visual PET image interpretation. FDG-PET may be useful in identifying patients who may not benefit from temporal lobe resection. (Supported by University of Washington Radiology Research Fund.) 1.227 DYSTONIC POSTURING IN SEIZURES OF MESIOTEMPORAL ORIGIN: ELECTROCLINICAL CORRELATIONS AND METABOLIC PATTERNS 1 Violeta Rusu, 1,2 Francine Chassoux, 1 Elisabeth Landre, 2 Viviane Bouilleret, 1 Francois Nataf, 1 Bertrand C. Devaux, and 2 Franck Semah (1 Department of Neurosurgery, Sainte-Anne Hospital, Paris, and 2 Service Hospitalier Frederic Joliot, CEA, Orsay, France) Rationale: To analyze metabolic data according to electro-clinical features in seizures of mesiotemporal lobe (MTL) origin associated with dystonic posturing (DP). Methods: Thirty-six patients (17 males, 19 females; mean age, 34 years; range, 14–49) had presurgical investigations for MTLE. Inclusion criteria consisted of at least one usual seizure recorded during surface or depth EEG and no early secondary generalization during the recorded seizure. All patients had (a) isolated unilateral hippocampal sclerosis demonstrated by magnetic resonance imaging (MRI), (b) mesiotemporal origin of their seizures confirmed by video-EEG recordings; and (c) interictal 18 FDG-PET scan. In addition, 11 patients underwent a stereo-EEG. Thirty-one patients were operated on with favorable outcome in most cases (of 26 patients with a follow-up of ≥2 years, 21 were in class I, four in class II, and one in class III, according to Engel’s classification. 18 FDG-PET scans were performed using a high-resolution head dedicated PET camera (ECAT 953/31B Siemens) with 5.8-mm inplane and 5-mm axial resolution, allowing 31 transverse sections of the brain, spaced 3.37 mm apart. We analyzed electroclinical findings in patients with DP or without, and compared interictal hypometabolism in each group with 10 control subjects using SPM99 software (The Wellcome Trust, U.K.). Results: DP was found in 20 patients (55%) of this series (group A), and 16 had no DP (group B). In group A, we observed three types of DP characterized by a gradual increase in duration and complexity : DP type 1 corresponded to the most subtle and brief DP, type 2 corresponded to the most typical DP pattern, whereas bilateral but asymmetric DP was found in DP type 3. DP was contralateral to the epileptogenic focus in 95% of the cases. Group A patients had significantly higher frequency of head deviation, salivation, somatomotor manifestations, secondary generalization, severe consciousness clouding, and prolonged postictal confusion when compared with group B patients. Ictal discharge patterns in group A mainly consisted of fast rhythmic activity involving frontal or suprasylvian areas, whereas slow rhythmic activity restricted to the temporal areas was observed in group B. A large temporal and extratemporal hypometabolism including basal ganglia was found in group A, whereas it was restricted to the anteromesial part of the temporal lobe and anterior part of insula in group B. Hypometabolism involved the lateral part of the putamen in DP type 1, a larger putaminal area associated with frontal cortex involvement in DP type 2; the most severe hypometabolism involving the putamen associated with perisylvian and parietal cortex involvement was observed in DP type 3. Conclusions: Our data suggest that DP results from involvement of both subcortical and cortical areas implied in motor functions. Moreover, different networks may be involved according to the DP type. 1.228 REFRACTORY EPILEPSY AND BASAL GANGLIA: THE ROLE OF SEIZURE FREQUENCY 1 Viviane Bouilleret, 2 Maria-Jo˜ ao Ribeiro, 2 Marina Mantzarides, 2 Francine Chassoux, 2 Franck Semah, 3 Antoine Depaulis, and 4 Arnaud Biraben (1 EFSN, CHU Bicetre, Bicetre, Paris, 2 Service Hospitalier Frederic Joliot, CEA, Orsay, 3 U398, INSERM, Strasbourg, and 4 Service de Neurologie, CHU Pontchaillou, Rennes, France) Rationale: Studies in animals models and epilepsy patients have suggested that circuits of the basal ganglia may control epileptic seizures and that striatal dopaminergic neurotransmission plays a key role in seizure interruption. The aim of this study was to analyze if duration and seizure frequency are associated with reduction of striatal dopamine in drug-
81
resistant patients by using positron emission tomography (PET) with [18 F]fluoro-L-DOPA. Methods: PET [18 F]fluoro-L-DOPA studies were performed in adult patients with drug-resistant epilepsy. The epilepsy patients were divided in three groups: group 1 patients with ring chromosome 20 (r20) epilepsy (n = 15; 21.8 ± 5.4 years old); group 2 patients with resistant generalized absence epilepsy (n = 10; 32.3 ± 11.4 years old); and group 3 patients with pharmacoresistant temporal lobe epilepsy with hippocampal sclerosis (n = 9; 35.2 ± 10.3 years old). The [18 F]fluoro-L-DOPA uptake constant (Ki, min−1 ) in the caudate and putamen nuclei was calculated using a multiple-time graphical analysis. For each nucleus, Ki values were average between both hemispheres and expressed as mean ± SD. The Ki values for each epileptic group were compared with those of a group of healthy volunteers (n = 10; 45.1 ± 16.5 years old) using a ANOVA test with repeated measures. Results: A decrease of mean Ki values were observed in the caudate and putamen in all groups of patients with refractory epilepsy (Table 1). This decrease of [18 F]fluoro-L-DOPA uptake is more marked for r20 epilepsy, where epilepsy is characterized by repeated and long-lasting seizures (group 1, p < 0.0001 for both nuclei). In group 2, where seizures occur once or twice a day, a significant decrease in uptake was also found (p < 0.001 and p = 0.02, for caudate and putamen, respectively). In the last group, where seizure events occur less than once a day, no significant differences were observed (group 3, p > 0.05, for both nuclei).
TABLE 1. Ki values in the striatum of patients and controls
Group 1 Group 2 Group 3 Controls
Caudate
Putamen
0.0102 ± 0.0013 0.0109 ± 0.0013 0.0119 ± 0.0009 0.0127 ± 0.0013
0.0100 ± 0.0015 0.0105 ± 0.0019 0.0114 ± 0.0010 0.0117 ± 0.0010
Conclusions: The decrease of [18 F]fluoro-L-DOPA uptake observed in the striatum of patients with refractory epilepsy seems to be related to the seizure frequency, suggesting that dopaminergic striatal dysfunction may impair the mechanisms controlling seizures. We cannot exclude the possibility that the antiepileptic drugs had some consequences on [18 F]fluoro-L-DOPA uptake. However, considering that the number of treatments was similar in the three groups, the influence of antiepileptic drugs does not seem predominant in the decrease of Ki. 1.229 IS F-18-FDG OR FLUMAZENIL PET INDICATED IN PRESURGICAL EVALUATION OF PATIENTS WITH LESIONAL TEMPORAL LOBE EPILEPSY? 1 Rene M.C. Debets, 3 Emile F.I. Comans, 2 Demetrios N. Velis, 2 Walter Van Emde Boas, 3 Dick Veltman, and 3 Adriaan A. Lammertsma (1 Department of Neurology and 2 Department of Clinical Electroencephalography and Epilepsy Monitoring Unit, Stichting Epilepsie Instellingen Nederland, Heemstede, and 3 Department of Nuclear Medicine–PET Centre, Vrije Universiteit Medical Centre, Amsterdam, Netherlands) Rationale: Patients with refractory temporal lobe epilepsy who demonstrate convergent data of noninvasive presurgical evaluation are considered excellent surgical candidates. However, some presurgical candidates have mesial temporal sclerosis (MTS) or a temporal lesion with nonconvergent EEG data, including seizure semiology and interictal and ictal EEG recordings. Is positron emission tomography (PET) indicated in these patients and should F-18-fluorodeoxyglucose (FDG) PET or flumazenil PET be preferred to improve presurgical workup and possibly prevent intracranial EEG recordings? Methods: We studied 10 patients with refractory epilepsy; eight had MTS, and two demonstrated a low-grade temporal lesion with nonconvergent EEG data. All 10 patients were evaluated with dynamic FDG and flumazenil PET. One patient had intracranial EEG recordings before temporal lobe resection. Follow-up after surgery varied from 1 to 4 years, and outcome was scored according to the Engel classification. Epilepsia, Vol. 44, Suppl. 9, 2003
82
AES PROCEEDINGS
Results: Seven of 10 patients had unilateral MTS; one had asymmetric bilateral MTS. Two patients demonstrated a low-grade temporal lesion, probably a dysembryoplastic neuroepithelial tumor (DNET). EEG studies of these patients were inconclusive and suggested bilateral epileptogenic zones in three and ipsilateral to the MRI abnormality, a posterior temporal or extratemporal epileptogenic regions in six, whereas one patient had inconclusive EEG results. All 10 patients demonstrated hypometabolic zones convergent with the lesional temporal lobe. Flumazenil PET was abnormal in nine patients, and in two of these nine patients, the area of decreased benzodiazepine-receptor binding was less extensive compared with the area of hypometabolism in FDG-PET. Nine patients were operated on without and one after intracranial EEG recordings. Postoperative outcome was favorable (Engel classes 1 and 2) in seven and less favorable (Engel class 3) in three patients. Conclusions: In 10 patients with a temporal lesion (MTS or DNET), EEG studies were insufficiently localizing, lateralizing, or even inconclusive, and they were considered poor surgical candidates. In these patients, FDG and flumazenil PET supported presurgical lateralization and localization of the epileptogenic temporal lobe in a way that made surgery possible without intracranial EEG recordings in nine of these 10 patients. Only one patient needed confirmation with intracranial EEG recordings before surgery. Flumazenil PET was negative in one patient and therefore slightly less sensitive compared with FDG-PET. This study supports the idea that either one of these PET techniques is sufficient to present the zone of functional deficit related to the epileptogenic zone in patients that otherwise would be rejected for surgery because of inconclusive results of noninvasive EEG recordings or needed intracranial EEG recordings. 1.230 IN VIVO EVIDENCE FOR ENDOGENOUS OPIOID RELEASE AFTER SPONTANEOUS SEIZURES: AN [11 C]Diprenorphine PET Study 1,2 Alexander Hammers, 3 Marie-Claude Asselin, 3 Rainer Hinz, 2 David J. Brooks, 1,2 John S. Duncan, and 1,2 Matthias J. Koepp (1 Department of Clinical and Experimental Epilepsy, Institute of Neurology, UCL, and 2 Clinical Sciences Centre, MRC Cyclotron Building, and 3 Imaging Research Solutions Limited, Hammersmith Hospital, London, United Kingdom) Rationale: To determine the role of endogenous neuropeptide release for the termination of spontaneous seizures in patients with refractory temporal lobe epilepsy (TLE). Methods: We performed paired positron emission tomography (PET) studies with the nonspecific opioid receptor ligand [11 C]diprenorphine (DPN) in four patients with TLE: two patients with normal magnetic resonance imaging (nMRI) and ictal left frontotemporal EEG changes, and two patients with right hippocampal sclerosis (rHS) and concordant ictal EEG changes in one, but contralateral ictal EEG changes in the other patient. The first DPN-PET scan (postictal) was performed between 1.5 and 15 h after the last spontaneous complex partial seizure (CPS); the second PET scan (interictal) was performed after a seizure-free interval of 7–20 days. Eleven healthy controls were studied twice to establish test–retest variation. All subjects had high-resolution MRI and quantitative [11 C]DPN PET on a Siemens/CTI ECAT HR++ scanner. Spectral analysis and metabolite-corrected arterial plasma input functions were used to produce parametric images of DPN volume-of-distribution (Vd). Individual parametric images were spatially normalised to a DPN template in standard stereotaxic space, using Statistical Parametric Mapping (SPM99) software, and difference images calculated. A reduction in DPN binding in the postictal PET scan would be suggestive of endogenous opioid release. Results: Test–retest variation of global DPN-Vd for the 11 controls was 0.33 ± 9.5% (range, −16.8% to +14.5%). All four patients showed postictally decreased opioid binding compared with baseline in the temporal lobe concordant with the side of ictal EEG abnormalities. Quantification of changes showed a 7 to 13% decrease in the presumed epileptogenic hippocampus postictally compared with baseline. Conclusions: Our in vivo findings of reduced opioid receptor binding are further evidence for endogenous opioid release after spontaneous seizures at critical sites in the brain involved in the termination of these seizures. (Supported by Medical Research Council and the National Society for Epilepsy.)
Epilepsia, Vol. 44, Suppl. 9, 2003
1.231 FLUMAZENIL PET VERSUS INTRACRANIAL EEG FOR LOCALIZATION OF TEMPORAL AND EXTRATEMPORAL SEIZURE-ONSET ZONE IN ADULTS 1 Christian Vollmar, 1 Stephan Arnold, 2 Rainer Linke, 3 Alexander Drzezga, 4 Peter Bartenstein, and 1 Soheyl Noachtar (1 Department of Neurology, and 2,3 Department of Nuclear Medicine, University of Munich, Munich, and 4 Department of Nuclear Medicine, University of Mainz, Mainz, Germany) Rationale: [11 C]Flumazenil-PET (FMZ-PET) has been reported to localize the seizure-onset zone, particularly in pediatric epilepsy patients. Aim of the study was to correlate localization of reduced γ -aminobutyric acid (GABA)-receptor binding by FMZ-PET with the results from invasive EEG recording, using 3D image registation of FMZ-PET and magnetic resonance imaging (MRI) in adult temporal and extratemporal epilepsy patients. Methods: We investigated 18 patients (13 male, five female, aged 15–52 years; mean age, 36 years) with temporal (n = 5) and extratemporal (n = 13) epilepsy. All patients underwent cranial MRI, CT, and FMZ-PET scans and and invasive monitoring with subdural grid and strip electrodes. MRI showed lesions in 15 of the 18 patients. In all patients, the epileptogenic zone was identified by subdural EEG recording. So far, 15 patients underwent resective epilepsy surgery. Analysis of FMZ-PET images was performed by automated quantitative analysis based on a set of 63 regions of interest and visual inspection by four different observers. PET findings were rated abnormal with a left–right asymmetry of >10% for automated quantification or after consensus of the observers for visual inspection. The exact anatomic relation between subdural electrodes and FMZ-PET findings was determined three-dimensionally by image coregistration and volume rendering of MRI, CT, and FMZ-PET scans. Results: Automated quantification of FMZ-PET localized the epileptogenic zone correctly in six of 18 (33%) patients and revealed falsepositive findings in 13 patients (72%). Consensus of visual evaluation detected the correct seizure-onset zone in nine patients (50%), falsepositive findings were reported in eight cases (44%). Color-encoded 3D rendering of coregistered MRI and FMZ-PET data helped to identify the correct epileptogenic zone in four additional cases (22%) and ruled out false-positive findings in five patients (28%). The results were similar in temporal and extratemporal epilepsies. Conclusions: 3D rendering of coregistered images improved interpretation of FMZ-PET studies in 50%, primarily because it allows discrimination of FMZ-PET hypointensities into enlarged CSF space and true reductions of benzodiazepine-receptor density in morphologically normal brain. 1.232 MODELING OF NORMAL AND ICTAL PERFUSION PATTERNS IN SPECT USING CORRESPONDENCE ANALYSIS 1,2 Christophe Grova, 1 Arnaud Biraben, 1 Pierre Jannin, 1 Bernard Gibaud, and 1 Scarabin Jean-Marie (1 Laboratoire IDM, Universite de Rennes 1, Rennes, Bretagne, France; and 2 Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada) Rationale: Characterizing brain perfusion interindividual functional variability in single-photon emission computed tomography (SPECT) is a key issue to better understand the physiology and physiopathology to model any perfusion pattern and to support diagnosis. Volume of Interest (VOI)-based anatomic standardization analysis has been widely used to study normal perfusion patterns in SPECT (1). Principal component analysis has also been proposed to study functional variability in healthy subjects (2). The purpose of this study was to explore the structure of statistical dependencies, inherent to a perfusion pattern as observed in a population of SPECT data, using correspondence analysis (CA). We studied two perfusion patterns: (a) normal perfusion in healthy subjects, and (b) ictal perfusion in mesial temporal lobe epilepsy (MTLE) patients. Methods: SPECT scans from a group of 27 healthy subjects (12 men and 15 women, aged 20–56 years) were used to study normal perfusion. We selected 10 operated seizure-free MTLE patients (six men and four women, aged 19–43 years), who showed a typical MTLE ictal SPECT pattern (3) during presurgical investigation. VOIs specifically
AES PROCEEDINGS designed to study MTLE were used to automatically perform perfusion measurements, using a previously proposed anatomic standardization method (4). For each population, we then used CA and hierarchical clustering (5) to explore statistical dependencies between the shapes of perfusion measurements distributions within each VOI. Results: For the normal group (a), the first three principal components were selected to recover 49% of global inertia, highlighting three groups of anatomic structures: internal structures (caudate nucleus, thalamus, lenticular nucleus), mesiotemporal structures (amygdala, hippocampus), and remaining cortex. Within each of these groups, perfusion measurements may then be considered as statistically dependent. For the patient group (b), the first two principal components were considered sufficient to extract main information (70.9% of global inertia). Similar to healthy subjects, the analysis highlighted three groups corresponding to internal structures, temporal structures, and remaining cortex. Nevertheless, the temporal pole and the lenticular nucleus, respectively, classified in the remaining cortex group and in the internal structures group within the healthy subjects analysis, appeared now as most relevant structures of the temporal structures group. Conclusions: These results revealed a structure of dependencies between perfusion measurements, highlighting the relevant role of the temporal pole and the lenticular nucleus in ictal perfusion. This suggests that CA is a promising approach to model functional variability among a population of SPECT data. REFERENCES 1. 2. 3. 4. 5.
Van Laere K, et al. N Engl J Med 2000;28:873–87. Pagani M, et al. N Engl J Med 2002;29:67–75. Ho S, et al. Epilepsia 1996;37:788–95. Grova C, et al. CARS conference 2002;450–5. Lebart L, et al. Wiley series in probability and math statistics. New York.: John Wiley, 1984.
1.233 SEQUENTIAL SUBSTRACTION ICTAL SPECT COREGISTERED TO MRI IN PRAXIS-INDUCED EPILEPSY 1 Stephanie Jacques, 2 Jean-Paul Soucy, 1 Dang Khoa NGuyen, 2 Louis Laflamme, 1 Isabelle Rouleau, 1 Helene Cossette, 1 Suzanne Brideau, 1 Jean-Marc Saint-Hilaire, and 1 Patrick Cossette (1 Neurology and 2 Nuclear Medicine, CHUM-Hopital Notre-Dame, Montreal, Quebec, Canada) Rationale: The term praxis-induced seizures has been recently proposed for a variety of reflex epilepsies induced by complex cognitive and motor tasks. Little is known on the pathophysiology of these rare epileptic syndromes. Methods: We report a 25-year-old woman with reflex seizures systematically induced by goal-oriented oculomotor tasks. The ictal events were documented by continuous video-EEG monitoring. A total of three subtraction ictal SPECTs coregistered to magnetic resonance imaging (MRI) (SISCOM) were performed. Results: Three seizures were recorded. The most effective triggers were the manipulation of electronic devices, such as cellular phones or pocket computer games. Seizures were stereotyped, and consisted of a sensory aura involving the left hand, followed by clonic movements of the left hemiface and arm, along with left cephalic deviation lasting 30 s. The patient remained conscious throughout the seizure. Ictal EEG shows rhythmic theta activity starting over the right posterior quadrant, and rapidly spreading to both posterior regions. Interictal EEGs were normal. A SISCOM study was performed for each ictal event (n = 3), respectively at 14, 11, and 1 s. Early injected SISCOM (1 s) revealed an activation of the right occipital lobe and ipsilateral rolandic region. In turn, later injections (11 and 14 s) resulted in activation of the right rolandic region or anterosuperior portion of the right temporal lobe. Conclusions: Sequential analysis of SISCOM in praxis-induced seizures suggested an initial activation of the right occipital lobe, rapidly followed by spreading to the ipsilateral rolandic and temporal lobe regions. Visual input combined with motor activity and decision making were essential to triggering seizures. This study gives some insight into the neuronal network involved in this rare form of epilepsy.
83
1.234 CLINICAL PERISYLVIAN SYNDROME WITH AND WITHOUT MORPHOLOGIC CHANGES 1 Jan-Anders Ahnlide, 2 Kristina K¨ all´en, 3 Bo Geijer, and 1 Ingmar Rosen (1 Clinical Neurophysiology, 2 Neurology, and 3 Radiology, University Hospital in Lund, Lund, Sweden) Rationale: In presurgical evaluation of patients with intractable partial epilepsy, ictal SPECT co-registered with magnetic resonance imaging (MRI) has been shown to provide accurate localizing information in search for the epileptogenic region. We report two patients with refractory partial epilepsy from our epilepsy surgery program that had normal MRI scans at primary evaluation. The patients had similar seizure semiology, including perioral sensations, impaired speech, and tonic posturing of the arms. Methods: The ictal and interictal SPECT image were co-registered using a surface matching technique (Analyze AVW). The images were normalized, and the interictal image was subtracted from the ictal image. To measure the statistical deviation from the normal random intensity difference between the images, the standard deviation of the difference image was calculated in each cerebral hemisphere (excluding cerebellum), using volumes defined in the MRI and a mass significance (Bonferroni) correction based on a cubic voxel with 7-mm side was applied. The SD of the hemisphere with the least variance was used as normal reference. Results: Patient 1: Subtraction images showed a marked ictal hyperperfusion (5.1 SD) in the left perisylvian region and a mild hyperperfusion in the corresponding region on the right side. Apart from a bilateral hyperperfusion in cerebellum, no other significant hyperperfusion was seen. At a reevaluation of the MRI examination after the reconstruction of transaxial MP-RAGE images, areas of cortical dysplasia outlining the sylvian fissures bilaterally were identified. Patient 2: Subtraction images showed a hyperperfusion (3.9 SD) in the left perisylvian region and a mild hyperperfusion in the corresponding region on the right side. There was also a hyperperfusion in the left thalamus and a mild hyperperfusion in the cerebellum with right predominance. Reevaluation of the MRI showed mild cerebellar atrophy, but no cerebral pathology. Conclusions: Our report demonstrates that subtraction ictal SPECT co-registered to 3D MRI improves diagnosis of subtle brain malformations and helps localize the seizure focus in certain intractable epilepsy patients. The method might help avoid unnecessary invasive procedures, and thus minimize the risks for complications during presurgical evaluation. It can serve as a diagnostic tool for congenital perisylvian syndrome patients with minor perisylvian malformations and mild clinical signs.
1.235 REGIONAL CEREBRAL BLOOD FLOW PATTERNS IN UNILATERAL MESIAL TEMPORAL LOBE EPILEPSY: STATISTICAL PARAMETRIC MAPPING OF ICTAL AND INTERICTAL SPECT 1 Seung Bong Hong, 1 Woo Suk Tae, 1 Eun Yeon Joo, 1 Dae Won Seo, 2 Seung Chyul Hong, 3 Moon Hyang Lee, and 1 Jee-Hyun Kim (1 Neurology, 2 Neurosurgery, and 3 Pediatrics, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea) Rationale: To investigate the regional cerebral blood flow (rCBF) of patients with mesial temporal lobe epilepsy (mTLE), we performed the statistical parametric mapping (SPM) analysis with ictal singlephoton emission computed tomography (SPECT) of patients and interictal SPECT of the patients and normal subjects. Methods: Nineteen normal subjects and 38 patients with mTLE (22 left, 16 right) were included. All underwent brain SPECT. For SPM analysis, all SPECT images were spatially normalized with linear 12 parameters affine transformation and then smoothed with 14-mm FWHM (full width at half maximum). The left (mean age, 30.0 ± 8.75 years; seven men, 13 women) and right (mean age, 28.4 ± 10.30; eight men, eight women) mTLE groups were statistically compared with the normal subjects (mean age, 34.1 ± 9.88; nine men, 10 women) by SPM t test, and paired t test was performed between ictal and interictal SPECTs of
Epilepsia, Vol. 44, Suppl. 9, 2003
84
AES PROCEEDINGS
the patients. The positive and negative contrasts were used to display the regions of hypo- or hyperperfusion, respectively. The significance level was set to false discovery rate corrected p < 0.05, and spatial extent was set to k > 50. Results: In interictal state, the ipsilateral hippocampus, both thalami, pericentral gyri, left insula, and both medial frontal lobes showed hypoperfusion, whereas both posterior lateral temporal areas showed hyperperfusion. In ictal state, the rCBF of ipsilateral temporal lobe to the epileptic focus and both prefrontal white matters increased, whereas both medial frontal lobes showed hypoperfusion. In the right mTLE, the left hippocampus and insula showed hypoperfusion during interictal state. In paired t test, ipsilateal temporal lobe, hippocampus, thalamus, putamen, insula, and both precentral gyri showed hyperperfusion. Conclusions: Surprisingly, the hypoperfusion and hyperperfusion patterns of patients with mTLE were similar between interictal and ictal states. These findings indicate that occurrence and propagation of epileptic discharges happen not only in the ictal state but also in the interictal period. The hypoperfusion patterns suggest that the corticothalamohippocampoinsula circuit was impaired during the interictal state, whereas temporal and prefrontal regions showed hypofunction during ictal period.
1.236 EXCITATORY AND INHIBITORY NETWORK INTERACTIONS DURING LOSS OF CONSCIOUSNESS IN TEMPORAL LOBE EPILEPSY 1 Seung Bong Hong, 1 Woo Suk Tae, 1 Eun Yeon Joo, 1 Dae Won Seo, 2 Seung Chyul Hong, 3 Moon Hyang Lee, and 1 Jee-Hyun Kim (1 Neurology, 2 Neurobiology, 3 Diagnostic Radiology, and 4 Pediatrics, Yale University School of Medicine, New Haven, CT) Rationale: Temporal lobe seizures cause diverse ictal behaviors including loss of consciousness, dystonic movements, and neuroendocrine changes, suggesting that widespread networks beyond the temporal lobes are involved. To elucidate these networks, we used single-photon emission computed tomography (SPECT) ictal–interictal imaging to compare complex partial seizures (CPSs), with impaired consciousness, with simple partial seizures (SPSs), with spared consciousness, in patients with mesial temporal epilepsy. Methods: Patients with surgically confirmed mesial temporal lobe epilepsy were selected based on mesial temporal sclerosis on pathology, and lack of recurrent seizures with ≥1-year follow-up. CPSs were studied during the late ictal and early postictal periods, respectively, with Tc99m-HMPAO injections 60–90 s after seizure onset (n = 10) and at >90 s after seizure onset (n = 8). Patients with SPSs were also studied (n = 6). Group analyses were done in SPM99 using a paired t test model. The extent threshold below which clusters were rejected was 125 voxels (1 cc) and height threshold, p = 0.01 (z = 2.33). To study network involvement, correlation analysis was done between brain regions. Results: Significant cerebral blood flow (CBF) increases in the CPS 60- to 90-s group were found in the temporal lobe and the basal ganglia ispilateral to seizure onset. In the >90-s group, the temporal lobe was no longer activated, but significant hyperperfusion was present in midline subcortical structures with limbic connections, including the medial thalamus, hypothalamus, and upper brainstem. In both CPS groups, there was marked hypoperfusion of the bilateral frontoparietal association cortex. In contrast, individual patients with SPSs had CBF increases confined to the temporal lobe. The group data from SPS patients revealed no significant network involvement of midline subcortical structures or of the association cortex, as was seen in CPSs. Correlation analysis across all patients (n = 24) revealed that hyperperfusion in the ipsilateral medial thalamus was strongly correlated with hypoperfusion in the ipsilateral frontal association cortex (r = 0.67; p < 0.001). Conclusions: Abnormal excitatory and inhibitory network mechanisms may lead to behavioral phenomena associated with partial seizures. Spread of seizure activity from the medial temporal lobe to midline subcortical structures, especially during the late ictal and early postictal periods, may be crucial for these long-range network effects. We propose that during CPSs, abnormal inhibition of higher order association areas as well as abnormal excitation of medial diencephalon and upper brainstem may result in disruption of normal information flow, and thus lead to loss of consciousness.
Epilepsia, Vol. 44, Suppl. 9, 2003
1.237 UTILITY OF SPECT SCANS IN NONEPILEPTIC SEIZURES Paul B. Pritchard III, Kris Topping, and Mark T. Wagner (Department of Neurology, Medical University of South Carolina, Charleston, SC) Rationale: Nonepileptic seizure (NES) is frequently the diagnosis among patients who undergo extended VEEG monitoring, representing ≤ 50% of cases evaluated in our inpatient unit. We routinely analyze recorded clinical events, ictal and interictal EEG, postictal serum prolactin levels, and neuropsychological study to differentiate between epileptic seizures (ESs) and NESs, each of which has limitations in discerning the difference. Very few reports have addressed the possible application of single-photon emission tomography (SPECT) in this setting, including one series of 11 patients (Spanaki MV, Spencer SS, Corsi M, et al. J Neuroimaging 1999;9:210–6). Methods: This study represents a retrospective analysis over 28 months of 60 patients in whom a diagnosis of NESs was made after extended inpatient monitoring, based on VEEG analysis, postictal serum prolactin studies, and neuropsychological evaluation. We reviewed adults with NESs who underwent ictal SPECT, interictal SPECT, or both. SPECT images of the brain were obtained and reviewed in sagittal, coronal, and transverse planes after injection with 30 mCi of technetium 99m Neurolite. Studies were interpreted by radiologists who were aware of ictal versus interictal status of the patient but were blinded as to seizure type. Results were correlated with NES type and demographic data. Results: Seventeen (11 women, six men) of 60 patients with NES had SPECT in the ictal (two), interictal (four), or both (11) states. Patients ranged in age from 22 to 72 years (mean, 42 years). NES types included seven with generalized convulsions, four with staring spells, four with focal motor activity and secondary generalization, and two with focal motor activity alone. Both patients with ictal SPECT alone had normal studies. Among those with interictal SPECT only, two of four were abnormal, demonstrating mildly increased radiotracer uptake in both frontal lobes. Eleven had ictal and interictal SPECT, of which eight were normal for both studies. Two of the ictal studies showed increased focal uptake compared with co-registered interictal SPECT, and in each case (one frontal, one parietal), the focal radiotracer uptake was contralateral to focal motor activity. Conclusions: Most ictal SPECT scans for patients with NESs were normal, as expected. In two cases, there was increased radiotracer uptake contralateral to focal motor activity, even though all other parameters spoke for NESs. It is possible that intense motor activity can in itself induce increased radiotracer uptake in SPECT scans.
1.238 COMPARISON OF SISCOM AND INTRACRANIAL EEG LOCALIZATION IN NEOCORTICAL EPILEPSY 1 Shobhit Sinha, 2 Eric M. Rohren, 1 Elson L. So, 1 Gregory D. Cascino, and 2 Brian P. Mullan (1 Division of Epilepsy and Section of Electroencephalography, and 2 Department of Diagnostic Imaging, Section of Nuclear Medicine, Mayo Clinic & Mayo Medical School, Rochester, MN) Rationale: Subtraction ictal SPECT co-registered to magnetic resonance imaging (MRI) (SISCOM) improves the determination of the epileptogenic zone in patients with neocortical epilepsy. However, very few studies have compared SISCOM with intracranial ictal EEG localization. In this study, we compared the relation of the SISCOM focus with that of intracranial EEG localization by using three-dimensional (3D) volumetric brain maps with co-registered SISCOM and intracranial electrode images. Postsurgical outcome in these intractable neocortical epilepsy patients was also assessed. Methods: Patients with neocortical epilepsy who underwent SISCOM studies and intracranial EEG monitoring were retrospectively reviewed. The 3-D volumetric brain maps with co-registered SISCOM and intracranial electrode images were constructed utilizing the Analyze software (Mayo Foundation). The images were reviewed in a blinded fashion separate from the clinical history, video-EEG findings, and postsurgery outcome. The Seizure Frequency Scoring System was used to quantify presurgical and postsurgical seizures. Excellent outcome was defined as postsurgical score of ≤4 (seizure-free, auras only, or fewer than three
AES PROCEEDINGS nocturnal seizures per year); whereas favorable outcome was defined as an improvement of ≥2 points (>75% reduction in seizure frequency). Results: Eighteen patients were identified (eight males, 10 females), with a mean age of 27.9 ± 14.7 (±1 SD) years. Mean duration of epilepsy at presentation was 18.7 ± 11.4 years. No lesion was found in the MRIs of 10 patients. The intracranial EEG seizure started from electrodes overlying the SISCOM focus in 10 (56%) patients, and in the vicinity (within 3 cm) of the SISCOM focus in another three patients. Among five patients with discordant findings, the intracranial EEG seizure commenced in the same lobe as the SISCOM focus in four patients. Concordance between the SISCOM focus and the intracranial EEG did not vary according to the timing of the ictal SPECT injection or the presence of MRI lesion. Resective epilepsy surgery was performed in 15 of 18 patients. The SISCOM focus was resected in all but two patients. Ten patients had ≥1 year postoperative follow-up. Outcome was excellent in five, favorable in three, and poor in two patients. Conclusions: The SISCOM focus and the intracranial EEG seizure focus are generally concordant in majority (72%) of intractable neocortical epilepsy patients. Further study is needed to assess the effect of the concordance versus the extent of surgical resection on the outcome of resective epilepsy surgery. 1.239 EVOLUTION OF SPECT PERFUSION CHANGES DURING TEMPORAL LOBE COMPLEX PARTIAL SEIZURES 1 Wim Van Paesschen, 2 Patrick Dupont, 1 Guido Van Driel, 3 Hubert Van Billoen, and 2 Koen Van Laere (1 Neurology, 2 Nuclear Medicine, and 3 Radiopharmacy, Universitair Ziekenhuis Gasthuisberg, Leuven, Belgium) Rationale: During temporal lobe complex partial seizures (TL-CPS) associated with hippocampal sclerosis (HS), there is hyperperfusion in the ipsilateral (IL) temporal lobe, and hypoperfusion in the frontal lobes and contralateral (CL) cerebellum (Van Paesschen et al. Brain 2003;126:1103–11). Our aim was to determine (a) whether similar changes were present during TL-CPS associated with etiologies other than HS, and (b) whether there were differences in ictal perfusion changes between a group with early versus late ictal SPECT injection. Methods: We studied patients with refractory temporal lobe epilepsy (TLE) associated with HS and other etiologies. All had an interictal and ictal SPECT with injection during a CPS. Ictal SPECT injection started ≥30 s before the end of the CPS. Images were normalised and co-registered. Using statistical parametric mapping (SPM99), brain regions with significant (corrected p < 0.05) ictal perfusion changes were determined. We compared ictal perfusion changes of (a) patients with HS (n = 29) versus other etiologies (n = 9), and (b) patients with ictal SPECT injections that started within 15 s of seizure onset (n = 15) (group 1) versus injections that started after 30 s of seizure onset (n = 10) (group 2). Results: Ictal perfusion changes did not differ between patients with HS and other etiologies. Group 1 had IL anterior temporal lobe hyperperfusion and (pre)frontal lobe hypoperfusion. Group 2 had propagation of IL temporal hyperperfusion toward the posterotemporal regions, and concomitant hypoperfusion of IL posterofrontal and parietal regions (Fig. 1. Left, group 1; right, group 2; black, hyperperfusion; white, hypoperfusion). Group 2 had significant hyperperfusion of the IL thalamus and putamen, and CL parahippocampal gyrus, but not group 1. Group 1 showed weak hypoperfusion (uncorrected p < 0.001) of the CL cerebellum, but not group 2.
85
Conclusions: Ictal perfusion changes during TL-CPS associated with HS are comparable with the perfusion changes during TL-CPS associated with other etiologies. Posterior propagation of hypoperfusion toward IL posterofrontal and parietal regions concomitant with posterior propagation of ictal temporal hyperperfusion is consistent with the hypothesis that the hypoperfusion represents an ictal surround inhibition. CL cerebellar hypoperfusion occurs during the earlier phase of TL-CPS, whereas hyperperfusion of IL putamen and thalamus and CL parahippocampal gyrus occurs during a later stage of TL-CPS (Fig. 1). [Supported by Grant Research Fund Katholieke Universiteit Leuven Interdisciplinair Onderzoeksprogramma (IDO)/99/5.]
1.240 CORTICAL AND THALAMIC FMRI ACTIVATION DURING SPIKE AND WAVE IN IDIOPATHIC GENERALIZED EPILEPSY Y. Agha Khani, A.P. Bagshaw, C.G. B´enar, C. Hawco, M. Veilleux, E. Andermann, F. Andermann, F. Dubeau, and J. Gotman (Department of Neurology and Neurosurgery, Montreal Neurological Hospital and Institute, Montreal, Quebec, Canada) Rationale: We used continuous EEG–functional magnetic resonance imaging (fMRI) to evaluate the regions activated as a result of generalized spike–wave (GSW) discharges in patients with idiopathic generalized epilepsy. Methods: We selected 23 patients with active GSW. During EEGfMRI study, EEG (21-channel) was recorded in a 1.5-T MR. Artefacts from the imaging sequence were removed from the EEG (Hoffmann et al. Magn Reson Med 2000;44:791). Eight patients were excluded because of lack of GSW during the study and one for technical reasons. Areas activated were found by statistical comparison to a model including five hemodynamic response functions: the standard response to auditory stimuli, peaking at 5.4 s (Glover NeuroImage 1999;4:16), and four gamma functions peaking at 3, 5, 7, and 9 s. Activations were considered significant if three contiguous voxels had a t stat ≥3.0. The highest t stat for each activation in different models was determined. Activations in the brainstem, basal ganglia, and cerebellum were excluded. Results: Fourteen patients were evaluated (13 women; mean age, 36; age at seizure onset, 10). Two had no fMRI activation. Cortical activation was largely symmetrical in the 12 other patients. It was diffuse in three, and multiple regional in nine: mesial and lateral frontal in nine, parietal convexity in nine, mesial parietal in eight, mesial occipital in seven, occipital convexity in two, mesial central in six, central convexity in five, mesial or lateral temporal in six, and insula in three. Activation was almost equally distributed over anterior and posterior head regions in five subjects, with anterior predominance in five and posterior in two. We found negative and positive cortical activations: predominantly negative in seven patients (mean volume of negative activation, 164.9 vs. 20.4 cc for positive activation), and predominantly positive in five (mean volume of positive activation, 233.9 vs. 65.5 cc). Maximum t stat varied from 3.5 to 14.3 for positive and from −3.4 to −17.6 for negative activations. All 12 patients showed thalamic activation, bilateral in 10, positive in eight, negative in three, and both positive and negative in one. Thalamic activations had positive maximum t stat of 4.3 to 10.5 (volume, 0.4–38.9 cc), and negative max t stat of −3.2 to −5.6 (volume, 0.4–38.9 cc). Conclusions: GSW resulted in fMRI cortical responses that were largely symmetrical and involved primarily frontal and parietal regions, but also central and occipital regions. The thalamus was involved in all patients showing cortical activation. A surprising result was the presence of positive and negative activations in cortex and thalamus. Whereas positive activation results primarily from increased synaptic activity, excitatory or inhibitory, negative activation is thought to result from decreased neuronal input, which may also play a significant role in GSW. (Supported by Canadian Institute of Health Research.)
1.241 ANALYSIS OF INTERICTAL EEG-FMRI DATA USING A RANGE OF HAEMODYNAMIC RESPONSE FUNCTIONS Andrew P. Bagshaw, Christian-G. B´enar, Yahya Agha Khani, Colin Hawco, Fran¸cois Dubeau, Bruce Pike, and Jean Gotman (Neurology and Neurosurgery, Montreal Neurological Institute, Montr´eal, Quebec, Canada)
Epilepsia, Vol. 44, Suppl. 9, 2003
86
AES PROCEEDINGS
Rationale: Simultaneous measurement of electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) shows great promise as a clinical tool and as a method to study the origin of interictal spiking (1,2). However, a serious impediment to its routine use is the relatively low percentage of patients who show fMRI activations despite having interictal discharges while in the scanner. The purpose of the current work was to attempt to increase the number of patients with significant activations by analyzing the fMRI data with a range of modeled hemodynamic response functions (HRFs). Methods: fMRI images were acquired in one of two 1.5-T MR scanners (Vision and Sonata, Siemens, Germany) using echo-planar imaging. Twenty-one channels of EEG were recorded simultaneously using the EMR amplifier (Schwarzer, Germany). Statistical processing of the fMRI images produced maps of the t statistic, allowing positive and negative activations to be detected (3). Thirty-four image sets were selected from a total scanned population of 42 patients, on the basis of a clinical diagnosis of focal epilepsy and more than five spikes during the fMRI session. The initial analysis was performed using the measured HRF to brief auditory stimuli which peaks at 5.4 s (4). The data were reanalyzed with four alternative HRFs modeled as single gamma functions peaking at 3, 5, 7, and 9 s. The source of epileptic activity was determined from EEG and clinical data. Results: Twelve data sets (35%) showed activations that were in agreement with the known source of the spikes and had a higher t value when using one of the single gamma functions than when using the original HRF. Two sets demonstrated areas of activation that were also consistent with the focus but which had a higher t value when using the original HRF. In three sets, the original and new analyses were very similar, and in four sets, areas of activation not consistent with the known source were found when using one of the single gamma function HRFs. In the remaining 13 (38%) sets, no activation was seen. Thirteen data sets showed areas of negative activation. All but two, both from the same subject, also had some areas of significant positive activation. In 10 of the 13, the most significant negative cluster had a maximum t value when using an HRF that peaked later than the original. In comparison, the positive activations had maximum t values that were evenly spread across the five HRFs. Conclusions: The results suggest that an exploratory approach based on using several HRFs with different latencies can provide new information concerning the source of interictal epileptic spikes. However, misrepresentation of the form of the HRF is not the major reason for the high percentage of subjects who do not show fMRI activation after interictal discharges.
REFERENCES 1. 2. 3. 4.
B´enar CG, et al. NeuroImage 2002;17:1182. Lemieux L, et al. NeuroImage 2001;14:780. Worsley KJ, et al. NeuroImage 2000;15:1. Glover GH. NeuroImage 1999;9:416.
1.242 FUNCTIONAL MRI ANALYSIS OF THE PREICTAL STATE 1 Paolo Federico, 2 Regula S. Briellmann, 2 David F. Abbott, and 2 Graeme D. Jackson (1 Department of Clinical Neurosciences, University of Calgary, AB, Canada; and 2 Brain Research Institute, University of Melbourne, Victoria, Australia) Rationale: The transition from interictal to ictal discharges is poorly understood. One possibility involves the brain changing into a facilitating state promoting seizures. Nonlinear mathematical analysis of EEG frequency components has confirmed the presence of a preictal state in focal epilepsy; however, spatial resolution was limited. We report on functional magnetic resonance imaging (fMRI) analysis of the preictal state in three patients with focal epilepsy. Methods: Patient 1 was a 22-year-old man with daily complex partial seizures originating from the left face sensory area, based on video-EEG monitoring (VEM), positron emission tomography (PET), and singlephoton emission computed tomography (SPECT). MRI showed no lesion. Surgical resection of suspected focus resulted in seizure cure, confirming clinical localization. Patient 2 was a 14-year-old girl with frequent nocturnal frontal lobe seizures based on VEM, with MRI showing a Epilepsia, Vol. 44, Suppl. 9, 2003
possible dysplastic lesion in the right superior frontal gyrus. Ictal SPECT showed hyperperfusion in the same area. Patient 3 was an 11-year-old girl with frequent nocturnal complex partial seizures originating from the right supplementary sensorimotor area based on VEM. MRI showed no obvious abnormality, but ictal SPECT showed hyperperfusion in the left (contralateral) motor/premotor area. Blood oxygen level–dependent (BOLD) fMRI was performed at 3 T, while the patients slept. Imaging was stopped at the clinical onset of a seizure. Block-design analysis was carried out comparing BOLD signals for the 1 min before seizure onset (“task”) with a span of 1 minute beginning 5 min before seizure onset (“rest”). Individual subject maps of BOLD signal changes were based on a t test comparison between the two preictal periods (task vs. rest). Results: One typical complex partial seizure was captured in each patient. In patient 1, a striking preictal BOLD signal increase was seen over the left (ipsilateral to seizure focus) frontocentral region, maximal at the seizure focus. No significant BOLD signal decreases were observed. Patient 2 showed a robust preictal BOLD signal increase over the left (contralateral to seizure focus and lesion) posterior frontal region, as well as a focal BOLD decrease at the seizure focus on the right. In patient 3, a preictal BOLD signal increase was seen in the premotor area contralateral lateral to the presumed seizure focus on the right. Notably, this BOLD increase colocalized precisely with the site of hyperperfusion seen on an ictal SPECT study. No significant BOLD decreases were seen in this patient. Conclusions: We demonstrate that the preictal state is associated with robust BOLD signal changes, although they were not consistent between patients. These findings support the concept of a preictal physiological state, although the mechanisms through which the transition to seizures occur remain unknown. (Supported by National Health and Medical Research Council of Australia, Canadian Institutes of Health Research, Alberta Heritage Foundation for Medical Research.) 1.243 FUNCTIONAL MRI ANALYSIS OF EPILEPTIFORM DISCHARGES AND SLOWING IN JUVENILE MYOCLONIC EPILEPSY 1 Graeme D. Jackson, 1 Tony B. Waites, 1 Angelo Labate, and 2 Paolo Federico (1 Brain Research Institute, University of Melbourne, Victoria, Australia; and 2 Department of Clincial Neurosciences, Univeristy of Calgary, AB, Canada) Rationale: We have previously shown that the posterior cingulate may have a role in the generation of spike-and-wave discharges (SWDs) in persistent childhood absence epilepsy. Using event-related functional magnetic resonance imaging (fMRI), we sought to document regional changes in blood oxygen level–dependent (BOLD) activity associated with spontaneous generalized SWDs as well as slow waves in a patient with juvenile myoclonic epilepsy. Methods: A 29-year-old man with a clinical diagnosis of juvenile myoclonic epilepsy and frequent SWDs was studied. EEG was recorded continuously inside a 3-T MR scanner for 30 min, and the timing of occurrence of SWDs and slow-wave events were noted. These events were considered separately for analysis. An event-related analysis of the two groups of events was carried out using SMP99. The statistical threshold for activation was set at p < 0.005, with a minimum affected cluster size of 10 pixels. Results: During the study, the subject had 30 SWDs (4.5-Hz spike and polyspike and wave) as well as five bursts of 2- to 3-Hz semirhythmic generalized slowing. Significant SWD-related BOLD signal decreases were seen in the posterior cingulate. Some SWD-related BOLD increases were seen in the depths of the precentral sulci. No thalamic BOLD signal changes were seen with SWDs. In contrast, generalized slowing was associated with significant BOLD signal increases in the thalamus and brainstem reticular formation bilaterally. No significant BOLD signal decreases were seen with generalized slowing. Conclusions: We have confirmed our previous observation that the posterior cingulate may have a role in the generation of SWDs in generalized epilepsies. We have also shown that the thalamus and brainstem reticular formation activation is seen with generalized slowing but not with SWDs. These results suggest that posterior cingulate deactivation and brainstem/thalamic activation may have different, but complementary roles in the electroclinical phenomena of SWDs and generalized slowing. (Supported by National Health and Medical
AES PROCEEDINGS Research Council of Australia, Canadian Institutes of Health Research, Alberta Heritage Foundation for Medical Research.)
1.244 EPILEPTIFORM LOCALIZATION FROM SIMULTANEOUS EEG–FMRI RECORDINGS 1 Michiro Negishi, 2,3 Hal Blumenfeld, 4 Dennis D. Spencer, and 1,4 R. Todd Constable (1 Diagnostic Radiology, 2 Neurology, 3 Neurobiology, and 4 Neurosurgery, Yale University, New Haven, CT) Rationale: Combination of electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) for epileptiform localization has been an active area of research (EEG-triggered fMRI: e.g., Neurology 1996;47:89–93; EEG-fMRI simultaneous recording: e.g., Proc ISMRM 1998;168). However, a better understanding of the difference between EEG and fMRI signals is necessary to fully benefit from the both modalities. We present a comparison of EEG source estimation and fMRI spatial parameter mapping from data that are simultaneously collected from epilepsy patients. Methods: Five runs of 10- to 15-min EEG-fMRI recordings are conducted for each epilepsy patient. The fMRI parameters are 1.5-T (GE) echo-planer, TR = 3 s, TE = 50 ms, 64 × 64 × 18 voxels. The EEG data are sampled at 500 Hz (Neuroscan 22 to 32 channels system). Epileptiforms from each subject are averaged to obtain a time-course of a spike–wave pair and are subjected to source localization using the MUSIC algorithm (IEEE Trans Antenas Propagat 1986;AP-34:276–80). From standardized fMRI volumes, eight volumes including one that contains the onset of each spike are selected, and a spatial parametric map (SPM99, Hum Brain Mapp 1995;2:189–210) is computed using epileptiform periods as regressors. Results: EEG source localization maps and an fMRI spatial parameter map from a patient are shown as an example. The EEG source estimation on the spike part of the epileptiforms (Fig. 1A) results in an activation in the left dorsal frontal cortex, whereas that on the wave part (B) results in the right ventral frontal cortex. The SPM result (Fig. 2) shows an activity in the left parietal lobe [(x, y, z) = (−16, −52, 36), left precuneus] as
87
the only statistically significant activity (p = 0.026). There is a weaker left activity closer to the EEG spike localization. There is also a weak right activity in the parameter map, which is inferior compared with the EEG wave localization. Precision of EEG wave localization may be low, considering the spatially widespread wave activities. Conclusions: We conclude that (a) EEG recording during fMRI scanning can be used to detect epileptiforms; and (b) EEG and fMRI data can complement each other in the analysis of the onset and propagation of epileptiforms (Figs. 1 and 2). (Supported by the National Institutes of Health grants NS40497 and NS38467.) 1.245 INTERICTAL EEG-CORRELATED FUNCTIONAL MRI: A STUDY OF 50 PATIENTS WITH LOCALISATION-RELATED EPILEPSY Afraim Salek-Haddadi, Beate Diehl, Martin Merschhemke, Khalid Hamandi, Louis Lemieux, and David R. Fish (Department of Clinical & Experimental Epilepsy, UCL Institute of Neurology, London, United Kingdom) Rationale: EEG-correlated functional magnetic resonance imaging (MRI) allows physiological imaging of spontaneous epileptiform activity, in vivo, across the entire brain, and with superior spatiotemporal resolution. Our aim was to characterise and map blood oxygen level– dependent (BOLD) signal changes linked to interictal epileptiform discharges (IEDs), in a large group of patients with focal epilepsy. Methods: Fifty patients with localisation-related epilepsy, independent of aetiology, and frequent IEDs on recent EEG, were studied on a GE Horizon 1.5-T scanner using whole-brain EPI (TE/TR 40/3000, 64 × 64 matrix); 700 scans were acquired continuously over 35 min. Ten channels of scalp EEG, plus ECG, were recorded simultaneously using an MRcompatible system, developed in-house, with on-line pulse and imaging artefact subtraction (1). IEDs were classified, labeled, and used to perform an event-related analysis of the fMRI data using SPM99. Images were realigned and smoothed. Event-related responses were modeled both flexibly and using a canonical haemodynamic response function (HRF). Results were thresholded for multiple comparisons. Results: Good-quality EEG was obtained in all patients, and IEDs were successfully captured from 29. Of these, 12 (41%) had BOLD activations concordant with electroclinical data across a range of pathologies. Four (14%) showed activation of uncertain significance, and in seven (24%), no activation was observed. In this group, there was a tendency to abnormal background rhythms, head motion, and subtle myoclonus. In four patients (14%), IEDs were too frequent for BOLD changes to be observed, and in two patients, the study was terminated because of seizures. The HRF in patients with activation was in broad keeping with the “canonical” response, although some intersubject variability was evident (Fig. 1). EEG/functional magnetic resonance imaging in a 33-year-old woman with left hippocampal sclerosis: blood oxygen level– dependent activation, overlayed onto orthogonal slices (left), linked to runs of left temporal (F7 ) spikes (see right). Fitted haemodynamic response function is shown in lower middle.
Epilepsia, Vol. 44, Suppl. 9, 2003
88
AES PROCEEDINGS
Conclusions: Our experience indicates that simultaneous EEG/fMRI can provide novel localising information as to the irritative zone in selected patients: those with frequent stereotyped high-amplitude unifocal IEDs distinguishable from the background. Future studies should address the pathophysiology of IEDs using more sophisticated models or consider alternative paradigms based on quantitative rather than qualitative EEG measures. [Supported by Medical Research Council (U.K.).]
REFERENCE 1. Allen PJ, Josephs O, Turner R. A method for removing imaging artifact from continuous EEG recorded during functional MRI. Neuroimage 2000;12:230–9.
1.246 FUNCTIONAL MR STUDY FOR LANGUAGE ORGANIZATION IN LEFT MEDIAL TEMPORAL LOBE EPILEPSY PATIENTS 1 Chun Kee Chung, 2 In Chan Song, 1 So Hee Kim, and 2 Kee Hyun Chang (1 Neurosurgery and 2 Diagnostic Radiology, Seoul National University Hospital, Seoul, Republic of Korea) Rationale: Functional magnetic resonance imaging (fMRI) using four language tasks was used to determine difference between normal controls and left medial temporal lobe epilepsy patients. Methods: Normal volunteers (n = 12; mean age, 25 years) and left temporal lobe epilepsy patients (n = 12; mean age, 26 years) were examined on a 1.5-T MR unit (GE Horizon Echospeed) using a conventional head coil. All subjects had strong right-handedness on Edinburgh handedness inventory. All patients had left hippocampal sclerosis on conventional MRI. A single-shot GRE-EPI sequence (TR/TE/flip angle = 3,000ms/50ms/90, FOV = 240 mm, matrix = 64 × 64, slice thickness/gap = 5 mm/0 mm, 20 axial slices) was used. A flow-sensitive conventional gradient-echo sequence (TR/TE/flip angle = 50 ms/4 ms/60) was used for high-resolution anatomic images. Four different wordgeneration paradigms were used; noun, verb, adjective, and adverb. Simple eight phrases/sentences were presented at a rate of one per 3 s during each activation period of 24 s. Rest periods consisted of reading pseudowords, to say the meaningless letters during 24 s. All tasks consisted of 48 phases including three activation and three rest periods. All tasks were visual-guided by LCD projector. All EPI data sets were realigned and normalized to the EPI template image included in the SPM99 package. Data sets were then smoothed with a gaussian filter (FWHM = 6 mm). For group analysis, random-effect model was used, and the activation maps were generated with a height threshold of uncorrected p = 0.05 and an extent threshold of k = 100 voxels. For individual analysis, correlation coefficient analysis was performed using the reference function of a box car (p < 0.05). Results: Normal volunteers used left middle frontal, left parietal, and right medial frontal areas more actively than left medial temporal lobe epilepsy patients in all four language tasks. In comparison, left medial temporal lobe epilepsy patients used left occipital, left superior temporal, and right frontal areas more actively. However, this difference depends on the language tasks. Conclusions: Left medial temporal lobe epilepsy patients had different language organization from the normal volunteers. 1.247 REDUCED DEGREE OF LANGUAGE DOMINANCE IN LEFT HEMISPHERE LOCALIZATION-RELATED EPILEPSY ASSESSED BY A FMRI READING-COMPREHENSION TASK 1 William D. Gaillard, 1 Bonnie C. Sachs, 1 Lyn M. Balsamo, 2 Ben Xu, 1 Madison M. Berl, 2 Cecile B. Grandin, 1 Phillip L. Pearl, 1 Joan A. Conry, 1 Steven L. Weinstein, 3 Frank J. Ritter, 2 Susumu Sato, and 2 William H. Theodore (1 Neurology, Children’s National Medical Center, Washington, D.C.; 2 Clinical Epilepsy Section, NINDS, NIH, Bethesda, MD; and 3 Minnesota Epilepsy Group, St. Paul, MN) Rationale: Functional magnetic resonance imaging (fMRI) provides a noninvasive means to assess the association of epilepsy and the reorganization of language networks. We investigated the extent of language
Epilepsia, Vol. 44, Suppl. 9, 2003
dominance in patients with left hemisphere localization-related epilepsy using a fMRI reading-comprehension task. Methods: We studied 30 patients [24 right-handed, three left-handed, three ambidextrous; 17 M, 13 F; aged 8.7–56 years (mean age, 21.1 years; mean age at seizure onset, 9.6 years)], and 32 healthy volunteers [all right-handed; 17 M, 15 F; aged 7.9–34.8 years (mean age, 22.4 years)], using whole-brain 1.5-T fMRI (EPI BOLD) and a covert reading task— naming to description—employing a box car design. Data were analyzed with a region-of-interest analysis from t-maps. The number of activated voxels was determined in inferior frontal gyrus (IFG), mid-frontal gyrus (MFG), and Wernicke’s area (WA), using a semiautomated program. An asymmetry index (AI) was calculated [(L−R)/(L+R)] for each region at t = 4. fMRI language laterality was defined as |AI| >0.20. Groups were compared using unpaired t tests. Results: Seven of 30 patients had atypical (bilateral or right hemisphere dominant) language representation (23.0%). Regional AIs for patients were less than controls (IFG, 0.80 vs. 0.43; MFG, 0.94 vs. 0.34; WA, 0.84 vs. 0.47; all p < 0.005). Excluding atypical language patients (IFG, 0.13; MFG, −0.04; WA, −0.02), AIs for left hemisphere focus/left hemisphere dominant patients remained lower than for controls (IFG, 0.51; MFG, 0.45; WA, 0.59; all p < 0.02). Motion measures did not differ between groups. Conclusions: Patients with left hemisphere localization-related epilepsy have a greater likelihood of atypical language representation. Lower regional AIs in left dominant patients suggests increased recruitment of homologous right areas for language processing compared with healthy controls. (Supported by NINDS KO8-NS1663, NICHD P30HD40677, NINDS R01 NS44280, Clinical Epilepsy Section, NINDS, NIH.) 1.248 CAN FMRI AT 3 TESLA PREDICT EPILEPSY SURGERY OUTCOME? 1,2 Jerzy P. Szaflarski, 3 Scott K. Holland, 3 Vincent J. Schmithorst, 2,4 Paula K. Shear, 1 William T. Cahill, 1 Michael D. Privitera, 1 David M. Ficker, 1 Jennifer Cavitt, and 2,5 Stephen M. Strakowski (1 Department of Neurology and 2 Center for Imaging Research, University of Cincinnati, 3 Imaging Research Center, Children’s Hospital Medical Center, and 4 Department of Psychology and 5 Department of Psychiatry, University of Cincinnati, Cincinnati, OH) Rationale: This preliminary study aimed to assess the ability of functional magnetic resonance imaging (fMRI) to lateralize language and memory in epilepsy patients undergoing presurgical evaluation and to compare the outcomes based on memory lateralization with either intracarotid amobarbital procedure (IAP) or fMRI. Methods: Thirty-seven healthy and seven epilepsy subjects underwent fMRI using two fMRI-language [verb generation (VG), semantic decision/tone decision (SDTD)] and two fMRI-memory tasks [picture encoding (PE) and word retrieval (WR)] presented in block-design paradigm along with control tasks. Images were acquired on a 3-T Bruker MRI scanner. Regions of interest (ROIs) for each task were defined based on regional activations in healthy subjects identified from the global composite map for each task (through cross-correlation analysis). The ROIs were then used to calculate language and memory laterality indices [LI = (L−R)/(L+R)] in healthy and epilepsy subjects. We accepted fMRI of memory LI ≤ 1SD to be symmetric (b), LI between 1 and 2 SD to be mildly (m) lateralizing, and LI ≥ 2SD to be strongly (s) lateralizing. LI for IAP of language and memory were calculated based on hemispheric scores using the formula above. Language and memory lateralization with IAP and fMRI and the predictions of surgical outcome based on memory lateralization with either IAP or fMRI were compared. Results: In healthy subjects, the LI for VG = 0.19 ± 0.08; for SDTD = 0.26 ± 0.12; for PE = −0.05 ± 0.06; WR = −0.01 ± 0.09. In epilepsy patients, there was agreement between language lateralization with IAP and fMRI (for VG R = 0.853; p = 0.015; for SDTD R = 0.91; p = 0.004). All patients underwent surgical resection. In the first four patients, IAP of memory was strongly lateralizing (predicting seizure-free outcome), whereas fMRI was symmetric in one patient and mildly lateralizing (PE) or symmetric (WR) in another patient (both had not-seizure-free outcome). In two other patients, fMRI was strongly lateralizing and predicted seizure-free outcome. Outcome data are not yet available for the last three patients (see Table 1).
AES PROCEEDINGS
89
TABLE 1. IAP : Language 0.53 0.69 0.82 0.8 1 −0.2 1
fMRI : VG 0.24 0.22 0.37 0.14 0.19 −0.3 0.23
fMRI : SDTD 0.22 0.18 0.4 0.23 0.26 −0.28 0.26
IAP : Memory
fMRI : PE
0.23 (L) 1 (L) 1 (L) −0.62 (R) −0.18 (R) −0.68 (R) 0.23 (L)
0.06 (mL) 0.00 (b) 0.08 (sL) 0.03 (mL) −0.09 (b) 0.13 (sL) −0.05 (b)
fMRI : WR
Seizure onset
Outcome (Engel)
a
R MT R MT R MT LT L MT L MT L FT
I III I III
0.08 (b) 0.22 (sL) −0.01 (b) a
0.25 (sL) −0.02 (b)
IAP, intracarotid amobarbital procedure; fMRI, functional magnetic resonance imaging; VG, verb generation; SDTD, semantic decision/tone decision; PE, picture encoding; WR, word retrieval. a Excessive motion artifact; L, left; R, right. Conclusions: Preliminary data suggest that fMRI of language and memory at 3T may augment presurgical staging in patients with refractory epilepsy. Although fMRI of memory was better in predicting seizure-free outcome than IAP in our study, larger groups of patients need to be evaluated before fMRI replaces IAP. [Supported by The Neuroscience Institute, Cincinnati, OH (J.P.S.), NIH RO1-HD38578-03 (S.K.H.).] 1.249 FUNCTIONAL MRI IN ASSESSMENT OF LANGUAGE DISTRIBUTION IN REFRACTORY PATIENTS UNDER INVASIVE MONITORING Kenou van Rijckevorsel, C´ecile Grandin, and Marianne de Tourtchaninoff (Reference Center for Refractory Epilepsy, Universit´e Catholique de Louvain, Brussels, Belgium) Rationale: Functional MR imaging (fMRI) is considered a valid and noninvasive method in assessing language localization in epilepsy patients before neurosurgery. This test could be repeated and adapted to the cognitive level of the patient. It also gives more precise information than the Wada test, and the results can be combined with those of other investigations such as imaging modalities and EEG. Methods: From our series of 35 epilepsy patients who underwent a complete presurgical evaluation [EEG-video monitoring, with subdural grids in five of them, structural MRI, interictal fluorodeoxyglucose– positron emission tomography (FDG-PET), cognitive tests, IQ, and fMRI language study], we selected six patients in whom the fMRI study was crucial for the surgical decision. This group (four M and two F, aged 24– 49 years) included four patients with a cryptogenic epilepsy, three with a frontal focus (two left and one right), and one with a right parietotemporal focus, one patient with a left temporal (T1) tumor, and one patient with a dual pathology (a post-traumatic lesion on the right parieto-temporal region and a right mesiotemporal sclerosis). All patients with an invasive monitoring underwent a 3D GRE T1 -weighted MR sequence with the subdural grids in place. The fMRI protocol included at least a structural T2 -weighted sequence and whole-brain axial gradient echo echo planar sequences for the functional tasks (TR/TE 3000/50 ms; voxel size, 3.6 mm isometric). The language paradigms were all covert and consisted of a verbal fluency task, a reading task, and an auditory task. Each paradigms alternated six epochs of rest with six epochs of task condition, for a total duration of 4 min 48 s. Statistical t maps of brain activation were generated with the SPM software and superimposed on the T2 -weighted images (slices) and on the 3D T1 -weighted images with the grids (surface rendering), after coregistration and with the knowledge of the results of the EEG monitoring. Results: The patient with the left T1 tumor has his Wernicke area on left T2 and T3. He has been operated on without any sequelae. One patient with left frontal focus has his epileptogenic zone near and on Broca area and was considered a candidate for VNS instead of ressection. One patient had a bilateral localization of the Wernicke area but more anterior than the parietotemporal focus. One patient with a right parietotemporal focus has his language exclusively lateralized to the left. One patient with a left frontal lobe epilepsy has been improved since his invasive monitoring and prefers to wait before surgery. The last patient with a right frontal focus will be operated on in July 2003. Conclusions: fMRI is reliable and useful for language areas assessment. It can be superimposed to other presurgical technicques and gives
more precise information than the Wada test for language. fMRI should be part of the presurgical evaluation for refractory epilepsy patients or patients with brain tumor to assess the risk and/or the possibility to operate on these candidates. 1.250 LANGUAGE LATERALIZATION IN FOCAL EPILEPSY: INCREASED REGIONAL CALLOSAL THICKNESS WITH STRONGLY LATERALIZED FUNCTION 1 Amanda G. Wood, 2 Michael M. Saling, 3 Graeme D. Jackson, and 1 David C. Reutens (1 Medicine, The University of Melbourne, Austin and Repatriation Medical Centre, Heidelberg, 2 Psychology, The University of Melbourne, Parkville, and 3 Brain Research Institute, Austin and Repatriation Medical Centre, Heidelberg, Victoria, Australia) Rationale: Corpus callosum size correlates with a variety of indices of laterality such as handedness and sex. These relations are said to reflect the role of the callosum in language representation, but few studies to date have examined this proposition directly. Furthermore, studies of the corpus callosum tend to make a priori decisions about its regional topography or make relatively coarse measurements that preclude a refined understanding of the relation between the structure and any functional correlate. The present research examined the relation between functional magnetic resonance imaging–determined language representation in patients with focal epilepsy and regional callosal morphology. Methods: Functional magnetic resonance imaging of language was performed using an orthographic lexical retrieval paradigm and a laterality index was derived from the number of significantly activated pixels in the left and right frontal lobes. All subjects underwent high-resolution T1 -weighted MR images. The images were placed in a standard stereotaxic coordinate space and, using a semiautomated algorithm, the corpus callosum was segmented, and callosal thickness was measured at 39 equally spaced nodes along its length. The latter measurement took into account the inherent shape of the callosum. Results: Correlational analysis of the absolute value of the laterality scores, which reflect the degree of asymmetry rather than its direction, and callosal-thickness measurements showed significant positive relations at nodes 11 to 13 and node 30. The former region corresponds to the anterior midbody, whereas the latter is within the isthmus. Conclusions: The findings support the notion that interhemispheric fibers that traverse the corpus callosum inhibit homologous cortical regions during well-lateralized tasks. The isthmus of the corpus callosum has previously been implicated in language processing. Our data provide support for this topographic specialization within the callosum.
Neuropharmacology 1.251 USEFULNESS OF LEVETIRACETAM IN EPILEPTIC SYNDROMES WITH CONTINUOUS SPIKES AND WAVES DURING SLEEP: PRELIMINARY RESULTS Alec Aeby, Denis Verheulpen, and Van Bogaert Patrick (Pediatric Neurology, Hopital Erasme, Brussels, Belgium) Rationale: Children with continuous spikes and waves during sleep (CSWS) display neurologic or neuropsychological dysfunction Epilepsia, Vol. 44, Suppl. 9, 2003
90
AES PROCEEDINGS TABLE 1. Patient no./age (yr)/sex Etiology Neurologic status Concomitant AEDs % CSWS before/after LEV Clinical improvement
1/4/F
2/9/F
3/10/F
4/14/M
Cryptogenic Psychomotor delay, ataxia VPA-CLB 95/92 Yes
Cryptogenic IQ = 50 VPA-CLB 93/96 No
Cryptogenic IQ 0.21 (Stapleton FB et al. Pediatrics 1982;69:594–7). Three of 40 children (7.5%) with baseline hypercalciuria developed nephrocalcinosis/lithiasis, necessitating discontinuation of TPM. An additional five children with baseline hypercalciuria did not progress to nephrocalcinosis/lithiasis. There was no statistically significant difference in the mean urine Ca/Cr ratio, age at initiation of TPM therapy, or length of time taking TPM between patients with nephrocalcinosis/lithiasis and those with hypercalciuria alone. Thirty-one children had baseline serum HCO3 levels before the initiation of TPM. The mean ± SEM serum HCO3 level decreased from 24.3 mEq/L ± 0.8 before TPM to 22.8 mEq/L ± 0.8 at 2–6 months after initiation of therapy (p = 0.047). Conclusions: 1. Hypercalciuria is a common finding in children starting TPM therapy for epilepsy. 2. Nephrocalcinosis/lithiasis occurred in 7.5% of our patients taking TPM for seizures, as opposed to the 1.5% frequency of nephrolithiasis reported among adults taking TPM. However, there is no medical literature regarding the incidence of nephrocalcinosis in adults taking TPM. 3. We observed a decrease in serum HCO3 levels after 2–6 months of TPM therapy, consistent with its inhibitory effect on carbonic anhydrase. 4. The long-term renal response to hypercalciuria and increased urinary HCO3 from TPM may be associated with nephrocalcinosis and/or nephrolithiasis in children taking this drug.
1.252 NEPHROLITHIASIS AND NEPHROCALCINOSIS IN CHILDREN TAKING TOPIRAMATE FOR SEIZURES 1,4 Sarah M. Barnett, 2,4 Anthony H. Jackson, 3,4 Gregory L. Braden, 5 Jane L. Garb, 2,4 Herbert E. Gilmore, 2,4 Beth A. Rosen, and 2,4 Dina H. Kornblau (1 Pediatrics, 2 Pediatrics and Neurology, and 3 Renal Division, Baystate Medical Center, Springfield, 4 Tufts University School of Medicine, Boston, and 5 Baystate Medical Center, Springfield, MA)
Rationale: Zonisamide (ZNS) is FDA approved for treatment of focal seizures in adults and has been reported to be effective in treatment of seizures in children. We sought to determine the efficacy of ZNS after 12 months of treatment in children categorized by seizure type, cognitive status, concomitant anticonvulsants (AEDs), and prior AED use. Methods: This is a retrospective study of 35 children (ages 8 months to 22 years; mean age, 9 years) with intractable seizures treated with ZNS. Seizure frequency was calculated before and after 12 months of ZNS therapy. Charts were reviewed for seizure type (focal, generalized, mixed, or infantile spasms), presence of cognitive impairment (no special education vs. special ed.), concomitant AEDs (range, 0–5), and number of prior AEDs (range, 1–12). Results: Of 35 patients, five (14%) were classified as having focal seizures, 13 (37%) had generalized seizures, 14 (40%) had mixed seizure types, and three (9%) had infantile spasms. At 12 months, 16 children were no longer taking ZNS because of lack of efficacy or side effects. Good to excellent seizure control (50–100% reduction) was attained in two (40%) patients with focal seizures, seven (54%) with generalized seizures, five (36%) with mixed seizures, and one (33%) with infantile spasms. Good to excellent control was achieved in five of 11 (45%) patients with no/mild cognitive impairment and in 10 of 22 (45%) with moderate/severe cognitive impairment. (No cognitive information on two
Rationale: As a new antiepileptic drug, topiramate (TPM) effectively treats multiple types of seizures in children; known adverse effects include cognitive deficits and somnolence. Notably, in adult patients treated with this drug, nephrolithiasis is estimated to occur in 1.5% of patients [Wasserstein AG et al. Epilepsia 1995;36(suppl 3):S153]. However, the frequency of nephrolithiasis in children taking TPM is unknown. Additionally, little is known about the frequency of nephrocalcinosis in both children and adults treated with TPM. Possible mechanisms for nephrocalcinosis/lithiasis include hypercalciuria and increased urinary bicarbonate (HCO3 ) excretion from the inhibitory effect of TPM on carbonic anhydrase. Methods: We conducted a retrospective chart review of 40 consecutive children who initiated TPM therapy for seizures between January 1997 Epilepsia, Vol. 44, Suppl. 9, 2003
1.253 EFFICACY OF ZONISAMIDE AT 12 MONTHS IN CHILDREN CLASSIFIED BY SEIZURE TYPE, COGNITIVE STATUS, AND PRIOR AND CONCOMITANT ANTICONVULSANT DRUG USE 1 Marjorie E. Bunch, 2 David E. Mandelbaum, 1 Steven L. Kugler, 1 Anuradha Venkatasubramanian, and 1 Jan B. Wollack (1 Pediatrics, UMDNJ-Robert Wood Johnson Medical School, New Brunswick, NJ; and 2 Clinical Neurosciences, Brown Medical School, Providence, RI)
AES PROCEEDINGS children). Among those children taking ZNS and one other drug, five of nine (56%) had excellent (>90%) control, and two of nine (22%) had poor (6 months. Major side effects were only anorexia or insomnia. The average dosage was ∼1,200 mg/day. An average of 3.2 drugs had failed before patients tried Felbatol. No serious hepatic or hematologic effects were noted. Felbatol was able to be used as monotherapy in 27% of these patients, and seizure control improved in 11of 15 patients. Conclusions: FBM is still a viable choice in refractory epilepsy in children. Long-term use was very safe in our clinical series, with no
91
hepatic or bone marrow injuries noted. (Supported by MedPointe Pharmaceuticals.) 1.255 OXCARBAZEPINE-RELATED HYPONATREMIA IN PEDIATRIC PATIENTS: BRAIN TUMORS ARE A RISK FACTOR Jennifer Disabato, Paul Levisohn, and Pramote Laoprasert (Division of Child Neurology, the Children’s Hospital, University of Colorado Health Sciences Center, Denver, CO) Rationale: Although risk factors for oxcarbazepine (OXC)-related hyponatremia (HN) have been determined for adult patients, risk factors for HN in children treated with OXC have not been defined. Two children with a history of brain tumors developed HN, defining some risk factors for OXC-associated HN in children. Methods: The charts of two patients treated at our institution in whom HN developed were reviewed. Results: Two patients, both boys, developed clinically significant HN while receiving OXC for complex partial seizures. Both had developed brain tumors in the remote past but were free of active disease. Both were neurologically impaired as a result of the tumors and treatment. Both patients were mildly symptomatic with altered mental status. Patient 1 was treated with increased salt intake and has remained on OXC. His most recent serum Na+ was 137 mM. OXC was discontinued in patient 2, with his most recent Na+ also 137 mM. Conclusions: HN occurs in ∼2.5% of adults treated with OXC. Risk factors include older age and treatment with diuretics or other sodiumwasting drugs. Most cases occur within 3–4 months of initiation of OXC. We have seen only two patients with OXC-related HN. Common features include remote history of brain tumors and panhypopituitarism requiring supplementation. Of note is the relatively delayed onset of HN in one patient. In one patient, HN resolved without medication discontinuation. Although uncommon, HN may occur in pediatric patients receiving OXC. Sequelae of CNS tumors appear to be a risk factor, and such patients should be followed up for possible development of HN (Table 1). 1.256 NEUROBEHAVIORAL ADVERSE EVENTS (AES) WITH TOPIRAMATE MONOTHERAPY IN CHILDREN AND ADOLESCENTS 1 Dennis J. Dlugos, 2 Liza Squires, and 2 Steven Wang (1 Division of Neurology, The Children’s Hospital of Philadelphia, Philadelphia, PA; and 2 Johnson & Johnson Pharmaceutical Research & Development, Raritan, NJ) Rationale: Neurobehavioral effects of topiramate (TPM) have been the subject of considerable discussion since early studies of high-dose TPM as add-on therapy in adults. However, such adverse events (AEs) were generally lower in children, perhaps due to the use of relatively lower doses and more gradual titration. TPM as monotherapy in newly diagnosed epilepsy allows neurobehavioral AEs to be evaluated without antiepileptic drug (AED) cotherapy as a confounder, albeit without placebo to control for background events. We focus on neurobehavioral
TABLE 1. Clinical characteristics Patient 1 11 year
Age Tumor Treatment
Astrocytoma Chemotherapy
Seizures
Complex partial seizures with secondary generalization Clonazepam, hydrocortisone, growth hormone, levothyroxine 20 mg/kg/day 9 month 123 mM
Concomitant medications OXC dose Time from initiation to HN Lowest serum Na+
Patient 2 16 year Malignant ependymoma Chemotherapy, radiation, bone marrow transplantation Complex partial seizures Gabapentin, testosterone, hydrocortisone, levothyroxine 25 mg/kg/day 2 month 120 mM
OXC, oxcarbazepine.
Epilepsia, Vol. 44, Suppl. 9, 2003
92
AES PROCEEDINGS TABLE 1.
Concentration/attention difficulty Mood problems Memory difficulty Cognitive problems NOS Psychomotor slowing Aggressiveness Depression Language problems
43 mg Flex (n = 95)
204 mg Flex (n = 14)
364 mg Flex (n = 77)
% Patients (N) 98 mg Fixed (n = 30)
197 mg Fixed (n = 29)
Overall
5 (5) 2 (2) 1 (1) 1 (1) 2 (2) 1 (1) 0 0
7 (1) 0 14 (2) 0 0 7 (1) 0 7 (1)
10 (8) 8 (6) 3 (2) 6 (5) 3 (2) 1 (1) 3 (2) 1 (1)
7 (2) 3 (1) 13 (4) 0 7 (2) 3 (1) 0 0
21 (6) 7 (2) 0 3 (1) 3 (1) 7 (2) 10 (3) 7 (2)
9 (22) 4 (11) 4 (9) 3 (7) 3 (7) 2 (6) 2 (5) 2 (4)
AEs reported in children. AEs in TPM trials are coded with a WHOART dictionary modified to reflect symptoms reported in early add-on TPM studies. Such terms as concentration/attention or memory difficulty and speech/language problems are unique to the TPM dictionary compared with the conventional WHOART dictionary. Methods: Three randomized double-blind trials evaluated TPM monotherapy in newly diagnosed epilepsy. In two studies, patients were randomized to 50 or 500 mg/day (25 or 200 if ≤ 50 kg) or to 50 or 400 mg/day. Dose reductions were allowed to manage AEs (“flex” dosing), although dose reductions before patients achieved 200 mg/day were not allowed in patients randomized to TPM 400. In a third study, patients were randomized to fixed doses of the investigator’s choice of carbamazepine (CBZ) or valproate (VPA) or to 100 or 200 mg TPM; dosage reductions were not allowed. Study duration was 4–6 months after the last patient randomized. Data are reported for the intent-to-treat population. Results: Of 1,131 patients, 245 were children/adolescents (6– 15 years). Table 1 presents mean doses, patient number, and neurobehavioral AEs of special interest. Conclusions: Although numbers are relatively small in individual groups and subject to considerable variation by an event’s chance occurrence in even one patient, as demonstrated by the variability in memory difficulty, data suggest a pattern in which higher doses, particularly if fixed, are less well tolerated. However, 50 and 200 mg TPM are generally associated with a low incidence of neurobehavioral AEs when doses are adjusted according to patient response. The recommended initial target dose, based on efficacy and tolerability, is 100 mg. Even in a fixed-dose study, 100 mg TPM was well tolerated. The incidence of neurobehavioral AEs with this initial target dose would be expected to be even lower with flex dosing. Studies in larger pediatric populations are warranted. (Supported by Johnson & Johnson Pharmaceutical Research & Development.)
1.257 OXCARBAZEPINE TOLERABILITY IN CHILDREN YOUNGER THAN 12 MONTHS Korwyn L. Williams, Atefeh Hosseini, and Angus A. Wilfong (Section of Pediatric Neurology, Department of Pediatrics, Baylor College of Medicine, Houston, TX) Rationale: Oxcarbazepine (OXC) is one of the newer antiepileptic drugs (AEDs) indicated for adjunctive therapy in the pediatric population and is generally well tolerated. Methods: We reviewed the electronic medical database to identify children who had been placed on OXC by age 12 months to assess safety and tolerability in this age group. All children had been evaluated by a pediatric neurologist and diagnosed with epilepsy. Where available, patient characteristics were reviewed, including age at time of OXC; ILAE classification; electroencephalographic (EEG) signature; dose; side effects; and concomitant AED use. Results: Twenty-seven patients were identified, whose age at the time of OXC initiation ranged from 1 week to 12 months. Thirteen of the patients were designated as cryptogenic localization-related; 12 were symptomatic localization-related; one was cryptogenic generalized; and one was symptomatic generalized. The maximum dose in each patient ranged from 12 to 70 mg/kg/day. Sixteen of the patients had never been Epilepsia, Vol. 44, Suppl. 9, 2003
exposed to an AED. Eight had normal EEGs; five had focal epileptiform abnormalities; and eight had multifocal epileptiform abnormalities. Two reported sleepiness as a side effect; one was taking 60 mg/kg/day and concomitant phenobarbital (PB) and primidone (PRM), whereas the other was having daily infantile spasms. One patient was switched to another AED, because of an asymptomatic hyponatremia. One patient was thought to be irritable, but had been chronically malnourished and abused. Conclusions: Overall OXC appears to be well tolerated in patients younger than 1 year. 1.258 EFFECT OF LEVETIRACETAM ON BEHAVIOR AND ALERTNESS IN CHILDREN WITH REFRACTORY EPILEPSY 1 Lieven G. Lagae, 1 Gunnar Buyse, and 2 Berten Ceulemans (1 Paediatric Neurology, University Hospitals KULeuven, Leuven, and 2 Epilepsy and Revalidation Centre, Pulderbos, Belgium) Rationale: Levetiracetam (LEV) is a promising antiepileptic drug (AED) in childhood epilepsy, with proven efficacy both in partial (Glauser, 2002; Wheless, 2002) and generalized seizures (Lagae, 2003). We evaluated the side effects of LEV in a cohort of 39 children with refractory epilepsy. Because alertness, communication, interaction, and behavior are important issues in the quality of life in children with refractory epilepsy and mental retardation, special attention was paid to these particular items. Methods: In 39 children with seizures that were refractory to at least one AED, LEV was introduced as add-on medication. LEV dosage started at 10 mg/kg/day with increases every 4–7 days of 10 mg/kg up to a maximum of 60 mg/kg/day. Maximum dosage depended on individual efficacy and tolerability. In the first 12 weeks, baseline AEDs remained unchanged. At 16–20 weeks after introduction, we assessed both efficacy and side effects, using a standardized questionnaire. Results: The median age of the patients was 5 years (range, 5 months– 16 years). Thirty of 39 children were taking two or more AEDs at the time of inclusion. Epilepsy syndromes included West syndrome (n = 2), Lennox–Gastaut syndrome (n = 7), cortical malformations (n = 9), and postlesional epilepsy (n = 3). Thirty-four of 39 had moderate or severe mental retardation. The median LEV end dosage was 34 mg/kg/day (range, 10–60 mg/kg/day). Overall, 40% of the children had a seizure reduction of >50%. Three patients became seizure free. At 16 weeks, 31 of 39 children were still taking LEV. Side effects were rare: somnolence (n = 2), aspecific headache (n = 1), and increased behavioral problems (n = 2). In only one patient was this the cause of early withdrawal of LEV. These five patients were all taking a LEV dosage >40 mg/kg/day. In nine children, the parents reported a significant increase of alertness, making communication and interaction with these children clearly better. This positive effect was not related to LEV dosage or control of the seizures. In four of these nine children, the parents decided to continue the medication solely because of this positive side effect, and despite a relatively poor seizure control. In 13 children, the parents reported a better behavior, again not related to seizure control or LEV dosage. Conclusions: We conclude that LEV has a positive efficacy and safety profile in children with refractory epilepy syndromes. In a substantial percentage of the children, a clear positive effect was reported on alertness and behavior, thereby increasing interaction and communication.
AES PROCEEDINGS 1.259 EFFICACY OF LEVETIRACETAM AT ONE YEAR IN CHILDREN CLASSIFIED BY SEIZURE TYPE, COGNITIVE STATUS, AND PRIOR ANTICONVULSANT DRUG EXPOSURE 1 David E. Mandelbaum, 2 Marjorie Bunch, 2 Steven L. Kugler, 2 Anuradha Venkatasubramanian, and 2 Jan B. Wollack (1 Clinical Neurosciences, Brown Medical School, Providence, RI; and 2 Pediatrics, UMDNJ-Robert Wood Johnson Medical School, New Brunswick, NJ) Rationale: Levetiracetam (LEV) is FDA approved for treatment of focal seizures in adults and has been reported to be effective in the treatment of seizures in children. We sought to determine the efficacy of LEV at 1 year in children classified by seizure type, cognitive status, concomitant medications, and prior anticonvulsant (AED) use. Methods: This is a retrospective study of 59 children (ages 9 months to 23 years; mean age, 11 years) with intractable seizures treated with LEV. Charts were reviewed for seizure type (focal, generalized, or mixed), cognitive function (no special education vs. special ed), concomitant AEDs (range, none to five), and number of prior AEDs (range, 1–12). Seizure frequency was determined before and after 12 months of LEV therapy. Results: Of 59 patients, 15 (25%) had focal seizures, 29 (49%) had generalized seizures, 13 (22%) had mixed seizure types, and two were not classified. At 12 months, 18 children were no longer taking LEV because of lack of efficacy or side effects. Good to excellent seizure control (50– 100% reduction) was attained in six (40%) patients with focal seizures, 16 (55%) with generalized seizures, and eight (61%) with mixed seizures. Good to excellent control was achieved in seven of 15 (47%) patients with no/mild cognitive impairment and in 23 of 43 (53%) patients with moderate/severe cognitive impairment. (No cognitive information in one child). Good to excellent control was accomplished in 16 of 24 (67%) patients with prior exposure to fewer than four AEDs and in 15 of 35 (43%) previously treated with four or more drugs. Among those children taking LEV and one other drug, eight of 20 (40%) had excellent (>90%) control, and three of 20 (15%) had poor (50% reduction in seizure frequency, including 14 (21%) seizure free; eight (12%) for ≥6 months. Fifteen (23%) had no significant improvement, four (6%) had increased seizures; 10 (15%) patients reported adverse events, leading to withdrawal in eight (12%); there were no serious adverse events. Conclusions: LEV appears effective and well tolerated in a variety of intractable childhood epilepsies. Further systematic clinical audit of additional patients is under way, which we hope will allow subgroup analysis. (This preliminary work has received no external funding. We will be seeking funding to extend the data collection and analysis.)
1.261 TOPIRAMATE AS FIRST-LINE THERAPY: FINDINGS FROM CHILDREN/ADOLESCENTS WITH NEWLY DIAGNOSED EPILEPSY 1 Jong M. Rho, 2 Santiago Arroyo, 3 Liza Squires, 3 Steven Wang, and 3 David Jacobs (1 Department of Pediatrics, ZC 4482, University of California, Irvine Medical Center, Orange, CA; 2 Medical College of Wisconsin, Milwaukee, WI; and 3 Johnson & Johnson Pharmaceutical Research & Development, Raritan, NJ) Rationale: Most double-blind, randomized controlled trials of antiepileptic drug (AED) monotherapy in newly diagnosed epilepsy have been conducted in adults; data for children are limited, especially with newer AEDs. A recent dose-comparison study of topiramate (TPM) as monotherapy in untreated epilepsy enrolled patients age 6 years or older, thereby providing an opportunity to evaluate the efficacy of TPM as firstline therapy in the subset of children/adolescents with newly diagnosed epilepsy. Methods: Children and adults in whom epilepsy diagnosis had been confirmed ≤3 months before study entry were eligible provided they had one or two partial-onset (POS) or primary generalized tonic–clonic (PGTC) seizures in the 3-month retrospective baseline. Patients who had previously achieved seizure remission with AED therapy but relapsed when off AEDs were also eligible. Any AED being used for temporary/emergency purposes was withdrawn during a 7-day openlabel phase in which patients received 25 mg/day TPM. After randomization to double-blind treatment, patients remained in the study until their first POS or GTCS or until 6 months after the last patient was randomized. Results: Among 470 randomized patients, 151 were children/ adolescents (≤15 years), randomized to 50 mg/day TPM (n = 74) or 400 mg/day TPM (n = 77). Median time since epilepsy diagnosis, 1 month. Median treatment duration, 12 months. Kaplan–Meier analysis for time to first seizure showed a significant (p = 0.002) treatment effect favoring TPM 400. The treatment effect emerged within the first week of double-blind treatment and was statistically significant with 100 mg/day TPM. Based on Kaplan–Meier analyses, 6-month seizurefree rates were TPM 400, 90%; TPM 50, 78% (p = 0.04); 12-month seizure-free rates, TPM 400, 85%; TPM 50, 62% (p = 0.002). Cumulative incidence (≤2 years) for the most common adverse events were headache (21%), upper respiratory infection (17%), appetite decrease (13%), somnolence (12%), weight loss (12%), and dizziness (11%). Of these, only weight loss was dose-related, occurring more frequently in older children (16% of children 12–15 years old vs. 7% of children 50%), leading to toxicity in some patients. Conversely, a successful therapy with LTG and corresponding adequate SLs was possible in some other patients only after withdrawal of OXC. Illustrative case reports including the determination of glucuronidated LTG-metabolites in the urine in one patient will be presented. Conclusions: OXC and LTG are new AEDs with a hitherto widely neglected new spectrum of possible pharmacokinetic interactions. Clinicians must be careful with a combined use of OXC and LTG, especially Epilepsia, Vol. 44, Suppl. 9, 2003
96
AES PROCEEDINGS
when adding or withdrawing OXC in patients already treated with or remaining on LTG. REFERENCES 1. May TW, Rambeck B, J¨urgens U. Influence of oxcarbazepine and methsuximide on lamotrigine concentrations in epileptic patients with and without valproic acid comedication: results of a retrospective study. Ther Drug Monit 1999;21:175–81. 2. Sabers A, Buchholt JM, Uldall P, et al. Lamtotrigine plasma levels reduced by oral contraceptives. Epilepsy Res 2001;47:151–4.
1.268 MORNING VERSUS EVENING DOSING OF DIVALPROEX EXTENDED-RELEASE TABLETS DOES NOT MAKE A DIFFERENCE: A BIOPHARMACEUTIC AND SAFETY COMPARISON 1 Sandeep Dutta, 2 Ronald C. Reed, 1 Alex H.C. Chun, 3 Yiming Zhang, and 4 John H. Cavanaugh (1 Clinical Pharmacokinetics, 2 Neuroscience, 3 Statistics, and 4 Clinical Pharmacology, Abbott Laboratories, Abbott Park, IL) Rationale: Enteric-coated divalproex sodium (DVP; Depakote) is well known for the treatment of epilepsy. A novel once daily DVP extended-release (ER) tablet formulation (Depakote ER) has recently been approved. Informal clinician surveys indicate that nighttime daily dosing of DVP ER, rather than morning daily dosing, is frequently used to minimize the potential for adverse effects (AEs). To date, no information exists concerning DVP ER absorption profiles with nocturnal dosing. Methods: In a two-period, repeat-dose, randomized crossover study in 16 healthy subjects, DVP was given either as twice-daily 500-mg DVP or as a once-daily 1,000-mg DVP ER in the morning or evening. Valproic acid (VPA) concentration–time profiles were used to assess pharmacokinetics. Statistical analysis of pharmacokinetic parameters included ANOVAs. Safety was evaluated based on AEs, physical examinations, vital signs, and laboratory assessments. Results: The time of DVP ER dosing (i.e., morning vs. evening) had no significant effect on VPA bioavailability. All regimens were equivalent. Once-daily DVP ER significantly (p < 0.05) reduced mean peak levels (Cmax ) but produced similar mean trough levels (Cmin ) compared to the twice-daily DVP regimen. Mean steady-state VPZ exposure (AUC0−24 , mg/h/L), Cmax (mg/L) and Cmin (mg/L) were 1,771, 87.0, and 55.5 for 1,000-mg once-daily DVP ER morning dosing; 1,728, 85.0, and 57.4 for 1,000-mg once-daily DVP ER evening dosing; and 1,773, 100, and 54.0 for 500-mg twice-daily DVP regimens, respectively. AE frequency was similar between DVP ER and DVP groups for asthenia (n = 4), headache (n = 3), and somnolence (n = 3). However, gastrointestinal AEs occurred in two subjects taking DVP ER, but in seven with DVP. All AEs were mild or moderate; none caused withdrawal from the study. There were no significant differences in the pharmacokinetic parameters (p > 0.05) and safety between subjects who received the ER formulation in the morning and those who received the ER formulation in the evening. Conclusions: The results demonstrate the extended-release characteristics of the DVP ER formulation. This study indicates that once-daily treatment with DVP ER, given either in the morning or evening, may be a suitable replacement for twice-daily dosing with DVP. (Supported by Abbott Laboratories.)
1.269 ABSENCE OF EFFECT OF ADJUNCTIVE SPM 927 ON CONCOMITANT AED PLASMA CONCENTRATIONS IN SUBJECTS WITH PARTIAL SEIZURES 1 Nathan B. Fountain, 2 Rolf Horstmann, 2 Willi Cawello, 3 Pamela Doty, and 3 G. David Rudd (1 Department of Neurology, University of Virginia, Charlottesville, VA; 2 Clinical Pharmacology, Schwarz Biosciences GmbH, Monheim am Rhein, NRW, Germany; and 3 Clinical Development Neurology, Schwarz Biosciences Inc, Research Triangle Park, NC)
Epilepsia, Vol. 44, Suppl. 9, 2003
Rationale: SPM 927 (formerly harkoseride) was previously reported to reduce seizure frequency in a phase 2, multicenter, open-label trial in subjects with uncontrolled partial seizures taking one (25%) or two (75%) concomitant AEDs (study SP607). We report the steady-state plasma concentrations of concomitant AEDs during study SP607 to determine whether efficacy was due to elevations of these levels. Methods: The trial consisted of a 4-week baseline, a maximum 8week titration, and a 4-week maintenance period. SPM 927 started at 100 mg/day (50 mg b.i.d.) and increased weekly by 100 mg/day until either 600 mg/day or the maximum tolerated dose (MTD) was reached. Concomitant AED doses were constant throughout. To determine baseline steady-state values, AED levels were obtained at visit 1 and 30 min before the first dose of SPM 927 at visit 2. To determine whether SPM 927 influenced these levels, samples were again obtained at Cmax for SPM 927 (∼2–4 h after dose) at the end of the titration phase and at the end of the maintenance phase. Mean steady-state plasma levels were calculated for each concomitant AED. Plasma levels were measured by validated analytic methods using appropriate ranges of calibration. Samples were analyzed only for subjects with at least one pretreatment and two posttreatment samples. Differences between pretreatment and posttreatment AED levels were analyzed using ANOVA. Results: At least one dose of SPM 927 was taken by 91 subjects. A ≥50% reduction in seizure frequency occurred in 33%, and seven subjects were seizure free throughout the maintenance period, as reported previously. The median MTD for SPM 927 was 300 mg/day; ∼50% of subjects had an MTD of 400–600 mg/day. Mean (±SD) baseline AED plasma levels were carbamazepine (CBZ; 9.6 ± 2.9 mg/ml); CBZ-10,11epoxide (2.1 ± 0.8 mg/ml); gabapentin (15.6 ± 14.4 mg/ml); lamotrigine (7.5 ± 5.6 mg/ml); levetiracetam (34.1 ± 30.6 mg/ml); monohydroxy derivative of oxcarbazepine (26.6 ± 12.0 mg/ml); phenytoin (14.2 ± 6.5 mg/ml); and zonisamide (23.6 ± 11.8 mg/ml). ANOVA did not reveal any differences between plasma concentrations at visit 1 and plasma concentrations at the end of titration and the end of maintenance (p > 0.05). Conclusions: Plasma concentrations of concomitant AEDs remained stable during adjunctive treatment with SPM 927, suggesting that the observed seizure reduction in this trial is not the result of elevations of concomitant AED levels. However, these results require confirmation from an ongoing placebo-controlled trial with larger numbers of subjects. (Supported by Schwarz Biosciences.)
1.270 EFFECT OF DEINDUCTION ON VITAMIN D SERUM CONCENTRATIONS AFTER WITHDRAWAL OF CARBAMAZEPINE AND PHENYTOIN 1,2 Barry E. Gidal, 1 Mary E. Elliott, 3 Flavia M. Pryor, 3 Lauren Morgenroth, and 3 R. Eugene Ramsay (1 School of Pharmacy and 2 Department of Neurology, University of Wisconsin, Madison, WI; and 3 Department of Neurology, University of Miami, Miami, FL) Rationale: Osteoporosis and osteomalacia have been reported in patients with epilepsy, particularly those receiving enzyme-inducing AEDs (EIAEDs) such as phenytoin (PHT) or carbamazepine (CBZ). Although the mechanism(s) responsible for these osteopathies remain uncertain, an interaction with vitamin D has been proposed, whereby increased degradation secondary to enzyme induction occurs, resulting in vitamin D deficiency. Vitamin D is metabolically activated and degraded through the actions of three CYP450 isozymes. Studies of AED-mediated changes in vitamin D, however, have had inconsistent results. One recent study reported increased bone turnover, but no significant effect on vitamin D in patients treated with CBZ. Moreover, many studies are cross-sectional and do not address time-course issues or potential differential AED effects on vitamin D. It may instead be useful to explore potential drug-mediated effects by measuring vitamin D concentrations after deinduction, and conversion to a non-EIAED such as lamotrigine (LTG). Methods: A pilot study of changes in vitamin D serum concentrations in patients being converted from either CBZ or PHT monotherapy to LTG monotherapy was conducted. Patients selected for the study were on stable doses of either CBZ or PHT and were also taking LTG. Reduction of the EIAED was carried out weekly. Serum concentrations of 25hydroxy vitamin D (25-OHD) were measured before and 30 days after
AES PROCEEDINGS complete withdrawal of the EIAED. No attempt was made to alter diet or sun exposure during this period. 25-OHD was measured by ELISA. Results: Four patients were evaluated (three men and one woman). Two received CBZ (mean age and dose were 60.5 years and 450 mg/day), and two received PHT (mean age and dose, 42.5 years and 250 mg/day) at baseline. For the patients receiving CBZ, mean 25-OHD at baseline and after withdrawal of the EIAED was 19.1 versus 17.1 ng/ml, respectively (changes of +0.2 and −0.3 ng/ml for these two patients). For the PHT patients, mean baseline versus withdrawal 25 OHD values were 18.6 versus 22 ng/ml, respectively (changes of +3.9 and +2.9 ng/ml for these two patients). These changes represent an average 0.45% decline for the CBZ patients, versus an average 18.45% increase for PHT patients. Conclusions: These preliminary findings are consistent with previous observations that patients treated with EIAEDs tend to have low 25OHD serum concentrations. In addition, our observations suggest that a differential effect of CBZ versus PHT may exist. In the case of CBZ, no appreciable change was noted in 25-OHD, whereas for PHT patients, withdrawal resulted in marked increase at 30 days. Further studies will be conducted to verify these apparent differences, as well as to more fully describe the time-course of change.
1.271 SPM 927 DOES NOT INTERACT WITH VALPROIC ACID AND CARBAMAZEPINE 1 Rolf Horstmann, 1 Rainer Bonn, 1 Willi Cawello, 2 Pamela Doty, and 2 David Rudd (1 Clinical Pharmacology, Schwarz Biosciences GmbH, Monheim am Rhein, NRW, Germany; and 2 Clinical Development Neurology, Schwarz Biosciences Inc., Research Triangle Park, NC) Rationale: SPM 927 is a drug in clinical development for the treatment of epilepsy and neuropathic pain. Essential clinical pharmacologic data have been established in a series of phase 1 trials. Clinical development in epilepsy requires information about possible drug–drug interactions between SPM 927 and other antiepileptic drugs (AEDs). Methods: Two trials were designed to evaluate the influence of SPM 927 on valproic acid (VPA, trial SP601) and carbamazepine (CBZ, trial SP618) in 16 and 20 subjects, respectively. VPA was titrated up to 300 mg b.i.d. over 17 days, CBZ up to 200 mg b.i.d. over 22 days, and SPM 927 up to 200 mg b.i.d. over 7 (SP601) and 8 days (SP618). Two additional trials were designed to evaluate the influence of VPA (trial SP602) and CBZ (trial SP603) in 16 and 20 subjects, respectively: 200 mg b.i.d. SPM 927 was administered over 22 (SP602) or 17 (SP603) days; VPA was titrated up to 300 mg b.i.d. over 13 days or CBZ up to 200 mg b.i.d. over 8 days. Each trial was an open-label, randomized, multiple-dose trial. Study participants were healthy white male subjects (18–45 years), all with normal body weight. Results: SP601: Steady-state AUC (µg/h/ml) and Cmax (µg/ml) of VPA were 433 ± 84 and 41 ± 7.9 for VPA alone and 449 ± 89 and 42 ± 8.6 for VPA together with SPM 927. The point estimates (combination of SPM 927+AED/sole treatment ratio) were 1.04 and 1.01; the 90% confidence intervals were 99–109% and 97–107%, respectively. SP618: Steady-state AUC (µg/h/ml) and Cmax (µg/ml) of CBZ were 53 ± 8 and 4.7 ± 0.7 for CBZ alone and 57 ± 9 and 5.1 ± 0.8 for CBZ together with SPM 927. The point estimates were 1.07 and 1.09, the 90% confidence intervals were 104–111% and 104–114%, respectively. SP602: Steady-state AUC (µg/h/ml) and Cmax (µg/h/ml) of SPM 927 were 83 ± 14 and 9.5 ± 1.3 for SPM 927 alone and 83 ± 14 and 9.7 ± 1.2 for SPM 927 together with VPA. The point estimates were 1.0 and 1.01; the 90% confidence intervals were 98–103% and 96–107%, respectively. SP603: Steady-state AUC (µg/h/ml) and Cmax (µg/ml) of SPM 927 were 76 ± 17 and 8.5 ± 1.6 for SPM 927 alone and 82 ± 18 and 9.4 ± 2.2 for SPM 927 together with CBZ. The point estimates were 1.08 and 1.1; the 90% confidence intervals were 100–116% and 101–119%, respectively. The elimination half-life (t1/2 ) and Tmax of SPM 927, VPA, and CBZ remained unchanged after sole and combined treatment. Conclusions: The 90% confidence intervals of the AUC and Cmax point estimates were within the generally accepted bioequivalence ranges of 80–125%. SPM 927 had no influence on rate and extent of absorption of VPA or CBZ; VPA or CBZ did not influence the rate and extent of absorption of SPM 927. Thus, there is no indication for a drug– drug interaction between SPM 927 and VPA or CBZ. Pharmacokinetic data from longer-term studies of patients with epilepsy should be used
97
to confirm the absence of interaction observed in these phase 1 trials. (Supported by Schwarz Biosciences GmbH, Alfred-Nobel-Strasse 10, D-40789, Monheim am Rhein, NRW, Germany.) 1.272 ARE FDA BIOEQUIVALENCE STANDARDS SUFFICIENT FOR “NARROW THERAPEUTIC RATIO” AEDS? PHARMACOKINETIC COMPARISONS OF EXTENDED-RELEASE AND IMMEDIATE-RELEASE CARBAMAZEPINE FORMULATIONS 1 Gregory L. Krauss, 2 Hamzeh Fayez, and 1 Miller Akemi (1 Department of Neurology and 2 Department of Clinical Pharmacology, Johns Hopkins University, Baltimore, MD) Rationale: FDA bioequivalence standards may miss clinically important differences in drug delivery for AEDs that have narrow therapeutic ratios (NTRs). An example of the limitations of the current standards is the observation that many patients treated with extended-release carbamazepine (ER-CBZ; Carbatrol) have fewer CNS side effects than patients treated with immediate-release carbamazepine (IR-CBZ). The differences in tolerability between these formulations may be related to intra- and interindividual variability of absorption and concentrations over time. We compared pharmacokinetic (PK) studies of ER-CBZ and IR-CBZ to see whether these differences may be missed by current bioequivalence standards. Methods: The PK profile of ER-CBZ (Carbatrol) was modeled using individual data from three single-dose studies of Carbatrol provided by Shire Pharmaceuticals (Shire studies 107, 108, 112). The PK profile of IR-CBZ was modeled using data from 10 published studies. Steadystate elimination rate constant (Kel ) and absorption rate constant (Ka ) for Carbatrol and IR-CBZ were determined and used for PK modeling. We performed PK modeling for both formulations using a one-compartment model with a lag absorption phase (WinNonlin pharmacokinetic software). The Kel for Tegretol and Carbatrol were not significantly different, and therefore the mean Kel value of 0.04 was used to simulate drug delivery and produce the PK profile of both formulations. Results: Absorption constants (Ka ) for IR-CBZ derived from published PK studies were fivefold (Ka mean, 0.27–0.6) those for ER-CBZ (Carbatrol) (Ka mean, 0.07–0.10), reflecting the extended-release properties of Carbatrol. Individual variability in absorption was markedly increased, however, for IR-CBZ compared with Carbatrol: the standard deviation for individual absorption for IR-CBZ was ∼5 times that of ER-CBZ (Carbatrol). Pharmacokinetic modeling demonstrated similar total AUC/24 h for both formulations; however, fluctuations in daily concentrations (Cmax – Cmin )/Cav for IR-CBZ TID were twice those for ER-CBZ b.i.d. Conclusions: Individual absorption of IR-CBZ is highly variable compared with ER-CBZ and probably accounts for improved tolerability of ER-CBZ compared with IR-CBZ. Current FDA bioequivalence standards do not recognize the importance of uniform delivery for AEDs with narrow therapeutic ratios (Fig. 1). (Supported by Shire Pharmaceuticals, U.S.)
Epilepsia, Vol. 44, Suppl. 9, 2003
98
AES PROCEEDINGS
1.273 PHARMACOLOGICAL AND PATHOLOGICAL SIGNIFICANCE OF MDR1 EXPRESSION IN HUMAN EPILEPTIC BRAIN 1 Nicola Marchi, 1 Luca Cucullo, 2 Gabriel Moddel, 3 Annamaria Vezzani, 4 Christoph Baumgartner, 4 Susanne Pirker, 5 Thomas Czech, 5 Klaus Novak, 6 Giorgio Lo Russo, 6 Laura Tassi, 6 Carlo Galli, 1 Kerri Hallene, 1 Kelly Kight, 3 Massimo Rizzi, 3 Silvio Caccia, 3 Giovanna Guiso, and 1 Damir Janigro (1 Cerebrovascular Research and 2 Neurological Surgery, The Cleveland Clinic Foundation, Cleveland, OH; 3 Neuroscience, Mario Negri Inst, Milano, Italy; 4 Neurology and 5 Neurosurgery, University of Vienna, Vienna, Austria; and 6 Epilepsy Surgery, Claudia Munari, Milano, Italy) Rationale: Increased MDR1 expression of in endothelial cells (ECs) of epileptic human blood–brain barrier (BBB) supports the hypothesis that lack of appropriate pharmacodistribution leads to suboptimal CNS antiepileptic drug (AED) levels. MDR1 expression in astrocytes from epileptic brain is associated with loss of p53 signaling, suggesting that MDR1confers resistance to apoptosis. We investigated the functional significance of MDR1 expression in parenchymal cells from drug-resistant epileptic human brain. We tested the hypothesis that whereas MDR1 expression at the BBB correlates with poor AED penetration, neuronal and glial MDR1 expression relates to abnormal sensitivity to apoptosis. Methods: Brain samples were taken during surgical resections in patients (n = 14) administered with carbamazepine (CBZ). Astrocyte cultures were established from cortical tissue of epileptic (n = 8) or nonepileptic (n = 2) patients. Control and epileptic glial cells were exposed to therapeutic concentration of phenytoin (PHT) or doxorubicin, in the presence or absence of MDR inhibitors. Results: The brain-to-plasma ratio for CBZ was 1.3 ± 0.07 (n = 14) compared with distribution across the rodent BBB of 2.4 ± 0.4), suggesting limited CNS drug access in epilepsy patients. MDR1 mRNA expression in resected tissue revealed higher levels in samples with the largest accumulation of AED (p = 0.08). Levels of P-gp in epileptic versus normal glia revealed a fivefold increase in epileptic tissue. P-gp immunocytochemistry in brain sections from eight drug-refractory patients showed staining of ECs and parenchymal cells; 64% of neurons (n = 264) weere P-gp+ as compared with 91% of astrocytes (n = 107). Cultured epilpetic astrocytes exhibited a significant increase in AED extrusion as compared with astrocytes isolated from nonepileptic brain (n = 3), demonstrating a decreased PHT uptake by epileptic astrocytes compared with nonepileptic glia (15.03 ± 0.38 vs. 17.43 ± 0.39 fmol/cell). MDR inhibitors abolished these differences, indicating that active drug extrusion by epileptic glia was due to MDR. Cryopreserved sections from human epileptic tissue were immunostained for P-gp, and nuclear integrity was assessed by DAPI. Regardless of cell type, MDR1 expression was associated with DNA integrity: 85% of glia and 66% of neurons expressing MDR1 were devoid of DNA damage. Conclusions: These results suggest that AED extrusion by epileptic glia may ultimately enhance interstitial drug levels. Thus the diminished AED penetration into the brain may be determined by P-gp overexpression by epileptic BBB ECs, whereas P-gp in glia and/or neurons may subserve a novel role independent of drug resistance related to neuron/glia protection from apoptotic cell death. (Supported by HL51614, NS43284, NS38195.)
1.274 SECONDARY INEFFICACY IN LEVETIRACETAM: TOLERANCE? Heinz Joachim Meencke, Hans Beatus Straub, Christoph Dehnicke, Martin Merschhemke, and Doris Meincken-Jaeggi (Epilepsie-Zentrum Berlin-Brandenburg, Ev. Krankenhaus Koenigin Elisabeth Herzberge, Berlin, Germany) Rationale: Pharmacologic tolerance is predominantly associated with benzodiazepine drug therapy. In levetiracetam (LEV), several experimental data support the suggestion of pharmacologic tolerance in this drug. To obtain clinical evidence for this phenomenon is difficult, but nevertheless, the course of clinical treatment many times raises the question, whether the loss of efficacy in this drug is related to the development of tolerance. In this study we analyzed the different factors of secondary inefficacy and the contribution of tolerance to this effect. Epilepsia, Vol. 44, Suppl. 9, 2003
Methods: We analyzed retrospectively the patient charts from 205 patients with LEV therapy. The seizure frequency was established by the personal diary of the patients. Secondary inefficacy was analyzed in the group of patients with a minimum of 50% improvement in seizure frequency with LEV. Results: From 205 patients with LEV, we had 107 (52%) responders. From these 107 responders, 34 (32%) showed secondary inefficacy. The detailed analysis of the 34 patients with supposed secondary inefficacy showed the following results: 17 had a change in the basic medication, indicating a primary inefficacy of LEV. Eight cases had observational artefacts. Four cases had noncompliance, and only five cases had suggested development of tolerance. That indicates that only 4.8% of the responders had the suggestion of tolerance. These cases showed typical aspects of tolerance (e.g., with renewed improvement of seizure control after elevation of the drug dosage). Conclusions: The verification of drug tolerance in clinical settings is difficult. The subanalysis of secondary inefficacy showed that only 4.8% of responder-patients are supposed to develop tolerance. This shows the need for detailed analysis of the therapeutic course before diagnosing drug tolerance. 1.275 CONVERSION TO LAMOTRIGINE MONOTHERAPY FROM VALPROATE IN PATIENTS WITH EPILEPSY: A PRELIMINARY ANALYSIS 1 Mark Sale, 2 Alain Vuong, 2 Sheri Hoyler, 2 John A. Messenheimer, and 3 Anne E. Hammer (1 Pharmacokinetic Modeling and Simulation, 2 Epilepsy Clinical Development and Medical Affairs, and 3 Biomedical Data Sciences, GlaxoSmithKline, Research Triangle Park, NC) Rationale: Lamotrigine (LTG) and valproate (VPA) treat a broad range of seizures. Some patients are converted from VPA to LTG for efficacy or tolerability reasons. This conversion is complicated by the inhibitory effect of VPA on the metabolism of LTG. This open-label study assessed serum concentrations in patients following a dosing-conversion algorithm derived from pharmacokinetic modeling. The study provides guidelines to enable maintenance of stable LTG concentrations during conversion from VPA to LTG monotherapy in patients with epilepsy. Methods: All patients (older than 16 years, receiving VPA monotherapy, and candidates for conversion to LTG due to poor seizure control or intolerable side effects) received LTG (Lamictal) with a total study duration of ≤19 weeks. Study phases consisted of Screening (≤1 week), LTG Escalation (8 weeks), VPA Withdrawal (0–6 weeks based on initial VPA dose), and LTG Monotherapy (4 weeks). Dose conversion was conducted in four steps: (a) escalate LTG to 200 mg/day according to package insert for add-on to VPA; (b) once at 200 mg/day LTG, gradually decrease VPA to 500 mg/day by ≤500 mg/week dose decrements, and hold stable for 1 week; (c) decrease VPA to 250 mg/day and increase LTG to 300 mg/day, holding stable for 1 week; and (d) discontinue VPA,
TABLE 1. Mean weekly trough serum concentrations Phase LTG Escalation LTG Escalation LTG Escalation Variable VPA withdrawal Variable VPA withdrawal Variable VPA withdrawal Variable VPA withdrawal Fixed VPA withdrawal Fixed VPA withdrawal LTG Monotherapy LTG Monotherapy LTG Monotherapy LTG Monotherapy
Study week
n
4 6 8 9 10 11 12 13 14 15 16 17 18
54 57 54 27 14 6 4 44 44 38 37 33 34
Mean LTG Mean LTG dose (mg/day) trough (µg/ml) 25 48 95 200 200 200 200 203 295 386 484 490 493
1.3 4.2 7.9 7.5 7.9 9.5 9.7 7.2 8.7 9.0 8.1 7.3 7.2
Duration of variable VPA withdrawal phase depended on initial VPA dose. LTG, lamotrigine; VPA, valproate.
AES PROCEEDINGS and make weekly adjustments in LTG based on clinical response. LTG and VPA trough serum concentrations were collected 3 times during escalation of LTG and weekly thereafter. Results: Seventy-seven patients enrolled (mean age, 34 years; female, 60%; seizures: partial, 72%; and generalized, 55%, with some patients having both. Of the 77 patients, 29 prematurely discontinued. Overall compliance was excellent (>90%) for most patients completing the study. Mean weekly trough serum concentrations from evaluable patients are summarized in Table 1. Conclusions: The data show that serum LTG concentrations were held relatively stable while converting from VPA to LTG monotherapy by using the specified four-step dosing algorithm. (Supported by GlaxoSmithKline Research and Development.) 1.276 OXCARBAZEPINE METABOLITE (MHD) CONCENTRATION IN SALIVA OFFERS AN ALTERNATIVE TO SERUM FOR THERAPEUTIC MONITORING 1 Michael V. Miles, 4 Peter H. Tang, 2 Melody A. Ryan, 3 Shellee A. Grim, 2 Toufic A. Fakhoury, 2 Robert J. Baumann, 1 Richard H. Strawsburg, and 1 Ton J. DeGrauw (1 Pediatric Neurology, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH; 2 Neurology and 3 College of Pharmacy, University of Kentucky Chandler Medical Center, Lexington, KY; and 4 Pathology & Lab Medicine, University of Cincinnati Medical Center, Cincinnati, OH) Rationale: Using saliva in place of serum for therapeutic monitoring has several advantages: increased ease of collection; deceased risk of infection; and decreased discomfort. This option becomes particularly important for infants and children, patients with limited physical mobility, persons who cannot drive, and for patients in rural or remote settings with limited access to health care facilities. The purpose of this study was to determine the correlation between saliva and serum concentrations of 10-11-dihydro-10-hydroxy-carbamazepine (MHD), the active metabolite of oxcarbazepine (OXC), and to assess the feasibility of salivary MHD determination for routine therapeutic drug monitoring. Methods: Patients with epilepsy who were taking OXC and attended the neurology clinic at the University of Kentucky Medical Center (UKMC) were eligible for this study. IRB approval was obtained. A minimum of 0.25 ml of saliva was collected into a microtube. Blood was collected by a phlebotomist. Serum was sent for routine MHD analysis at UKMC (reference lab A). Saliva and serum specimens were stored at −20◦ C until MHD analysis at a reference laboratory at Cincinnati Children’s Hospital Medical Center (reference lab B). MHD concentrations were determined by validated HPLC methods at both reference laboratories. Results are expressed as mean ± SD. Results: Patient ages ranged from 3 to 51 years (22.5 ± 14.0 years). High patient and family acceptance of saliva collection was observed. Serum results from reference lab A (n = 20) ranged from 7 to 48 µg/ml (24.8 ± 11.8 µg/ml). Serum results from reference lab B (n = 20) ranged from 5.8 µg/ml to 48.4 µg/ml (25.9 ± 11.8 µg/ml). Saliva concentrations (n = 20) ranged from 4.7 to 42.6 µg/ml (23.9 ± 10.4 µg/ml). There was a strong correlation between the serum results reported by reference laboratories A and B (r = 0.970). The saliva-to-serum concentration ratios from reference lab A ranged from 0.67 to 1.23 (0.98 ± 0.14), and there was a strong correlation between saliva and serum values (r = 0.951). Reference lab B saliva-to-serum concentration ratios ranged from 0.72 to 1.13 (0.93 ± 0.12), and a similar strong positive correlation was evident (r = 0.960). Conclusions: Saliva collection is simple and well tolerated by children and adults. Results indicate that the mean concentration of MHD in saliva is ∼95% of the serum concentration, although some interpatient variability is apparent. Results from two reference laboratories indicate a strong correlation between saliva and serum MHD concentrations. This study shows that saliva MHD concentration monitoring is feasible, and may provide a useful alternative to serum testing for certain patients. Further validation is needed to prove reliability and clinical usefulness. 1.277 ORAL ABSORPTION KINETICS OF LEVETIRACETAM: EFFECT OF MIXING WITH FOOD OR ENTERAL NUTRITION FORMULAS
99
1,2 Barry E. Gidal, 2 Raj D. Sheth, 1 Paul R. Hutson, 2 Heather A. Stanko, and 2 Jennifer Parnell (1 School of Pharmacy and 2 Department of Neu-
rology, University of Wisconsin, Madison, WI) Rationale: Levetiracetam (LEV) is a structurally novel antiepileptic drug (AED) that has a favorable pharmacokinetic (PK) profile including negligible protein binding and linear elimination kinetics. LEV does not appear to cause or be subject to significant PK interactions. LEV is highly water soluble, and is rapidly (Tmax ∼1 h) and extensively absorbed (F > 95%). Although previous studies have suggested that food does not impair the overall absorption of LEV, there is a lack of data regarding concomitant administration with enteral nutrition formulation (ENF). Given the utility of this agent, it is likely to be used in a variety of patient populations including children and the elderly. It may be speculated that alternative administration techniques such as crushing and mixing with various semisolid foods and/or ENF may be required. The objective of this study was to evaluate the oral absorption of LEV after concomitant administration with these vehicles. Methods: This study used an unblinded, randomized, crossover design. After an overnight fast, a single dose of 500-mg LEV (Keppra) given as either as an intact tablet with 120 ml water (control, phase I), or after being crushed and mixed with 4 oz applesauce (phase II), or 120 ml ENF (Sustacal, phase III). A 7-day washout period occurred between phases. Serial blood samples were obtained over 24 h to determine LEV serum concentration–time profile (GC NPD). Calculated PK variables included area under the concentration–time curve (AUC), max serum concentration (Cmax ), and time to max serum concentration (Tmax ) using noncompartmental methods. Statistical analysis was conducted with ANOVA. Results: Ten adult healthy volunteers (six women/four men)were evaluated (mean age and weight were 28.9 years and 78.6 kg, respectively). AUC values for phases I–III were 191.9 ± 50.2, 165.7.2 ± 43.4, and 168.3 ± 43.9 µg/h/ml, respectively, and did not significantly differ (p = 0.38). Tmax did not significantly differ between any phases (1.08 ± 0.65, 1.32 ± 0.75, and 1.62 ± 0.73 h; p = 0.25). Cmax values for phase I–III were 14.8 ± 5.6, 12.1 ± 2.8, and 10.8 ± 2.0 µg/ml, respectively. No significant differences were noted between control and any other study phase, but a nonsignificant trend toward a lower Cmax (−27%, p = 0.07) after coadministration with the ENF (phase III). Mean LEV serum concentrations at 10 h after dose, however, were essentially the same for each study phase (phase I–III: 3.9, 4.1, 4.0 µg/ml, respectively). Long-term stability of LEV in these various formulations was not assessed. Conclusions: These data suggest that the overall rate and extent of LEV oral absorption is not impaired after crushing and mixing the tablet formulation with either a food vehicle or a typical ENF product. This method of LEV administration is therefore acceptable from a kinetic perspective. Our data do suggest slightly reduced peak serum concentrations after mixing with EF. The clinical significance of these modest changes is doubtful. (Supported by UCB-Pharma.) 1.278 CLINICAL EXPERIENCE WITH LAMOTRIGINE FOR PARTIAL SEIZURES IN CHILDREN Mary R. Andriola, Lourdes Bello, and Scott Pearlman (Neurology Stony Brook University Epilepsy Management Program, State University of New York at Stony Brook, Stony Brook, NY) Rationale: Lamotrigine (LTG) is a broad-spectrum antiepileptic drug (AED) initially approved in the United States for adjunctive use in patients 16 years or older with partial epilepsy and for patients age 2 years or older with generalized seizures associated with Lennox–Gastaut syndrome. Approval was expanded this year to pediatric patients 2 years or older with partial epilepsy with or without secondary generalization. The objective of this study was to evaluate our clinical experience with LTG in children with partial epilepsy in an academic referral center. Methods: This retrospective chart analysis included 13 children who received LTG as adjunctive therapy for partial epilepsy from November 2001 to February 2003. LTG was slowly titrated up and given as follows: LTG + enzyme-inducing AED without valproate (VPA): 0.6 mg/kg/day in one or two divided doses, increased up to maintenance doses of 5– 10 mg/kg/day. LTG + AED regimen with VPA: 0.15 mg/kg/day in one or two divided doses increased up to maintenance doses of 1–5 mg/kg/day. Patients were evaluated regarding efficacy and side effects.
Epilepsia, Vol. 44, Suppl. 9, 2003
100
AES PROCEEDINGS
Results: The study included nine boys and four girls, ranging in age from 2 to 14 years; mean age, 8 years. Twelve patients were initially taking one or two AEDs. One patient was started as monotherapy. The most common concurrent AEDs were VPA, four; carbamazepine, two; and topiramate, two. Nine patients (69%) had ≥50% seizure reduction with the addition of LTG. Four (31%) were able to achieve conversion to monotherapy. Four (31%) remain seizure free after 14 months’ followup. Two patients discontinued the medication because of lack of efficacyand two because of adverse effects. No patients developed a rash. Five patients (38%) experienced side effects that included drowsiness (n = 2), behavioral changes (n = 1), nausea (n = 1), anorexia (n = 1), ataxia (n = 1), and double vision (n = 1). Conclusions: LTG was safe, effective, and well tolerated in our pediatric patients with partial epilepsy. It was useful as monotherapy, with complete seizure control achieved in some patients. Helpful dosing and the new FDA approval should allow greater use of LTG in this age group.
1.279 CLINICAL EXPERIENCE WITH LEVETIRACETAM IN INTRACTABLE EPILEPSY James D. Bruno, Deborah A. Galloway, Hasan M. Mousli, and Mary R. Andriola (Neurology, State University of New York at Stony Brook, Stony Brook, NY) Rationale: Levetiracetam (LEV) is an antiepileptic drug (AED) approved as adjunctive therapy in the treatment of partial-onset seizures in adults with epilepsy. The purpose of this study was to review our clinical experience with LEV in the various epilepsies in adults and children referred to an academic center. Methods: Adult patients with intractable epilepsy despite adequate exposure to other AED(s) were started on LEV, 500 mg b.i.d., and titrated up to 3 g/day. The children were started on LEV, 250 mg b.i.d., and titrated up to 1.5 g/day. These patients were followed up for efficacy (≥50% seizure reduction) and adverse events (AEs) from April 2000 until April 2003. Mean duration of treatment at the last clinic visit was 12.2 months. Results: Twenty-four patients who ranged in age from 6 to 88 years were entered into the study. Seizures were classified as intractable primary generalized (n = 2) or partial onset (n = 21). One patient had Lennox–Gastaut syndrome. Associated conditions included schizencephaly, stroke, cerebral palsy, and astrocytoma. Six patients had vagal nerve stimulators (VNS). All patients were taking one to three AEDS when LEV was added. Efficacy of ≥50% seizure control was reported in eight (33%) of the 24 patients using LEV as adjunctive therapy to one, two, or three other AED(s). Of those patients showing improvement, one had generalized epilepsy, and seven had partial-onset epilepsy. One patient with partial onset achieved 100% seizure control. No patients demonstrated worsening of their seizures while taking LEV. No patients with VNS showed benefits from LEV. Eleven patients discontinued the drug because of lack of efficacy. Five patients discontinued the drug because of AEs. One child experienced night terrors. Three patients reported confusion and agitation. One patient reported rash. Six patients reported symptoms of lethargy. Conclusions: LEV was well tolerated in the majority of patients (58%) in this varied population of adults and children with intractable generalized and partial-onset epilepsy. Approximately eight (33.3%) of the 24 patients had had significant decrease in seizure severity and frequency. The most commonly reported AEs were lethargy and confusion.
1.280 CLINICAL EFFICACY OF LAMOTRIGINE, TOPIRAMATE, GABAPENTIN, AND LEVETIRACETAM IN A LARGE GROUP OF STRONGLY THERAPY-RESISTANT PATIENTS WITH EPILEPSY Martijn Engelsman and Nine de Beer-Pawlikowski (Neurology, Stichting Epilepsie Instellingen Nederland, Heemstede, Netherlands) Rationale: In the last decade, a number of new antiepileptic drugs (AEDs) became available in the Netherlands indicated for add-on therapy in refractory partial epilepsy. Our aim was to compare in retrospect the efficacy, side effects, reasons for dropout, and reactions of differ-
Epilepsia, Vol. 44, Suppl. 9, 2003
ent seizure types to lamotrigine (LTG), topiramate (TPM), gabapentin (GBP), and levetiracetam (LEV). Methods: In the Utrecht outpatient ward of a Dutch epilepsy centre, we followed a total of ∼600 patients with chronic epilepsy taking the four mentioned drugs. Many patients used more than one of these drugs in different periods (and sometimes in combination). Data that were collected on each period consisted of seizure type and frequency, comedication at start of each therapy, current comedication status at that time, dosage of the drug to be studied, response to therapy, side effects, and duration of therapy. The follow-up is currently still going on. There were 456 patients taking LTG, 219 taking TPM, 166 taking GBP, and 239 taking LEV. Results: On average, patients were taking two other AEDs at the start of the add-on therapy. In the LTG group, 59 patients became seizure free (7.7%); in the TPM group, 21 (9.6%); in the GBP group, seven (7.3%); and in the LEV group, 30 (12.5%). The preliminary analysis of the data indicates that there is no association between favorable outcome and specific seizure types. One hundred forty-eight stopped using LTG (32.5%), 126 TPM (57.5%), 96 GBP (57.8%), and 85 LEV (35.6%), either because of lack of efficacy or side effects. Conclusions: There are considerable differences in efficacy and side effects between the four studied drugs. Only patients who were not responders to treatment with LTG (the first available) were selected for treatment with TPM, GBP, or LEV, respectively.
1.281 TOLERABILITY AND SAFETY OF TOPIRAMATE AS FIRSTLINE MONOTHERAPY IN 1,000+ EPILEPSY PATIENTS 1 Edward Faught, Jr.,2 Liza Squires,2 Steven Wang, and 2 Ulrich Thienel (1 Department of Neurology, University of Alabama School of Medicine, Birmingham, AL; and 2 Johnson & Johnson Pharmaceutical Research & Development, Raritan, NJ) Rationale: Pharmacodynamic interactions may contribute to adverse events (AEs) when antiepileptic drugs (AEDs) are assessed as add-on therapy. Monotherapy studies provide a more accurate profile of AEs, particularly CNS and neurobehavioral effects. Topiramate (TPM) was evaluated as monotherapy in three randomized double-blind trials in which >1,000 patients with newly diagnosed epilepsy received 50– 500 mg/day TPM. Pooled safety information provides a dataset for evaluating the tolerability and safety of TPM monotherapy. Methods: Three randomized double-blind trials evaluated TPM monotherapy in patients with newly diagnosed epilepsy. Target TPM doses were 50, 100, 200, 400, or 500 mg/day (one study used 25 or 200 mg/day for patients ≤ 50 kg). TPM titration was brisk. Depending on dose assignment, the 25-mg/day starting dose was increased to 50 and then 100 mg/day TPM. For groups assigned to higher doses, TPM dose was then increased weekly in 100-mg/day increments. In two dosecomparison trials (50 vs. 400 and 50 vs. 500 mg/day), dose reductions were allowed to manage side effect; dose reductions were not allowed in one study. Pooled safety data (AEs, clinical laboratory data, and other safety data) were analyzed. Results: Safety data were collected for 1,131 patients: 50 (n = 359), 100 (n = 210), 200 (n = 199), 400 (n = 236), or 500 (n = 127) mg/day TPM. Median age was 27 (6–85) years. Double-blind treatment durations ranged from 6 to 9 months (mean) and extended up to 2.2 years. AEs most likely to occur during titration (cumulative incidence, ≤ 2.2 years) were paresthesia (25%), fatigue (16%), dizziness (13%), somnolence (13%), and nausea (10%); AEs most commonly occurring during maintenance were headache (20%), appetite decrease (11%), and weight loss (11%). Fewer than 10% of patients reported adverse cognitive effects. Overall, 17% of patients discontinued because of AEs, 12% if the protocol allowed dose reduction. No noteworthy changes from baseline to final visit were observed in hepatic, renal, and hematologic laboratory values. Percentage decreases in baseline body weight ranged from 0.2% with 50 mg/day TPM to 3.5% with 400 mg/day TPM. Conclusions: TPM monotherapy is well tolerated, with CNS effects, particularly cognitive complaints, being substantially lower than those with TPM as add-on therapy. Paresthesia is the most common side effect, but seldom requires discontinuation of TPM. TPM has a favorable tolerability and safety profile for long-term use. (Supported by Johnson & Johnson Pharmaceutical Research & Development.)
AES PROCEEDINGS 1.282 DEMOGRAPHIC AND CLINICAL CHARACTERISTICS OF PATIENTS ENROLLED TO DATE IN A MULTICENTER, OPEN-LABEL STUDY TO ASSESS THE TOLERABILITY, SAFETY, AND EFFECTIVENESS OF SWITCHING FROM IMMEDIATE-RELEASE CARBAMAZEPINE TO EXTENDEDRELEASE CARBAMAZEPINE (CARBATROL) IN EPILEPSY SUBJECTS WITHIN A COMMUNITY-BASED POPULATION 1 David Ficker and 2 Charles W. Gorodetzky (1 Neurology, University of Cincinnati Medical Center, Cincinnati, OH; and 2 Medical and Scientific Services, Quintiles, Inc., Kansas City, MO) Rationale: We examined the demographics and clinical characteristics, including seizure frequency, carbamazepine (CBZ) dose and range, as well as epilepsy duration, in patients enrolled to date in a multicenter, open-label study assessing the efficacy and safety of switching from an immediate-release carbamazepine (Tegretol) to an extended-release CBZ (Carbatrol). Methods: Patient data on demographic information, medical and medication background, and seizure history were collected and analyzed. Patients were asked to indicate the frequency and types of seizures experienced during the past month and to disclose their epilepsy duration. Investigators gathered information on the subject’s prior use of CBZ including dose/regimen, duration of use, and current brand/formulation. Inclusion criteria for this study required subjects, male or female, to be 12 years or older at the time of study entry; female subjects may not be pregnant; and women of childbearing potential agree to protect against conception. In addition, subjects should be diagnosed with history of partial epilepsy, currently prescribed IR-CBZ for ≥3 months before screening, and are not taking more than one concomitant antiepileptic drug (AED). All statistical analyses were carried out using SAS Windows (version 8.0). Results: Of the 236 patients currently enrolled to date in the study, 51.3% are female. Subject race was 73.7% white, 12.5% black, 8.2% Hispanic, 2.6% Asian/Pacific islander, 2.2% Native American, and 0.9% listed as other. The average age among enrolled subjects was 42.6 ± 16.8 years, and the mean CBZ dose at screening was 737.0 ± 284.5 mg/day. The mean seizure frequency for all patients was 1.95 ± 5.97 per month, and complex partial seizures were the most frequently reported (1.33 ± 5.18); 66.3% of patients had epilepsy for >10 years, 13.1% had epilepsy for 6 to 10 years, 16.1% had epilepsy for 1 to 5 years, and 4.5% had epilepsy for 0.05 for all variables tested); 31 (9.1%) patients reported using CAM known to have an adverse effect on seizure control (e.g., St. John’s Wort, evening primrose). Alternative medicines that were anecdotally reported to be helpful in seizure control included melatonin, aloe vera, valerian, and Chinese herbal medicine. Patients also reported improvements in seizure control after undergoing aromatherapy, homeopathy, Reiki, and reflexology. These improvements were subjective, and no attempt to formally assess the effect of these therapies on seizure control was made. Conclusions: Although one in three patients reported use of CAM at some time, the number of patients who had used CAM for seizure control was small. No patient demographic predicted use of complementary medicine. As some patients found certain therapies had a positive effect on their seizures, further investigation of these may be warranted. (Supported by the Wigan and Leigh Epilepsy fund, K.R.E.)
Epilepsia, Vol. 44, Suppl. 9, 2003
106
AES PROCEEDINGS
1.297 ADULT EXPERIENCE WITH DIAZEPAM RECTAL GEL Toufic A. Fakhoury (Department of Neurology, Chandler Medical Center, Lexington, KY) Rationale: The safety and efficacy of diazepam (DZP) rectal gel (Diastat) has been established in both adults and children. However, most of the reported experience with treatment has focused on the pediatric age group. This study examines the use of DZP rectal gel among adult patients and its role in adult seizure management. Methods: We reviewed the charts of adult epilepsy patients to identify those who had used DZP rectal gel for breakthrough seizures. We collected information on diagnosis, DZP rectal gel dose, frequency of treatment, and reasons for use, as well as efficacy and safety. Results: Twenty patients received at least one treatment of DZP rectal gel. There were seven males and 13 females, and the age range was 20–56 years, with a mean age of 37 years. Thirteen patients had partial epilepsy, and seven had generalized epilepsy, including four with Lennox–Gastaut syndrome. The dose of DZP rectal gel used ranged from 10 to 20 mg. The medication was administered for treatment of clusters of seizures in 10 patients, prolonged seizures in seven patients, and for both types in three patients. The frequency of use ranged from once every 4 months, for patients with well-controlled epilepsy, to weekly. DZP rectal gel was effective for 19 patients (95%). All patients experienced somnolence, but no serious adverse event was reported. Conclusions: DZP rectal gel is effective and well tolerated in adult patients. As a rescue medication for breakthrough seizures, DZP rectal gel is a valuable component of seizure management for a diverse group of adults with well-controlled or refractory epilepsy syndromes. (Supported by Xcel Pharmaceuticals.) 1.298 PATIENT SATISFACTION WITH DIAZEPAM RECTAL GEL Bruno V. Gallo (Department of Neurology, University of Miami, School of Medicine, Miami, FL) Rationale: Patient satisfaction with a treatment regimen is an important factor in epilepsy management, especially with regard to compliance. This survey was conducted to assess the level of patient satisfaction with DZP rectal gel (Diastat) as a rescue medication for seizure emergencies. Methods: Adult patients from a general adult neurology practice who received at least one dose of DZP rectal gel to abort a prolonged seizure, or acute repetitive seizures were interviewed. Patients were interviewed at clinic visits or via phone, using questionnaires to assess the efficacy and satisfaction with the therapy. Seizure control, seizure duration, quality of life, and overall satisfaction with the medication were evaluated. Seizure control was assessed in terms of seizure frequency and intensity. Attempts were also made to have family evaluate whether the therapy was effective in truncating and aborting previously well-documented flurries of seizures. Results: Review of patient charts yielded 12 patients that met inclusion criteria for the study. All patients interviewed reported excellent seizure control. For patients who experienced acute repetitive seizures, all events treated with DZP rectal gel were aborted within 15 min of administration (100% efficacy). Several patients were initially resistant to the rectal route of delivery, but became more satisfied with the medication based on education from their physician and their own experience with its efficacy in terminating seizures. Overall, the patients interviewed were satisfied or very satisfied with DZP rectal gel as a rescue medication for breakthrough seizures. Conclusions: Patient satisfaction, an important factor in quality of life, was high for DZP rectal gel. DZP was 100% effective against acute repetitive seizures. Patient opinion of the drug appears to be driven by its efficacy and by education provided by their physician. 1.299 VALPROATE ACTS AS AN APOPTOTIC AGENT IN THE OVARY: A FINDING OF POSSIBLE IMPORTANCE FOR THE TREATMENT OF OVARIAN CANCER 1 Ewa L. Gregoraszczuk, 2 Erik Tauboll, 1 Anna Kolodziej, 3 Malgorzata Kajta, and 4 Erik Ropstad (1 Department of Animal Physiology, Institute of Zoology, Jagiellonian University, Krakow, Poland; 2 Department of Neurology, Rikshospitalet, University of Oslo, Oslo, Norway; Epilepsia, Vol. 44, Suppl. 9, 2003
3 Department of Endocrinology, Institute of Zoology, Jagiellonian University, Krakow, Poland; and 4 Department of Reproduction and Forensic Medicine, Norwegian School of Veterinary Science, Oslo, Norway)
Rationale: Several antiepileptic drugs (AEDs) are associated with anticancer activity. At the same time, many AEDs alter endocrine function with valproate (VPA) inducing hyperandrogenism. Changes in sex steroid hormone levels are known to affect apoptosis in endocrine tissue and androgens are proapoptotic in the ovary. It has also been shown recently that VPA reduces estrogen-induced cell growth in human breast cancer cells. Does VPA affect cell proliferation and apoptosis in the ovary? Methods: Small and medium-sized ovarian follicles were obtained from pig ovaries on days 8–10 and 14–16 of the oestrus cycle. Cocultures of theca and granulosa cells were prepared and cultured for 48 h in control medium or VPA (100 or 250 µg/ml). Cell apoptosis was studied by measuring caspase-3 activity, and cell viability was estimated by the Alamar Blue test. Results: VPA significantly and dose-dependently increased caspase-3 activity in both small and medium follicles, suggesting an apoptotic effect of the drug. In small follicles, caspase-3 activity increased with ∼50% at 100 and 200% at 250 µg/ml of VPA. In medium follicles, caspase-3 activity increased with ∼40% at 100 and 115% after 250 µg/ml. VPA had no effect on follicular cell proliferation and viability evaluated by the Alamar Blue test. Conclusions: VPA increased caspase-3 activity in a dose-dependent manner and at therapeutic drug concentrations, suggesting an apoptotic effect of the drug. Our previous animal experiments have shown a disruption of follicular steroidogenesis induced by VPA, resulting in an increased androgen-to-estrogen ratio and low estradiol levels. The possibility exists that an apoptotic action of VPA in the ovary might be linked to the steroidogenic effects of the drug. This will have wide clinical implications, possibly also for the treatment of cancer in endocrine tissue. 1.300 CLOBAZAM AS ADD-ON THERAPY IN PATIENTS WITH REFRACTORY TEMPORAL LOBE EPILEPSY AND HIPPOCAMPAL ATROPHY Claudio M. Ferreira, Maria A. Montenegro, Fernando Cendes, Li M. Li, and Carlos A.M. Guerreiro (Depertment of Neurology, University of Campinas, Campinas, Sao Paulo, Brazil) Rationale: Clobazam (CLB) is a benzodiazepine with excellent antiepileptic action; however, it is not considered first-line therapy in the treatment of epilepsy. The objective of this study was to evaluate the efficacy of CLB as add-on therapy in adults with refractory temporal lobe epilepsy associated with hippocampal atrophy. Methods: This was a retrospective study, conducted at the epilepsy clinic of our university hospital. We evaluated the patients that were exposed to CLB as add-on therapy from a group of 100 consecutive patients with temporal lobe epilepsy and hippocampal atrophy. CLB was prescribed based on the concept of maximum tolerated dose. Results: Seventy-eight patients met the inclusion criteria (51 women), ages ranging from 16 to 76 years (mean, 42.2). Dosage of CLB ranged from 5 to 60 mg/day (mean, 22.6 mg/day). CLB was used from 1 month to 8 years (mean, 29 months). Sixteen (20.5%) patients were seizure free, 20 (25.5%) had >75% improvement in seizure control, eight (10%) had >50%, and three (4%) had 1 year in 30 (68%) patients. Conclusions: CLB has an excellent efficacy even for patients who would be considered candidates for epilepsy surgery. That is, CLB should be considered as first-line add-on therapy in patients with refractory temporal lobe epilepsy and hippocampal atrophy. 1.301 MIGRAINE PROPHYLAXIS WITH TOPIRAMATE: RESULTS OF DOUBLE-BLIND, PLACEBO-CONTROLLED, DOSE– RESPONSE TRIALS 1 Carolyn E. Hart, 2 David W. Dodick, 3 Jan L. Brandes, 4 John F. Rothrock, 5 David Jacobs, 5 Walter Neto, 5 Suman Bhattacharya, and 5 Jennifer Schmitt (1 Department of Neurology, Mecklenburg Neurological
AES PROCEEDINGS Associates, Charlotte, NC; 2 Mayo Clinic, Scottsdale, AZ; 3 Nashville Neuroscience Group, Nashville, TN; 4 University of South Alabama, Mobile, AL; and 5 Johnson & Johnson Pharmaceutical Research & Development, Raritan, NJ) Rationale: Given the ability of anticonvulsants such as topiramate (TPM) to reduce neuronal hyperexcitability, it is reasonable to expect that at least some may be useful in paroxysmal disorders other than epilepsy (e.g., migraine). In double-blind, placebo-controlled trials, 44–48% of valproate (VPA)-treated patients had ≥50% migraine reduction versus 14–21% of placebo-treated patients (23–34% net increase in responder rate with VPA treatment). VPA and TPM share similar broad-spectrum profiles in epilepsy, but have different mechanisms of action. Nonetheless, in open-label studies, TPM, like VPA, prevented migraine headache. Two double-blind, placebo-controlled dose–response trials assessed the potential usefulness of TPM as migraine prophylaxis. Methods: TPM was evaluated in the absence of other preventive medications in adolescents/adults (12–65 yrs) with an established history of migraine (≥6 months), with or without aura, for whom no more than two preventive regimens had failed. In the prospective 28-day baseline after other migraine preventives had been discontinued, patients had to have at least three to 12 migraines. Patients were randomized to placebo, 50, 100, or 200 mg/day TPM. The 8-week titration phase was followed by an 18-week maintenance phase. The 25-mg/day starting dose was increased weekly in 25-mg increments to assigned target dose or maximum tolerated dose. Primary efficacy variable was mean change from prospective baseline in monthly migraine frequency during double-blind treatment. Secondary efficacy measures included responder rate (% patients with ≥50% seizure reduction). Results: The 469 patients were randomized in study 1 (placebo, n = 115; TPM 50, 117; TPM 100, 125; TPM 200, 112) and 468 in study 2 (placebo, 114; TPM 50, 117; TPM 100, 120; TPM 200, 117). Both studies showed significant differences between placebo and TPM 100 (study 1, p < 0.001; study 2, p = 0.008) and TPM 200 (p < 0.001, both studies), but not TPM 50, in mean migraine frequency reduction. In study 1, responder rates were placebo, 23%; TPM 50, 36% (p = 0.04); TPM 100, 54% (p < 0.001), and TPM 200, 52% (p < 0.001). In study 2, responder rates were placebo, 23%; TPM 50, 39% (p = 0.01); TPM 100, 49% (p < 0.001), and TPM 200, 52% (p < 0.001). Discontinuations due to adverse events (AEs) (combined studies): placebo, 11%; TPM 50, 17%; TPM 100, 23%; TPM 200, 27%. The most common adverse AEs (≥10% incidence) with TPM 100 as target dose: paresthesia (TPM 100, 48%; placebo, 6%); fatigue (13%, 10%), and appetite decrease (13%, 6%). Conclusions: TPM 100 mg/day as a target dose is effective and well tolerated for migraine prophylaxis. With TPM 100 mg, the net difference from placebo in responder rate was 26% and 31%, which compares favorably with VPA. (Supported by Johnson & Johnson Pharmaceutical Research & Development.) 1.302 ZONISAMIDE SAFETY AND TOLERABILITY IN AN ADULT OUTPATIENT CLINIC POPULATION Laura Hershkowitz and Kimberly Konieczki (Northshore Clinical Associates, Erie, PA) Rationale: Zonisamide (ZNS) is a broad-spectrum antiepilepsy drug (AED) with multiple mechanisms of action. In controlled clinical trials, 12% of ZNS-treated patients discontinued the drug for adverse events.
107
This retrospective chart review investigates the safety and tolerability of ZNS in outpatients with a variety of seizure types. Methods: Charts of patients from an outpatient practice who received ZNS between January 2000 and April 2003 were reviewed for ZNS dosage, use of other AEDs, length of time on ZNS, adverse events, and reason for discontinuation of ZNS (where applicable). Results: Charts for 95 patients with a variety of seizure types were reviewed. Mean patient age was 34.1 years (SD, 16.1). ZNS dosages ranged from 100 to 800 mg/day. Patients received a mean of one other AED (SD, 0.61). Most commonly used other AEDs included lamotrigine (n = 21), phenytoin (n = 15), and divalproex sodium (n = 15). Patients received ZNS for a mean of 58.1 weeks (SD, 47.3). Twenty-one patients discontinued ZNS without restarting it, 10 for adverse events, nine for inefficacy/seizures, and five for other/unspecified reasons (some patients cited more than one reason for discontinuation). One additional patient elected to discontinue ZNS but restarted the drug 2 days later. The most common adverse events were cognitive effects (n = 4), weight loss (n = 3), behavioral effects (n = 2), gastrointestinal effects (n = 2), and mood effects (n = 2). Two additional patients reported weight loss, but this was not considered an adverse event, as the patients were overweight. One patient reported rash and discontinued ZNS after 7 days. No patients reported kidney stones. Conclusions: The tolerability of ZNS and the adverse events reported in this outpatient seizure clinic population were remarkably similar to what was reported in controlled clinical trials. Notably, only one patient experienced rash, and no patients developed kidney stones while receiving ZNS. ZNS was safe and well tolerated in this cohort of patients. (Supported in part by Elan Pharmaceuticals, Inc.) 1.303 DIFFERENTIAL ANTIEPILEPTIC DRUG EFFECTS ON SEXUAL FUNCTION AND REPRODUCTIVE HORMONES: A COMPARISON BETWEEN LAMOTRIGINE AND ENZYMEINDUCING ANTIEPILEPTIC DRUGS 1 Andrew G. Herzog, 1 Frank W. Drislane, 1 Donald L. Schomer, 4 Page B. Pennell, 3 Kevin M. Kelly, 5 Edward B. Bromfield, 1 Erin L. Farina, and 2 Cheryl A. Frye (1 Neurology, Beth Israel Deaconess Medical Center, Boston, MA; 2 Psychology, University at Albany, Albany, NY; 3 Neurology, Alleghany General Hospital, Pittsburgh, PA; 4 Neurology, Emory Clinic, Atlanta, GA; and 5 Neurology, Brigham & Women’s Hospital, Boston, MA) Rationale: Sexual dysfunction is common in men with epilepsy and has been related to lower serum bioactive testosterone (BAT) levels. Enzyme-inducing antiepileptic drugs (EIAEDs) are thought to play a role. Little is known, however, about the effects of non-EIAEDs. This investigation compared sexual function and gonadal steroid levels among men with localization-related epilepsy (LRE) who take lamotrigine (LTG), EIAEDs, no AEDs, and normal controls. Methods: Sixty-three men with LRE (EIAEDs, n = 36; LTG, n = 18; no AEDs, n = 9) and 18 normal controls, between ages 18 and 50 years, had blood samples assayed for luteinizing hormone, follicle-stimulating hormone, prolactin-releasing hormone, BAT, bioactive estradiol (BAE), sex hormone–binding globulin (SHBG), BAT/BAE, and gonadal efficiency (BAT/LH). Sexual interest and function were assessed by standardized questionnaire (S-score). Results: S-scores, BAT, BAT/BAE, and BAT/LH were significantly lower, and SHBG was significantly higher in the EIAED group than in
TABLE 1. S-score (/20) Controls (n = 8) 18.8 [18.2–19.4] LRE/no meds 17.3 [14.6–20.0] (n = 9) LTG (n = 18) 18.3 [17.5–19.1] LRE/EIAED 16.2∗∗∗# [15.0–17.4] (n = 36)
SHBG (µg/dl)
BAT (ng/dl)
BAE (pg/ml)
BAT/BAE
BAT/LH
0.83 [0.65–1.05] 0.85 [0.55–1.10]
307 [265–350] 236∗ [152–320]
25.1 [21.7–28.5] 26.6 [20.3–32.9]
129 [107–147] 93∗ [63–123]
70.9 [54.6–87.2] 64.3 [48.1–90.5]
0.85 [0.66–1.04] 1.47∗∗∗†# [1.21–1.73]
262 [215–309] 211∗∗∗# [180–242]
21.7 [18.4–25.0] 128 [105–151] 23.5 [21.2–25.8] 98∗∗# [83–113]
58.1 [48.4–67.8] 44.7∗∗∗† [36.8–52.6]
S-score, standardized questionnaire; SHBG, sex hormone-binding globulin; BAT, bioactive testosterone; BAE, bioactive estradiol; BAT/LH, gonadal efficiency; LRE, localization-related epilepsy; LTG, lamotrigine; EIAED, enzyme-inducing antiepileptic drugs. Values are means and 95% CI comparison to control: ∗ p < 0.05; ∗∗ p < 0.01; ∗∗∗ p < 0.001 comparison to LRE/no meds: † p < 0.05; comparison to LRE on LTG: # p < 0.05. Epilepsia, Vol. 44, Suppl. 9, 2003
108
AES PROCEEDINGS
any other group (Table 1). There were no significant differences in these measures between men on LTG and controls or untreated men with LRE. S-scores fell below the control range in 15 of 63 (23.8%) men with LRE, including 12 of 36 (33.3%) on EIAEDs, one of 18 (5.5%) on LTG, and two of nine (22.2%) on no AEDs (χ 2 : p < 0.01). BAT was below the control range (50% seizure reduction), fair (50% seizure reduction. Side effects were reported in 21 (34%) in the form of decreased appetite, weight loss, and sedation. Most of the side effects were transient; however, TPM had to be discontinued in seven children because of excessive weight loss (6%) and seizure worsening (5%). Follow-up renal ultrasound was performed on 34 (55%) children and was always normal. Conclusions: TPM is a very effective AED with broad spectrum of antiepileptic activities. Most side effects were transient; however, careful monitoring of the body weight is recommended. 1.305 LEVETIRACETAM MONOTHERAPY FOR ELDERLY PATIENTS WITH EPILEPSY Suzanne C. Koopmans, Michelle L. Apperson, Sarah T. Farias, Anna V. Marquez, Anthony R. Lima III, and Taoufik M. Alsaadi (Department of Neurology, University of California at Davis, Sacramento, CA) Rationale: To examine the efficacy and tolerability of levetiracetam (LEV) as monotherapy in elderly patients. Elderly patients may have increased sensitivity to the effects of antiepileptic drugs (AEDs) because of age-dependent changes in pharmacokinetics and pharmacodynamics. These may include a lower degree of drug binding to plasma proteins and decreased metabolic and renal drug clearance. LEV has been approved as adjunctive treatment for adults with partial-onset seizures. LEV has a favorable pharmacokinetic profile that includes minimal protein binding, lack of hepatic metabolism, and twice-a-day dosing. These features and others make it ideal for use as monotherapy in elderly patients. Methods: We retrospectively reviewed the charts of all of our elderly patients with a confirmed diagnosis of epilepsy. Thirty-three patients were identified. Of these, 15 patients began LEV either as first-line therapy or were converted to LEV monotherapy after prior AEDs failed. We reviewed demographic data, diagnostic evaluation for epilepsy, seizure Epilepsia, Vol. 44, Suppl. 9, 2003
types, and seizure frequency before and after initiation of LEV monotherapy. We compared baseline seizure counts from the 2 months before initiation of LEV treatment to seizure counts at 6 months of follow-up after LEV was started and maintained as therapy. Adverse effects while on LEV were noted. Results: We identified 15 patients, age 62–92 years (mean, 66 years) with a history of partial seizures with and without secondary generalization. The duration of epilepsy ranged from 1 to 39 years (mean, 12.6 years; median, 2.5). Five of these patients began LEV as first-line therapy. Two of these patients had liver disease and wished to avoid AEDs that required hepatic metabolism. The remaining three patients chose to start LEV for its favorable pharmacokinetic and side-effect profile. Ten patients converted to LEV monotherapy after their initial trials of AEDs failed, which included phenytoin, carbamazepine, divalproex, and lamotrigine. Six patients had an identifiable cause to their seizures: Three secondary to head trauma, two from strokes, and one due to Alzheimer disease. Of the 15 patients, 14 continued on LEV monotherapy for ≥6 months. One patient was lost to follow-up. Eight patients (57%) became seizure free. Four patients who began LEV as monotherapy became seizure free; the remaining four patients who began LEV as add-on therapy became seizure free. Three patients (21%) had a >50% seizure reduction, and three patients (21%) had >75% reduction of seizures. Total dosages used to control seizures were 500–3,000 mg/day (mean, 1,839.2 mg/day). One patient reported being dizzy within days after starting LEV, but was able to continue on LEV. No other major side effects were reported. Conclusions: LEV monotherapy can be effective and well tolerated in elderly patients with epilepsy. A prospective, large, double-blind study is needed to confirm this.
1.306 LEVITERACETAM MONOTHERAPY 1 Beth Korby, 1,2 Patricia E. Penovich, 1,2 John R. Gates, 1,2 Deanna L. Dickens, and 2 Gerald L. Moriaraty (1 Minnesota Epilepsy Group, PA of United Hospital and Children’s Hospital and Clinics, St. Paul, St. Paul, and 2 Department of Neurology, University of Minnesota, Minneapolis, MN) Rationale: Leviteracetam (LEV) has been approved for adjunctive use in refractory partial epilepsy. Its pharmacokinetic advantages and unique mechanism of action have led to increasing use. Sporadic reports of LEV efficacy in monotherapy led us to review our experience over the last 3 years in a large population of patients with refractory epilepsy, new-onset epilepsy and epilepsy secondary to newly diagnosed CNS neoplasm. Methods: A retrospective chart review of patients receiving LEV was performed for monotherapy patients. Data were tabulated for age, gender, IQ, behavioral and psychogical history, medication history, seizure frequency, adverse effects (AEs), LEV doses, and serum levels. Results: Thirty-six patients reached LEV monotherapy (LEV-M): three (8%) de novo, 33 (92%) transitioned from another AED. Six (17%) had primary CNS neoplasms. Thirty-two (89%) remain on LEV-M, one required additional AED, three dropped because of AEs. Doses ranged from 250 to 3,000 mg per day (mean, 1,389; median, 1,000). Serum levels ranged from 2 to 36 µg/ml (mean, 15.9; median, 12.5). Seizure freedom is present in 26 (72%) over 3–35 months (median, 19; mean, 17). Tumor patients are seizure free in 83%. AEs were reported in 42% (n = 15) at some time, resolving with lowering dose in three (8%). Only three (8%) discontinued LEV due to AEs. Reported AEs were hand tremor, 27%; tiredness or weakness, 20%; ataxia, irritability, decreased memory, cognitive decline, 13% each (two each); single reports of tinnitus, headache, weight gain, decreased libido, increased dreaming, increased depth perception, dizziness. Fourteen patients (39%) had a history of behavioral or psychological problems before LEV therapy; two of these patients reported additional problems while taking LEV-M, whereas 12 did not. Six patients with depression and mood swings reported improved psychiatric status. Conclusions: LEV-M offers significant benefit for patients with epilepsy. Seizure control can be excellent. Most AEs are temporary and well tolerated. Those with a history of behavioral or psychiatric problems did not experience intolerance to LEV at any increased rate. A subgroup with depression and mood swings improved. LEV monotherapy may be
AES PROCEEDINGS achieved without difficulty in patients with refractory epilepsy and in patients with primary CNS neoplasms.
1.307 DIAZEPAM RECTAL GEL USE IN A RESIDENTIAL FACILITY FOR MENTALLY RETARDED ADULTS Carole Lane and James Valeriano (Allegheny General Hospital, Allegheny Neurological Associates, Pittsburgh, PA) Rationale: Diazepam (DZP) rectal gel (Diastat, Xcel Pharmaceuticals) is chiefly regarded as a useful out-of-hospital rescue medication for prolonged or repetitive seizures, because of its ease of use by nonprofessional caregivers. However, it is also useful in an inpatient setting. The data presented here reflect our experience with DZP rectal gel in a residential facility for mentally retarded adults. Methods: Patient charts were reviewed to identify patients who had received DZP rectal gel. Demographic and background seizure data, duration of intermittent DZP rectal gel use, clinical impression of efficacy, and adverse events were collected. Results: Sixteen patients received at least one dose of DZP rectal gel. The study group included five female and 11 male patients aged 14 –72 years, with median age 33.5 years. Doses, for the 13 patients who had dose information available, ranged from 10 to 20 mg. Five patients have been using DZP rectal gel intermittently for ≥5 years. Efficacy impressions were available for 15 patients. In all cases, DZP rectal gel was effective for stopping seizures, with a mean efficacy impression of 81%. In five patients (33.3%), it was 100% effective over the period of use. Seizures were terminated within 10 min, with some patients responding in as little as 2–3 min. Only one patient experienced respiratory compromise and was not intubated. Another was sometimes lethargic after DZP rectal gel administration. Conclusions: DZP rectal gel was safe and effective over a period of 2–6 years in a residential facility for mentally retarded adults. Its rapid efficacy and ease of administration make DZP rectal gel useful for medically supervised patients as well as independently living people with epilepsy. (Supported by Xcel Pharmaceuticals.)
1.308 PHYSICIAN CLINICAL EXPERIENCES AND IMPRESSIONS WITH DIAZEPAM RECTAL GEL 1 Anthony Marmarou and 2 John M. Pellock (1 Department of Neurology and Neurosurgery, and 2 Department of Neurology, Medical College of Virginia, Virginia Commonwealth University, Richmond, VA) Rationale: By allowing patients and caregivers to treat seizure emergencies without seeking additional medical attention, diazepam (DZP) rectal gel (Diastat) has great potential to improve patient and caregiver confidence and quality of life. Additionally, use of the drug can decrease the time to treatment in seizure emergencies, resulting in reduced risk of neuronal damage. This survey study evaluates physician clinical experiences with and impressions of DZP rectal gel. Methods: Surveys containing questions regarding the use and effectiveness of DZP rectal gel were sent to 41 sites (all were members of the International Epilepsy Consortium). Survey responses were tabulated and summarized. Results: Thirteen sites returned their surveys, representing data from 33 physicians. Regarding the availability of DZP rectal gel in their region, 10 respondents indicated it was available in emergency rooms (ERs), seven respondents reported it was available on hospital floors, and four respondents said it was used by emergency squads. It is also used in epilepsy-monitoring and intensive care units. Respondents indicated a variety of seizure emergencies for which DZP rectal gel is used, including acute repetitive seizures, prolonged seizures, febrile seizures, clusters, status epilepticus, and during medication changes or withdrawal. Mean respondent estimate of the percentage of patients in which DZP rectal gel is effective was 80.1% (range, 50–96%; n = 13). Respondent estimates indicated that a mean of 80.9% (range, 50–90%; n = 8) of patients and 85.2% (range, 75–100%; n = 10) of caregivers using DZP rectal gel experienced improvement in control in their lives. All respondents felt that DZP rectal gel improves the quality of life for patients and/or their caregivers. Additionally, all respondents reported that use of DZP rectal
109
gel reduced the number of ER visits for their patients; mean estimated reduction was 69.5% (range, 25–93%; n = 11). Conclusions: These survey results demonstrate positive clinical experiences with, and favorable physician impressions of, DZP rectal gel. DZP rectal gel allows for decreased time to treatment in seizure emergencies, and fewer ER visits. The cost savings, improved control, and quality-of-life improvements associated with DZP rectal gel use are significant for patients with a variety of seizure emergencies. (Supported by Xcel Pharmaceuticals.) 1.309 ENHANCEMENT OF ANTIEPILEPTIC DRUG TREATMENT OF EXPERIMENTAL STATUS EPILEPTICUS THROUGH INHIBITION OF P-GLYCOPROTEIN 1,3 Andrey M. Mazarati, 1,2 Claude G. Wasterlain, and 2 Steve Shinmei (1 Neurology, D. Geffen School of Medicine, UCLA, 2 Research, West LA VA Medical Center, and 3 Brentwood Biomedical Research Institute, Los Angeles, CA) Rationale: P-glycoprotein (PGP), encoded by a multidrug resistance gene mdr-1, has been implicated in drug resistance in epilepsy. Overexpression of mdr-1 and PGP were reported both in epilepsy patients and animal epilepsy models. Based on our previous findings that the efficacy of several antiepileptic drugs (AEDs) decreased during the course of status epilepticus (SE), we examined whether pharmacologic inhibition of PGP improves anticonvulsant treatment of intractable SE. Methods: SE was induced in male Wistar rats by kainic acid (KA, 0.5 µg) injected into the ventral hippocampus. Diazepam (DZP, 10 mg/kg), fos-phenytoin (FOS, 50 mg/kg), or phenobarbital (PB, 60 mg/kg) was administered i.v. 30 or 90 min after KA. A PGP inhibitor cyclosporin A (CSA, 75 mg/kg i.p.) was injected 30 min before KA in the experiments where AEDs were given after 90 min of seizures. Electrographic activity was recorded for 10 h after injection of KA, using AcqKnowledge 3.7 software (Biopac Systems), and analyzed off-line. Results: SE lasted between 10 and 16 h, with the time spent in seizures 4–7 h, and a number of seizure episodes, 250–400. When administered 30 min after KA, all three AEDs stopped SE within 25 min. When injected after 90 min of seizures, all drugs attenuated SE by shortening its duration to 4–6 h, total seizure time to 2–3 h, and a number of seizures to 105–210 (p < 0.05 vs. control). CSA alone did not affect SE. PHT injected 90 min after KA to CSA-treated animals stopped SE within 35 min, with time spent in seizures ≤10 min and seizure count six to 11 (p < 0.05 vs. CSA-free PHT treatment). In contrast, failure of DZP to treat SE after 90 min of seizures was not reversed by CSA. After PB administration 90 min after KA to CSA-treated animals, both time spent in seizures (0.75–2 h) and seizure count (54–110) were significantly lower than in CSA-free PB group (p < 0.05), whereas SE duration remained unchanged. Immunohistochemical studies revealed PGP staining in the astrocytes after 90 min of seizures. Conclusions: During its course, KA-induced SE became resistant to the action of AEDs. Intractable phase of SE correlated with the de novo expression of PGP in the astrocytes. CSA administration restored the effectiveness of PHT during intractable phase of SE, suggesting the involvement of PGP in the resistance to PHT. Resistance to DZP did not depend on PGP, and likely resulted from γ -aminobutyric acid (GABA)receptor failure, as it had been shown previously (Kapur et al., 1996). PHB, a PGP-regulated GABAergic agent, showed partial recovery of anticonvulsant efficacy under conditions of CSA treatment, suggesting contribution of both pharmacokinetic and pharmacodynamic components in its failure during intractable phase of SE. Our data are useful for understanding the mechanisms of drug resistance in epilepsy, and for the development of novel approaches to the optimization of epilepsy treatment. (Supported by grants NS 43409 and NS 13515 from NIH and VHA Research Service.) 1.310 THE ANTICONVULSANT FLUOROFELBAMATE PROVIDES PROTECTION AGAINST DEPOLARIZATION-INDUCED CA1 NEURONAL INJURY 1,2 Kimberly L. Panizzon, 1,2 Roi Ann Wallis, and 3 Duane Sofia (1 Neurology, VA Greater Los Angeles Healthcare System, North Hills, 2 Neurology, UCLA, Los Angeles, CA; and 3 Carter-Wallace Inc., Cranbury, NJ) Epilepsia, Vol. 44, Suppl. 9, 2003
110
AES PROCEEDINGS
Rationale: Felbamate (FBM) is a dicarbamate anticonvulsant showing high efficacy in reducing seizures, particularly for Lennox–Gastaut syndrome. FBM also exhibits neuroprotective properties. However, hematotoxicity and hepatotoxicity have limited the use of FBM. This toxicity is likely related to the formation of oxidative metabolites. Fluorofelbamate is a newly developed anticonvulsant structurally related to FBM that exhibits similar anticonvulsant efficacy. However, fluorofelbamate is less likely to produce oxidative metabolites, because of its fluoridated structure. We have previously found that FBM provides excellent protection against CA1 depolarization-induced injury. Therefore, we assessed whether fluorofelbamate would also be neuroprotective against depolarization-induced injury. Methods: Using paired rat hippocampal slices, we monitored the CA1 orthodromic and antidromic population spike (PS) amplitude during depolarization injury with and without fluorofelbamate treatment. To induce depolarization injury, slices were exposed to 25 mM KCl for 8 min. Treatment with fluorofelbamate was begun 30 min before KCl exposure and continued for the first 15 min of recovery. Final recovery was assessed 60 min after exposure to KCl. Results: Fluorofelbamate supplied dose-dependent neuroprotection of CA1 PS amplitude in hippocampal slices at concentrations of ≥25 mg/L (p < 0.05). At a fluorofelbamate concentration of 100 mg/: CA1 orthodromic and antidromic PS amplitude recovered to 94 ± 2% and 96 ± 2%, compared with unmedicated slices, which recovered to 13 ± 3% and 14 ± 1%, respectively. The EC50 for fluorofelbamate was 54 and 53 mg/L for strong neuroprotection against CA1 depolarization injury. By comparison, the EC50 for FBM against this injury was 138 and 132 mg/L. Conclusions: These results demonstrate that the anticonvulsant fluorofelbamate provides protection against CA1depolarization-induced neuronal injury. This neuroprotection appears to be more potent than that seen for FBM in this model. These data also suggest that fluorofelbamate may be a useful in the preventing or limiting of brain injury from status epilepticus. (Supported by VA Research Services.)
1.311 RESPIRATORY ADVERSE EVENTS ASSOCIATED WITH DIAZEPAM RECTAL GEL: A SUMMARY OF THE CLINICAL EXPERIENCE 1 John M. Pellock and 2 Shlomo Shinnar (1 Comprehensive Epilepsy Institute, Virginia Commonwealth University, Medical College of Virginia, Richmond, VA; and 2 Comprehensive Epilepsy Management Center, Montefiore Medical Center, and Albert Einstein College of Medicine, Bronx, NY) Rationale: Respiratory depression is rare after administration of rectal diazepam (DZP). In four studies involving >500 patients treated with DZP rectal gel, no episodes of respiratory depression were reported. However, the possibility of respiratory depression as a consequence of home administration of DZP rectal gel remains a concern among clinicians and caregivers. Methods: All respiratory adverse events (those termed respiratory depression, hypoventilation, dyspnea, or apnea) spontaneously reported to Xcel Pharmaceuticals, which markets DZP rectal gel, were reviewed to evaluate the incidence of respiratory depression and the possible association of each of these adverse events with the administration of DZP rectal gel. Results: As of February 1, 2003, >1.3 million DZP rectal gel syringes were prescribed, and it is estimated that >1 million doses have been given. Reports of respiratory adverse events were filed for six patients. Two subjects had a history of respiratory depression and apnea associated with their seizures, which also occurred with rectal DZP gel use. One patient became unresponsive and had shallow breathing without cyanosis, which spontaneously resolved before arrival in the emergency department. One consumer-filed report observed only that the patient had difficulty breathing that started a few minutes after DZP rectal gel administration, and was attended to by emergency personnel with full recovery. Two subjects had more serious respiratory adverse events. One required respiratory support after administration of a total of 20 mg DZP rectal gel, a dose well in excess of that recommended for his age and weight. The other was hospitalized for respiratory distress secondary to an anaphylactic reaction after DZP rectal gel administration. The patient tested positive for opiates, which were not prescribed. The patient’s history was Epilepsia, Vol. 44, Suppl. 9, 2003
remarkable for a similar reaction, also involving opiates, which resulted in cardiorespiratory arrest. It is unknown whether benzodiazepines were administered before the first event. Conclusions: Reports of respiratory depression with DZP rectal gel are very rare. In the majority of cases, other factors were also involved. While relying on reported adverse events may underestimate the total number of respiratory adverse events, the extreme rarity of these reports, even with widespread clinical use, is reassuring. Respiratory depression can also occur with seizures, and in one recent study of the prehospital treatment of seizures (Alldredge et al., N Engl J Med 2001;345:631–7) was as common in the placebo group as in the i.v. benzodiazepine-treated group. When used as recommended, DZP rectal gel has an extremely low rate of respiratory depression. (Supported by Xcel Pharmaceuticals.) 1.312 IMPACT OF SEIZURE ACTIVITY ON PHENYTOIN CONCENTRATIONS IN THE FOCUS REGION OF AMYGDALAKINDLED RATS Heidrun Potschka, Steffen Baltes, and Wolfgang L¨oscher (Department of Pharmacology, Toxicology, and Pharmacy, Veterinary School, Hannover, Germany) Rationale: Decreased access of antiepileptic drugs (AEDs) to the epileptic focus region due to enhanced multidrug transporter expression at the blood–brain barrier is considered to be one mechanism involved in drug-refractoriness of epilepsy. Investigations in different rodent models of epilepsy have demonstrated induction of multidrug transporter expression in response to seizure activity. Furthermore transporter-mediated efflux of phenytoin (PHT) at the blood–brain barrier has been demonstrated for two multidrug transporters. The aim of the present study was to evaluate the impact of overexpression of multidrug transporters and of an acute leakiness of the blood–brain barrier in response to seizures, which has been reported repeatedly, on concentrations of PHT in the epileptic focus region. Methods: Bipolar electrodes and guide cannulae were implanted side by side in the amygdala of female Wistar rats. The animals were kindled by daily constant-current electrostimulation. Microdialysis experiments were performed in fully kindled rats. PHT was administered 14 days or 2 h after a single generalized seizure induced by stimulation via the depth electrode. PHT plasma and amygdala dialysate concentrations were repeatedly analyzed by HPLC. In additional experiments, PHT was administered 10 min before a generalized seizure, which was elicited during the microdialysis procedure, or during a self-sustained status epilepticus induced by prolonged electrostimulation of 30 min. Results: The PHT amygdala dialysate-to-plasma ratio proved to be 50 mg/L at doses listed. Conclusions: These simulations suggest that no significant perturbation in plasma [VPA] is likely when converting DVP q12h to daily ER “all at once” 12 h after the a.m. or p.m. DVP dose. Waiting 24 h to convert from DVP to ER results in too large a decrease in plasma [VPA] and is not recommended. For daily ER dosing, there is no apparent advantage in converting DVP to ER in small steps; no tapering of DVP with concurrent ER upward dose titration is needed. (Supported by Abbott Laboratories.)
1.314 USE OF SUSTAINED-RELEASE BUPROPION IN PATIENTS WITH EPILEPSY 1 Stanley R. Resor, Jr., and 2 Louise D. Resor (1 Neurology, Columbia University, New York City, NY; and 2 Neurology, Stamford Hospital, Stamford, CT) Rationale: Depression is more common among patients with epilepsy than it is in the general population. Bupropion hydrochloride, is an effective adjunctive and first-line antidepressant. Its use is not associated with most of the side effects that limit the usefulness of selective serotonin reuptake inhibitors. Moreover, it can be helpful in smoking cessation and painful neuropathies. However, it can cause seizures in patients without risk factors for epilepsy. The sustained-release preparation appears less likely to cause seizures but is still contraindicated in patients with a seizure disorder. Nevertheless, some patients who have seizures while taking bupropion appear to do well when switched to the sustained-release preparation, and some patients with epilepsy treated with anticonvulsants (AEDs) appear to tolerate bupropion without an exacerbation of their seizures. We attempted to find clinical evidence to support the prohibition of sustained-release bupropion in patients with epilepsy.
111
Methods: We reviewed the records of all patients with epilepsy treated by us over the past 5 years and identified 28 (six with idiopathic primary generalized epilepsy and 22 with localization-related epilepsy) who had used sustained-release bupropion in divided doses. We reviewed the patient’s charts and when necessary spoke with the patients by phone. Results: One patient with 38 years of uncontrolled myoclonic and absence seizures reported that a single 100-mg tablet increased the frequency of her myoclonus. The five other patients with generalized epilepsy whose seizures were all easily controlled (three with juvenile myoclonic epilepsy, one with awakening grand mal, and one with absence and grand mal) have remained seizure free for 2 months to 2 years on 150 mg b.i.d. Two patients with intractable weekly partial seizures tried the drug to stop smoking. For one it worked without incident the first time, but the following year, at the same 300-mg dose, she had more simple partial seizures. The other patient tried it three times; the second time she had more complex partial seizures. Another patient with intractable weekly seizures due to an oligodendroglioma reported more “warnings” but no seizures. The remaining 19 patients including 11 with uncontrolled partial seizures and two with tumors reported that bupropion (200 mg in three and 300 mg in 16) had no effect on their seizures when taken for periods of 2 weeks to 2 years. Conclusions: In this small group of patients with epilepsy, more than half of whom had uncontrolled seizures, sustained-release bupropion given in divided doses ≤300 mg per day caused no convulsive seizures and an increase in seizures associated with impairment of consciousness in only one patient (and then only on one of the three times she tried it). Depression and tobacco addiction are serious conditions. When they coexist with epilepsy, the use of bupropion may be a reasonable consideration. ∗
Abstract 1.315 has been withdrawn
1.316 DOES VALPROATE RESPONSE IN JUVENILE MYOCLONIC EPILEPSY DEPEND ON FREQUENCY OF SEIZURES AND DURATION OF EPILEPSY? Patr´ıcia S. Sousa, Paulo B.N. Liberalesso, Fl´avia S. Miyshiara, Eliana Garzon, Am´erico C. Sakamoto, and Elza M.T. Yacubian (Unidade de Pesquisa e Tratamento das Epilepsias, Escola Paulista de Medicina, Universidade Federal de S˜ao Paulo, S˜ao Paulo, Brazil) Rationale: Juvenile myoclonic epilepsy (JME) is an idiopathic generalized epileptic syndrome with age-related onset. It is characterized by myoclonic jerks on awakening, generalized tonic–clonic seizures (GTCSs), and absence seizures. Myoclonic jerks, the hallmark of the syndrome, are often omitted or forgotten by the patients who only seek medical assistance after the first GTCS. The ideal treatment consists of valproate (VPA) associated with avoidance of precipitant factors (APF), which is considered effective in ∼90% of the cases. We retrospectively assessed the response to VPA in 50 patients who had neither correct syndromic diagnosis nor seizure control. Methods: Fifty patients were included, 30 men, aged between 12 and 52 years, evaluated as outpatients between 1998 and 2003. We estimated the number of GTCSs (>20 or 6 months in four. Of the 11 patients (22%) who had persisting seizures despite VPA+ APF, the duration of the disease varied from 2 to 23 years (13.7 ± 8.3); four patients (36.3%) had 20 GTCSs. Factors possibly related with persistence of seizures included lack of compliance (eight), abnormal
Epilepsia, Vol. 44, Suppl. 9, 2003
112
AES PROCEEDINGS
lifestyle (10) insufficient treatment (one,) and intolerable side effects (one). Conclusions: Little is known about the factors involved in JME response to VPA+APF. In this series the duration of the disease and number of GTCSs did not influence the response to treatment, even when the diagnosis was established about one decade after seizure onset and use of other therapeutic schemes. Despite these factors, in one third of the patients, control of the seizures was reached in 12 points). Depression (r = −0.72), seizure bothersomeness (r = −0.54), seizure severity (r = −0.37), and days disabled with seizures (r = −0.65) were significantly correlated with poorer HRQOL in all domains than found for nondepressed patients (all subscales, p < 0.0001). Conclusions: Clinical depression is significantly associated with poorer HRQOL among people with all types of seizures. Categorization of people by presence of depression was important in understanding HRQOL among all seizure types. Diagnosis and treatment of depression could be an important contribution to wellness of people with epilepsy whether seizures are controlled. (Epilepsy Impact Project supported by Glaxo SmithKline.) 1.334 WHAT PARENTS AND CHILDREN WORRY ABOUT WITH A NEW DIAGNOSIS OF EPILEPSY: THE BENEFIT OF EDUCATIONAL GROUPS Sandra Cushner-Weinstein, Steven Weinstein, Leslie Bethke, Jay Salpekar, Nancy Elling, Phillip Pearl, Joan Conry, Madison Berl, Marian Kolodgie, Audrey Scully, and William Gaillard (Department of Neurology, Children’s National Medical Center, Washington, D.C.)
117
Rationale: Parents of children diagnosed with epilepsy worry and their concerns evolve throughout the course of the condition. Children also have their own sets of concerns. To educate parents and their children effectively about epilepsy, information should be provided within the context of their needs at the time of diagnosis. Condition-specific groups that address universal and specific concerns may help to increase knowledge and allay worries. Methods: The 113 families, in groups, attended a 2-h New Onset Seizure educational clinic (NOC), within 6–8 months of their child’s initial diagnosis of epilepsy. The 65 parents of the 113 families completed a mailed questionnaire about worries and concerns with eight questions, seven yes/no, and one rating level of worry (10 = most). The 30 children (ages 5–10) also returned an open-ended questionnaire. Information gathered helped address their concerns and worries. After the NOC, attendees completed a seven-question evaluation ranking their level of understanding on a 5-point scale, from strongly disagree (1) to strongly agree (5). Results: Of the 65, 58% returned their mailed questionnaire, rating their level of worry as 7–10 (30% selected 10); 72% described their child’s seizure, but only 32% identified its medical name; 74% were concerned that the seizures would affect their child’s ability to learn and/or behavior; 52% did not know the medication side effects to monitor or how the medication worked; 67% believed their child’s activities would need to be restricted; 43% were uncomfortable explaining seizures to their child; 55% felt that other family members would have difficulty adjusting to their child’s condition. An open-ended question about other concerns led to 49 different responses. The most prevalent concerns in order of frequency were medication effects, ability to outgrow seizures, psychological impact of seizures, drug interactions, dealing with schools and teachers, and what causes and increases seizures. The 113 families completed a training evaluation. The following percentages of parents rated 4 or 5 (strongly agree) to a better understanding of: seizure recognition (90%); the effect of seizures on activities and safety (92%); and what to do the when their child has a seizure (89%). 99% would recommend the program to other parents. Children worried about death, feeling different, the uncertainty of when a seizure would occur, and where and how to tell others, peer’s reactions, being picked on, blood work, would they stop breathing during a seizure, and if their parents would leave. Conclusions: Condition-specific educational groups for parents of similar-aged children with epilepsy can increase knowledge and allay worries when relevant concerns are addressed. A questionnaire can highlight these concerns to assist with teaching. Evaluations can provide feedback regarding the parents’ level of understanding.
1.335 UNDERSTANDING THE LINK BETWEEN THE BURDEN OF EPILEPSY AND DEPRESSION ON SELF-MANAGEMENT IN ADULTS WITH EPILEPSY 1 Colleen A. DiIorio, 2 Patricia O. Shafer, 1 Richard Letz, 2 Donald L. Schomer, 3 Thomas Henry, 1 Katherine Yeager, and1,2,3 Project EASE Study Group (1 Rollins School of Public Health, Emory University, Atlanta, GA; 2 Department of Neurology, Beth Israel Deaconess Medical Center, Boston, MA; and 3 Department of Neurology, Emory University, Atlanta, GA) Rationale: Personal and environmental factors are critical influences on epilepsy self-management. The burden of epilepsy from depression, stigma, and negative expectations may affect a person’s coping, compliance, and quality of life. Understanding their impact on self-management practices will help identify poor self-managers and those at risk for psychosocial difficulties. Methods: This study is part of a larger study of epilepsy selfmanagement conducted at three centers in Boston and Atlanta. Participants completed three assessments each 3 months apart addressing demographics, seizures, and personal and environmental factors associated with self-management of epilepsy. The presence of depression, perceived stigma, lack of social support, and negative outcome expectancies and were correlated to self-management practices using descriptive statistics, including correlations. Structural equation modeling is being conducted to evaluate the fit of these variables into a model of epilepsy self-management.
Epilepsia, Vol. 44, Suppl. 9, 2003
118
AES PROCEEDINGS
Results: The 314 adults were mean age 40, 50% female, 80.3% white. Mean age at seizure onset was 22 years, with seizure duration of 17 years; 76% had a seizure in past year, but only 35% perceived poor control of seizures. Depression scores using CES-D scale correlated highly with negative outcome expectancies for seizures (0.500∗∗ ) and treatment (−0.418∗∗ ), lack of social support (−0.507∗∗ ), low selfefficacy (−0.467∗∗ ), stigma (0.428∗∗ ), and low patient satisfaction (−0.347∗∗ ). While outcome expectancies for management were not related, medication and stress self-management practices did correlate with depression. Modeling showed that outcome expectancies appear to influence stigma, stigma links to depression, and depression is a weak predictor of self-management. The strongest link to self-management was mediated through self-efficacy. Further analysis is under way to examine the group of patients with these negative consequences in more detail. (∗∗ Significant at 0.01 level) Conclusions: This study highlights the intricate relations among personal and environmental negativity and self-management practices. Mood disorders in epilepsy are fairly common, but are often not diagnosed or treated adequately, adding to the burden of epilepsy. The weak relation between depression and self-management may contribute to these problems being too easily overlooked in clinical practice. Further analysis of these factors may improve ability to assess and influence self-management behaviors and depression. (Supported by grant number: R01-NR04770 from the National Institute of Nursing Research and in part by grant number: M01-RR01032 from the National Institutes of Health to the Beth Israel Deaconess Medical Center GCRC.) 1.336 SEIZURE FREQUENCY BEFORE AND AFTER SEPTEMBER 11, 2001 Karen E. Eck, Tracey A. Perrine, and Julianne Costabile (Neurology, Columbia University, New York, NY) Rationale: Patients with epilepsy frequently identify stress as a precipitating factor in seizure occurrence. September 11, 2001, was considered to be a significantly stressful event for most individuals within the New York Metropolitan area. To date, inconclusive data evaluate how different types of stress, described as significant life events or daily stressors, play a role in seizure frequency. Methods: We retrospectively evaluated 31 individuals, both surgical and nonsurgical (14 males, 17 females) from the Multi-Center Study of Outcomes of Epilepsy Surgery (Columbia University site), who had seizures during the investigative time period. The study population was age 22–66 with refractory epilepsy who had seizures and/or auras 3 months before 9/11 and 3 months after 9/11. The Outcome Study protocol required patients to record seizures and isolated auras as separate events thereby enabling us to evaluate them separately. Three patients were excluded due to a combination of epileptic and nonepileptic seizures. Results: Three months before 9/11, the nonsurgical group experienced a mean seizure frequency of 16.1. In the same group, 3 months after 9/11, there was a mean seizure frequency of 19.7 (p = 0.184). The mean seizure frequency for the surgical group was 2.4, 3 months before 9/11, and the mean seizure frequency was 3.4 after 9/11 (p = 0.279). Auras in the pre-9/11 period, for the nonsurgical group, was a mean of 5.3. Auras had a mean increase of 9.4 after 9/11 (p = 0.216). In the surgical group, there was mean aura frequency of 3.2 before 9/11 and a mean aura frequency of 1.8 after 9/11, respectively (p = 0.192). Conclusions: There was no statistically significant difference in seizure and/or aura frequency during the investigative time period for the study population. There was a trend, however, toward an increase in seizures during the post-9/11 period. It is worthwhile to note that two patients (one surgical, one nonsurgical) had seizures on 9/12. Additionally, during the routine study interviews, many patients reported stress as seizure provoking. Stress is very individualized, and it is not clear that the modest change in seizure frequency was related to the events of 9/11. The results indicate that further study is warranted to investigate the effects of this common phenomenon on seizures and epilepsy. 1.337 PSYCHOSOCIAL OUTCOMES IN CHILDREN AND ADOLESCENTS 2 TO 4 YEARS AFTER EPILEPSY SURGERY: HAS ANYTHING CHANGED?
Epilepsia, Vol. 44, Suppl. 9, 2003
1,3 Irene
Elliott, 4 Lucyana Lach, and 2,3 Mary Lou Smith (1 Department of Nursing, 2 Department of Psychology, 3 Division of Neurology, and 4 Department of Social Work, McGill University, Montreal, Quebec, Canada) Rationale: Phase one of our longitudinal study identified few psychosocial changes in children and adolescents 1 year after surgery when compared with a group of youth with intractable seizures. The purpose of phase two was to determine if changes in psychosocial well being occur beyond the second year after surgery. Methods: Data were obtained from our larger, prospective study. A subset of youth (19 surgery; 13 nonsurgery) completed the Piers-Harris (PH) and their mothers (20 surgery; 12 nonsurgery) completed the Child Behavior Checklist (CBCL) at three points in time: before surgery or baseline (time 1), and at 1 year (time 2) and 2–4 years (time 3) after surgery or a comparable follow-up time for the nonsurgery group. At baseline, groups were comparable in age, age at seizure onset, number of medications, and IQ. Repeated measures analyses were used to obtain Group × Time interactions. Linear regression analyses were undertaken within the surgical group to explore the extent to which baseline scores and seizure outcome predicted change (calculated as difference between time 3 and time 1 scores). Results: Fifty percent of this surgery group subset were seizure free at time 3. On most CBCL measures, changes over time were seen to a similar degree in both the surgery and nonsurgery groups. Two Group × Time interactions were observed, with the surgery group showing improvements in the CBCL social problems subscale [F(2, 60) = 3.56, p < 0.035] and the CBCL social sub-scale [F(2, 58) = 4.74, p < .012]. No significant differences were obtained for the PH measures. Baseline (time 1) performance predicted change on most CBCL and PH scales and subscales. Seizure outcome (seizure free, not seizure free) predicted change toward improvement on the CBCL social subscale (adjusted R2 = 0.335, p < 0.008) in the seizure-free group. Conclusions: This study emphasizes the importance of a comparison group and longitudinal assessment to assess changes over time. Except for improvements observed on the CBCL social problems and social scales, changes found within the surgical group on other CBCL scales were seen to a similar degree in the nonsurgical group. Of particular interest, youth did not report any significant changes or interactions on the PH scales. The hoped for improvements in behavior and emotional functioning anticipated by families and youth may not be realized. These findings provide direction for discussions with youth and their families regarding realistic expectations of surgery outcomes. (Supported by the Ontario Mental Health Foundation.) 1.338 THE OTHER CLINICAL EXPRESSIONS OF DEPRESSION IN EPILEPSY: WHAT THE AVAILABLE DIAGNOSTIC CRITERIA DO NOT DETECT 1 Andres M. Kanner, 1 Joanne J. Wuu, 2 Bruce Hermann, 3 Kimford J. Meador, 4 John Barry, and 5 Frank Gilliam (1 Neurological Sciences, Rush Medical College and Rush-Presbyterian St. Luke’s Medical Center, Chicago, IL; 2 Neurology, University of Wisconsin School of Medicine, Madison, WI; 3 Neurology, Georgetown University School of Medicine, Washington, D.C.; 4 Psychiatry, Stanford University School of Medicine, Palo Alto, CA; and 5 Neurology, Washington University School of Medicine, St. Louis, MO) Rationale: Various authors have questioned the sensitivity of the diagnostic criteria suggested by the Diagnostic and Statistical Manual of Mental Disorders (Third and Fourth editions) to identify certain types of depressive disorders in epilepsy. The purpose of this study was to determine whether patients with epilepsy and no identifiable current mood disorder by DSM-IV criteria present symptoms of depression, irritability, and anxiety, and if so, to establish whether their presence has a direct impact on these patients’ quality-of-life ratings. Methods: We created a 46-item mood disorder instrument (Neurologic Disorders Depression Inventory-Epilepsy) with the aim of identifying symptoms of depression, anxiety, irritability, psychosis, and physical symptoms. Each item was scored on a 1–4 Likert Scale; 205 adult outpatients were prospectively evaluated at five university-based epilepsy centers. All patients completed the NDDI-E, a comprehensive set of mood and medication toxicity instruments and the Quality of Life in Epilepsy
AES PROCEEDINGS Inventory-89 (QOLIE-89). A current mood disorder was identified with the structured Clinical Interview for DSM-IV Axis I Disorders (SCID). NDDI-E items were grouped into five domains: mood, irritability, anxiety, psychosis, and physical symptoms. Forty-three patients met DSM-IV criteria for one of the current mood-disorder categories and were excluded from the study; 162 patients were included. Patients with at least one positive item (scored as 3 or 4) in the three domains of mood, anxiety, and irritability were coded as “symptomatic” and the rest “not symptomatic.” We compared the QOLIE-89 scores between symptomatic and nonsymptomatic patients and investigated whether the “item load” (number of items per domain) had a significant impact on the QOLIE-89 scores. Results: There were 71 symptomatic (44%) and 91 nonsymptomatic patients (56%). Symptomatic patients experienced a median of seven mood items, four anxiety items, six irritability items, and three psychosis items. The QOLIE-89 scores of the symptomatic patients were significantly lower (68.5 ± 13.3) than those of nonsymptomatic (76 ± 14.7, p = 0.0002, Wilcoxon Rank-Sum). Among the 71 symptomatic patients, bigger item loads were significantly associated with worse QOLIE-89 scores for all three domains: mood (one to four vs. five to 16 items, 73.7 ± 12.2 vs. 64.7 ± 12.8, p = 0.004), anxiety (one to three vs. four to eight items; 74.1 ± 11.0 vs. 62.1 ± 12.8, p = 0.0002), and irritability (one to three vs. four to 13 items; 75.1 ± 11.7 vs. 65.6 ± 13.0, p = 0.0037). Conclusions: These data show that in epilepsy patients, the available DSM-IV criteria of mood disorders are insensitive in identifying relatively frequent symptoms of depression that have a negative impact on quality of life. We are investigating whether these symptom clusters lead to homogeneous clinical pictures, as suggested by Kraepelin (1923) and Blumer (1995). (Supported by Glaxo-Smith-Kline.) 1.339 TREATMENTS AND PERCEPTIONS OF EPILEPSY IN KASHMIR AND THE UNITED STATES: A CROSS-CULTURAL ANALYSIS Alisa Khan, Victoria Huerter, and Elizabeth A. Thiele (Pediatric Epilepsy Program, Massachusetts General Hospital, Boston, MA) Rationale: Treatments for and perceptions of epilepsy have been found to vary across cultures. The objective of this study was to compare social perceptions of epilepsy as well as attitudes and practices surrounding conventional and alternative treatments in a developed Western and a developing Eastern society. Methods: Interviews were conducted with 25 Kashmiri and 25 American epilepsy patients and their families. Patients were interviewed about perceptions of epilepsy they held or had encountered, alternative treatment they had considered or sought, and the ways in which the disease had affected their quality of life. Alternative and conventional healthcare providers were also interviewed. Results: The use of alternative therapies, particularly spiritual ones, in conjunction with conventional therapies was more prevalent in Kashmiri than in American society. Although Kashmiris utilized conventional medicine as a primary treatment, spiritual remedies were often concurrently sought, and herbal therapies were frequently used as adjunctive therapy. Psychiatrists, general practitioners, and neurologists were primarily responsible for providing conventional treatment for Kashmiri epilepsy patients. American patients, however, tended to rely predominantly on conventional treatments by neurologists; reliance on alternative therapies was found to be comparatively minimal. Although Kashmiri physicians prescribed a similar regimen of traditional antiepileptic drugs as did their American counterparts, including phenytoin, valproic acid, carbamazepine, and phenobarbital, a wider range of drugs and treatments, such as the vagus nerve stimulator and the ketogenic dDiet, were available and used by the latter. And although patients in both societies tended to defer to their physicians to determine the cause of their illness, Kashmiri patients were more likely to attribute it, at least in part, to spiritual causes. Quality of life for Kashmiri patients was found to be poorer in many respects, including greater absence of educational and occupational opportunities, higher prevalence of feelings of stigmatization, and lesser tendency for openness with others about their illness. Additionally, psychosocial support, while limited in both societies, was less readily available in Kashmir. Conclusions: The results of this study affirm the need for sensitivity and understanding of the cultural, social, and spiritual aspects of the
119
commonly held views and practices of patients. Such cross-cultural understanding can help to ensure optimal patient treatment and care within and across individual societies. 1.340 EPILEPSY, PSYCHIATRIC DISORDERS, AND FAMILY FUNCTIONING 1 W. Curt LaFrance, Jr., 3 Andrew S. Blum, 2 Christine Ryan, 2 Joan Kelley, and 2 Gabor I. Keitner (1 Neurology and Psychiatry, 2 Psychiatry, and Neurology, Brown Medical School, Rhode Island Hospital, Providence, RI) Rationale: Research shows that epilepsy affects quality of life (QoL) in patients with epilepsy and their family members. QoL consists of health, family life, community, economics, and personal development variables. Studies of family functioning revealed that family members of patients with epilepsy scored in the unhealthy range on family dimensions. We sought to assess the impact of family functioning and QoL on individuals with epilepsy and their families. Methods: The first 10 patients to date in an ongoing study, between the ages of 18 and 90 years, diagnosed with epilepsy by history, physical, and electroencephalography (EEG), were administered a Family Assessment Device (FAD), Quality of Life in Epilepsy (QOLIE-31), Psychiatric Diagnostic Screening Questionnaire (PDSQ), and Seizure Diary on entry into the Rhode Island Hospital Comprehensive Epilepsy Program. In addition, the individual’s first-degree relatives (and/or primary caretaker) completed the FAD to assess perception of family functioning and was asked to assist in documenting the seizure frequency and current and past treatment history. We compared family means of nine patients with epilepsy and their family member with nine age/sex/marital status– matched patients with acute phase unipolar depression and their family, versus nine matched patients with acute phase bipolar disorder and their family, versus 23 normal controls and their family. Results: Patients with epilepsy scored in the unhealthy range in the following areas: roles (four of nine), affective involvement (three), and behavioral control (four). T-test comparisons revealed that baseline epilepsy family means were healthier than family means for patients with depression and bipolar disorder in the areas of communication, affective responsiveness, and behavioral control. Family means for the epilepsy group were healthier than normal controls in problem solving, communication, and affective responsiveness. Conclusions: The preliminary findings reveal that individual patients with epilepsy report dysfunction within specific elements of family functioning. The disparate combined mean scores between epilepsy, clinical, and control families may be indicative of a subset of patients with epilepsy and their families with healthier family functioning in various subscales. This may be related to (a) sample size, or (b) selective bias in research-cooperative patients. Alternatively, it may reflect family dynamic differences between phase of illness status, and/or illness severity in patients with well-controlled or intractable epilepsy. This study is part of an ongoing assessment of family functioning, QoL, and seizure frequency in patients with epilepsy. (Supported by Rhode Island Hospital.) 1.341 STIGMA OF EPILEPSY IN THE WORKPLACE IS NOT ASSOCIATED WITH VISIBILITY OF THE ILLNESS Blagovest G. Nikolov, Susanne M. Vera, Luydmila Jovine, and Cynthia L. Harden (Comprehensive Epilepsy Center, Weill Medical College at Cornell University, New York, NY) Rationale: To compare perceptions that co-workers have of three chronic neurologic illnesses: epilepsy, multiple sclerosis, and depression based on the visibility of the disease-disabling symptoms. Methods: We distributed a 31-question survey to employees in two New York City companies. Three vignettes were presented describing new employees who eventually revealed that they had either epilepsy; depression, or multiple sclerosis. Multiple sclerosis was presented as a visible illness, depicting a person with a disability, whereas the persons in the epilepsy and depression vignettes had no visible disabling features, and no seizure was described in the epilepsy vignette. Perceptions were assessed by a 3- or 4-degree (choices) scale. Results: Seventy-four of 200 surveys were returned (37% response). Forty-four percent of respondents reported that they realized the subject Epilepsia, Vol. 44, Suppl. 9, 2003
120
AES PROCEEDINGS
of the vignette had MS before it was revealed to them, compared with 28% with depression and 12% with epilepsy. Forty-two percent of respondents reported “much more” or “somewhat more” anxiety at the idea of interacting with the person with epilepsy, compared with 40% with depression and 30% with MS. Twenty-one percent of respondents reported that they were “very worried” that the person with epilepsy would have sudden, unpredictable behavior, compared with 15% for depression and 8% for MS. Ten percent of respondents thought that the depressed subject could “very much” have avoided getting depression, compared with 3% for MS and 7% for epilepsy. However, 43% of respondents reported that they thought epilepsy would have no effect on job ability, compared with 32% for depression and 16% for MS. Conclusions: Although well-controlled epilepsy is an invisible illness, it produced the highest level of anxiety in the workplace among the illnesses evaluated. Visibility of a neurologic illness is associated with the impression that the job is at risk for being performed inadequately, but among epilepsy, depression, and MS, the visible and known illness produced less anxiety in the workplace. The “blame” of being afflicted with neurologic illness also be based on the social perception of the disease. 1.342 A QUANTITATIVE STUDY OF ANXIETY AND DEPRESSION SYMPTOMS IN PATIENTS PRESENTING TO A FIRSTSEIZURE CLINIC 1,5 Rosemary J. Panelli, 1 Susan Moore, 2,3 Wendyl J. D’Souza, 2 Daniel Ghougassian, and 2,4 Terence J. O’Brien (1 School of Social and Behavioural Sciences, Swinburne University of Technology, 2 Department of Neurology, The Alfred, 3 Department of Clinical Neurosciences, St. Vincent’s Hospital, University of Melbourne, 4 Department of Medicine, University of Melbourne, and 5 Epilepsy Foundation of Victoria, Melbourne, Victoria, Australia) Rationale: Elevated rates of anxiety and depression have been linked to epilepsy. However, little is known about their incidence or severity in patients with a suspected first seizure. This study measured anxiety and depression symptoms in patients referred to a First Seizure Clinic. Methods: Patients presenting to the First Seizure Clinic of a large metropolitan hospital were offered enrollment in a prospective longitudinal Quality of Life study. Patients were predominantly referred from the hospital’s Accident and Emergency Department. A self-administered baseline questionnaire, including the Hospital Anxiety and Depression Scale (HADS), was completed before the consultation. Patients were seen by an epileptologist who diagnosed the event as a seizure or a nonseizure according to the International League Against Epilepsy classification. Results: Eighty-four patients completed the baseline questionnaire. Sixty-one percent were subsequently diagnosed as having had a seizure, and 39% were diagnosed as having had a probable nonepileptic event, 46% of which were syncopal. Anxiety scores at HADS case level (11– 21) were measured in 30% of nonseizure patients and 29% of seizure patients. Depression scores at HADS case level (11–21) were measured in 27% of nonseizure patients and 14% of seizure patients. The scores did not differ significantly between the seizure and the nonseizure groups for anxiety (p = 0.763) or depression (p = 0.065). Conclusions: The data indicate high levels of depression and anxiety in patients referred to a First Seizure Clinic, regardless of whether they are subsequently confirmed to have had a seizure. This has important implications for the psychosocial management of this patient group. (Supported by Australian Research Council; Epilepsy Foundation of Victoria.) 1.343 COST OF ANTIEPILEPTIC DRUGS: RELATION TO SEIZURE AND SOCIOECONOMIC VARIABLES AND QUALITY OF LIFE 1 Elia M. Pestana, 1 Nancy Foldvary-Schaefer, 1 Diana Marsilio, 2 Christopher Burant, and 1 Harold H. Morris (1 Neurology, The Cleveland Clinic Foundation, and 2 Bioethics, Case Western Reserve University, Cleveland, OH) Rationale: The introduction of new antiepileptic drugs (AEDs) and other new techniques for the diagnosis and treatment of epileptic seizures
Epilepsia, Vol. 44, Suppl. 9, 2003
has increased the cost of epilepsy. This study aimed to determine the relation between seizure severity/frequency, socioeconomic factors, and coverage of medical expenses with the monthly cost of AEDs. Methods: The 288 patient with epilepsy admitted for video-EEG (VEEG) monitoring at the Cleveland Clinic Foundation (CCF) were studied. Patients completed a questionnaire regarding demographics, seizure variables, socioeconomic variables, health care utilization, coverage for medical expenses, and self-rating quality of life (QOL) that was created at the CCF. The monthly cost of AEDs was compared (Pearson χ 2 ) or correlated (Pearson’s R coefficient) with the following variables: seizure severity/frequency, seizure and epilepsy classification, health care utilization, major daily activities, coverage for medical expenses, discontinuation of AEDs because of cost, and self-reported QOL. Differences were established when >95% of significance was obtained. Results: No correlation between seizure variables and monthly cost of AEDs was found. Patients who had a higher monthly cost of AEDs were more likely to be on polytherapy (χ 2 = 85.68, p = 0.000) (Pearson’s R = 0.213; p = 0.000), including three or more drugs (χ 2 = 99.65; p = 0.000; Pearson’s R = 0.191; p = 0.001) and have more adverse effects due to AEDs (Pearson’s R = 0.220; p = 0.000). Patients with higher monthly cost of AEDs were also more likely to have income from the government (SSI or SSD; χ 2 = 15.34; p = 0.009; Pearson’s R = 0.236; p = 0.000) and assistance from federal government programs to pay for medications (χ 2 = 28.13; p = 0.000; Pearson’s R = 0.274; p = 0.000). This group also have the poorest self-reported QOL (Pearson’s R = −0.144; p = 0.023). Current disability and (χ 2 = 10.43; p = 0.064) and discontinuation of AEDs due to cost (χ 2 = 10.25; p = 0.069) correlated weakly with monthly cost of AEDs. A positive correlation between higher monthly cost of AEDs and disability (Pearson’s R = 0.163; p = 0.008) and higher monthly cost of AEDs and discontinuation of AEDs due to cost (Pearson’s R = 0.152; p = 0.016) was found. Conclusions: In this sample of patients with medically refractory epilepsy, cost of AEDs was not related to seizure severity/frequency. Rather, monthly AED expenses were related to polytherapy, ability to pay for medications, and QOL. Federal government programs cover medication expenses in patients with higher monthly AED cost.
1.344 PATIENT EXPECTATIONS AND PREDICTORS OF QOL IN EPILEPSY Linda M. Selwa, Beth Malow, Daniela Minecan, Mihaela Mihaescu, Barbara Schauble, and Beydoun Ahmad (Neurology, University of Michigan, Ann Arbor, MI) Rationale: Quality of life (QOL) in epilepsy in previous studies has not been dependant on severity of seizure disorder or medication effects. We surveyed 90 consecutive patients to determine what their expectations were for epilepsy visits and to evaluate the impact of several features of their illness on QOL. Methods: We surveyed 90 epilepsy patients identified through clinic rosters at the University of Michigan epilepsy program. The patients were asked nine multiple-choice questions about their expectations and limitations, one open-ended question about their fears, and completed a QOLIE 10 questionnaire. Demographic information was collected both directly from patients and from their medical records. Seven epilepsy faculty were asked to predict the patient responses to multiple-choice questions. Results: Average QOL was 2.5/5 in this group where 71% had at least monthly seizures. Most were not married and had no children; 44% were unemployed. The majority were taking a single AED. There was no correlation between QOL and gender, age, or number of AEDs. The number of monthly seizures did not negatively correlate with QOL. QOL scores tended to improve with employment and marital status. Patients feared injuring self or others and losing independence most. The majority of patients refused to answer questions about driving. Consistent compliance with medications was reported by 70% of patients. Physicians were not able to predict a strong desire for improved employment and a very common avoidance of childbearing in our population. Conclusions: Our population of refractory epilepsy patients has a relatively poor QOL that does not depend on the apparent severity of their epilepsy. Issues such as vocational rehabilitation and fears about childbearing were not expected by epilepsy faculty. Patients are reluctant
AES PROCEEDINGS to discuss driving issues with their physicians. Patients’ level of fear about injury and even death are substantial. Efforts to improve QOL including vocational discussions and education about pregnancy may be helpful.
1.345 FACTORS ASSOCIATED WITH MATERNAL ADAPTATION TO A CHILD’S EPILEPSY Cheryl P. Shore and Joan K. Austin (School of Nursing, Indiana Univeristy, Indianapolis, IN) Rationale: Mothers of children who have seizure disorders are at risk for a poor adaptation related to the challenges of managing a condition that can be unpredicable, difficult to control, and often associated with a stigma. The purpose of this study was to investigate the influence of child characteristics, maternal characteristics, and maternal perceptions of the impact of the child’s seizure disorder on maternal adaptation to the child’s epilepsy. Maternal adaptation was conceptualized as low level of negative affect, high confidence in seizure-condition management, and maintenance of family leisure activities. Perceptions of impact were proposed to mediate the effects of child and maternal characteristics on maternal adaptation. Methods: Research participants were recruited from a large Midwestern university medical center, private practices, and school nurses. The children were required to have a diagnosis of a seizure disorder for ≥6 months before enrollment. The sample for this study consisted of 156 mother–child dyads. Data were collected by trained research assistants in structured interviews. Mulitiple regression was used to test the relations among the variables. Three preliminary sets of regressions were carried out as a prerequisite to testing for mediation. First, indicators of maternal adaptation were regressed on child and maternal characteristics variables. The proposed mediators (maternal perceptions of the seizure condition’s impact) were regressed on child and maternal characteristics variables in the second step. Third, indicators of maternal adaptation were regressed on the proposed mediators. Eight potential mediated relations were identified by this method. To test each of the eight relations, both the independent variable (child or maternal characteristic) and the proposed mediator (maternal perception) were entered into the same regression simultaneously. Mediation was supported if the influence of the independent variable decreased to a nonsignificant level. Results: None of the eight relations identified as potentially mediated were supported statistically. The effects of child characteristics, maternal characteristics, and maternal perceptions of impact of the seizure condition were direct rather than mediated. Maternal learned helplessness was associated with all outcome variables. Child internalizing behavior problems were associated with maternal depressive mood, and maternal family satisfaction was associated with confidence in seizure-condition management. Seizure severity and maternal family satisfaction were associated with maintenance of family leisure activities. Conclusions: The results of this study suggest that maternal adaptation to a child’s chronic seizure condition is complex and includes multiple factors in addition to the characteristics of the child’s seizure condition. Mothers who perceive themselves as helpless, are dissatisfied with their nuclear famlies, and have children with more behavior problems are at risk for a poor adaptation. (Supported by NR04536.)
1.346 NOW WE LAY THEM DOWN TO SLEEP: ETHICAL ISSUES WITH THE USE OF PHARMACOLOGIC COMA FOR ADULT STATUS EPILEPTICUS 1 Katrina A. Bramstedt, 2 Harold Morris, and 2 Adriana Serje (1 Department of Bioethics and 2 Department of Neurology, Cleveland Clinic Foundation, Cleveland, OH) Rationale: The use of pharmacologic coma (PC) to treat status epilepticus (SE) is often successful, but not always. Currently there are no established guidelines for the length of PC trial in an attempt to achieve seizure control, thus the potential arises for ethical dilemmas. Methods: Using two case examples, we explore three concepts regarding the use of PC in the treatment of SE: (a) SE as a terminal condition;
121
(b) PC resulting in permanent unconsciousness; and (c) using PC for extended periods to alleviate SE. Results: With respect to a patient’s Advance Directive or Living Will, these three concepts can pose ethical complexities for the medical team. Such health care–planning documents or other prior statements of treatment preferences can be relevant to PC therapy due to the notions of unconsciousness, cognitive appreciation, and life support. Conclusions: Because of the risks of long-term PC therapy, this intervention should be reflected on in conjunction with the patient’s best interests if the patient’s preferences for treatment are not known. We argue that in the setting where PC therapy is not reversing the patient’s clinical course and only offering to sustain organic life, it is ethically appropriate to discontinue such therapy and provide the patient comfort care. Similarly, if PC therapy is not expected to improved the patient’s clinical course and is only expected to sustain organic life, then it is ethically appropriate not to offer such therapy to patients.
1.347 RECRUITMENT RATES OF PEDIATRIC PATIENTS AND FAMILY MEMBERS IN A GENETIC STUDY OF HUMAN EPILEPSY 1 Theresa M. Scattergood, 2 Dennis J. Dlugos, 3 Michael R. Sperling, 1 Thomas N. Ferraro, 1 Wade H. Berrettinni, and 1 Russell J. Buono (1 Psychiatry, University of Pennsylvania, 2 Neurology, Children’s Hospital of Philadelphia, and 3 Neurology, Thomas Jefferson University Hospital, Philadelphia, PA) Rationale: The ability to recruit sufficient numbers of participants into genetic research studies is critical for success in identifying genes related to disease. The purpose of this study is to quantify the percentage of patients and family members who agree to participate by signing informed consent (or verbalizing assent), and giving a peripheral blood sample. Reasons for refusal to enroll will be reported. These data will help in design of future genetic studies and may increase future participation rates. Methods: This is a retrospective analysis of enrollment rates and reasons for refusal to participate in a study entitled “Genetic Influences on Human Epilepsy,” which is currently in progress. Patient charts were screened to determine if study inclusion criteria were met, and the attending physician would introduce the study coordinator at the end of the office visit. Patients and family members then chose whether to participate in the study. Results: Over an 18-month period, 1,570 neurology outpatient clinical records at the Children’s Hospital of Philadelphia were screened for this project. Ninety-nine patients (ages 3 –20) met inclusion criteria and were asked to participate. Of those 99 patients, 80 agreed to enroll (81%), and 19 refused (19%). The major reasons cited for patient refusal included fear of blood draw (42%), a sense of being too overwhelmed to consider participation (16%), and none given (42%). All 80 pediatric patients who agreed to enroll have had blood drawn. In addition, 115 family members of patients who agreed to enroll were asked to participate. To date, 111 (96%) have agreed to enroll, and four (4%) refused participation for unknown reasons. Conclusions: The rate of participation is 81% for pediatric patients and 96% for family members. Because fear of “a needle” is a major reason for refusal in this population, an alternative would be to obtain a DNA sample from a buccal swab. Although the DNA is of lesser quality and quantity when isolated this way, it could increase participation rates in this population significantly. Because newer AEDs require few blood draws for clinical purposes, this may be an even more important issue in future years. A more detailed analysis of the factors influencing decision to refuse to participate is warranted. Parental age, education, ethnic background, and other factors may play a significant role and should be systematically examined in research studies. (Supported by NS-R01-400396 to RJB and the Univ. of PA Center for Neurobiology and Behavior.)
1.348 COMPLICATIONS AND OUTCOMES IN 1,428 SURGICAL PROCEDURES FOR EPILEPSY: A COMPREHENSIVE REVIEW OF CONTEMPORARY EPILEPSY SURGERY FROM ONE EPILEPSY CENTER
Epilepsia, Vol. 44, Suppl. 9, 2003
122
AES PROCEEDINGS
Alyson Silverberg and Werner K. Doyle (Comprehensive Epilepsy Center, New York University Medical Center, New York, NY) Rationale: A retrospective review of a large contemporary surgical series from a single center by one surgeon is presented. The series spans 8 years. Demographics, outcome, efficacy, complications, patient selection criteria, general surgical technical descriptions, and surgical philosophy are discussed. Methods: Data are from 944 craniotomies, 88 bilateral invasive surveys, 69 one-stage resections, 369 two-stage procedures, 42 three-stage procedures, 296 vagal nerve stimulator operations, and 38 stereotactic procedures. The 695 unique patients underwent 480 craniotomies. Average age for patients with craniotomy was 31 ± 12 years, and 47% were female. Outcome reporting uses minimally modified Engel scores. Results: One-stage resections (primarily anterior medial temporal lobectomy with hippocampectomy) when stratified by Engel scores 1/2/3/4, respectively, comprised 91/6/4/0% of the entire group, with 76% follow up. Two-stage procedures, involving invasive VEEG before resection, had respective outcome scores 64/14/14/8% with 81% follow up. Three-stage procedures yielded outcome stratification 50/19/28/3% with 76% follow up. One-, two-, and three-stage operations were associated with decreases of three full Engle grades between preop and postop states in 83%, 53%, and 41%, respectively. Demographics of the patient group without outcome statistics because of outstanding follow-up data are not different from demographics of the reported outcome group. Demographics and complications are reported for all patients. The majority of follow data is >2 years out from surgery. In this entire series, no mortality occurred, and important morbidity was infrequent. The entire cranial series had 15 infections and nine hemorrhages, representing respective relative risk of 1.0%/1.8% for hemorrhage and 1.6%/3.1% risk for infection per craniotomy/per patient. Other common morbidities are subdural hygromas, hydrocephalus, memory deficits, and nerve III injury with respective incidence of 0.4%/0.8%, 0.1%/0.2%, 0.5%/1.0%, and >0.1%/>0.1%. Bone resorption requiring cranioplasty occurred in three patients. No vascular injury with symptomatic or functional results occurred. Conclusions: There was no mortality in this entire series, and important morbidity and complications were infrequent. The most common morbidities were hemorrhage and infection. Single- and multistage procedures have acceptable important risks and excellent to good seizure control.
1.349 ASSESSMENT OF SUBJECTIVE MEMORY COMPLAINTS IN EPILEPSY SURGERY CANDIDATES William B. Barr, Dolly Varghese, Chris Morrison, and Orrin Devinsky (NYU Comprehensive Epilepsy Center, NYU School of Medicine, New York, NY) Rationale: Reports of memory disturbance are common in patients with epilepsy. Studies have shown that subjective ratings of memory disturbance are highly influenced by mood and other psychological factors. Subjective ratings are found to be less related to scores on objective tests of memory. The goal of this study was to reinvestigate this topic with a new instrument that has been modified for use in this population. Methods: We administered the Memory Complaints Inventory (Green and Allen, 1996), modified for use in patients with epilepsy (MCI-E), to 55 subjects (27 male, 26 female) undergoing VEEG monitoring. All were found to have electrophysiologic evidence of frontal (n = 8) or temporal lobe (n = 45) onset, as determined by VEEG recordings of ictal and interictal epileptiform abnormalities. The mean age of the sample was 36.5 years (range, 17–62). Mean years of education was 14.3 years (range, 8–18). Full-Scale IQ for the sample was 98.3 (range, 70–136). The MCI-E consists of 65 items, consisting of ratings on a 5-point scale (0 to 4). Responses are typically divided among 10 subscales. Analyses for this study were limited to the total score, which represents a sum of the item ratings from each of the 65 questions. Results of preliminary analyses indicated that the internal reliability of the instrument was excellent (Coefficient alpha = 0.964). Subjects were also administered a battery of neuropsychological tests including standard measures of language (verbal fluency and naming) and memory (verbal and nonverbal). They also completed inventories of mood [Beck DepresEpilepsia, Vol. 44, Suppl. 9, 2003
sion Inventory (BDI), Beck Anxiety Inventory (BAI)] and quality of life (QOLIE-10). Results: The mean total score on the MCI-E was 68.1 (SD, 44.3). No differences were found in patients with a temporal lobe versus frontal lobe seizure onset. There was no relation with the laterality of the seizure focus. Relations with measures of mood and neuropsychological functioning were assessed with Pearson correlations. The total scores were found to correlate highly with scores on the BDI (r = 0.705; p < 0.001) and the QOLIE-10 (r = 0.600; p < 0.001) and to a lesser extent with the BAI (r = 0.488; p < 0.01). Significant correlations were also found with scores on measures of verbal and nonverbal recall (California Verbal Learning Test, delayed recall; r = −0.400; p < 0.01; Brief Visuospatial Memory Test–Revised, delayed recall, r = −0.332; p < 0.05). No relations were found between MCI-E scores and other measures of verbal (WMS-R, Logical Memory I & II) and nonverbal (Rey Figure Recall) memory. There were no relations with measures of verbal fluency or naming. Conclusions: Similar to previous studies, we found that scores on MCI-E correlate highly with self-reports of mood and quality of life. However, we also found that scores on this instrument have some relation to indices of objective memory performance. These results indicate that the MCI-E holds promise as a valid measure of subjective memory impairment in patients with epilepsy.
1.350 ROUTE LEARNING AND TEMPORAL LOBE EPILEPSY Bell, 1 Jana Jones, 2 Christian Dow, 1 Russ Hansen, 1 Austin Woodard, 1 Raj Sheth, 1 Paul Rutecki, 2 Michael Seidenberg, and 1 Bruce Hermann (1 Neurology/Neuropsychology, University of Wisconsin, Madison, WI; and 2 Psychology, Finch University of Health Sciences/Chicago Medical School, North Chicago, IL)
1 Brian
Rationale: There is a well-established relation between verbal memory and left hippocampal integrity in patients with temporal lobe epilepsy (TLE). A reliable association between visual memory test performance and the right hippocampus has not been found. Barrash et al. (2000) created a real-life route learning test (RLT) and administered it to a large group of patients with different types and sites of focal brain lesions. Right hippocampal injury was among the lesions associated with impaired route learning in their investigation. In this study, we assessed TLE patients and controls on a similar RLT. Methods: The two groups did not differ significantly in gender ratio, age, education, or IQ. The participants were led by the examiner on a fixed route through unfamiliar hospital corridors at a leisurely pace for ∼4 min. The layout of the University of Wisconsin Hospital is topographically complex. The participants were asked to pay close attention to the route so that they subsequently could follow it independently while making as few errors as possible. They were informed that they would be timed, but that accuracy was more important than speed. After the demonstration, the participant led the way through the same route 3 times, or twice if performance on the first two trials was flawless. When an error occurred, the examiner immediately pointed out the correct turn. Number of errors and time to completion were recorded for each trial. Results: Across the three RLT trials, the patients committed significantly more errors and were slower to complete the route (p < 0.05). Within the TLE group, RLT errors correlated best with tests measuring visual-perception (Judgment of Line Orientation test and WAIS-III PIQ) and memory (WMS-III delayed visual memory and immediate and delayed auditory memory indexes). Speed on the RLT was associated with a wider range of cognitive measures. These results are based on data from 10 controls and 12 TLE patients. Additional data, including correlations between hippocampal volumes and RLT performance, will be reported for a larger sample of patients who will be returning for a longitudinal study. Conclusions: In this study, we tested TLE patients and controls on an RLT that involved traversing a relatively complex path through hospital hallways after one demonstration of the route by the examiner. There were significant group differences in total errors and mean time to completion. Among the patients, RLT errors correlated best with tests measuring visual-perceptual ability and auditory and visual memory. These results suggest that topographic learning impairment is among the memory deficits characteristic of TLE. The incidence of RLT impairment among the patients and the relation between RLT ability and
AES PROCEEDINGS left and right hippocampal volumes, age of onset of epilepsy, etc., will be reported. (Supported by NIH grants K23 NS42251, 2RO1-37738, and MO1 03186.)
1.351 SUBJECTIVE AND OBJECTIVE MEMORY AND ATTENTION IN PATIENTS WITH TEMPORAL LOBE EPILEPSY (TLE) Thomas Bengner, Meline Stoy, and Heinz Joachim Meencke (Neuropsychology Unit, Epilepsy Center Berlin-Brandenburg, Berlin, Germany) Rationale: To describe underlying factors of subjective and objective memory and attention. To investigate whether they can distinguish patients with a left or right temporal lobe epilepsy (TLE; L-TLE, R-TLE). To correlate subjective and objective measures of attention and memory. To look at the influence of demographic variables, personality traits, and psychic distress on subjective complaints and objective results. Methods: The 70 L-TLE (59%) and 49 R-TLE (41%) were examined preoperatively with tests on verbal intelligence, verbal and figural memory, and attention. Questionnaires on complaints about memory (severity and frequency) and attention deficits (anergia, mental slowing, practical slowing). Education, training, and occupation were recorded. In 21 L-TLE and 16 R-TLE, questionnaires on personality (Freiburger Personality Inventory, FPI) and psychic distress (Symptom Check List 90 R, SCL) were available. Results: A nonorthogonal analysis (eigenvalue >1.5) distinguished a verbal memory factor, a subjective complaint factor (comprising memory and attention), an educational, and a figural memory factor (explaining 52% of the variance). To distinguish L- and R-TLE, a discriminant analysis was applied to the highest loading variables of each of these factors (verbal delayed recall, anergia, graduation, delayed figural recognition). The combination of delayed verbal recall and subjective anergia led to the best differentiation of L-TLE and R-TLE (Wilks lambda, 0.92 and 0.93; p < 0.01). L-TLE patients performed worse in delayed recall for verbal material after 30 min than did R-TLE but had less anergia (p < 0.05). Nevertheless, only 65% of the cases were classified correctly (79% of LTE, 47% of RTE). A weak correlation was found between subjective and objective measures of memory (r = 0.20 to 0.35; p < 0.05). Correcting for age, employed patients were better than nonworking patients in subtests of verbal and figural memory and had less frequent memory deficits (n = 91; p < 0.05). In the subsample of 37 patients, anergia (and other cognitive complaints) correlated with the FPI scales contentment (r = 0.55; p < 0.001), ambition (r = 0.54; p < 0.001), and the number of symptoms (SCL, r = −0.35; p < 0.05), whereas delayed verbal recall (and other objective measures) did not show a correlation. Number of symptoms was higher in R-TLE (p < 0.05). Conclusions: Measures of subjective memory and attention both loaded highly on the same factor. Thus, we conclude that they measured complaints about cognitive deficits without distinguishing between memory and attention. Although statistically significant, the value of objective and subjective measures in distinguishing R-TLE and L-TLE is clinically insufficient. Especially R-TLE patients were classified on a chance level. Nevertheless, a subjective measure of anergia significantly contributed to the distinction. In accordance with former studies, we found an influence of daily demands and ambition, life contentment, and the number of psychiatric symptoms on the subjective feeling of cognitive deficits.
1.352 EEG PATTERNS IN MILD COGNITIVE IMPAIRMENT Rabia B. Choudry, Sigmund Jenssen, and Carol F. Lippa (Neurology, Drexel University College of Medicine, Philadelphia, PA) Rationale: Mild cognitive impairment (MCI) is defined as patients with memory deficits who do not fulfill the criteria, but may develop Alzheimer disease. Whereas Mini Mental Status Examination, neuropsychology, and neuroimaging are widely used, the role of EEG has not been established in the diagnosis of MCI or its progression to dementia. Our objective is to define EEG patterns consistent with MCI. Methods: We reviewed 83 charts of patients with MCI. We compared the MMSE scores to the presence or absence of slowing on EEG by using simple logistic regression.
123
Results: Eighteen of the 83 patients (21.7%) had EEGs. Mean age was 59 years (34–78). Median MMSE score was 28 (25–30). Six EEGs (33%) showed focal slowing, one with sharp waves. Three (16.7%) had only diffuse slowing. Two (11%) had increased beta. Nine (50%) were normal. No correlation was found between MMSE scores and the presence or absence of slowing on EEG. Conclusions: Of EEGs, 50% showed diffuse or focal slowing. There is a risk of overrepresentation of abnormal studies because only 21.7% of patients had EEGs. Larger studies are needed to better define the EEG characteristics of patients with MCI to assist in the early diagnosis of dementia. 1.353 MEMORY FUNCTIONS IN PATIENTS WITH NEOCORTICAL TEMPORAL LOBE EPILEPSY Nicole Maineri, Julia Lima Espirito Santo, Ana Paula Alves Pereira, and Mirna Wetters Portuguez (Neurology Service-Neuropsychology Unit of Epilepsy Surgery Program, Hospital S˜ao Lucas da PUC, Porto Alegre, Rio Grande do Sul, Brazil) Rationale: Patients with medically refractory epilepsy have surgery as their last resource to control seizures. Presurgical evaluation and the surgical treatment of neocortical epilepsy is one of the most challenging areas in epilepsy surgery (Hong, 2002). In some cases, these patients have nonlesional hippocampi revealed by magnetic resonance imaging (MRI), and most of them have memory problems. Our aim is to investigate if nonlesional hippocampi and memory impairment are associated. Methods: In the study, 20 patients were included with neocortical temporal lobe epilepsy (NTLE) and without detectable hippocampal lesions on MRI who underwent a complete investigation, including neuropsychological assessment, to detect any possible association between memory deficits/impairment and age at seizure onset, frequency, gender, and location of epileptogenic hemisphere. Descriptive analysis and Fisher’s exact tests were used as well. Results: Significant associations within memory deficit and patients with NTLE (p = 0.00006) and nonlesional hippocampi were found. Also, memory-impairment type (verbal or visual) when separately compared, showed no significant association. Higher memory impairment (85%) predisposition was shown when three tests were statistically combined. Memory impairment and age at seizure onset reported expressive association. The patients with age at seizure onset younger than 15 years had a higher predisposition to demonstrate memory impairment (p = 0.031). Conclusions: The results suggests that NTLE patients with normal MRI showed higher predisposition to have memory deficit associated with age at seizure onset younger than 15 years. Therefore, seizure’s early appearance in these patients seems to be an important fact in memory impairment establishment. [Supported by Neurology Service– Neuropsychology Unit of Epilepsy Surgery Program, Hospital S˜ao Lucas da Pontif´ıcia Universidade do Rio Grande do Sul (PUCRS).] 1.354 NAMING DEFICITS IN TEMPORAL LOBE EPILEPSY: COGNITIVE AND NEURAL RELATIONS 1 Anna R. Giovagnoli, 1 Elena Colombo, 2 Alessandra Erbetta, and 3 Giuliano Avanzini (1 Neurology and Neuropathology, 2 Neuroradiology, and 3 Neurophysiology, Carlo Besta National Neurological Institute, Milano, Lombardy, Italy) Rationale: Naming disturbances may be caused by lesions in different dominant hemisphere regions (e.g., the temporal pole and the lateral temporal cortex). In left temporal lobe epilepsy (TLE), naming deficits have been often interpreted as lexical–semantic disturbances and have been usually associated with hippocampal sclerosis. On this basis, questions have been raised whether naming impairment consistently reflects semantic memory impairment and may be caused by temporal lobe lesions of different nature and location. Methods: Eighty-two TLE patients with lateral (12 left, nine right) or mesial temporal lesions (30 left, 31 right) were compared with 35 healthy controls. Hippocampal sclerosis was determined in 16 left and 18 right TLE patients; low-grade gliomas, cavernous angiomas, and neural migration disorders were documented in the other patients. Picture naming (PN) required the naming of line-drawings of 40 living and 40
Epilepsia, Vol. 44, Suppl. 9, 2003
124
AES PROCEEDINGS
nonliving things. Verbal and visual Pyramid and palm trees test (PPTT) assessed material-specific semantic memory. Results: Separate Mann–Whitney tests comparing left TLE patients and controls revealed significant deficits in the cortical lateral (PN: U = 95; p = 0.0047; verbal PPTT: U = 129.5; p = 0.015), and mesial lesion subgroups (PN: U = 162; p = 0.0003; verbal PPTT: U = 372; p = 0.018; visual PPTT: U = 325.5; p = 0.0059). Subsequent comparisons showed that left hippocampal sclerosis patients were impaired in all three tests (PN: U = 162; p = 0.0038; verbal PPTT: U = 225; p = 0.040; visual PPTT: U = 190; p = 0.012), whereas patients with other hippocampal lesions were only impaired in PPTT (verbal: U = 126; p = 0.031; visual: U = 111; p = 0.020). Compared with controls, right TLE patients with cortical lateral lesions were impaired in visual PPTT (U = 78.5; p = 0.013), whereas patients with mesial lesions did not show any deficits. In left TLE patients, PN scores were significantly related with PPTT scores. Conclusions: The data suggest that left TLE associated with either lateral or mesial temporal lobe lesions may significantly affect naming and that naming deficits reflect semantic memory impairment. The association of naming deficits and material-specific semantic deficits with lateral temporal lobe lesions is in accord with the role played by the temporal cortex in semantic storage and retrieval. By contrast, the impairment of both verbal and visual semantic memory in patients with left mesial temporal lobe lesions suggests implication of particular left hippocampal functions (e.g., the processing of semantically meaningful verbal or visual stimuli). The significant naming impairment in patients with left hippocampal sclerosis (in contrast with normal performances in patients with other hippocampal lesions) suggests that this type of lesion plays a specific causal role: it could interfere with the functions of distant brain areas through hippocampal–cortical pathways involved in lexical retrieval, in line with its effects on other cortical cognitive functions.
1.355 ANALYSIS OF FIGURAL MEMORY PERFORMANCE IN PERSONS WITH EPILEPSY 1,2 Deborah S. Hoffnung,2 Wm. Drew Gouvier, and 3,4 Kenneth R. Perrine (1 Comprehensive Epilepsy Center, Long Island Jewish Medical Center, New Hyde Park, NY; 2 Department of Psychology, Louisiana State University, Baton Rouge, LA; 3 Department of Psychology, Queens College/City University of New York, Flushing, and 4 Department of Neurological Surgery, Cornell-Weill Medical School, New York, NY) Rationale: Despite the widespread use of the Rey–Osterrieth Complex Figure (ROCF) and similar figural reproduction tests as measures of nonverbal memory, conflicting findings have been reported among persons with epilepsy. To date, the ROCF has shown only limited value in the identification of nonverbal memory impairment in presurgical patients with right temporal lobe epileptic (TLE) foci. In this study, the Boston Qualitative Scoring System (BQSS) was used to examine whether qualitative features of ROCF performance could discriminate between those with right and left TLE. It was hoped that using this new system would improve lateralization of seizure focus in a clinical sample. Methods: This study included patients undergoing preoperative evaluation for anterior temporal lobectomy at the Comprehensive Epilepsy Center at Long Island Jewish Medical Center. Individuals were classified as to side of primary seizure focus based on a combination of procedures including routine EEG, ictal and interictal continuous videoEEG recording, neuroradiologic evaluation, and, when relevant, eventual surgical side. Individuals with complex partial seizures of temporal origin and idiopathic etiology were included. Concurrent with inpatient video/EEG monitoring or during a scheduled outpatient evaluation, participants were administered the Copy, Immediate, and 30-min Delayed Recall trials of the ROCF as well as additional nonverbal memory and visuoperceptual tasks. The ROCF productions were scored using a traditional quantitative method as well as the BQSS scoring technique. Results: Forty-five individuals were included in this study; 25 individuals with left temporal lobe epilepsy (L TLE) and 20 individuals with right temporal lobe epilepsy (R TLE). As predicted, seizure groups did not differ on a standard quantitative scoring system for the ROCF. Contrary to prediction, the right TLE group did not perform more poorly on BQSS measures of quality or organization, and they did not have greater difficulty recalling the figure after a delay. There was a trend toward
Epilepsia, Vol. 44, Suppl. 9, 2003
poorer performance by the right TLE group on two BQSS scales, those quantifying the presence or absence of elements of the figure. ROCF performance was more strongly correlated with measures of visuoperception than with additional measures of nonverbal memory. Conclusions: The ROCF does not seem a promising task for the measurement of nonverbal memory, at least in an epilepsy sample, given its lack of association with other measures of nonverbal memory and its repeated failure reliably to differentiate between R and L TLE groups. However, figural reproduction tasks do appear to be useful for the evaluation of visuoperception in this population, and a scoring system for the ROCF that evaluates aspects of visual analysis may prove valuable in differentiating right from left temporal lobe dysfunction. (Supported by the Epilepsy Foundation through the generous support of the American Epilepsy Society.) 1.356 VERBAL MEMORY OUTCOME AFTER LEFT ATL IN PATIENTS WITH NORMAL PREOPERATIVE VERBAL MEMORY AND LEFT MESIAL TEMPORAL LOBE SCLEROSIS 1 Roy Martin, 2 Anthony Lagalbo, 1 Stephen Sawrie, 1 Jorge Burneo, 1 Robert Knowlton, 1 Melissa Mendez, 1 Avinash Prasad, 1 Edward Faught, and 1 Ruben Kuzniecky (1 Neurology and 2 Psychology, University of Alabama at Birmingham, Birmingham, AL) Rationale: Previous studies have shown that normal preoperative verbal memory (VM) is a risk factor for VM decline after left anterior temporal lobectomy (ATL). This pattern of outcome is typically associated with resection to normal or minimally atrophic left mesial temporal lobe structures. However, the presence of a more severe degree of left mesial temporal sclerosis (MTS) is thought to reduce the risk to postoperative VM. To date, limited information is available regarding relative risk to VM in patients with both normal preoperative VM and left MTS. This study investigated VM change after left ATL in patients who had preoperative MRI-verified evidence of exclusive left MTS and normal baseline VM. Methods: Two groups of left MTS patients were identified: those exhibiting normal preoperative VM (NVM; n = 17) and those with impaired preoperative verbal memory (ImVM; n = 29). Patients were classified as having normal memory if performances on both Acquisition (learning across trials 1–5) and Retrieval (long-delayed free recall) portions of the California Verbal Learning Test were above a T-score of 40 (>16%ile). Verbal memory outcome was established by incorporating standardized regression-based change scores (SRB). Change scores are expressed in z-score units. Results: Patients were similar across age, gender, education, FullScale IQ, age at seizure onset, and seizure duration (years). All patients had FSIQ >70 and were right-handed. Wada testing was administered to 19 patients (10 NVM, nine ImVM patients). All patients were left speech/language dominant. Postoperatively, groups demonstrated similar levels of decline on the CVLT Acquisition variable (p > 0.05; SRB z-score change: NVM = −1.05, ImVM = −1.02). However, a statistical trend was found (p < 0.09) in which the NVM group displayed poorer outcome on the CVLT Retrieval outcome variable (SRB z-score change: NVM = −2.50, ImVM = −1.55). Postoperative seizure outcome did not correlate with either CVLT Acquisition or Retrieval SRB change scores. Ipsilateral Wada recognition memory score correlated significantly with the CVLT Retrieval SRB change score (r2 = −0.48; p < 0.04), but not for the contralateral score (r2 = 0.01, NS). Conclusions: These results suggest that the presence of preoperative MRI-verified left MTS does not exclude the risk for postoperative VM decline in patients with normal preoperative VM function. Our findings indicate that the structural integrity of the mesial temporal lobe is not an exclusive factor in determing risk to VM decline after left ATL. Our findings lend support to the notion of functional adequacy of the ipsilateral mesial temporal lobe in explaining verbal memory outcome.
1.357 COGNITIVE ANALYSIS IN SPANISH-SPEAKING TEMPORAL LOBE EPILEPSY PATIENTS 1,2,3,4 Silvia A. Oddo, 1,2,4 Patricia L. Solis, 1,2,4 Damian Consalvo, 1,2,4 Brenda Giagante, 1,2,3,4 Walter H. Silva, 1,2,4 Luciana D’Alessio,
AES PROCEEDINGS 1,2,4 Estela Centurion, 1,2,4 Patricia Saidon, and 1,2,3,4 Silvia Kochen [1 Epilepsy Center, Neurology Division., Ramos Mejia Hospital, 2 Celullar Biology and Neuroscience Institute “Prof Dr. Eduardo De
Robertis” Buenos Aires University (UBA), 3 CONICET, Health Ministry, and 4 CEFYBO, Buenos Aires University, Capital Federal, Buenos Aires, Argentina] Rationale: The aim of this study was to analyze the basal cognitive status in temporal lobe epilepsy patients with hippocampal sclerosis (HE), to determine their cognitive profile and to correlate memory deficits with MRI lesion. Methods: We selected 71 pacientes (p.) with temporal lobe epilepsy and HE. All the patients were evaluated with a Neuropsychological Protocol that had been designed specifically by our group. This protocol includes the assessment of intelligence, attention, handedness, verbal memory, visual memory, language, and executive function. Test results were standardized for their analysis. We used a χ 2 test, a Pearson and Spearman correlation test, and a multiple regression logistic model. Results: From the total population evaluated, 66% (47 p.) have memory deficits. Patients were divided in four groups for their analysis: Patients without memory deficits (n = 25), verbal memory deficit patients (n = 21), visual memory deficit patients (n = 17), and both memorydeficit patients (n = 8). We found a correlation of 70% between MRI and material-specific memory deficits. In a 46% (33 p.) we found language deficits, especially anomias. Executive-function deficits were observed in 25% (18 p.) of the evaluated patients. Seventy-five percent of the p. were refractory to medical treatment, and they were surgery candidates. Conclusions: Of the studied patients, the neuropsychological profile was characterized by material-specific (verbal/visual) memory deficits and, less frequently, language and executive-function deficits. Correlation with material-specific memory deficits and HE was observed in a high percentaje of the patients. We also will discuss some aspects of cerebral plasticity and memory reorganization related especially to language. Neuropsychological evaluation contributes important information on the epileptogenic zone localization in temporal lobe epilepsy patients. 1.358 NEUROPSYCHOLOGICAL MORBIDITY IN CHRONIC TEMPORAL LOBE EPILEPSY 1 Temitayo Oyegbile, 2 Christian Dow, 2 Michael Seidenberg, 1 Brian Bell, 1 Jana Jones, 1 Austin Woodard, 1 Paul Rutecki, and 1 Bruce Hermann (1 Department of Neurology, Unversity of Wisconsin, Madison, WI; and 2 Department of Psychology, Chicago Medical School, North Chicago, IL) Rationale: The purpose of this cross-sectional investigation was to (a) characterize the nature and extent of cognitive morbidity in chronic temporal lobe epilepsy (TLE), (b) determine whether cognitive morbidity progresses in association with increasing years of epilepsy, and (c) identify factors that may accelerate or slow progressive cognitive effects. Methods: Healthy controls (n = 85) and patients with TLE (n = 96), ranging in age from 14 to 60 years, underwent comprehensive neuropsychological assessment of intelligence, language, perception, verbal and nonverbal memory, psychomotor processing, and speeded fine-motor dexterity. Twenty cognitive indices were derived for each subject from this test battery. Test scores were converted to adjusted (age, gender, education) z-scores, and an impairment index computed for each subject, defined as the proportion of test scores exceeding z = −1.5. Results: Three sets of findings were evident: (a) Chronic TLE was associated with widespread cognitive morbidity. At the group level, TLE patients exhibited significantly poorer performance across all cognitive domains compared with controls, with several cognitive abilities (motor speed, mental flexibility, naming) at least as impaired as memory function. At the individual subject level, an average of 37% of test scores were abnormal per TLE patient compared with 7.8% for controls (p < 0.001). (b) Evidence of cognitive progression was obtained. Increasing duration of epilepsy was significantly associated with a larger proportion of abnormal test scores per patient (r = 0.40; p < 0.001). This relation remained significant after controlling for several factors (e.g., polytherapy, seizure frequency, and other clinical seizure features). (c) Some factors that appear to attenuate the adverse effects of chronic epilepsy on cognition. Among patients with greater cerebral reserve (defined as 12+ years
125
of education), the adverse effects of increasing duration of epilepsy on cognition were significantly attenuated (r = 0.19, NS) compared with those observed among patients with less cerebral reserve (r = 0.53, p < 0.001). Conclusions: These findings provide a greater understanding of the nature and extent of cognitive morbidity associated with chronic TLE. Cognitive morbidity can be widespread, affect several cognitive domains at least as severely as memory, with an average of 37% of test scores abnormal per individual TLE patient. Cognitive morbidity increases with the chronicity of epilepsy and, consistent with human and animal findings, factors were identified that appear to attenuate the degree of associated cognitive morbidity. (Supported by NIH 2RO1-37738 and MO1 03186.) 1.359 MEMORY LATERALIZATION AFTER ANTERIOR TEMPORAL LOBECTOMY AS A FUNCTION OF VERBALIZABILITY AND FAMILIARITY Alana E. Salvucci, Christina J. Chiew, and Guila Glosser (Department of Neurology, University of Pennsylvania, Philadelphia, PA) Rationale: Memory dysfunction is a known consequence of anterior temporal lobectomy (ATL). The memory loss pattern has been reported to differ between patients with surgery in the language dominant versus the non–language dominant hemisphere, though the factors underlying lateralized memory differences continue to be debated. A left hemisphere deficit for memory for familiar verbal material (such as real words) has been well demonstrated, and though less consistent, a right hemisphere deficit for memory for unfamiliar nonverbal material (such as unfamiliar faces) has also been shown. It is unclear from these results whether apparent lateralization is a function of the verbalizability of the to-belearned information or whether it is due to exposure or familiarity of the information before learning. Methods: To assess the effects of verbalizability and familiarity on memory, a completely crossed design used four types of materials; familiar verbal stimuli consisted of real abstract English nouns, unfamiliar verbal stimuli consisted of pronounceable pseudowords (PWs), familiar nonverbal stimuli consisted of photographs of familiar but unnamable people taken from the popular media, and unfamiliar nonverbal stimuli consisted of photographs of unknown people. LATL and RATL patients who were left hemisphere language dominant and healthy controls completed four paired-associate learning followed by yes-no recognition memory tests. Results: Significant lateralized effects of verbalizability were found on memory accuracy discrimination. LATL patients were impaired for both real words and PWs compared with other groups. RATL and LATL groups were equally impaired compared with controls for familiar and unfamiliar faces. There were no hemispheric differences in familiarity of the to-be-learned information in memory discrimination, nor were there effects of familiarity for discrimination of verbal and nonverbal materials. With respect to response bias, or the strategy that was adopted in conditions of response uncertainty, RATL patients and controls showed more liberal response bias for unfamiliar compared with familiar information, whereas LATL patients’ response bias was more liberal for familiar information, especially real words. Conclusions: These results support a partial material-specific account of memory lateralization. There was clear left hemisphere superiority for verbal discrimination (words and PWs) and involvement of both left and right hemispheres in nonverbal face memory processing. Familiarity did not play a role in lateralization of memory accuracy. Response bias, on the other hand, was related to prior familiarity with to-be-learned information and was selectively affected by LATL lesion. Thus, three factors were identified as underlying lateralization of memory functions in temporal lobectomy patients; verbalizability of the to-be-learned material, prior familiarity with the material, and response strategies. 1.360 RECENT VERSUS REMOTE AUTOBIOGRAPHIC MEMORY DURING THE WADA TEST: SUPPORT FOR A CONSOLIDATION MODEL OF MEMORY 1,2 Tara Tamny-Young,1 Marla J. Hamberger, and 2,3 Jeffrey J. Rosen (1 Department of Neurology, Columbia Presbyterian, 2 The Graduate
Epilepsia, Vol. 44, Suppl. 9, 2003
126
AES PROCEEDINGS
Center, The City University of New York, and 3 The City College, The City University of New York, New York, NY) Rationale: Some investigators assert that medial temporal structures initially play a critical part in memory, but that a subsequent consolidation process results in the formation of remote cortical memories that no longer rely on this region (Scovile & Milner, 1957; Squire & Alvarez, 1995). However, others have proposed a multiple trace model of memory whereby medial temporal structures remain critical in the retrieval of remote memories indefinitely (Nadel & Moscovitch, 1997). To test these theories, we examined recent and remote autobiographic memory during unilateral hemispheric anesthesia after intracarotid amobarbital injections. We reasoned that greater difficulty with recent than with remote recall after injections contralateral to known medial temporal damage (i.e., producing a temporary state of bilateral medial temporal dysfunction) would support a consolidation theory of memory. In contrast, comparable difficulty recalling recent and remote events would support a multiple trace theory. After injections ipsilateral to known medial temporal damage, patients were expected to recall both recent and remote autobiographic events. Methods: Sixteen temporal lobe epilepsy patients (11 left, five right) with unilateral medial temporal damage underwent intracarotid amobarbital testing both ipsilateral and contralateral to the hemisphere of suspected seizure onset. Two types of temporally distinct autobiographic memory questions were used: “recent” items, which assessed memory for events that occurred within the last 5 years (e.g., What memorable event occurred during your last vacation?), and “remote” items, which assessed memory for events that occurred >5 years ago (e.g., What memorable event occurred during grammar school?). Recent and remote autobiographic memory item scores for both ipsilateral and contralateral injections were compared via paired-samples t tests. Results: There was no significant difference between recent and remote autobiographic recall after ipsilateral injections. As expected, after injections contralateral to medial temporal damage, recall of recent events was poorer than memory for remote events (t = 2.33, p < 0.04). Conclusions: These results support the consolidation theory of memory suggesting that medial temporal structures play only a temporary role in autobiographic memory retrieval, whereas over time, memory retrieval is supported by structures outside the medial temporal region. (Supported by the Epilepsy Foundation through the generous support of the American Epilepsy Society.) 1.361 LONGITUDINAL NEUROPSYCHOLOGICAL CHANGES IN PATIENTS WITH INTRACTABLE TEMPORAL LOBE EPILEPSY 1 James R. White, 1 Thomas E. Beniak, 4 Debra J. Matchinsky, 1 David B. Lund, 2 Christine M. Cox, 1 Thaddeus S. Walczak, 1 Teresa A. Tran, 1 Jeanne L. Beattie, 1,3 Ilo E. Leppik, 3 John O. Rarick, and 1 Robert J. Gumnit (1 Neurology, MINCEP Epilepsy Care, 2 Surgery, Neurosurgical Associates, Ltd., and 3 Experimental and Clinical Pharmacology, University of Minnesota, Minneapolis, and 4 Psychology, North Hennepin Community College, Brooklyn Park, MN) Rationale: Material-specific memory deficits are commonly identified in patients with intractable temporal lobe epilepsy (TLE), but the extent and time course of cognitive deterioration over time has not been well established. We studied cognitive changes in patients with intractable TLE with serial neuropsychological testing available before anterior temporal lobectomy (ATL). The objectives of this study were to (a) describe the proportion of patients who experience >1 stdev decline in memory, and (b) to identify independent variables associated with memory decline in patients with intractable TLE before surgery. Methods: Forty consecutive patients (20 left TLE, 20 right TLE) were identified who met the following inclusion criteria: (a) age >16 years at the time of the initial test; (b) two neuropsychological tests >1 year apart available before ATL (test 1 and test 2); and (c) all patients underwent ATL after test 2. Data from test 1 and 2 were analyzed. Average length of time between test 1 and 2 was 4.4 years (range, 1.1–12.5 years). Informed consent for the use of clinical information was obtained from each patient. Results: Fifteen of 40 patients (38%) experienced >1 stdev decline on at least one memory test (eight left TLE, seven right TLE). Multiple
Epilepsia, Vol. 44, Suppl. 9, 2003
regression analyses identified several significant models. LEFT TLE PATIENTS: Higher frequency of complex partial seizures (CPSs) (p = 0.003), higher initial test score (p < 0.0005), and older age at initial test (p < 0.0005) were associated with greater decline in verbal memory. Higher frequency of CPSs (p = 0.039), and higher initial test score (p = 0.023) were associated with greater decline in nonverbal memory. RIGHT TLE PATIENTS: Higher initial test score (p < 0.0005), older age of onset of epilepsy (p = 0.024), and greater anticonvulsant toxicity (p = 0.015) were associated with greater decline in verbal memory. Higher initial test score (p = 0.047) was associated with greater decline in nonverbal memory. These models explained between 25.3 and 88.0% of the variance. Conclusions: This study demonstrated that more than one third of patients with intractable TLE experienced significant memory decline over a few years. The independent variables associated with memory decline included higher frequency of seizures, higher initial test score, and older age at initial test. Our study differs from previous longitudinal studies by demonstrating a highly significant correlation between frequent CPSs and memory deterioration in patients with TLE. Early and effective control of seizures in patients with TLE is likely to be important in preventing cognitive deterioration. (Supported by MINCEP Epilepsy Care.) 1.362 REMEMBERING STORIES VERSUS WORDS BEFORE AND AFTER SURGERY: A SPECIFIC PATTERN IN RIGHT TEMPORAL PATIENTS? G. Wicky, V. Sziklas, S. Tugulea, and M. Jones-Gotman (Neurology & Neurosurgery, McGill University, Montreal, Quebec, Canada) Rationale: Many studies have shown that memory for words is sensitive to resection from the left temporal lobe. The Abstract Word List learning test (AWL) and the Rey Auditory Verbal Learning Test (RAVLT) are currently used in the Montreal Neurological Institute as measures of verbal learning and memory. In addition, learning and memory for the content of a short prose passage are evaluated with a Story Learning test. These measures allow us to better differentiate patient groups. However, we have noted that some patients showed an unexpected pattern: they demonstrated a good retention for abstract and concrete words but a deficit in memory for the story. The aim of the present study was to determine which group(s) of patients exhibits this particular memory deficit. A further aim was to determine the specific validity of these three tests in patients with epilepsy before and after surgery. Methods: We studied patients who were admitted in the last 10 years in our institute with either unilateral or bilateral epilepsy of either the temporal or frontal lobe. All were administered the AWL, RAVLT, and Story Learning tests before or/and after surgery. Selection criteria were: age between 16 and 56 years, IQ of ≥75, right-handed, with left hemisphere speech dominance. Variables of interest were the side of epileptic focus (EEG) and lesion (MRI), the site of eventual resection, and preversus follow-up performance. The RAVLT measure was the delayedrecall score 30 min after five learning trials. The AWL measure was the delayed recall score 24 h after four learning trials. The Story Learning retention measure was the 24-h delayed recall of a short prose passage that had been learned to a strict criterion. Results: Data analyzed to date show that left temporal (LT) patients perform significantly worse than right temporal (RT) patients on recall of concrete and abstract words in preoperative and follow-up assessment. Moreover, before and after surgery, the AWL is clearly the more sensitive and specific measure. Patients with frontal-lobe focus showed no retention deficits at either time, except a few left frontal cases with mild difficulty retaining abstract words. Interestingly, in this sample, we found no significant difference between LT and RT patients on story recall preoperatively but a clear difference after surgery. Further, a cluster analysis showed that the unexpected pattern (good recall for abstract and concrete words but deficit in story recall) appeared only in some patients with clear RT focus. The reason for this surprising result will be explored. Conclusions: We have uncovered a specific deficit in certain RT patients of forgetting a story (while showing the expected good memory for word lists). We also demonstrate that the AWL is a more sensitive measure than RAVLT in LT epilepsy. Our results underline the importance of examining different aspects of verbal learning and memory with specific
AES PROCEEDINGS tools to increase the discriminative power of memory tests. (Supported by Canadian Institutes of Health Research.)
1.363 COGNITIVE FUNCTION IN NEWLY DIAGNOSED PATIENTS WITH ADULT-ONSET FOCAL EPILEPSY 1,2 Reetta Kalviainen, 1 Sari Piiroinen, 1 Tuuli Salmenpera, and 1 Marja Aikia (1 Department of Neurology, Kuopio University Hospital, and 2 Department of Neurology, University of Kuopio, Kuopio, Finland) Rationale: Cognitive dysfunction in chronic drug-resistant focal epilepsy has long been recognized. The purpose of the present study was to evaluate neuropsychological performance of the patients at the time of the diagnosis of epilepsy. Methods: The cognitive performance of 199 patients with newly diagnosed focal epilepsy was studied before starting the antiepileptic drug (AED) treatment. The patients are consecutive patients from our hospital district of 250,000 inhabitants, fulfilling the criteria of age between 15 and 65 years and of full intelligence quotient (IQ) ≥85. Patients with progressive diseases as well substance abuse were excluded. Altogether 77 patients with only a single epileptic seizure and a group of 47 healthy volunteers served as control groups. The patients and controls were tested with comprehensive neuropsychological battery. Differences between three groups were analyzed by one-way analysis of variance with Scheffe’s post hoc test; p < 0.05 was considered statistically significant. Results: The mean age and education between groups did not differ. Although subjects with IQ 1 year. Methods: Our series was made up by 16 patients [50% male, aged from 18 to 54 years (mean, 36 years)] submitted to temporal lobectomy (62.5% left) and seizure free for 1–7 years (mean, 2.06 years). The Social Adjustment Scale-Self-Report (SAS) was applied on these patients. This scale consists of 54 questions that evaluate performance, quality of personal interrelationship, and personal satisfaction. It comprises seven specifics areas employment status; social and recreational activities; relationship with extended family members; role as a family member; marital role; parental role; and economic independence. We also evaluated patients for learning and memory skills, using verbal and visual stimuli (Rey Auditory Verbal Learning Test–RAVLT and Rey Visual Design Learning Test–RVDLT), and mental flexibility (Wisconsin Card Sorting Test–WCST). All results were compared with normative data. Results: Patients had low scores on all SAS subfactors (p < 0.001). The subfactor employment status showed significant correlation with delayed recall of verbal and visual stimuli (p = 0.03) and with the number of categories achieved on WCST. Results were not influenced by other variables such as age, gender, and foci lateralization as well as by seizure control, because all patients were seizure free. Conclusions: Our findings suggest that the level of social adjustment remains impaired even when total seizure control, the main goal of epilepsy surgery, is achieved. However, we noted that our patients with higher scores for memory and mental flexibility presented better employment status. We believe our findings emphasize the need and importance of a neuropsychological rehabilitation program, as part of epilepsy surgery programs, and as a complementary postsurgical treatment aiming to provide improvement of cognitive and social skills.
1.368 EFFECTS OF MOOD ON QUALITY OF LIFE IN TEMPORAL LOBE EPILEPSY 1 Erica K. Johnson, 1 Jana E. Jones, 1 Brian Bell, 2 Christian Dow, 1 Austin Woodard, 1 Paul Rutecki, 2 Michael Seidenberg, and 1 Bruce P. Hermann (1 Neurology/Neuropsychology, University of Wisconsin, Madison, WI; and 2 Psychology, Finch University of Health Sciences/Chicago Medical School, North Chicago, IL) Rationale: Interictal depression is known to be a common comorbid psychiatric disorder in chronic epilepsy, including temporal lobe epilepsy (TLE). Recent research has suggested that in addition to traditional clinical variables such as seizure frequency, comorbid depressive symptoms may be a major contributor to poorer health-related quality of life (HRQOL) in epilepsy. This investigation attempted to (a) replicate the reported relation between interictal depressive symptoms and reduced HRQOL, (b) characterize the relative explanatory power of interictal depressive symptoms versus other pertinent clinical epilepsy (e.g., seizure severity, seizure frequency, duration of epilepsy, monotherapy vs. polytherapy) and sociodemographic factors (gender) known to influence HRQOL in epilepsy; and (c) determine whether interictal depressive symptoms exert comparable effects on other measures of psychosocial status such as life satisfaction. Methods: Eighty-eight subjects with TLE completed measures of HRQOL (Quality of Life in Epilepsy-89 [QOLIE]) and subjective handicap (Subjective Handicap in Epilepsy Scale [SHE]) from which dependent measures of overall quality of life (QOLIE-89) and life satisfaction (SHE) were derived. Predictor variables included severity of depressive symptoms (Symptom Checklist-90-Revised), seizure severity during the past 6 months (Liverpool Seizure Severity Scale), duration of TLE (years), overall seizure frequency (daily, weekly, monthly, yearly), monotherapy vs. polytherapy, and gender. Stepwise regression analysis was performed to control for collinearity. Results: The stepwise regression analysis for HRQOL (QOLIE-89) was significant (R2 = .59; p < 0.001). Poorer HRQOL was found to be associated with increased depression (p < 0.0001), severity of seizures (p < 0.002), and increasing duration of TLE (p < 0.002). The stepwise regression analysis for life satisfaction as measured by the SHE was significant (R2 = 0.37; p < 0.01). Poorer life satisfaction was found to be associated with increased depression (p < 0.0001). No other variables were significant.
AES PROCEEDINGS Conclusions: Interictal depressive symptoms, a common comorbidity in chronic epilepsy, are a significant predictor of HRQOL and life satisfaction among patients with TLE, replicating and extending previous findings (Gilliam, 2002). In addition, chronicity of epilepsy and seizure severity, but not seizure frequency, predict HRQOL. In addition to highlighting the adverse effects of chronic and self-reported severity of seizures, these findings reinforce the importance of identifying and treating interictal depression and depressive symptoms in epilepsy. (Supported by NIH NS 2RO1-37738 and MO1 RR03186.) 1.369 SUSTAINED POSITIVE MOOD AND QUALITY-OF-LIFE EFFECTS ASSOCIATED WITH THE USE OF LAMOTRIGINE 1 Alexandra E. McBride, 2 Faith M. Gunning-Dixon, 2 Roseanne A. Pinti, 1 Eileen M. Licari, 1 Alan B. Ettinger, 2 Ken R. Perrine, and 1 Nancy Matarese (1 Neurology, Long Island Jewish Medical Center, and 2 Psychiatry, Long Island Jewish Medical Center, New Hyde Park, NY) Rationale: Most studies of antiepileptic medications are conducted under somewhat artificial experimental conditions. This study assesses seizure efficacy and the mood and quality-of-life effects of lamotrigine (LTG) in a real-world clinical setting. Methods: This is an ongoing prospective observational study. We recruited adult patients treated at the outpatient centers of the Long Island Jewish Comprehensive Epilepsy Center who were identified as appropriate candidates for LTG treatment by their epileptologists. Patients completed the Quality of Life in Epilepsy-31 inventory (QOLIE-31), the Profile of Mood States survey (POMS), and a seizure severity scale at baseline, 2 months, 6 months, and 1 year after beginning LTG, with the dosage titrated according to concomitant antiepileptic medications. Results: Of 60 patients enrolled, nine have completed the baseline evaluation and 1-year assessment after LTG initiation. Five patients had >50% improvement in seizure control. Four patients had no change in seizure frequency. Data from completion at 6 months reported earlier found patients demonstrated significant improvement (p < 0.05) in the QOLIE-31 scores, and significant improvement on subtests assessing emotional well-being, increased energy level, and increased social interaction. On the POMS, subsets showing significant improvement included decreased anger, increased vigor, and decreased fatigue. In the nine patients at 1 year, who completed the QOLIE-31, there was significant improvement (p < 0.013) in the QOLIE-31 Overall Score. There was also significant improvement on subtests assessing emotional well-being, increased energy level, improved cognitive function, and decreased worry about seizures. In the eight patients completing the POMS, subsets demonstrating significant improvement included decreased anger, increased vigor, decreased fatigue, and decreased depression. From the 6-month to the 1-year interval, a trend toward significance (p < 0.06) was found in POMS subsets on increased vigor and decreased anger, and significant change was noted in decreased anxiety. Conclusions: Data from this ongoing study indicate positive mood and quality-of-life effects that are sustained after the introduction of LTG. At 1 year, patients had improvement in more QOLIE-31 subsets compared with 6 months, and POMS subsets found more significant improvement at 1 year compared with 6 months, suggesting not only maintenance of these positive effects, but improvement over time. (Supported by GlaxoSmithKline.) 1.370 QUALITY OF LIFE IN INTRACTABLE EPILEPSY: DOES AGE MATTER? 1 Jason M. Meckler, 1 Jerzy P. Szaflarski, 2 Magdalena Szaflarski, 1 Cynthia Hughes, 1 David M. Ficker, 1 William T. Cahill, 1 Jennifer Cavitt, and 1 Michael D. Privitera (1 Neurology and 2 Institute for Health Policy and Health Services Research, University of Cincinnati, Cincinnati, OH) Rationale: Health-related quality of life (HRQOL) in epilepsy is related to the severity of the disease, medication side effects, comorbidities, seizure freedom, and response to treatment. But the literature provides no clear statement about the role of age. In this study, we examine agerelated effects (patient’s age, age at seizure onset, and disease duration) on HRQOL in patients with epilepsy.
129
Methods: The 259 patients admitted to the Epilepsy Monitoring Unit of the University Hospital in Cincinnati between January 20, 2001, and April 28, 2003, completed epilepsy-specific quality-of-life instrument (QOLIE-89). All patients with monitoring-confirmed diagnosis of epilepsy were included in the analysis (n = 100; 64% female); patients with other diagnoses were excluded. Bivariate correlations (Pearson’s r for normally distributed and Spearman’s rho for not normally distributed variables) and simple linear regression were used to estimate the relations between age, epilepsy onset, and epilepsy duration and HRQOL. Results: There was no correlation (r = −0.01; p = 0.33) between age (mean, 37 years; SD ± 10) and HRQOL (mean 56; SD ± 17). Duration of epilepsy (mean, 18 years; SD ± 13) had a positive (rho = 0.26; p = 0.01) and age at seizure onset (mean, 19 years; SD ± 15) had a negative (rho = −0.32; p = 0.001) association with HRQOL. The regression results indicate that (a) adding 3 years to age at onset is associated, on average, with 1-point decrease in QOLIE-89 (QOLIE-89 = 62.65 − 0.35 × age of onset; R2 = 0.09) while (b) adding 3 years to duration of epilepsy is associated, on average, with 1-point increase in QOLIE-89 (QOLIE-89 = 49.05 + 0.34 × duration; R2 = 0.07). Conclusions: Although HRQOL and age are not correlated, other aspects of age in epilepsy appear to matter. In patients with intractable seizures, HRQOL improves with the duration of epilepsy. Also, HRQOL is better in patients who were younger at the onset of epilepsy. This may represent better coping or adaptive mechanisms in younger patients. 1.371 MEANINGFUL IMPROVEMENT IN QUALITY OF LIFE IN DIFFERENT PATIENT POPULATIONS 1 Ryan J. Punambolam, 1 Samuel Wiebe, and 2 Suzan Matijevic (1 Clinical Neurological Sciences and Epidemiology and Biostatistics, University of Western Ontario, and 2 London Health Sciences Centre, London, ON, Canada) Rationale: How much improvement in quality of life (QOL) do patients need to experience to consider that a new epilepsy treatment is worthwhile? Individual patients and clinicians judging the effects of epilepsy treatments need to know what is the minimum clinically important change (MIC) that a therapy can afford. Our purpose was to determine whether patients with varying epilepsy severity and at different stages of treatment would specify different thresholds of worthwhile improvement (MIC). Specifically, we tested the hypothesis that patients contemplating epilepsy surgery would require a smaller MIC than less refractory patients contemplating a change in medical therapy. Methods: Two cohorts of patients were identified: (a) patients undergoing inpatient evaluation for epilepsy surgery, and (b) outpatients contemplating a change in antiepileptic drugs. Researchers administered a questionnaire that assessed the MIC on a 7-point Likert Scale ranging from 1 (“Almost the same, hardly better at all”) to 7 (“A great deal better”) in three domains: overall QOL, seizure control, and work/social activities. Results: The medical cohort consisted of 22 males and 31 females, mean age, 32.7 (SD, 11.4) years. The surgical cohort included 40 males and 50 females, mean age, 36.3 (SD, 9.6) years. Both cohorts specified a median MIC of 4 “moderately better” for overall QOL and for seizure control (Interquartile Range (IQR) for difference, 0.35). Similarly, the MIC for work/social activities did not differ between the surgical (median 3, “somewhat better”) and medical cohorts (median 4, “moderately better”) (IQR for difference, 0.52). Conclusions: Patients require an improvement equivalent to somewhat or moderately better (3 or 4 in 7-point scale) to consider surgical or medical treatment as worthwhile. Contrary to our hypothesis, the surgical and medical populations specified similar MICs in overall QOL, seizure control, and work/social activities. The patient population and type of intervention do not seem to substantially influence the MIC. 1.372 REDUCTION OF DISTRESS, EVOLUTION IN PATIENT-RATED PRIORITIES AND IMPROVEMENT OF HEALTH-RELATED QUALITY OF LIFE DURING LEVETIRACETAM TREATMENT Joyce A. Cramer, Genevi`eve Van Hamm´ee, and Roy Beran (Department of Psychiatry and Neurology, Yale University School of Medecine, New Haven, CT; Clinical Epidemiology and Outcomes Research, R&D, UCB
Epilepsia, Vol. 44, Suppl. 9, 2003
130
AES PROCEEDINGS
Pharma SA, Braine-l’Alleud, Belgium; and Department of Neurology, University of NSW/Liverpool Hospital, Sydney, NSW, Australia) Rationale: To investigate the short-term effect of levetiracetam (LEV) on distress, prioritization of QOLIE-31-P domains, and health-related quality of life (HRQOL) in patients with refractory partial-onset seizures. Methods: Patients with uncontrolled partial seizures were enrolled in an open-label clinical trial, with LEV as adjunctive therapy. The QOLIE31-P questionnaire was completed by 99 patients at the end of an 8week baseline period and after a 16-week treatment period. Patients rated their level of distress associated with each QOLIE-31-P domain at baseline and at the end of treatment period. Comparison of scores (scale: 0–100) was performed by paired t test. Domains were ranked in order of importance before and after LEV therapy based on patient-rated priorities. Results: After 16 weeks of LEV adjunctive therapy, all QOLIE-31-P domains improved, except medication effects. A statistically significant improvement was observed in the total score 3.2 (p = 0.017), the health status item 6.6 (p = 0.001), and two domains: overall QOL 5.6 (p = 0.009) and social function 7.6 (p = 0.002). The energy/fatigue subscale also shows a trend toward improvement 3.7 (p = 0.07). All distress scores associated with the QOLIE-31-P domains decreased between baseline and follow-up: seizure worry –8.1 (p = 0.011), overall QOL –2.1 (ns), emotional well-being –2.8 (ns), energy/fatigue –7.4 (ns), cognitive functioning –2.3 (ns), medication effects –0.3 (ns), social function –6.6 (p = 0.025), and overall distress –4.3 (p = 0.030). Domains ranked as the most important at baseline were social function, seizure worry, overall QOL, and cognitive functioning. At follow-up, overall QOL, social function, and cognitive functioning were still ranked among the most important domains, whereas seizure worry moved from rank 2 to 6. Conclusions: Improvement in HRQOL was associated with a reduction of related distress. The diminished importance of “seizure worry” suggests confidence in LEV efficacy. Significant changes were observed among domains ranked as the most important to patients after a 16-week LEV therapy.
1.373 QUALITY OF LIFE BEYOND “THE LAST 4 WEEKS” 1 Samuel Wiebe, 1 Suzan Matijevic, and 2 GuangYong Zou (1 Clinical Neurological Sciences, London Health Sciences Centre, and 2 Epidemiology and Biostatistics, University of Western Ontario, London, ON, Canada) Rationale: Most instruments ask patients about their quality of life (QOL) in the last 4 weeks or less. Conversely, clinicians are interested in the effect of epilepsy therapies over months and years, not weeks. Researchers often report QOL results as the difference between two measures, each looking only at a few weeks of the patient’s life. The clinical meaningfulness of such data is uncertain in epilepsy, which is characterized by fluctuating seizure frequency, severity, seizure clustering, and medication side effects. The alternative, asking frequent 4-week questions, is impractical and burdensome. It is argued that recall bias could invalidate results from questionnaires spanning longer periods. However this has not been formally assessed. Based on previous work, we now examine whether commonly used QOL epilepsy tools with extended time frames of 2 and 4 months reflect the patients’ QOL during that time. Methods: Patients with epilepsy were seen at the London Health Sciences Centre. Two cohorts of 40 patients each completed valid, monthly questionnaires asking about their QOL in the last 4 weeks, for 2 months (cohort 1), and for 4 months (cohort 2). During the last month, patients also answered a set of the same questionnaires, but with an extended time frame, asking them to report QOL over the preceding 2 months (cohort 1) or 4 months (cohort 2). The questionnaires used were QOLIE-89, QOLIE-31, SF-36, and the Liverpool battery including Drug Toxicity, Impact of Epilepsy, and Seizure Severity. For each cohort, we used Pearson correlations to assess the association between (a) the scores from the typical monthly questionnaires and the extended time frame questionnaires (2 and 4 months), and (b) the scores from each typical monthly questionnaire and the last monthly questionnaire. The hypothesis that the first set of correlations was higher than the second set was tested by using a one-tailed type I error of 5%, as described by Meng et al. (Psychol Bull 1992;111:172–5).
Epilepsia, Vol. 44, Suppl. 9, 2003
Results: The mean age was 33.7 (SD, 13.1) in cohort 1, and 34.4 (SD, 12.6) in cohort 2. Thirty-eight percent and 33% were females, respectively, in cohorts 1 and 2. Twenty-eight of 36 correlations (78%) were higher for the extended time frame questionnaires than for the monthly questionnaires. In 13 (35%) this difference was statistically significant, favouring the extended time frame questionnaires. None of the correlations that were higher for the monthly questionnaires was statistically significant. Correlations varied somewhat among different QOL instruments, suggesting that results of extended time frame questions may vary depending on the domain of QOL being explored. Conclusions: The results suggest that questionnaires asking patients to rate their QOL over the preceding 4 months QOL (rather than the typical 4 weeks) may adequately reflect QOL during this time. Pending confirmation, this could have important implications for the assessment and interpretation of QOL in randomized trials of epilepsy interventions. (Supported by LHSC internal research fund.)
Pediatric Epilepsy 1.374 CLINICAL PATTERNS OF ADMITTED CHILDREN WITH FEBRILE SEIZURES Norah A. Al-Khathlan and Mohammed M.S. Jan (Pediatrics, King Abdulaziz University Hospital, Jeddah, Saudi Arabia) Rationale: Febrile seizures are the most common convulsive event in children younger than 5 years of age. Despite the progress in understanding this benign epilepsy syndrome, a wide variation in physician evaluation and management persists. Our objective is to evaluate the clinical approach and outcome of children admitted with febrile seizures in our institution. Methods: A series of 69 consecutive children with febrile seizures were identified by chart review from 1997 to 2001. Clinical, laboratory, EEG, and neuroimaging data were evaluated. One investigator retreived all the data by using a structured data collection form. Results: The children’s ages ranged between 7 and70 months (mean, 20; SD, 14), and 59.5% were males. The source of the febrile illness was evident in 65%, however, most admitted children (60 of 69) had atypical seizures (55%), were ill looking (24.5%), had febrile status (17.5%), or positive meningeal irritation signs (4%). Electrolyte abnormalities were uncommon (10%); however, complete blood count (CBC) was abnormal in 45% and increasing the likelihood of receiving i.v. antibiotics (p = 0.01). Lumbar puncture was performed on 75%, particularily those who had a first seizure compared with recurrent seizures (OR, 3.8; 95%CI, 0.9–15). Brain CT and electroencephalograms (EEG) were performed in 13% and 33%, respectively. Obtaining an EEG was less likely if the seizure was typical (13 vs. 50% in atypical, p = 0.002). Duration of hospitalization ranged between 1 and 14 days (mean, 4.7; SD, 3.2), and only one child had meningitis, which was predicted clinically. Conclusions: Pediatricians are selective in admitting children with febrile seizures to hospital. Admitted children frquently had atypical seizures, status, or were ill looking. The yield of investigations remain low and does not justify extensive workup.
1.375 COURSE AND PROGNOSIS OF PEDIATRIC EPILEPSY: FIVEYEAR FOLLOW-UP OF THE DUTCH STUDY OF EPILEPSY IN CHILDHOOD 1 Willem F. Arts, 1 Ada T. Geerts, 2 Boudewijn A. Peters, 3 Oebele F. Brouwer, 4 Hans Stroink, 5 Els A. Peeters, and 2 Cees A. Van Donselaar (1 Pediatric Neurology, Erasmus MC/Sophia Children’s Hospital, Rotterdam, 2 Neurology, University Medical Center, Utrecht, 3 Neurology, University Hospital, Groningen, 4 Neurology, Elisabeth Hospital, Tilburg, and 5 Pediatric Neurology, Juliana Children’s Hospital, The Hague, Netherlands) Rationale: If the prognosis of childhood epilepsy is known from the moment the patient is first seen, this may have important implications for the choice of the therapeutic strategy. The Dutch Study of Epilepsy
AES PROCEEDINGS in Childhood was designed to study the prognosis of childhood epilepsy in a prospective follow-up study of a large cohort; to compare terminal remission (TR) after 5 years with the course during the follow-up; to identify prognostically relevant variables; to develop prognostic models and study the reliability with which they indicate the individual prognosis; and to evaluate the concept of intractability in childhood epilepsy. Methods: The 453 children of the cohort of the Dutch Study of Epilepsy in Childhood were followed for 5 years. TR after 5 years (TR5) was compared with the result after 2 years and with the longest remission during the follow-up. Univariate and multivariate analyses using variables defined at intake and after 6 months of follow-up were performed to determine their prognostic relevance. Based on these, models were developed to determine the individual prognosis. Data about the use of antiepileptic drugs and the outcome of various epileptic syndromes were also studied. Results: Totally, 76% had a TR5 of >1 year; 64%, >2 years; and 14% had not had any seizure during the entire follow-up. Thirty-four of 108 children with TR5 >1 year (8% of the entire cohort) were considered really intractable according to predefined criteria. Significant variables in the multivariate analyses were nonidiopathic etiology, large number of seizures before intake, history of febrile convulsions, and the combination of epileptiform EEG and age younger than 4 years for intake variables only; and epileptiform EEG and age younger than 4, postictal signs and course of the disease during the first 6 months of follow-up for intake and 6-month variables combined. The resulting models predicted the outcome correctly in 72 and 75% of cases. Comparison of the course during the follow-up or terminal remission after 2 years on the one hand and 5-year outcome on the other demonstrates that improvement occurred more often than deterioration. Conclusions: The prognosis of childhood epilepsy is generally good, and tends to improve after a longer follow-up. A poor outcome is associated mainly with symptomatic or cryptogenic etiology, the combination of young age at onset and epileptiform EEG abnormalities, and the early course of the disease. With the help of the variables used, a prognosis can be calculated in individual cases with a reasonable sensitivity and specificity. This study seems to make the way for trials of individualized treatment strategies for children with epilepsy with varying prognosis. (Supported by Dutch National Epilepsy Fund, Janivo Foundation.)
1.376 INSTRUMENTS TO MEASURE STIGMA IN CHILDREN WITH EPILEPSY AND THEIR PARENTS: INITIAL DEVELOPMENT AND TESTING 1 Joan K. Austin, 2 David W. Dunn, 1 Jessica S. MacLeod, 3 Jianzhao Shen, and 3 Susan M. Perkins (1 School of Nursing, 2 School of Medicine, and 3 Division of Biostatistics, Indiana University, Indianapolis, IN) Rationale: Perceptions of stigma have been associated with poor psychosocial outcomes in children and adolescents with epilepsy. Few measures exist to measure stigma in children with epilepsy and their parents. To address this gap we developed two brief self-report scales to measure stigma in children with epilepsy and their parents: Parent Stigma Scale and Child Perceptions Scale. This study reports on their psychometric properties. Methods: Subjects were 170 children (ages 9–14 years) with epilepsy and their major caregiving parent. Data were collected as part of a larger study using computer assisted telephone interviews. Original items were revised based on recommendations from content experts. Internal consistency reliability was measured using coefficient alpha. One item from the parent scale and two items from the child scales were dropped because they reduced the reliability coefficient. Construct validity was explored using correlation. Results: Internal consistency reliability was found to be strong for the five-item parent scale (α = 0.79) and for the eight-item child scale (α = 0.81). Empirical support was found for hypothesized relationships in the predicted direction for both scales. Parent stigma scores were significantly correlated with seizure severity (r = 0.29, p = 0.0002), seizure management confidence (r = −0.23, p = 0.002), parent positive mood (r = −0.38, p < 0.0001), parent worry (r = 0.37, p < 0.0001), and impact on family leisure (r = −0.46, p < 0.0001). Child stigma scores were correlated with seizure severity (r = 0.24, p = 0.002), seizure management confidence (r = −0.27, p = 0.0004), child attitude toward
131
illness (r = −0.60, p < 0.0001), child worry (r = 0.52, p < 0.0001), child self-concept (r = −0.38, p < 0.0001), and child depression symptoms (r = 0.48, p < 0.0001). Conclusions: Both scales were found to have strong psychometric properties. The scales also are short, and items are easy to understand. These scales have potential for use in research or in the clinical setting to measure stigma. (Supported by NS 22416 and NR 04536.)
1.377 GROWTH MEASURED AS BODY MASS INDEX IN TOPIRAMATE-TREATED CHILDREN/ADOLESCENTS WITH REFRACTORY EPILEPSY 1 E. Martina Bebin, 2 Tracy A. Glauser, 3 Marilisa M. Guerreiro, 4 Liza Squires, and 4 Emmanuel Mohandoss for the INT-26 Investigators (1 Pediatric Neurology, University of Alabama at Birmingham, Birmingham, AL; 2 Children’s Hospital Medical Center, Cincinnati, OH; 3 Universidade Estadual de Campinas, Sao Paulo, Brazil; and 4 Johnson & Johnson Pharmaceutical Research & Development, Raritan, NJ) Rationale: During double-blind placebo-controlled trials in children/adolescents (2–16 years; n = 97), topiramate (TPM) was associated with temporary weight loss or slowed weight gain in some patients, although no patients discontinued for this reason. To prospectively characterize TPM effects on growth, we analyzed body mass index (BMI) data for children/adolescents with refractory epilepsy participating in a study of TPM as adjunctive therapy under conditions that more closely reflect clinical practice. Methods: The 554 children/adolescents (1–18 years) with seizures inadequately controlled with one or more antiepileptic drugs were enrolled in this prospective, open-label study and received TPM as adjunctive therapy; core study duration was 6 months. From weight and height measured at each study visit, BMI (kg/m2 ) was calculated for age-/gender-specific percentiles. To assess potential patterns in BMI percentile change, patients were grouped into strata representing underweight (motor), hypotonia, and mental retardation. Slightly 0.05). In the 1-month taper-off group, seizures recurred in 10 patients, as opposed to 12 in the 6-month taperoff group. The follow-up period ranged from 6 months to 4 years and 3 months (mean, 25 months). The analysis of seizure recurrence in each group showed no difference (p > 0.05). Conclusions: We conclude that the duration of AED tapering off does not influence the risk of seizure recurrence in childhood. (Supported by FAPESP.) 1.400 SEIZURES AND OTHER NEUROLOGIC COMPLICATIONS ASSOCIATED WITH RESPIRATORY SYNCYTIAL VIRUS BRONCHIOLITIS 1 Laura L. Sweetman, 1 Yu-tze Ng, 2 Ian J. Butler, and 1 John B. Bodensteiner (1 Child Neurology, Barrow Neurological Institute, St Joseph’s Hospital and Medical Center, Phoenix, AZ; and 2 Child Neurology, University of Texas-Houston Medical School, Houston, TX)
139
Rationale: To characterize further seizures and other neurologic complications associated with respiratory syncytial virus (RSV) bronchiolitis. Methods: The coded diagnoses of all RSV bronchiolitis patients younger than 14 years admitted to our institution over the previous 5 years were reviewed. Those patients with concomitant diagnoses consistent with neurologic disease underwent detailed chart review. Results: There were 964 patients (age 0–13 years) with a diagnosis of RSV bronchiolitis. Thirty-six of these patients had concurrent neurologic diagnoses. Eleven patients were excluded because their neurologic disorder was preexistent. A further six patients had probable simple febrile seizures, and seven patients had a history of seizure disorder and presented with breakthrough seizures in association with RSV infection. These 13 patients were also excluded from our study. Twelve RSV-positive patients (on nasopharyngeal aspirate immunoflourescence or cultures) had definite neurologic complications without another recognized cause. The average age of these patients was 7.9 months (range, 1–19 months), of whom six were boys. Seven of these patients had seizures (predominantly generalized tonic–clonic and one with status epilepticus), three had generalized encephalopathy (marked hypotonia and decreased responsiveness), of whom two also developed esotropia, and two patients developed isolated esotropia. Conclusions: We found an incidence of neurologic complications of 1.2% (0.7% seizures) in a total of 964 patients with RSV bronchiolitis. This percentage does not include presumed simple febrile seizures or exacerbations of preexisting seizure disorder (further 1.3%). These results are consistent with the findings of our previous study at another institution (1,2). Neurologic complications, predominantly encephalopathy, occur with RSV bronchiolitis, and physicians and other caregivers should be aware of this entity. REFERENCES 1. Ng YT, Cox C, Atkins J, Butler IJ. Encephalopathy associated with respiratory syncytial virus bronchiolitis. J Child Neurol 2001;16:105– 8. 2. Ng YT, Butler IJ. Respiratory syncytial virus. N Engl J Med 2001;345:1132
1.401 WHICH FACTORS PLAY A PIVOTAL ROLE ON DETERMINING THE TYPE OF PSYCHIATRIC DISORDER IN CHILDREN WITH EPILEPSY? 1 Sigride Thom´ e-Souza, 1 Evelyn Kuczinsky, 2 Patricia Rzezak, 2 Daniel Fuentes, 1 Lia Fiore, and 1 Kette Valente (1 Neurophysiology, Psychiatry, and 2 Neuropsychology, University of Sao Paulo, Sao Paulo, SP, Brazil) Rationale: Psychiatric disorders in children with epilepsy remain underdiagnosed and undertreated. This study was designed to (a) assess
TABLE 1. Characteristics of patients with RSV-related neurologic complications Patient
Sex
Age (mo)
1 2 3 4 5 6 7 8 9 10 11 12
M M F F M F M F M F M F
17 3 15 1 2 2 14 6 2 8 19 6
Neurologic complication GTC 2 × GTC 5 × Complex febrile seizures GTC 2 × Partial seizures GTC Status epilepticus Encephalopathy Encephalopathy, right esotropia Encephalopathy, left esotropia Bilateral esotropia Left esotropia
RSV positive
CSF
Investigations MRI/CT brain
EEG
IF IF IF IF IF IF IF, Culture IF Culture Culture IF IF
– N N – – – – – T – – –
N N N A N N A – A N – A
N FS N N ED ED – – – N – DS
RSV, respiratory syncytial virus; CSF, cerebrospinal fluid; M, male; GTC, generalized tonic–clonic seizure; IF, immunofluorescence; N, normal; FS, focal slowing; F, female; A, abnormal; ED, epileptiform discharges; T, traumatic; DS, diffuse slowing.
Epilepsia, Vol. 44, Suppl. 9, 2003
140
AES PROCEEDINGS
main psychiatric disorders in children/adolescents with epilepsy; (b) verify the age at onset compared with the age at diagnosis; and (c) determine which factors are correlated with the kind of manifestation. Methods: Fifty-six patients were referred from an epilepsy tertiary center to the Unit for Diagnosis and Treatment of Psychiatric Disorders in children with Epilepsy. These patients were evaluated by a multidisciplinary team in a 1-year period. Epileptic syndromes and seizures were classified according ILAE guidelines by an epileptologist. Neurophysiologic and neuroimaging studies were performed to determine etiologic diagnosis. Psychiatric interview was performed, and diagnosis was done according to DSM IV and CID-10. A neuropsychologist and a social worker also evaluated patients with a psychiatric disorder to assess cognitive and socioeconomic status. Results: Thirty-eight (50%) patients had a psychiatric diagnosis. The main disorders were depression, occurring in 11 patients (29%) and ADHD in 12 (31.7%). Other disorders included oppositional defiant disorder, somatoform disorder, and conduct disorder in 7.9%, PDD in 5.2%, obsessive-compulsive disorder, anxiety, speech retardation/learning impairment, and bipolar disorder in 2.6%. According to age, the main diagnosis at the moment of the first evaluation was ADHD (50%) younger than 6 years, ADHD (50%) and depression (27.3%) between the ages of 7 to 12 years, and depression (72.7%) from 13 to 17 years. The median age for diagnosis of ADHD was 6 years, and for depression, 15 years. Proper questioning of families and patients revealed that the onset of depressive symptoms occurred at the median age of 9 years, and ADHD, at 3 years. This difference was statistically significant for depression. Cognitive impairment, socioeconomic status, and polytheraphy did not present a significant correlation with the type of psychiatric disorder. Family history of depression was significantly higher in children with depression. Except for an earlier age at onset of seizures for children with ADHD, factors related with epilepsy type and severity were not important in determining the psychiatric disorder. Conclusions: In this series, the most frequent disorders were depression and ADHD. Age was the most important factor for determining of the type of psychiatric disorder, with predominance of ADHD in children and depression in adolescents. Analysis of these patients’ histories revealed that depression remained underdiagnosed and untreated for a longer period. Although these patients had an age-related manifestation, it remains to be determined whether irritability–hyperactivity in early childhood and depression in adolescents represent disorders in the same spectrum with an age-related expression and what is the impact of early diagnosis and treatment.
1.402 EVALUATION OF 22 PATIENTS WITH ELECTRICAL STATUS EPILEPTICUS DURING SLEEP Guzide Turanli, Aydan Degerliyurt, Dilek Yalnizoglu, Emel Erdogan, and Meral Topcu (Pediatric Neurology, Hacettepe University Medical Faculty, Ankara, Turkey) Rationale: Electrical status epilepticus during sleep (ESES) consists of sleep-induced paroxysmal EEG activity, lasting for months or years, which may appear continuously during sleep and is usually diffused bilaterally. Epilepsy is relatively easily controlled; however, the prognosis is complicated with cognitive and behavioral issues. ESES is frequently associated with behavioral disorders, hyperactivity, learning disorders, and in some cases, psychotic regression. This study aims to evaluate long-term clinical course of patients with ESES. Methods: Twenty-two patients diagnosed with ESES were evaluated. All patients had routine EEG with stage II sleep recordings, and/or prolonged video-EEG ≤3 h including sleep recordings. MRI was obtained in all patients, and SPECT was done in 12 patients. Follow-up EEGs under different treatment regimens were evaluated along with neuropsychological data. Results: Seventeen cases were classified as epilepsy with continuous spike–waves during slow sleep, three cases were diagnosed with Landau–Kleffner syndrome, one patient was diagnosed with benign childhood epilepsy with centrotemporal spikes, and one with progressive myoclonus epilepsy. Follow-up duration ranged between 1 and 14 years. All patients but one had seizures; in 17, the initial symptom was nocturnal seizures. Ten patients (45%) had language problems. MRI was abnormal in nine of 22 patients (41%), and interictal SPECT showed hy-
Epilepsia, Vol. 44, Suppl. 9, 2003
poperfusion in various regions in five of 12 (42%). EEG showed focal epileptiform activity in 13 of 17 patients when ESES temporarily ceased. Carbamazepine and phenobarbital resulted in increased seizures, onset of new type of seizures, and regression in language. Vigabatrin and topiramate failed to improve EEG in long-term follow-up. Clobazam and clonazepam improved EEG, seizure control, and behavioral and cognitive functions in ∼70% of patients; valproate was also effective but not as much as benzodiazepines. Psychometric evaluation showed significant decrease in IQ scores during ESES. Conclusions: Mental and behavioral disorders can persist even after ESES has ceased. The outcome seems to be correlated with the duration of ESES rather than the age at onset or associated seizure disorder. Therefore, early recognition and effective treatment of patients with ESES are necessary for favorable outcome. 1.403 FEBRILE SEIZURES AND SCHIZOPHRENIA 1 Mogens Vestergaard, 2 Carsten B. Pedersen, 3 Jakob Christensen, 1 Kreesten M. Madsen, 1 Jorn Olsen, and 2 Preben B. Mortensen (1 Department of Epidemiology and Social Medicine, 2 National Centre for Register-Based Research, and 3 Centre for Clinical Pharmacology at the Dept. of Pharmacology, Aarhus University, Aarhus, Denmark) Rationale: The impact of febrile seizures on the developing brain is not well known. Observational studies indicate that febrile seizures are benign, but neuroimaging studies have demonstrated acute swelling and edema of the hippocampus after prolonged febrile seizures, and experiments on animals suggest that prolonged febrile seizures may cause long-lasting modifications of channels, synapses, and neuronal networks within the hippocampus, leading to sustained dysfunction of these cells. The hippocampus may play an important role in the development of schizophrenia. Postmortem examinations and neuroimaging studies have shown that the volume of the hippocampus is reduced in people with schizophrenia. The reduced size of the hippocampus is present at the onset of the symptoms, indicating that it is not caused by the disease or its treatment. Thus, febrile seizures may increase the susceptibility to schizophrenia. Methods: We conducted a cohort study of all persons born in Denmark between January 1977 and December 1986, who were followed up from their fifteenth birthday until onset of schizophrenia, death, emigration, or December 31, 2001, whichever came first. The cohort was established by means of data from the Danish Civil Registration System. We obtained information on febrile seizures, epilepsy, and schizophrenia by linking the cohort with the two national registries: the National Hospital Register and the Psychiatric Central Register. Results: We followed up 558,958 persons for 2.8 million personyears and identified 952 persons who were diagnosed with schizophrenia. We found that febrile seizures were associated with a 44% increased risk of schizophrenia (adjusted relative risk, 1.44; 95% confidence interval, 1.07–1.95) after adjusting for various potential confounding factors such as age, sex, degree of urbanization at birth, and history of mental illness in a parent or sibling. The association between febrile seizures and schizophrenia was neither confounded nor modified by a history of epilepsy before the onset of schizophrenia. We found no increased risk of schizophrenia among siblings of children with febrile seizures (adjusted relative risk, 1.03; 95% confidence interval, 0.72–1.47) as compared with siblings of children without a history of febrile seizures. Conclusions: To our knowledge, we are the first to show an association between febrile seizures and schizophrenia. Our cohort study was based on data from nationwide registries, which provided virtually complete follow-up. The association may be due to a damaging effect of prolonged febrile seizures on the developing brain or to confounding by unmeasured factors. The absolute risk of schizophrenia after febrile seizures was small. [Supported by the Stanley Medical Research Institute and the Danish Research Agency (j.no: 22-02-0207). The Danish National Research Foundation funds the activities of the Danish Epidemiology Science Centre and the National Centre for Register-based Research.]
1.404 EPILEPSY DUE TO A DESTRUCTIVE BRAIN LESION CAUSED BY SCORPION STING
AES PROCEEDINGS 2 Ronan
Jose Vieira, 1 Leonardo Bonilha, 2 Enrico Ghizoni, 1 Fernando Cendes, and 1 Li Min Li (1 Laboratory of Neuroimaging and 2 Poison Control Center, State University of Campinas, Campinas, Sao Paulo, Brazil) Rationale: Symptomatic acute epileptic seizures may occur in ≤5% of patients who are victims of scorpion stings, especially children. The occurrence of a long-lasting brain lesion or development of epilepsy after a scorpion sting has never been observed. We report a case of a child who developed epilepsy because of an extensive hemispheric destructive brain lesion after a scorpion sting. Methods: We report a case of a child who developed epilepsy because of a destructive brain lesion after a sting by Tityus serrulatus. We review the biologic effects of the scorpion toxin and discuss the neurotoxicity of the venom in the setting of the development of a brain lesion and recurrent nonprovoked seizures. Results: A healthy 4-year-old child, living in the State of Bahia, in the Northwestern region of Brazil, was stung in her left middle finger by a brown scorpion. Her mother witnessed the event and described the features of T. serrulatus. She soon developed local pain and paresthesias followed by diaphoresis and somnolence. Approximately 24 h after the sting, she began to convulse. She was then taken to a hospital where she achieved suboptimal seizure control, with daily tonic–clonic seizures during the period of 1 week. During this period, and between the seizures, she stayed alert and conscious, with left hemiplegia. When the seizures began to abate, 1 week after the bite, she was discharged home. She persisted with almost daily tonic–clonic seizures for 2 years after the hospitalization, despite antiepileptic drugs. with dose adjustment, she currently has achieved almost complete seizure control. Today she is 15 years old, and her mother states that after the accident, she had difficulties learning and was not able to attend elementary school. She currently has a moderate global cognitive impairment and a mild left hemiparesis, and her seizures occur approximately once monthly. She underwent neurophysiologic and image investigation. Routine electroencephalogram (EEG) showed the presence of interictal spikes and diffuse slowing in the right brain hemisphere. Magnetic resonance imaging (MRI) showed a widespread destructive lesion of her right cerebral hemisphere affecting both the cortical and sub cortical structures. Conclusions: This case is a rare illustration of exogenous excitotoxicity-induced brain damage leading to recurrent seizures, and this may guide further studies regarding excitotoxicity and the development of recurrent seizures. [Study supported by FAPESP (00/04710-2).]
1.405 EEG ABNORMALITIES IN COMPLEX FEBRILE SEIZURES OF CHILDHOOD: RELATION WITH AGE AND TIMING OF EEG SINCE ICTUS Teresa M.J. Wawrykow, Asuri N. Prasad, and Charuta N. Joshi (Pediatrics, University of Manitoba, Winnipeg, MB, Canada) Rationale: This study examines the relationship of age and timing of postictal electroencephalogram abnormalities in children with complex febrile seizures (CFSs). Methods: A retrospective analysis of EEG records of children with CFSs was performed. EEG requisitions and reports, and medical records over a 10-year interval (1990–2000) were reviewed for patients meeting the standard criteria of CFSs: 202 records were identified; 28 were excluded because of previous history of afebrile seizure (nine), CNS infection (four), history of simple not complex febrile seizure (five), or incomplete information (nine). Clinical and EEG information in the remaining 175 patients was abstracted and entered into a database (MS Access 2000). Variables included were; in/out patient status, date of birth, age at time of first febrile seizure, seizure type (generalized or partial), description of the seizure, medical history, medications, date of EEG, EEG results, neurologic examination results, and family history. Statistical analysis was performed by using Student’s t test for continuous variables, and nonparametric tests (χ 2) for categorical variables. Results: There were 175 children with CFSs of whom 39.43% had abnormal EEG records. Abnormalities documented were epileptiform patterns in (40.58%), background slowing in (44.93%), or both (14.49%). Seizure types were categorized as focal (21.14%), focal secondary generalized (7.43%), generalized (69.71%), and unknown (1.71%). The mean
141
ages of patients [abnormal (19.75) vs. normal (15.72 months)] was higher in the patients with abnormal EEG results (p < 0.05). A family history of seizure was present in 41.14%. We examined factors predictive of an abnormal EEG in patients with CFSs. On univariate analysis, age older than 3 years (OR, 4.25; 95% CI, 1.30–14.68), timing of EEG ≤7 days (OR, 3.80; 95% CI, 1.90–7.66), and abnormal neurologic examination (OR, 4.84; 95% CI, 1.34–18.99) bear the strongest association. Conclusions: In children with CFSs, age of the patient at the time of EEG, timing of EEG after the ictal event, and the presence of neurologic abnormalities on examination bear a significant relation to the likelihood of finding of abnormalities on the EEG. (Supported by Children’s Hospital Foundation of Manitoba.)
Surgical Treatment 1.406 ISOLATED AURA, BUT NOT SUBCLINICAL SEIZURES, DURING NONINVASIVE VIDEO-EEG MONITORING PREDICTS SURGICAL OUTCOME IN MESIAL TEMPORAL LOBE EPILEPSY RELATED TO HIPPOCAMPAL SCLEROSIS Marino M. Bianchin, Tonicarlo R. Velasco, Veriano Alexandre, Jr., Roger Walz, Charles L. Dalmagro, Luciana M. Inuzuka, Vera C. TerraBustamante, Ricardo Guarnieri, Lauro Wichert-Ana, Antonio C. Santos, David Ara´ujo, Dr´aulio B. de Ara´ujo, Regina M.F. Fernandes, Jo˜ao A. Assirati, Jr., Carlos G. Carlotti, Jr., Luciano N. Serafini, H´elio R. Machado, Jo˜ao P. Leite, and Am´erico A. Sakamoto (CIREP—Centro de Cirurgia de Epilepsia, Hospital das Cl´ınicas de Ribeir˜ao Preto, Ribeir˜ao Preto, Sao Paulo, Brazil) Rationale: During noninvasive video-EEG (VEEG) for presurgical evaluation of mesial temporal lobe epilepsy associated with hippocampal sclerosis (MTLE-HS), some patients have subclinical seizures or isolated auras. These findings may occur because of more localized epileptic phenomena, suggesting a better surgical outcome. To evaluate this aspect, we carried out the present study. Methods: We reviewed clinical, imaging, VEEG, surgical, and anatomopathologic records of 234 consecutive patients with MTLE-HS submitted to anterior and mesial temporal lobectomy during 1995 to 2001. Presurgical parameters analyzed included: sex, ethnicity, age at surgery, age at epilepsy onset, the duration of epilepsy, history of initial precipitating insult, epilepsy duration until surgery, family history of seizures in the first-degree offspring, monthly complex partial seizure frequency impairing awareness in the year before surgery, distribution of interictal spikes, neuroimaging, and side of surgery. VEEG data were reviewed, especially regarding the presence of clinical auras and subclinical seizures during the period of the examination. The magnitude of association between the presurgical parameters and seizure outcome was measured by odds ratio (OR) and respective 95% confidence interval (CI). Crude and adjusted ORs were estimated by unconditional logistic regression. Statistical significance was determined by χ 2. Results: Overall, 82% of the patients became seizure free after anterior and mesial temporal lobectomy. We did not observe differences between patients seizure free (Engel class I) and those who remain with seizures among all the studied variables, except for isolated auras during the VEEG. Ninety-six percent of the patients that had isolated auras during video-EEG became seizure free when compared with 78% of the patients that had no auras. The chance of being seizure free in the group of the patients that had isolated auras during video-EEG increased ∼8 times (CI 95%, 1.84–33.87; p < 0.001). The observed association remain significant after the adjustment by multiple logistic regression. Conclusions: The presence of isolated aura, but not subclinical seizures, predicted a good surgical outcome. Our observation suggests that selected clinical or electrographic parameters easily observed during VEEG monitoring might be predictive of surgical outcome. (Supported by FAPESP 02/03743-0, and FAEPA-HC.) 1.407 AED WITHDRAWAL DURING VIDEO-EEG MONITORING IN CANDIDATES FOR EPILEPSY SURGERY 1,2 Addison R. Bolanos, 1 Daniel Narino, 1 Felipe Quiroga, 1 Margarita Benito, 1 Edwin Ruiz, 2 James J. Riviello, and 3 Sandra L. Helmers
Epilepsia, Vol. 44, Suppl. 9, 2003
142
AES PROCEEDINGS
(1 Neuroscience, Hospital Central de la Policia-HOCEN, Bogota, Colombia; 2 Neurology, Children’s Hospital, Boston, MA; and 3 Neurology, Emory University, Atlanta, GA) Rationale: There is a great deal information regarding seizure semiology in patient candidates for epilepsy surgey. However, there is little information regarding to protocol for antiepileptic drug (AED) withdrawal before admision and during video-EEG monitoring. The aim of this study is to analyze seizure occurrence after AED (abrupt) withdrawal during video-EEG monitoring in patient candidates for epilepsy surgery to establish the safety of a withdrawal protocol. Methods: Patient candidates for epilepsy surgery phase I or II admitted in the video-EEG monitoring unit between 2001 and 2002 were included in the protocol study. AEDs were withdrawn quickly in all the patients before admission. Thirty-nine patients with frontal and temporal lobe epilepsy were included. Results: Video-EEG monitoring lasted 3.5 days. During admission, patients had significantly more complex seizures and secondarily generalized seizures. The average of seizures was 5.4 per patient. There was a direct relation between AED withdrawal and seizure recurrence. The group with temporal lobe seizures was more prone to have early seizure manifestations and EEG discharges in comparison with the frontal group. There was no status epilepticus. We observed the presence of pseudoseizures or NES (nonepileptic seizures) in both groups. Conclusions: Withdrawal of AEDs increases early seizure frequency and severity in those patients who are candidates for epilepsy surgery during and before video-EEG monitoring, decreasing the number of days in LTM (long-term monitoring) and reducing epilepsy surgery expenses. However, further and well-controled studies are necessary to define a safe protocol. 1.408 FUNCTIONAL REORGANIZATION OF VITAL CORTEX IN PATIENTS WITH EPILEPTIC MALFORMATIONS OF CORTICAL DEVELOPMENT Jorge G. Burneo, Robert C. Knowlton, Elizabeth M. Bebin, and Ruben I. Kuzniecky (Neurology, University of Alabama at Birmingham, Birmingham, AL) Rationale: Surgical selection is evolving to a point at which localization of the epileptogenic zone and eloquent functional cortex before any surgery is vital. In patients with intractable epilepsy associated with malformations of cortical development (MCDs) in the rolandic regions, the presurgical evaluation is a challenge because unpredictable localization of the primary somatosensory function may exist. This localization is essential for optimal patient selection and surgical strategy. Experiments in animals and observations in humans indicate that the cerebral cortex has the capability to adapt to injury through a range of different mechanisms of changes, including functional reorganization of intact cortical areas. Evidence of brain reorganization has been observed in patients who have had strokes, limb amputation, and brain injury early in life. The main question to be answered is whether functional reorganization of somatosensory cortex exists in MCDs involving these cortical regions. This information was obtained by means of magnetoencephalography (MEG). Methods: Consecutive patients with medically intractable epilepsy and MCD in the rolandic regions were selected based on MRI findings. The patients underwent routine presurgical evaluation including intracranial EEG evaluation. Identified patients had a 148-channel whole-head MEG study to identify the localization of the somatosensory cortices. A validated localization of the sensory cortical function that is >2.0 cm distant (center-to-center) from expected anatomic localization (based on intrasubject control mapping in the contralateral nonlesional hemisphere) was considered representative of reorganized function. Additionally, disturbance of homuncular arrangement as compared with that expected on general anatomic rules and confirmed against the unaffected contralateral side was considered as reorganized function. Results: Seven patients were included in the study. Four patients had focal cortical dysplasia, two polymicrogyria and schizencephaly, and one had isolated polymicrogiria involving the rolandic areas. In the four patients with cortical dysplasias, the somatosensory cortices were identified outside the rolandic area and were considered to be reorganized. In the two patients with polymicrogyria and schizencephaly, the somatosensory
Epilepsia, Vol. 44, Suppl. 9, 2003
cortices remained in the rolandic areas if the anatomy was not distorted by the presence of the schizencephalic cleft, revealing no reorganization. In the patient with isolated polymicrogyria, the somatosensory cortex remained in the rolandic area. Conclusions: In patients with epileptic MCDs involving the rolandic areas, the somatosensory cortices may be reorganized if the MCDs are due to an abnormal neuronal and/or glial proliferation (i.e., cortical dysplasia with balloon cells), but may not in MCDs due to abnormal cortical organization (i.e., polymicrigyria). [Supported by an educational grant from the Epilepsy Foundation of America (J.G.B.).] 1.409 MEG AND FDG-PET IN THE PRESURGICAL EPILEPSY EVALUATION 1 Robert C. Knowlton, 1 Jorge Burneo, 1 Dalia Miller, 1 Roy Martin, 1 Avinash Prassad, 1 R. Edward Faught, 1 Richard Morawetz, 2 James Mountz, and 1 Ruben I. Kuzniecky (1 Neurology, UAB Epilepsy Center, and 2 Radiology, Division of Nuclear Medicine, University of Alabama at Birmingham School of Medicine, Birmingham, AL) Rationale: Patients without an unequivocal epileptogenic lesion identified on MRI remain a large challenge in presurgical epilepsy evaluations. FDG-PET and magnetoencephalography (MEG) are both established interictal noninvasive neuroimaging tools that can aid the goal of localization for surgery. However, the value of these imaging modalities in patients with nonlocalizing MRI (normal, questionable, or ambiguous abnormalities) is unclear. Methods: As part of a prospective multimodality imaging study in presurgical epilepsy evaluations, we reviewed the localization results of FDG-PET and MEG examinations acquired consecutively between September 2001 and January 2003 at UAB. Twenty-five patients had adequate imaging data sets and sufficient ultimate localizing information from video-EEG and other tests to allow comparison studies. FDG-PET was acquired in the resting condition with no evidence of seizure activity during the scan. PET analysis was based on inspection by two independent blinded reviewers with consensus on identification of relative focal deficits of FDG uptake. MEG was recorded with a 148-channel whole-head system with simultaneous scalp EEG using the standard International System of electrode placement. Epileptiform spikes and sharp waves were analyzed with a single equivalent current dipole model. Results: Of the 25 patients, the mean age was 13.9 years (range, 1– 56, 12 female). Only nine patients had localized potential epileptogenic pathology on MRI. See Table 1 for distribution of patients based on lobar localization and MRI classification. Note, five of the six patients with “lesions” had only subtle or questionable abnormalities on MRI, three suggestive of focal cortical dysplasia, and two with slight increase in T2 in the amygdala. Twenty patients had localizing MEG studies, and 18 had relative focal hypometabolism on FDG-PET (only one case with both negative MEG and PET studies). Concordant lobar localization was present between MEG and PET in 13 patients. Seven of the eight patients localized with MEG and not PET had normal MRI scans and extratemporal lobe epilepsy. All three patients with localized PET and negative MEG had frontal lobe epilepsy with colocalized subtle MRI abnormalities (as noted earlier), pathologically confirmed as cortical dysplasia. Conclusions: MEG and FDG-PET examinations are useful in the presurgical evaluation of patients with nonlocalized MRI whether they have temporal or extratemporal lobe epilepsy. Further study (with larger numbers) should be designed to determine if patient subgroups can be
TABLE 1. Epilepsy localization of patients according to MRI classification
Frontal Medial temporal Lateral temporal Frontal–parietal Temporal–parietal Total
Normal
Lesion
MTS
Ambiguous
7 5 1
3 2
3
2
3
1 3
1 13
6
Total 10 12 1 1 1 25
AES PROCEEDINGS identified who particularly benefit from one or the other modality, when the combination of the two is optimal, and when they are redundant. (Supported by NIH K-23 Award, R.C.K. EFA Clinical Epilepsy Fellowship Award, J.B.) 1.410 THE RELATION BETWEEN INTERICTAL AND ICTAL ACTIVITY IN LONG-TIME MONITORING P˚al Gunnar Larsson (Clinical Neurophysiology, The National Centre for Epilepsy, Sandvika, Akershus, Norway) Rationale: The focus-localization ability of interictal activity has been discussed. With the increasing possibilities with imaging methods, interictal activity and MRI findings have been suggested as sufficient for focus localization. This presentation gives the relation between ictal and interictal activity as recorded in the database at The National Centre for Epilepsy in Norway. Methods: Every long-time monitoring (LTM) since 1992 had the results coded in a database. A record holds two interictal and two ictal entries with lobe, side, and type of activity. Five years of recordings, 1995–2000, from the database has a total of 4,500 records. Of these 1,149 were complete and contained seizures. Results: A total of 611 LTMs showed both interictal and ictal activity, of which 457 were focal. Fifty-three percent of the patients had a temporal focus. Of the patients having a left focal interictal finding, 8.3% showed a clear right-sided focal ictal finding, whereas 11.5% of the right-sided interictal activity showed a clear focal left ictal finding. Conclusions: These findings suggest that some 10% of cases with focal interictal findings will have clear ictal findings on the opposite side. These findings are not compared with other methods such as MRI. It is unclear if the interictal or ictal activity gives the best results, but a combination of interictal and ictal activity gives more useful information. 1.411 THE PREDICTIVE VALUE OF MRI FINDINGS IN SEIZURE OUTCOME AFTER TEMPORAL LOBECTOMY 1 George L. Morris, 2 Wade M. Mueller, 3 Sara Swanson, and 3 David S. Sabsevitz (1 Regional Epilepsy Center, St Luke’s Medical Center, and 2 Neurosurgery and 3 Neuropsychology, Medical College of Wisconsin, Milwaukee, WI; and 4 Neuropsychology, Long Island Jewish Medical Center, New York, NY) Rationale: Successful epilepsy surgery of the temporal lobe has been associated with mesial temporal lobe MRI abnormalities. We retrospectively reviewed different findings to establish their value in predicting postoperative seizure outcome. Methods: We retrospectively created a database containing the preoperative and postoperative findings of >250 epilepsy surgeries beginning in 1991. We queried the database to identify patients with complete cranial MRI information and >6 months of postoperative seizure-frequency information. We specifically encoded MRI findings of hippocampal atrophy, temporal horn enlargement, and hippocampal sclerosis and analyzed their relation to seizure freedom or continued postoperative seizures. Results: Our database contained 267 surgical cases and 162 temporal lobectomies. Complete MRI and 6-month postoperative seizure outcome data was present in 140 patients. Seizure freedom was seen in 70% of the group overall. Seizure freedom with temporal horn enlargement, hippocampal atrophy, and hippocampal sclerosis were 68, 67, and 70%, respectively. Seizure freedom was statisitically significantly associated with temporal horn enlargement (p < 0.009, Pearson’s χ 2) and hippocampal sclerosis (p < 0.010). Assessments of hippocampal atrophy did not show a statistically significant prediction of seizure freedom (p = 0.847). Conclusions: We found hippocampal sclerosis and temporal horn enlargement associated with successful temporal lobe surgery for epilepsy. Success for the group was overall good, and the lack of an effect from hippocampal atrophy may reflect the qualitative versus quantitative techniques used when data collection started in 1990. We weight MRI findings for surgical selection today with these findings and believe quantitative MRI assessment of hippocampal volume would likely support its predictive value as well.
143
1.412 MEG, EEG, AND MRI CORRELATION IN A “PURE CULTURE” OF TEMPORAL LOBE EPILEPTIC PATIENTS 1 Tom´ as Ortiz, 1,2 Felipe Quesney, 3 Rafael Garc´ıa de Sola, 1 Alberto Fern´andez, 1 Fernando Maest´u, and 1 Carlos Amo (1 Centro MEG, Universidad Complutense de Madrid, Madrid, Spain; 2 Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada; and 3 Neurosurgery, Hospital de la Princesa, Madrid, Spain) Rationale: We investigated the localizing effectiveness of MEG in a group of patients with temporal lobe epilepsy (TLE) who remained seizure free for at least a year after anterior temporal lobectomy (“pure culture”). Methods: Nineteen patients (seven females; mean age, 30 years) were submitted to anterior temporal lobectomy including amygdalectomy and partial hippocampectomy after extensive presurgical investigations, which included long-term video-EEG monitoring (34 of 64 channels) with extracranial and foramen ovale (13 patients) electrodes, 3D-MRI (1.5–3 Tesla), interictal SPECT, PET (18-FDG), and MEG (148 channels) without anticonvulsant withdrawal. MRI findings were mesial temporal sclerosis (MTS), 12 (one with dual pathology); neoplasm, four; cortical gliosis, one; diffuse temporal lobe gliosis, one; and cavernoma, one. Results: Patients were classified in two groups: (a) MTS group (12 patients): MEG interictal epileptic discharges were recorded in six of 12 patients (50%). In four of these (33%), the epileptic activity was regional involving both mesial and neocortical temporal lobe structures. In two additional patients (16%), the epileptic abnormality had a diffuse temporal neocortical distribution. Bilateral independent epileptiform discharges were seen in three of the former patients (25%). No isolated MEG epileptic activity was recorded from mesial temporal lobe structures. (b) Temporal neocortical group (seven patients): Interictal epileptiform discharges were recorded in all seven patients with neocortical MRI lesions. Its distribution was strictly neocortical without mesial temporal involvement, and in three of them, it was bilateral. Conclusions: Our results show that MEG is a useful complementary technique in the presurgical evaluation of patients with TLE. Admittedly MEG has a higher yield in patients with temporal neocortical epilepsy as compared with patients with MTS. This probably represents a lower resolution of MEG in recording epileptiform potentials from mesial temporal structures, which are distant from the MEG sensors.
1.413 THE IMPACT OF SISCOM DATA REGARDING SURGICAL PLANNING AND UNDERSTANDING THE SPREAD OF NEURONAL ACTIVITY IN INTRACTABLE EPILEPSY PATIENTS 1 Jeffrey M. Politsky, 1 Anthony M. Murro, 1 Yong D. Park, 1 Ki H. Lee, 2 Mark R. Lee, 3 Jay J. Pillai, and 2 Joseph R. Smith (1 Neurology, 2 Neurosurgery, and 3 Radiology, Medical College of Georgia, Augusta, GA) Rationale: An ictal single positron emission computerized tomographic (SPECT) scan has come to represent a useful neuroimaging technique for seizure localization. The sensitivity of this data can be improved by combining it with a magnetic resonance image (MRI) of the brain in the format of subtraction ictal SPECT coregistered to MRI (SISCOM). The purposes of this study are first, to determine how SISCOM data affects surgical planning and outcome in patients with intractable epilepsy, and second, to improve understanding of transsynaptic activation (diaschisis) in patients with focal epilepsy. Methods: To determine the impact of SISCOM data on surgical planning and outcome, we reviewed the SISCOM data of 72 patients with intractable epilepsy. The SISCOM data, along with data acquired during continuous video (surface-scalp) electroencephalography (EEG) and intracranial EEG recording, were reviewed and analyzed for concordance. The influence of this information on surgical planning was also studied. To elucidate the pathway of neuronal activity, the same SISCOM data were reviewed, in a blinded fashion, by five of the authors. Each scan was interpreted based on a primary and secondary focus, in addition to evidence of thalamic, brainstem, and/or cerebellar activity. Results: Of the 72 cases investigated for surgical treatment of epilepsy, SISCOM and EEG data were concordant in 55% of the cases. These
Epilepsia, Vol. 44, Suppl. 9, 2003
144
AES PROCEEDINGS
patients either underwent surgery with a more confined intracranial electrode array, or were directed toward surgery without a phase II investigation. Of the 45% with discordant data, patients either underwent bilateral or broader unilateral electrode arrays, or were eliminated as surgical candidates. Decisions regarding electrode placement were based primarily on ictal EEG, secondarily on interictal EEG, and least of all, on SISCOM data. With regard to transsynaptic neuronal activity, there was evidence of cerebellar activation in 50%, thalamic activiation in 35%, and brainstem activation in 15%. Activation was ipsilateral to the presumed epileptic focus in nearly 60% and 75% of cases with cerebellar and thalamic activation, respectively. Conclusions: Cases with discordance between EEG and SISCOM data are more likely to (a) be excluded from surgery, (b) have broader intracranial electrode arrays, and (c) undergo more conservative resection margins. SISCOM data have the least weight in surgical decision making. Evidence of neuronal activation distant from the presumed epilepsy focus occurred in >50% of cases. The concept of neurophysiologic diaschisis may reflect pathways of augmented neuronal activity or recruitment, similar to neuronal plasticity, and may depend on the underlying pathology. The possibilities of a nonmotoric role for the cerebellum in epilepsy and a role for the thalamus in consciousness during an ictus are also discussed. [Supported by Internal funding (M.C.G.).] 1.414 A NEW METHOD TO IMPROVE MEG RESOLUTION IN TEMPORAL LOBE EPILEPSY 1,2 Felipe Quesney, 1 Alberto Fern´ andez, 1 Carlos Amo, 1 Fernando Maest´u, 3 Rafael Garc´ıa de Sola, and 1 Tom´as Ortiz (1 Centro MEG, Universidad Complutense de Madrid, Madrid, Spain; 2 Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada; and 3 Neurosurgery, Hospital de la Princesa, Madrid, Spain) Rationale: According to the literature, the localization of magnetoencephalography (MEG) dipoles to mesial temporal structures in patients with temporal lobe epilepsy (TLE) does not exceed 40%, and many of them do not present spikes during the MEG examination. We hypothesized that this limitation could be improved by analyzing the slow-wave activity so often seen in EEG recordings obtained in patients with TLE. Methods: MEG recordings were performed in 19 patients (seven females; mean age, 30 years) who remained seizure free after anterior temporal lobectomy including amygdalectomy and hippocampectomy. We performed a dipolar density analysis (DDA) of the delta and theta activity recorded during MEG tracings from the entire brain by using the method described by Fehr et al. (2001), which expresses results in dipoles/second. The dipole lateralization was calculated according to a logistic regression method. Traditional dipole source localization of the slow-wave activity was also calculated, and the results of both methods were compared with the dipole localization of the interictal epileptic discharges and with the MRI results. Results: In 17 of 19 patients (89%), the DDA of the slow-wave activity agreed with the traditional MEG dipole source localization of this activity and with the MRI findings. The most salient results were obtained in eight of 19 patients without MEG interictal spikes, in which the DDA and the source localization of the slow-wave activity were concordant with the MRI lesion. This method was also useful in five of 19 patients (26%) with bilateral slow-wave activity or with bitemporal epileptiform discharges in which the lateralization provided by DDA was concordant with the MRI. The morphologic appearance of the slow-wave activity was either monomorphic or polymorphic. A strong correlation of the slow-wave activity with the MRI lesion and not with the interictal spiking was documented. Although the DDA does not discern between mesial and neocortical dipole localization, traditional dipole source analysis documented a mesial temporal location of slow-wave dipoles in 42% of patients. Thus, both methods are complementary. Conclusions: DDA and logistic regression of the slow-wave activity significantly improves the localizing yield of MEG in TLE, reaching 89%. This is very useful, particularly in patients without interictal spiking during the MEG recording. The slow-wave activity seems to correlate more with the MRI lesion than with the interictal epileptic disturbance. 1.415 SEIZURES RECORDED IN PATIENTS WITH MESIAL TEMPORAL SCLEROSIS SEEN ON MRI: SHOULD WE “SCAN Epilepsia, Vol. 44, Suppl. 9, 2003
AND CUT” OR STILL RELY ON LONG-TERM VIDEO-EEG MONITORING? Paul C. Van Ness, Mark A. Agostini, Ramon R. Diaz-Arrastia, and Noel S. Baker (Neurology, University of Texas Southwestern Medical Center, Dallas, TX) Rationale: Because MRI is such a sensitive tool at detecting hippocampal atrophy (HA), many epilepsy centers use MRI as a screening test to select anteromesial temporal lobectomy candidates likely to have mesial temporal sclerosis (MTS). There is controversy about the need for or extent of video-EEG monitoring in patients with MTS, and there is financial pressure to minimize the associated expenses. Previous investigators have retrospectively analyzed surgical cases to determine if video-EEG made a difference in the surgical decision. In our series of cases with MTS based on MRI revealing HA, the results of video-EEG monitoring were reviewed to determine the types of events recorded during long-term monitoring because this more closely mimics the way patients are evaluated. Methods: We queried a prospectively maintained relational database for all EEG studies in patients age 14 years and older evaluated at a large metropolitan county hospital. We analyzed video-EEG reports in 1,458 unique patients seen during a 5-year period from 1998 to 2003. We reviewed data in 122 patients with MRI showing evidence for MTS on T2 and/or fluid-attenuated inversion recovery (FLAIR) image sequences and their 922 events recorded in the epilepsy-monitoring unit, excluding events in which we saw no EEG changes, clinical symptoms, or signs. Seizure types recorded and epilepsy classification were determined for each case. Results: Unsuspected bitemporal independent seizures were recorded in five patients (4%). Psychogenic nonepileptic seizures (PNESs) occurred in six patients (5%) with 10 seizures. Epileptic seizures were also recorded in five of six cases with PNESs. Unusual events occurred in 24 patients (20%) 46 times; there were no EEG changes, and the events were not clinically epileptic or psychogenic. The family or patient often thought these events were seizures, but they usually represented sleep, arousals, hypnic starts, atypical feelings, or other nonepileptic phenomena. Subclinical electrographic seizures occurred 32 times in six patients (5%). Isolated somatosensory auras occurred in three patients and five seizures and were not suspected at first in two of these patients from the initial history, because both also had more typical temporal lobe auras. Conclusions: In patients with MRI-documented MTSs, video-EEG can show that a substantial minority of medically refractory patients have nonepileptic events, bitemporal independent seizures, or seizures that suggest an extratemporal localization. The video-EEG monitoring results allowed physicians to optimize medical and psychological care and avoid potentially risky epilepsy surgery in poor surgical candidates. Some argue that patients with a good history and appropriate neuroimaging findings do not require video-EEG monitoring, but these data suggest that this type of presurgical evaluation can be justified. (Supported by Parkland Health and Hospital System, Dallas County, Texas, U.S.A.)
1.416 SUBTRACTION ICTAL SPECT CO-REGISTERED TO MRI (SISCOM) AND REOPERATION FOR INTRACTABLE EPILEPSY 1 Nicholas M. Wetjen, 2 Gregory D. Cascino, 2 Elson L. So, 4 A. James Fessler, 3 Brian P. Mullan, 1 W. Richard Marsh, and 1 Fredric B. Meyer (1 Department of Neurologic Surgery, 2 Department of Neurology, and 3 Department of Nuclear Medicine, Mayo Clinic, Rochester, MN; and 4 Department of Neurology, Washington University, St. Louis, MO) Rationale: Seizures persist or recur in 20–60% of patients who have undergone resection for intractable partial epilepsy. Techniques that predict likelihood of success for reoperation are needed because of the potential morbidity and cost associated with the presurgical evaluation and operative procedure. SISCOM has been shown to be a reliable indicator of the site of seizure onset in patients with intractable partial epilepsy. The diagnostic yield and prognostic importance of SISCOM in patients undergoing reoperation is not well known. Methods: We performed a retrospective study and investigated SISCOM in patients for whom surgery had failed and were being evaluated for reoperation. SISCOM images were analyzed in 59 consecutive patients who underwent video-EEG monitoring between January 1996 and October 1999 for possible reoperation. The mean age at the time of the
AES PROCEEDINGS imaging study was 29.7 years. Fifteen patients underwent subsequent prolonged intracranial monitoring. Medical records, imaging studies, and surgical reports for these patients at initial evaluation and follow-up were reviewed. Results: The SISCOM study revealed a hyperperfusion focus in 46 of 59 patients (78.0%). Twenty-eight of these 46 (60.9%) foci were in the vicinity of the previous surgical site, whereas 13 (28.2%) were remote but in the same hemisphere. The hyperperfusion focus was in the contralateral hemisphere in the remaining five patients (10.9%). The site of the ictal EEG onset zone was concordant with the SISCOM focus in 37 of 46 (80.4%). Thirty-one patients underwent reoperation and were followed up for a minimum of 6 months (mean, 24 months). Ictal onset with intracranial EEG correlated with the hyperperfusion focus in seven of 15 patients monitored (46.7%). Of the selected patients that underwent additional surgery without intracranial monitoring, 11 of 16 (68.8%) had extension of their previous temporal lobectomy. Five of these 11 (45.4%) patients had Engel class I–II at follow-up (average, 27 months). Seven of the nine patients (78.8%) with >3 years follow-up experienced Engel class I–II outcome. Conclusions: SISCOM may be useful in patients with intractable partial epilepsy for whom surgery has failed and who are being considered for reoperation. 1.417 THE UTILITY OF MAGNETOENCEPHALOGRAPHY IN PREDICTING THE EPILEPTOGENIC ZONE FOR PEDIATRIC EPILEPSY SURGERY Joyce Y. Wu, Susan Koh, Rinat Jonas, Raman Sankar, Maung Aung, Gary W. Mathern, and W. Donald Shields (Division of Pediatric Neurology, University of California, Los Angeles, Los Angeles, Scripps Clinic, La Jolla, and Department of Neurosurgery, University of California, Los Angeles, Los Angeles, CA) Rationale: MEG is a noninvasive method for localizing the epileptogenic zone in epilepsy patients. By detecting intracellular cerebral electrical activity, as opposed to extracellular activity seen on electroencephalograpy (EEG), MEG may offer more sensitive and superior localization for patients undergoing evaluation for epilepsy surgery. There is a paucity of literature on this topic as it applies to pediatric epilepsy patients. Methods: We analyzed retrospectively the records of 12 pediatric patients at the University of California at Los Angeles, whose standard presurgical tests (PET, MRI, video-EEG) revealed either discordant data or insufficient preoperative data for determining the extent of resection. MEG was utilized either to localize the epileptogenic zone for those patients with discordant data or to determine the extent of resection. MEG and electrocorticography (ECoG) results were compared. Postoperative seizure outcome was noted. Results: For four patients with discordant data, MEG aided in the localization of the epileptogenic zone. For eight patients with insufficient data, either due to multiple cortical tubers from tuberous sclerosis or due to porencephalic cysts from abscess, infarct, prior surgery, or meningitis, MEG was important in determining the extent of resection. For all 10 postoperative patients (presurgical evaluation not yet completed for two), MEG predicted preoperatively the area of resection as compared with ECoG. Furthermore, when compared with scalp ictal EEG, MEG defined a more restricted zone in 10 patients and a different zone in the remaining two patients. With a postoperative follow-up period ranging from 3 months to 1 year, seven of the 10 patients are seizure free, whereas the remaining three patients have seizure improvement (two with >50% seizure reduction, one unclear). Conclusions: MEG aids greatly in the presurgical evaluation of pediatric epilepsy patients, especially when standard presurgical testing yields discordant or insufficient data. In addition, MEG may offer superior identification of the epileptogenic zone compared with scalp EEG, in terms of both location and extent for determining resective therapy. 1.418 SENTINEL SPIKE IN PATIENTS WITH FRONTAL LOBE EPILEPSY (FLE) 1,2 Meire Argentoni, 1,2 Cristine M. Baldauf, 1,2 Carla Baise, 1,2 Cassio R. Forster, 1,2 Valeria Mello, 1,2 Leila Frayman, 1,2 Rodio Brandao, 1,2 Lauro
145
Ceda, 1,2 Joaquim O. Vieira, 1,2 Jose A. Burattini, 1,2 Pedro P. Mariani, and 1,2 Arthur Cukiert (1 Neurology and Neurosurgery, Hospital Brigadeiro, and 2 Neurology and Neurosurgery, Clinica de Epilepsia de Sao Paulo, Sao Paulo, SP, Brazil) Rationale: Patients with suspected FLE and normal MRI are often submitted to invasive recordings to define their epileptogenic areas. Invasive coverage may be unilateral, but is often bilateral because of the presence of intense secondary bilateral synchrony. A sentinelgeneralized spike (SS) is often seen preceding the eletrocorticographic (ECoG) ictal-onset pattern in these patients. Methods: Eleven patients with suspected FLE and normal MRI were submitted to invasive recordings using extensive coverage with subdural electrodes (six unilateral and five bilateral implants). In five patients, there was a generalized SS preceding the actual ECoG ictal-onset pattern. The prevailing ictal-onset pattern was a theta recruiting rhythm (seven of 11). The ictal phenomenon was synchronous with the ECoG ictal onset and not with the sentinel spike. All patients were submitted to extensive frontal lobe resection. Results: Five of the six patients without SS have been rendered seizure free by surgery. Only one of five patients with SS have been rendered seizure free, although all of them had ≥80% seizure frequency reduction. There was no significant morbidity or mortality. Conclusions: SSs are more frequently seen during ECoG. Surface video-EEG recordings rarely adequately define them. The presence of SSs seems to carry a poorer prognosis in relation to seizure outcome and might correlate to more widespread epileptogenesis.
1.419 INTRAOCCIPITAL LOCALIZATION OF INTRACTABLE OCCIPITAL EPILEPSIES: CLINICAL, NEUROIMAGING, AND SUBDURAL ELECTROENCEPHALOGRAPHIC CORRELATES Warren T. Blume, Lisa M. Tapsell, and Samuel Wiebe (Department of Clinical Neurological Sciences, London Health Sciences Centre– University Campus, London, ON, Canada) Rationale: Although several series (reviewed in Blume and Wiebe, 2000) described clinical phenomena among patients with occipitally originating seizures, whether intraoccipital site-specific semiologies occur is not clear. Moreover, Collins and Caston (1979) showed experimentally that a robust ictal discharge will propagate promptly throughout the occipital lobe, possibly obscuring ictal origin. Therefore, this study determined the proportion of intractable occipital epilepsies that could be localized to a single occipital surface, either mesial or lateral-inferior, using subdural EEG, neuroimaging, or both modalities. Second, whether ictal semiology or clinical examination can distinguish between seizures originating strictly mesially from those arising in other areas was sought. Methods: We included all patients whose subdurally recorded seizures arose from a single occipital lobe, or those with an occipital lesion whose removal reduced seizure quantity by >90%. Excluded were occipital seizure patients with a more active extraoccipital epileptogenic zone. Multiple surface occipital subdural electroencephalography was performed in 35 patients: 24 of 30 with focal occipital neuroimaging lesions, and all 11 without such. Results: Of 41 patients meeting criteria, a principal epileptogenic zone was established in 29 (71%): mesial surface in 18, lateral surface in 11, and both surfaces in 12 patients. Of 30 patients with a focal occipital neuroimaging abnormality, seizures arose from that single surface in 23 (77%), whereas six of 11 patients (55%) without neuroimaging focality had single-surface occipital originating seizures. Visual aura occurred in 35 (85%) of 41 patients. Of 31 patients with unformed phenomena, seizures arose from the mesial surface in 11 (35%), lateral surface in nine (29%) and simultaneously from both surfaces in 11 (35%). Of 18 patients with formed visual phenomena, seizures arose from the mesial surface in seven (39%), lateral surface in five (28%), and both surfaces in six (33%). Dyscognitive (partial complex) attacks occurred in 33 (80%) patients, and secondarily generalized tonic–clonic seizures in 35 (85%). Neither the visual aurae, dyscognitive, nor secondarily generalized phenomena distinguished patients whose seizures began from mesial, lateral, or both occipital surfaces. Although all nine patients with preoperative quadrantanopia or hemianopias had seizures arising from the mesial (six patients) or both mesial and lateral surfaces (three patients) as compared
Epilepsia, Vol. 44, Suppl. 9, 2003
146
AES PROCEEDINGS
with 20 of 30 patients with normal preoperative Goldmann visual fields, this relation did not achieve statistical significance (p = 0.0794, Fisher’s Exact). Optic atrophy precluded visual field testing in two patients. Conclusions: Unfortunately, ictal semiology and visual field testing failed to distinguish mesial, lateral, and both-surface occipital epileptogenesis groups. Nonetheless, single-surface occipital seizure origin could be discerned in 29 (71%) of 41 of these medically intractable patients.
1.420 IMPROVED MRI AND INTRACRANIAL EEG CORRELATION USING BRAINLAB PROTOCOL IN PEDIATRIC PATIENTS WITH CORTICAL DYSPLASIA Robert Flamini, Roger Hudgins, Susan Palasis, Raymond Cheng, Brian Philbrook, Jennifer Ganote, Thomas Burns, and Lynne Gilreath (Children’s Epilepsy Center, Children’s Hospital of Atlanta at Scottish Rite, Atlanta, GA) Rationale: Adequate electrode coverage of areas of abnormal cortical signal on MRI is essential when considering intracranial recordings for intractable epilepsy. Despite clear signal changes on MRI, the brain surface is commonly normal at the time of subdural electrode placement, particularly in children with cortical dysplasias. The BrainLab Imaging System improves accuracy for placement of recording electrodes. Methods: Four pediatric patients (ages 6 months to 8 years) underwent intracranial recordings (three extraoperatively, one intraoperatively) for delineation of epileptogenic zone using MRI-guided stereotactic placement with the BrainLab Computer Assisted Image Guided System. MRI images were acquired by using 3D FSPGR T1 sequence, T2 , and FLAIR preoperatively, and taken to the operating room for navigation using Laser Z-point guided coregistration with accuracy of 1.6–2 mm. Extraoperative cortical stimulation was completed to delineate eloquent cortex. Resection and multiple subpial transections were tailored for each patient. Results: All patients had excellent agreement between electrical and anatomic abnormalities based on stereotactic guided electrode placement. Both interictal and ictal recordings were located over the grid in the expected area of MRI abnormality based on laser-guided co-registration. This technique proved placement of the subdural electrodes was directly over the area of anatomical abnormality. The correlation between MRI and EEG allowed adequate delineation of epileptogenesis, followed by a tailored resection. This is reflected in excellent outcome in three of four patients (Engel IA, IB). The patient with poor outcome proved later to have Rasmussen encephalopathy, and underwent a functional hemispherectomy. Pathology confirmed cortical dysplasia in all patients, one of whom had co-diagnosis of tuberous sclerosis. Conclusions: MRI-guided stereotactic placement of subdural electrodes utilizing BrainLab protocol greatly improves correlation between MRI abnormalities and interictal as well as ictal EEG changes. Given the common normal appearance of cortical surface at the time of electrode placement, we currently use and recommend this technique to improve correlation between MRI and EEG for localization of ictal onset.
1.421 3D-RECONSTRUCTION OF MAGNETIC SOURCE IMAGE DATA FOR INTRAOPERATIVE NAVIGATION SYSTEM IN PEDIATRIC PATIENTS WITH INTRACTABLE EPILEPSY 1 Koji Iida, 1 Hiroshi Otsubo, 1 Yuuri Matsumoto, 2 Stephanie Holowka, 1 Rohit Sharma, 3 James T. Rutka, 1 Shelly K. Weiss, 2 Sylvester H. Chuang, 2 Nathaniel A. Chuang, and 1 O. Carter Snead III (1 Neurology, 2 Diagnostic Imaging, and 3 Neurosurgery, The Hospital for Sick Children, Toronto, ON, Canada) Rationale: Magnetic source imaging (MSI) is a combination of magnetic sources of evoked responses and epileptic discharges co-registered with structural magnetic resonance images (MRI). A combination of MSI data and intraoperative neuronavigation system has been developed to inform the exact location on the real-time functional mapping in patients with lesions around the motor cortex. We report a systematic study regarding clinical applications of this technique in surgical treatment for intractable epilepsy. Epilepsia, Vol. 44, Suppl. 9, 2003
Methods: We studied magnetoencephalography (MEG) to obtain somatosensory evoked fields (SEFs) and interictal spikes in 16 patients with (14 patients) or without lesions (two patients). MEG data were overlaid onto individual T1 -weighted MRI. We converted MSI data on a neurosurgical flameless stereotaxy system in the operating room. Central sulcus (CS) was identified using intraoperative somatosensory evoked potential (SEP) and/or cortical stimulation in cases of epileptic zone around the CS. Epileptic zone was demarcated by intraoperative electrocorticograpy (ECoG) and/or prolonged intracranial video-EEG (IVEEG) before the ECoG. We compared the epileptic zone on the exposed brain surface to 3D-reconstructed MEG epileptic zones on the intraoperative navigation system. We also analyzed the correlation of the SEF location on 3D-reconstructed MSI with the results of the SEP and/or cortical stimulation. Results: All 16 patients underwent cortical excision: in 11 patients, lesionectomy alone in four patients, and hemispherectomy in one patient. Intraoperative ECoG and IVEEG were performed in 12 patients. The epileptic zone was located around the sensorimotor cortex in six patients. SEF navigated the sensory cortex in five of the six patients. The navigated sensory cortex along with CS was highly correlated with the sensorimotor cortex identified by intraoperative SEP and/or cortical stimulation. MEG spike sources were obtained in 12 patients. Eight patients who had spikes on both MEG and intraoperative ECoG showed colocalized the epileptogenic zone. In the eight patients, the distribution of MEG spike sources were highly correlated with intracranial EEG spikes in the ipsilateral hemisphere with a mean of 97.7% (91.3–100%), excluding mirror foci. Conclusions: 3D-reconstrated MSI data for intraoperative navigation system can precisely localize functional and epileptic regions on the exposed brain surface. This technique allows the neurosurgeons to plan the optimal preoperative trajectory and to perform intraoperative demarcation of the epileptogenic region, especially around the functional cortex.
1.422 FALSE-POSITIVE LOCALIZATION OF LANGUAGE SITES DURING ELECTRICAL MAPPING NEAR EPILEPTIC FOCI Olgica Laban, William Barr, Werner Doyle, Souhel Najjar, and Manoj Raghavan (NYU Comperehensive Epilepsy Center, New York University Medical Center, New York, NY) Rationale: An assumption underlying electrical mapping of language cortex is that the excitatory interference produced by stimulation is local. Hyperexcitable cortical networks near epileptic foci may invalidate this assumption and lead to false-positive identification of language sites. We report a series of patients in whom this phenomenon was encountered, and identify features that allowed the recognition of such sites. Methods: We reviewed language-mapping studies performed in the context of two-stage epilepsy surgery over the past 3 years at our center. We identified nine patients in whom extraoperative electrical mapping identified positive language at one or more sites, which were subsequently excluded for language function, and resected without impairment. All these patients had left hemispheric seizure foci, and had undergone left craniotomies with placement of subdural grid and strips for intracranial localization of seizures foci before resective surgery. Language mapping using electrical stimulation was performed in all of these patients after seizure localization and resumption of antiepileptic medications. In all but one of these patients, specific sites were retested because of the positive language phenomena within the area of planned resection. Results: The patients ranged in age from 8 to 44 years. Seizure foci were in the left temporal lobe in seven patients, left frontal in one, and left temporoparietal in one. The IAP results were available for eight patients, and demonstrated left hemisphere language dominance in six, exclusive right language in one, and bilateral language in one patient. In seven patients, language impairment was produced at one or more sites near to, or overlapping the ictal focus. In three of these patients, auras, automatisms, or early ictal behavior were noted in the absence of afterdischarges in adjacent subdural contacts. Of the remaining two patients, both had anteromesial temporal seizure foci and showed language impairment at anterior sites. Of the nine patients, extraoperative retesting after achieving high anticonvulsant levels was able to exclude language at these sites in three. Retesting of the sites during awake surgery (with higher anticonvulsant levels), continuous language
AES PROCEEDINGS assessment during surgical manipulation of the area, and one instance focal cooling, allowed exclusion of language function in four of these patients. In two instances, resection was performed without further testing, and resulted in no impairment. Conclusions: We conclude that electrical stimulation near epileptic foci can yield false-positive localization of language sites. It becomes important to recognize such sites when the surgical plan for resection is significantly modified by the finding. Strategies such as retesting the sites electrically after achieving high anticonvulsant levels, or intraoperative language assessment during surgical manipulation, or focal cooling may provide ways to exclude language representation in these areas.
1.423 NORMAL BRAIN MRI AND TEMPORAL LOBECTOMY: OUTCOME AND CONCORDANCE IN PREOPERATIVE ASSESSMENT 1,2 William O. Tatum IV, 1,2,3 Selim R. Benbadis, 1 Leanne S. Heriaud, and 1,2 Fernando L. Vale (1 Department of Neurology, Tampa General Hospital, and 2 Department of Neurology and 3 Neurosurgery, University of South Florida, Tampa, FL) Rationale: During the presurgical evaluation for refractory localization-related epilepsy, high-resolution brain MRI plays a routine but central role. A nonlocalizing preoperative MRI often portends a poorer outcome for patients undergoing epilepsy surgery. We assessed preoperative concordance for predictive value in patients undergoing temporal lobectomy with normal high-resolution brain MRI. Methods: Nine patients with normal high-resolution brain MRI underwent temporal lobectomy (TL) for refractory epilepsy at our center by a single epilepsy surgeon. All had anatomic imaging obtained on a GE 1.5T scanner using a dedicated epilepsy protocol that included T1 standard sagittal images, T2 , and FLAIR standard axial images of the whole brain, with 1.5-mm slice thickness for 3-D SPGR coronals, FLAIR (4 skip 1.5), and inversion recovery T2 (4 skip 0) true coronal images. The presence of febrile seizures and seizure characteristics (psychic aura), unilateral interictal IEDs, localized ictal EEG with rhythmic theta discharges at onset, functional imaging with PET/ictal SPECT, and asymmetric IAP results were analyzed for concordance in those seizure free (SF) and not seizure free (NSF). Results: Five of nine patients became SF after TL with a mean followup of 1 year. Five required intracranial EEG, although only one was SF (two had two surgical procedures). Concordance was seen with unilateral temporal IEDs in five of five SF and two of four NSF, with ictal rhythmic theta discharges at onset in three of five SF and none of four NSF, with PET/ictal SPECT in four of five SF and four of four NSF, with febrile seizures in two of five SF and one of four NSF, with psychic auras in two of five SF and two of four NSF, and with significant IAP asymmetry scores in two of five SF and none of four NSF. The presence of four concordant localizing features for TLE predicted a class I outcome in every case and never resulted in surgical failure. Conclusions: Temporal lobectomy is an appropriate treatment for patients with refractory localization epilepsy, even when high-resolution brain MRI is normal. The need for intracranial electrodes commonly predicted an NSF outcome, whereas those with at least four preoperative concordant lines of evidence suggesting TLE predicted an SF outcome.
147
as a more efficient method of visualization. A simple but robust method to perform CMPR is presented. Methods: Imaging was done on a 0.5-T GE Signa SP. A fast gradient echo sequence (TR/TE = 18 ms/9 ms; slice thickness, 1 mm) was used. Imaging time was 10–12 min. Data were transferred to a PC workstation in DICOM format and processed using software written in MATLAB. One normal volunteer and three epilepsy patients with implanted subdural strip electrodes were studied. Processing took ∼8 min to construct multiple curvilinear surfaces of the entire brain. 1. Contours of the cortical surface on user-selected rectilinear slices were drawn. A polar coordinate system was used, where the slice axis was chosen as the Z direction. 2. Sp line fit for ρ = f (θ) was done for θ ranging from 0 to 2π at intervals of 0.01 radians for a smooth contour. 3. Taking z as the selected slice locations, fitted a surface of the form ρ = f (θ ) to the data in the nonuniformly spaced vectors and linearly interpolated this surface at points specified by the desired resolution in ρ, θ , and z directions, forming a grid surface. 4. Converted the grid surface in step 3 into cartesian coordinates (x , y , z ). Interpolated 3D MRI data IM = f (x, y, z) to find the pixel values of the function IM at points x , y , z on the curved surface. 5. To find inner layers of the cortex, changed ρ found in step 3 by a specified length and repeated step 4. Results: Fig. 1. Lateral curvilinear reformatting shows the cortical surface and subdural strip electrodes (solid lines). Fig. 2. Posterior view shows clearly the susceptibility artifacts created by three overlaying strip electrodes. Conclusions: CPMR of 3D MR data is useful in visualizing the underlying cortical surface of subdural electrodes. The method described is easily implemented on a standard PC, and shown to be efficient.
1.424 CURVILINEAR REFORMATTED 3D INTRAOPERATIVE MR DATA USING A SIMPLE ALGORITHM TO DEMONSTRATE SUBDURAL ELECTRODE STRIP PLACEMENT Janaka Wansapura, Mecheri Sundaram, James Williams, Otis Williams, and George Mandybur (Department of Neurosurgery, Department of Neurology, and Department of Radiology, University of Mississippi Medical Center, Jackson, MS) Rationale: To interpret EEG recordings of ictal foci, the positions of the subdural electrodes relative to the cortical surface are necessary. 3D-MRI has been used to estimate the electrode position by multiplanar reformatting but is limited because of the inherent complexity of the brain’s convolution patterns. Recently CMPR (curvilinear multiplanr reformating) using commercially available software has been introduced Epilepsia, Vol. 44, Suppl. 9, 2003
148
AES PROCEEDINGS TABLE 1. Patients with ictal electrodecremental events
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
EDE distribution
EDE evolution
Surgery
Pathology
Outcome
Rt. O. P. T Rt SFG Lt OP Lt F Rt SMA Lt FP Rt OPT Rt OrFr and SMA Lt P. F Rt P and F Lt ant interhemispheric Rt TO area Rt FT Lt PO area Lt. F.P area Lt F Rt O lobe Lt P lobe
TO focal beta SFG 2-3 focal spiking Focal beta P 25-Hz F SMA theta 15-Hz OT 9-Hz OT OrFr beta F gamma Rt F.P gamma 20–25 in Lt interhemispheric 9–11 Hz TO FT beta Focal O theta Focal F beta Focal F beta Focal O spiking Focal P spiking
PTO. resection F. resection O. topectomy. F resection SMA resection O. topectomy O topectomy OrFr resection F resection P resection and F MST SMA resection TO resection FT resection O lobectomy F lobectomy F resection O lobectomy Lesionectomy
Heterotopia Hamartoma Heterotopia Hamartoma Old stroke Normal Dysplasia Glial scarring Microdysgenesis Inflammation Dysplasia Nonspecific Nonspecific Heterotopia Glioma Inflammation Dysplasia Glioma
II I I I II IV II I IV II I IV III II I III II I
EDE, electrodecremental events Sz, seizure; Rt, right; Lt, left; O, occipital; F, frontal; P, parietal; T, temporal; SFG, superior frontal gyrus; SMA, supplementary motor area; OrFr, orbitofrontal. 1.425 INTRACRANIAL ICTAL DIFFUSE ELECTRODECREMENTAL EVENTS IN EXTRATEMPORAL EPILEPSY: FREQUENCY AND IMPLICATIONS 1 Megdad M. Zaatreh, 2 Dennis D. Spencer, 3 Brenda Wu, 3 Hal Blumenfeld, 3 Edward J. Novotny, and 3 Susan S. Spencer (1 Department of Neurology, University of North Carolina, Chapel Hill, NC; and 2 Department of Neurosurgery and 3 Departemnt of Neurology, Yale University, New Haven, CT) Rationale: Diffuse electrodecremental events (DEDE) are sometimes seen at seizure onset, and may be followed by focal rhythmic discharge. The clinical relevance of this initial ictal electrophysiologic pattern, particularly in terms of subsequent localization of the ictal onset region, is not well defined. We examined the association of initial ictal DEDE with surgical outcome in extratemporal epilepsy. Methods: All extratemporal epilepsy surgery patients who underwent long-term intracranial recording between 1990 and 1999 with minimum follow-up of 1 year were reviewed. Patients who had any seizures with DEDE at ictal onset were identified. Ictal DEDE was defined as diffuse background flattening coinciding with habitual ictal clinical behavior. Data were analyzed for DEDE frequency, electrophysiologic evolution, pathological substrate, and surgical outcome. Results: Fifty-one intracranial recordings of 44 patients with 398 seizures were identified. DEDE were seen in 18 of 51 recordings (35.3%) as the first ictal changes in some (six of 18) or most (12 of 18) of the seizures. No particular lobe or pathological substrates were more commonly seen in patients with DEDE. Patients with DEDE had similar long-term surgery outcome to the whole group studied. Diffuse EDE lasted between 2 and 10 s, after which focal evolution emerged in 12, and diffuse pattern persisted in six (Table 1). Patients with diffuse evolution had consistently poor resection outcome (p = 0.03). Conclusions: In patients with extratemporal epilepsy, DEDE at seizure onset were a common pattern. Persistence of diffuse changes with no focal progression predicted poor surgical outcome. Conversely, evolution to focal changes was as useful as initial focal change for localization of the ictal-onset area, as judged by surgical outcome.
1.426 SURGICAL APPROACH IN PATIENTS WITH MESIAL TEMPORAL SCLEROSIS AND GROSS NEOCORTICAL SCARRING 1 Warren W. Boling, 2 Adriana Palade, and 2 John Brick (1 Department of Neurosurgery and 2 Department of Neurology, West Virginia University, Morgantown, WV) Rationale: To investigate the appropriate surgical approach in temEpilepsia, Vol. 44, Suppl. 9, 2003
poral lobe epilepsy with mesial temporal sclerosis and prominent neocortical scarring. Methods: Four consecutive patients with prominent neocortical scarring and hippocampal sclerosis underwent surgery for intractable temporal lobe epilepsy at the West Virginia University Medical Center. Each patient had temporal lobe epilepsy defined by scalp EEG and seizure semiology. At surgery, extensive pial–dural adhesions over the temporal and frontal lobes were encountered on reflection of the dura that was not evident on the preoperative MRI. Electrocorticography was recorded in each patient, and a decision to perform a selective amygdalohippocampectomy or corticoamygdalohippocampectomy was based on results of the acute EEG recording (Fig. 1).
Results: Despite extensive residual neocortical scarring, all the patients have had a marked reduction in seizure frequency or become seizure free with a selective amygdalohippocampectomy or anterior temporal corticoamygdalohippocampectomy. Conclusions: Electrocorticography is a useful tool for formulating a resection strategy. Patients with extensive neocortical scar and temporal lobe epilepsy can realize an excellent result from surgery.
AES PROCEEDINGS 1.427 WHAT IS THE FUTURE OF INTRAOPERATIVE ELECTROCORTICOGRAPHY? 1,2 Arthur Cukiert, 1,2 Jose A. Burattini, 1,2 Pedro P. Mariani, 1,2 Joaquim O. Vieira, 1,2 Rodio Brandao, 1,2 Lauro Ceda, 1,2 Meire Argentoni, 1,2 Cristine M. Baldauf, 1,2 Carla Baise, 1,2 Cassio Forster, 1,2 Leila Frayman, and 1,2 Valeria A. Mello (1 Neurology and Neurosurgery, Hospital Brigadeiro, and 2 Neurology and Neurosurgery, Clinica de Epilepsia de Sao Paulo, Sao Paulo, SP, Brazil) Rationale: Intraoperative electrocorticography (iECoG) was the method of choice for the determination of the amount of tissue to be resected in the early days of epilepsy surgery. At that time, almost no neuroimaging was available, except for pneumoencephalogram. Modern neuroimaging, especially MRI, made it possible to diagnose preoperatively the majority of the epileptogenic lesions. However, 15–20% of the patients with refractory epilepsy still have normal imaging findings. We discuss the present role of iECoG based on a series of 726 patients submitted to surgery. Methods: We retrospectively reviewed the usefulness of iECoG in a series of 726 patients submitted to surgery for refractory epilepsy. There were 446 temporal, 75 frontal, 43 rolandic, 22 posterior quadrant, 15 parietal, nine occipital, and four insular cortical resections, 79 callosal sections and 33 hemispherectomies. Mean age at surgery was 24.7 years, and mean follow-up time was 3.1 years. iECoG was routinely used during the first 250 procedures. Results: iECoG findings did not correlate with seizure’s outcome after surgery in temporal lobe epilepsy. Although some of the early patients submitted to callosal sections received iECoG for documentation of the rupture of secondary bilateral synchrony, this also proved to have no correlation with outcome, and presently a maximized callosal section is preferred, without iECoG. Patients with normal neuroimaging invariably need prolonged recordings with invasive electrodes (cECoG) and iECoG is not needed. Until recently, iECoG continued to be used as a method for the determination of the additional cortical margins to be resected during lesionectomy. However, additional cortical resection is more often limited by vascular and functional boundaries then by iECoG findings. Conclusions: iECoG is being less and less used in epilepsy surgery. Although it is easy to be performed and analyze, decision making should be usually based on limited interictal data (30–45 min of recording) obtained in anesthetized patients. More important, iECoG did not correlate with outcome in the majority of the surgical modalities presently carried out for refractory epilepsy. The better comprehension of the epileptogenic lesions and their pathophysiology would certainly lead to the disappearance of iECoG as a routine neurophysiological tool. (Supported by Sao Paulo Secretary of Health).
1.428 DIFFERENCES IN ELECTROGRAPHICAL PATTERNS ON ECOG IN EIGHT CHILDREN WITH RASMUSSEN SYNDROME Susan Koh, Rinat Jonas, Joyce Wu, Raman Sankar, W. Donald Shields, and Gary Mathern (Division of Pediatric Neurology and Department of Neurosurgery, University of California, Los Angeles, Los Angeles, CA) Rationale: Rasmussen syndrome is a progressive illness leading to intractable seizures and eventual cognitive decline. Hemispherectomy is the only effective treatment. There is little literature available as to the electrographic pattern seen on electrocorticography (ECoG). Methods: Eight children (ages 3.5–14.6 at surgery) with Rasmussen syndrome and ECoG studies were reviewed. ECoG patterns were described as single-spike, runs of spikes, paroxysmal fast activity (PFA), slowing and attenuation, and spontaneous seizures. “Fast burners” (n = 3) were defined as patients who had status epilepticus and had seizure onset to time of surgery of 17 years) who underwent subdural recording to localize electrographic ictal onset. The number of contiguous electrode contacts (center-to-center interelectrode distance, 10 mm) involved at ictal onset was tabulated, as well as lobar localization and any associated structural lesion. Surgical resection was tailored according to ECoG results. The degree of postoperative seizure control was graded according to Engel classification. For six patients long-term (>5 years), outcome data were available. The remaining nine patients have been followed up for 5–24 months (short-term) after resection. Results: An average of four seizures per patient was analyzed. As expected, if fewer electrodes were involved in ECoG ictal onset, there was a greater chance for a favorable postoperative outcome, usually Engel class I (seizure free). In general, it appeared that ictal onsets encompassing seven or more ECoG electrode contacts were most often associated with an Engel class IV outcome (no worthwhile improvement). There Epilepsia, Vol. 44, Suppl. 9, 2003
152
AES PROCEEDINGS
was no significant difference in these results in comparing short- versus long-term follow-up. Conclusions: A small epileptogenic focus on ECoG results in better postoperative seizure outcome. This could be because larger foci have greater involvement of hyperexcitable subcortical networks rather than just a limited cortical mantle of epileptogenicity. A further implication of this study is that if broad regions of ictal onset are discovered on prolonged ECoG multiple subpial transactions or vagus nerve stimulation may offer equal or better treatment efficacy versus surgical extirpation of the entire ECoG focus. 1.437 EPILEPSY SURGERY IN THE ROLANDIC CORTEX: AN UPDATE David C. Diosy, Margarita Pondal, and Samuel Wiebe (Clinical Neurological Sciences, London Health Sciences Centre, London, ON, Canada; Department of Neurology, Hospital Severo Ochoa, Madrid, Madrid, Spain; Clinical Neurological Sciences, London Health Sciences Centre, London, ON, Canada) Rationale: To determine which factors are associated with functional outcome and seizure reduction in patients who underwent resective surgery or multiple subpial transections (MSTs) in the rolandic area. Methods: All patients who underwent rolandic cortical resection or MST from 1979 to 2003 at the London Health Sciences Centre were identified. Patients were included if they had seizures originating in the rolandic area recorded with subdural electrodes with adequate coverage, or if they had scalp-recorded seizures and a discrete epileptogenic lesion on MRI in the rolandic area. All patients had electrocorticography (ECoG) at the time of surgery. Data collected include preoperative and postoperative neurologic deficits, MRI findings, interictal and ictal scalp EEG, interictal and ictal subdural data, ECoG findings, type and extent of surgery, neuropathologic diagnosis, and follow-up seizure frequency. Results: Fifty-eight patients were identified, 33 men and 24 women. Etiologies included tumor (30), dysplasia (five), stroke (three), cavernoma (three), AVM (two), Rasmussen encephalitis (three), trauma (one), infection (one), cerebral palsy (one), porencephalic cyst (one), meningioangiomatosis (one), mitochondrial cytopathy (one), and undetermined (six). Patients were taking an average of 1.8 antiepileptic drugs (AEDs); range, 0–4. Scalp EEG telemetry was obtained in 49 of the 58 patients. Interictal spikes were detected from the central region in only 14 (28.6%), were multifocal in nine (18.4%), and there were no interictal spikes in 12 (24.5%). On scalp EEG focally originating seizures arose from the central region exclusively in 13 (26.5%) patients, independently from the central region, and other loci in three (6%), from other loci but not central in six (12%), multifocal onset in two (4%), ambiguous or diffuse in 10 (20.4%), and no seizures were captured in 17 (34.7%). Neuroimaging showed a focal lesion involving or abutting the rolandic area in most patients. MRI was normal in three, showed multifocal lesions in four, and a diffuse hemispheric abnormality in one. In patients from whom follow-up information was available (51 of 58), worthwhile improvement in seizures (Engel classes I–III) occurred in 38 (74.5%), but only 17 (33%) were completely free of disabling seizures. Better outcomes were more common when the entire epileptogenic MRI lesion could be removed. New deficits were seen in 50% of patients. Conclusions: Worthwhile seizure improvements can be achieved through well-planned resection and MST in the rolandic cortex. New deficits can be expected in 50% of patients, and can be moderate to severe in half to two thirds of these. Scalp EEG provided little useful information to guide surgery. Further studies are required to determine whether new techniques, such as functional neuroimaging and improved functional neurophysiology, may permit better delineation between eloquent and epileptogenic cortex. 1.438 EFFECT OF AMYGDALA RESECTION ON SEIZURE OUTCOME IN PATIENTS UNDERGOING SURGERY FOR MEDICALLY REFRACTORY TEMPORAL LOBE EPILEPSY 1 Cornelia Drees, 2 Joseph Kulas, 2 Gordon Chelune, 4 Richard A. Prayson, 1 Elaine Wyllie, 3 William Bingaman, 1 Hans O. L¨ uders, and 1 Imad Najm (1 Neurology, 2 Neuropsychology, 3 Neurosurgery, and 4 Pathology, Cleveland Clinic Foundation, Cleveland, OH)
Epilepsia, Vol. 44, Suppl. 9, 2003
Rationale: Anterior temporal lobectomy (ATL) is an effective treatment for patients with medically refractory temporal lobe epilepsy (TLE). Whether amygdala resection influences seizure outcome is not known. This may be relevant because the amygdala is one of the most sensitive structures from which to induce kindling (i.e., a process whereby repeated, intermittent sub-convulsant electrical stimulation, eventually elicits a generalized seizure, and eventually lowers the seizure threshold). Methods: We reviewed charts and postoperative magnetic resonance images of 178 patients who had undergone surgery for medically refractory temporal lobe epilepsy between 1990 and 2000. The linear extent of resection of each temporal gyrus, the fusiform gyrus, the parahippocampal gyrus, the hippocampus, and amygdala were measured. The patients were grouped into those without and those with complete amygdala resection. Seizure outcome was assessed ≥6 months after the procedure. Results: Of 178 patients, a group of 42 cases without ipsilateral amygdala resection (group A) and a group of 89 with complete amygdala resection (group B) were selected. The rest had a variable amount of amygdala resected and did not enter our analysis. The demographic data for both groups were comparable. In both groups, the majority of patients had undergone ATL (86% in group A, 97% in group B). There were more cortical resections in group A (12%). Pathology revealed a higher percentage of cortical lesions and heterotopia in group A (36 vs. 15% in group B) and a larger number of patients with hippocampal sclerosis (HS) in group B (65 vs. 40% in group A). The number of patients who were seizure free at 6 months’ follow-up was slightly higher in the group with complete amygdala resection (67 vs. 72% in group A); however, this difference was not statistically significant (p = 0.998). Conclusions: This retrospective study shows that patients who undergo surgery for TLE and have a complete amygdala resection may have a slightly better chance of postoperative seizure control. However, statistical analysis did not show any significant difference in seizure outcome between the two groups. A prospective study will be needed to validate these results.
1.439 SURGICAL TREATMENT OF SEIZURES IN TRISOMY 15: A CASE REPORT 1 Lawrence N. Eisenman, 1 Hrayr P. Attarian, 1 A. James Fessler III, 2 Jeffrey G. Ojemann, and 1 Frank Gilliam (1 Neurology and 2 Neurological Surgery, Washington University Medical School, St. Louis, MO) Rationale: Trisomy 15 has long been associated with mental retardation and seizures. It is well known that single gene defects can produce partial epilepsy. One of the best studied examples is autosomal dominant nocturnal frontal lobe epilepsy in which mutations affecting the neuronal nicotinic acetylcholine receptor (ACHR) produce a focal epilepsy syndrome. Interestingly, one of the candidate ACHR subunits is located on chromosome 15. Similarly, benign epilespy with centrotemporal spikes may also be linked to chromosome 15. Abnormal function of γ -aminobutyric acid (GABA) receptors has also been hypothesized to contribute to partial seizures. Of note, two GABA-receptor subunits are encoded on chromosome 15. Abnormal expression of any of these genes may result in an underlying substrate that is more prone to the development of seizures. More recently, partial trisomy 15 had been associated with temporal lobe abnormalities in one kindred, suggesting that focal developmental structural abnormalities are another mechanism by which genetic abnormalities can lead to partial seizures. Either of these mechanisms may contribute to the epilepsy associated with trisomy 15. Patients with chromosomal abnormalities are typically presumed to have diffuse abnormalities that would make them poor candidates for epilepsy surgery. The purpose of this report is to present a case of epilepsy in a patient with trisomy 15 successfully treated with epilepsy surgery. At the end of this event, particpants should be able to discuss the potential role of resective surgery in epilepsy related to a chromosomal abnormalities. Methods: The patient is a 21-year-old woman with a history of trisomy 15 diagnosed at age 10, mental retardation, and medically refractory seizures diagnosed at age 14, which, in retrospect, started in early childhood. She was having daily seizures including 30 episodes of status epilepticus in the 3 months before her initial evaluation at our center. MRI and 18 FDG-PET were negative. Scalp video-EEG monitoring suggested left neocortical temporal onset of seizures with left
AES PROCEEDINGS
153
TABLE 1. Postsurgical seizure reduction by surgery type Outcome Seizure free Auras only >90%↓ 50–90%↓ No reliable data Total
Total hemispherectomy
Subtotal hemispherectomy
Hemispherotomy
Total
1 0 1 0 1 3
4 0 4 1 1 10
11 1 2 0 1 15
16 1 7 1 3 28
frontotemporal interictal abnormalities. SPECT suggested lateral temporal onset. Intracranial monitoring confirmed left neocortical temporal seizure onset, and she subsequently underwent an extensive left temporal resection. Results: At the 1-year follow-up after surgery, she has had only a single seizure consisting of 30 s of staring and decreased responsiveness in association with a febrile illness. Her family feels that her speech, language, responsivenss, and mood have all improved after surgery. Conclusions: This case demonstrates that focal resection may be beneficial even in patients in which the presumed etiology involves a diffuse abnormality. Further study is suggested to better define the role of epilepsy surgery in patients with chromosomal abnormalities. This case also illustrates the complexities of epileptogenesis in which the diffuse pathology of a chromosomal abnormalitiy results in a focal epilepsy.
1.440 DOES SPARING AUDITORY NAMING SITES COMPROMISE SEIZURE OUTCOME? 1 Marla J. Hamberger, 1 Robert R. Goodman, 1 Guy M. McKhann, and 2 William T. Seidel (1 Neurology, Columbia University, New York, NY; and 2 Ortho McNeil Pharmaceutical, Raritan, NJ) Rationale: Visual naming sites identified by cortical stimulation mapping are typically spared from surgical resection to preserve postoperative language functioning. A common concern of the surgical team, however, is whether sparing functional tissue will reduce the likelihood of good seizure control or seizure freedom postoperatively, as potentially epileptogenic cortex might be excluded from the resection. In previous work, we identified auditory naming sites, distinct from visual naming sites, the latter of which are considered “essential” for normal postoperative language. As there has not yet been adequate evidence to suggest that auditory naming sites are critical as well, these sites have not routinely been spared from resection. To determine whether sparing auditory naming sites compromises seizure outcome, we compared seizure outcome in patients who had auditory naming sites spared with those who had these sites resected. Methods: Subjects were 16 left TLE patients who underwent preoperative cortical language mapping (seven intraoperative, nine extraoperative) using both visual and auditory naming tasks. Surgical resections were tailored to preserve visual but not auditory naming sites. Auditory naming sites were spared in seven patients, resected in six patients, and fell within 2 cm from the resection boundary in three patients. Visual naming sites were preserved in all patients. Engel’s seizure outcome ratings (i.e., class I–IV; Engel, 1993) were obtained for each patient 1-year postoperatively. To be conservative, analysis included only patients from the “spared” and “resected” groups, as all patients in the “middle” group (i.e., auditory naming sites < 2 cm from resection boundary) had class I outcomes. Fisher’s exact test was used to compare seizure outcome in “spared” and “resected” groups. Results: Comparison of seizure outcome between “spared” and “resected” groups showed no significant difference (p = 1.0). Conclusions: These preliminary results suggest that sparing auditory naming sites is not associated with poor seizure outcome. (Supported by The National Institute of Neurological Disorders and Stroke, grant number NS35140.)
1.441 PEDIATRIC HEMISPHERECTOMY OUTCOMES: COMPARISON OF THREE DIFFERENT PROCEDURES 1 Ugur Isik, 1 Edwin Trevathan, 2 Jeffrey G. Ojemann, 3 Robert T. Fitzgerald, 1 Susan T. Arnold, and 2 T.S. Park (1 Neurology and Pediatrics and 2 Neurosurgery, Washington University School of Medicine, and 3 Pediatric Epilepsy Center, St. Louis Children’s Hospital, St. Louis, MO) Rationale: Three decades of epilepsy surgery performed at Washington University Medical Center have allowed comparison of seizurereduction rates and complications associated with three different hemispherectomy procedures. Methods: Records of all patients who underwent a hemispherectomy procedure at Washington University/St. Louis Children’s Hospital were identified from medical record review. Prior to 1995, patients underwent total hemispherectomy or subtotal hemispherectomy. Beginning in 1995, all patients underwent hemispherotomy. Data were abstracted from medical records by using a standardized form. Data abstracted included type of surgery, etiology, pre/post-surgery seizure frequency, surgical complications, and motor outcomes. Postsurgical seizure reduction was calculated from the average pre- and postsurgical seizure frequencies documented in medical records. Results: From 1972 to 2002, 28 children (61% boys) underwent a hemispherectomy procedure (53% right hemisphere). Before 1995, three patients had a total hemispherectomy, and 10 had a subtotal hemispherectomy. Since 1995, only hemispherotomies have been performed (n = 15). Patients’ ages at surgery ranged from 2 months to 16 years. The identified etiologies were: stroke (12 of 28, 43%), cortical dysplasia (seven of 28, 25%), Rasmussen syndrome (four of 28, 14%), meningoencephalitis (three of 28, 11%), Sturge–Weber syndrome (one of 28, 4%), and traumatic brain injury (one of 28, 4%). The mean patient follow-up time was 3.6 years (range, 0.5–21 years). There were no postsurgical follow-up data for three patients. Sixteen of 25 patients (64%) were seizure free at last follow-up; seven of 25 (28%) had more than a 90% reduction is seizure frequency, and two of 25 (8%) had a 50–90% reduction in seizures after surgery. A summary of postsurgical seizure reduction by surgery type is presented in the Table. Arm/hand use, contralateral to surgery, improved in eight of 25 (32%), worsened in three of 25 (12%), and remained unchanged in 14 of 25 (56%) after surgery. Early postoperative complications (within the first week) were absent in five of 15 (33%) hemispherotomy cases, one of 10 (10%) subtotal hemispherectomy cases, and none of three total hemispherectomy cases. Overall, the most common early complications were fever (14 of 25) and subgaleal fluid collection (seven of 25). Late postoperative complications (>1 week) were absent in 13 of 15 (87%) hemispherotomy cases, five of 10 (50%) subtotal hemispherectomy cases, and two of three (67%) total hemispherectomy cases. Hydrocephalus, shunt-related complications, and precocious puberty were the most common late complications. Conclusions: Hemispherotomy appears to be as effective in reducing seizures and may have fewer associated postoperative complications compared with the older surgical procedures. 1.442 LONG-TERM OUTCOME OF EPILEPSY SURGERY IN TUBEROUS SCLEROSIS PATIENTS Randa G. Jarrar and Jeffery R. Buchhalter (Department of Neurology, Mayo Clinic, Rochester, MN)
Epilepsia, Vol. 44, Suppl. 9, 2003
154
AES PROCEEDINGS
Rationale: Tuberous sclerosis (TS) is an autosomal dominant disorder of cellular differentiation, proliferation, and migration. Seizures occur in 84% of patients carrying the TS mutation, and they can be difficult to control with antiepileptic medications. The role of surgery and long-term postoperative outcome in these patients is not clear. Methods: Between 1986 and 2002, 23 patients with definite and probable TS underwent epilepsy surgery at the Mayo Clinic-Rochester. Patients with possible TS were excluded, as were patients with a follow-up period of 85% patients could achieve reasonable seizure control. Statistical analysis revealed frontal lobe epilepsy, mild cortical dysplasia, and simple partial seizure for worse prognostic factors. Surgery to younger age patients did not always produce better outcome. 1.448 WADA TESTING PREDICTS MATERIAL-SPECIFIC MEMORY DECLINE AFTER EPILEPSY SURGERY IN CHILDREN 1,2 Gregory P. Lee, 1 Yong D. Park, 3 Lynn K. Blackburn, 1 David W. Loring, and 1 Morris J. Cohen (1 Department of Neurology, and 2 Department of Occupational Therapy, Medical College of Georgia, Augusta, GA; and 3 Pediatric Epilepsy Center, St. Louis Children’s Hospital, St. Louis, MO) Rationale: Several investigations have shown an association between Wada memory performance before surgery and material-specific (verbal and visuospatial) memory outcome after surgery on a group level in adult
Epilepsia, Vol. 44, Suppl. 9, 2003
156
AES PROCEEDINGS
epilepsy surgery patients. There continues to be uncertainty whether these results are applicable to children, and the degree to which Wada memory results may be used to predict memory decline in the individual patient. We examined the capacity of Wada memory asymmetries to predict postoperative memory decline in children undergoing epilepsy surgery using both group and individual levels of analyses. Methods: Seventy-five children who underwent Wada memory evaluation in both hemispheres, and who subsequently underwent some form of resective epilepsy surgery, served as subjects. Children were classified into one of two groups: (a) Wada Asymmetry group, memory after injection ipsilateral to the side of surgery was greater than memory after contralateral injection, or (b) No Asymmetry group, Wada memory after ipsilateral injection was less than or equal to memory after contralateral injection. There were no significant differences between groups with regard to age, handedness, gender, duration of seizure disorder, side of surgery, location (TL vs. ExTL) of surgery, IQ, preoperative verbal memory, preoperative visuospatial memory, or interval between surgery and postoperative memory assessment. Verbal memory was assessed using Children’s Memory Scale (CMS) Stories and visuospatial memory was assessed using either CMS, Visual Delay Index, or Rey Complex Figure. Results: There were significant interaction effects for both verbal (p = 0.04) and visuospatial (p = 0.04) memory tests. Verbal and visuospatial memory scores increased after surgery for children with Wada Asymmetries and decreased after surgery in children who showed No Wada Asymmetries. Group differences were also obtained when the number of individual cases showing postoperative memory change was examined. Significantly more children with Wada Asymmetries had improved verbal (74%) and visuospatial (66%) memory scores before to after surgery than children with No Wada Asymmetries. Significantly more children with No Wada Asymmetries had verbal (64%) and visuospatial (87%) memory scores decline after surgery. Conclusions: Wada memory asymmetries may be used to measure risk of material-specific memory decline in verbal and visuospatial memory functions after epilepsy surgery in children.
1.450 SURGERY FOR FRONTAL LOBE EPILEPSY: PROGNOSTIC FACTORS AND IMPACT ON QUALITY OF LIFE Jose-Victor Longo Martinez, Andre Palmini, Eliseu Paglioli-Neto, Jaderson Costa Da Costa, Eduardo Paglioli, Mirna Portuguez, Ligia Coutinho, and Sergio Raupp (Porto Alegre Epilepsy Surgery Program, Sao Lucas Hospital-PUCRS, Porto Alegre, RS, Brazil) Rationale: Although the frontal lobe epilepsies (FLE) are the second most common type of refractory epilepsy treated with surgery, the surgical outcome for FLE is invariably worse than that for temporal lobe epilepsy. The reasons for such outcome are still unknown. To address that, we analyze the outcome of surgery for FLE and evaluate the impact that the surgery had on the patient’s life. Methods: We analyzed 60 adult patients from the Porto Alegre Epilepsy Surgery Program who underwent to surgical treatment for refractory FLE. All patients had surgical procedures on the frontal lobe. At least 2 years of follow-up was considered to search the factors associated with a good or bad outcome. A questionnaire of quality of life was applied to these patients during the outcome examination. Frequency of seizures before and after surgery was compared during the outcome. Evaluation of the ECoG before and after resection and analyze of the area resected were performed for all patients. Results: We observed that the most conspicuous aspect associated with a good outcome was the absence of lesion remaining after the excision. The factors more clearly associated with a bad outcome were (a) the presence of abundant spiking in the postoperative ECoG, (b) abnormal tissue left after resection. The satisfaction of the patients with the surgery and improvement of the quality of life were associated with seizure-frequency reduction. Conclusions: Our results suggested that the presence of abundant spiking in the postoperative ECoG and abnormal tissue left after resection are aspects of poor outcome. Surgical treatment could provide a better quality of life for patients with refractory FLE.
1.449 VALUE OF INTERICTAL EEG AND EXTENT OF HIPPOCAMPAL RESECTION IN THE SURGICAL TREATMENT OF TEMPORAL LOBE EPILEPSY Li Min Li, Leonardo Bonilha, Eliane Kobayashi, Donizete Cesar Honorato, and Fernando Cendes (Laboratory of Neuroimaging, State University of Campinas, Campinas, Sao Paulo, Brazil)
1.451 THE SURGICAL OUTCOMES IN 204 INTRACTABLE EPILEPSY PATIENTS 1 Guoming Luan, 2 Yunlin Li, 3 Zhengrong Sun, 4 Li Yan, and 5 Zhuang Cui (1 Dept of Neurosurgery, Beijing Hospital, Beijing Neurosurgery Institute, 2 Dept of Neurosurgery, Beijing Hospital, 3 Dept of Neurosurgery, Beijing Tiantan Hospital, 4 Beijing Neurosurgery Institute, and 5 Dept of Neurosurgery, Beijing Hospital, Beijing, Beijing, China)
Rationale: Unilateral hippocampal atrophy is an indicator of good surgical prognosis in patients with temporal lobe epilepsy (TLE). Some patients, however, do not become seizure free after surgery. We assessed if (a) interictal EEG and (b) the extent of hippocampal and amygdala resection are associated with outcome. Methods: Thirty patients with TLE with unilateral or clearly asymmetric hippocampal atrophy who underwent surgical treatment were evaluated concerning preoperative clinical variables and interictal EEG abnormalities. Amygdala and hippocampal resection was evaluated by postoperative MRI. We compared seizure-free versus non–seizure-free patients, and patients with good outcome (Engel’s classes I and II) versus patients with poor outcome. Results: All patients underwent anterior temporal lobe resection plus amygdalohippocampectomy ipsilateral to side of hippocampal atrophy or predominant atrophy. Two patients (7%) had null resection of the amygdala, 20 (66%) had partial resection, and eight (27%) had complete resection. Three patients (10%) had null resection of the hippocampus, 11 (37%) had incomplete resection of the anterior portion of the hippocampus, 10 (33%) had complete resection of the anterior portion without resection of the posterior portion, and six (20%) had complete resection of the anterior portion with partial resection of the posterior portion. The extent of amygdala resection did not correlate with the outcome. Conversely, the extent of hippocampal resection showed significant association with the good outcome (p < 0.001) and seizure-free outcome (p < 0.001). Conclusions: Interictal EEG was not predictive of surgical outcome. The extent of hippocampal resection, in turn, was associated with postoperative seizure control. [Supported by FAPESP (00/04710-2).]
Rationale: To explore the possible factors affecting the surgical outcomes in patients with intractable epilepsy and to summarize the experiences of resecting the epileptic foci and/or combining with bipolarcoagulation methods. Methods: In 204 patients (males, 130; females, 74), age ranged from 3 to 51 years old. Age younger than 18 years was group I, and the others, group II. MRI, EEG, and V-EEG were used as preoperative evaluations for all patients; 12 cases were treated with purely bipolar coagulation on the epileptic discharges areas, seven cases with corpus callosotomy plus bipolar coagulation, nine cases with hemispherectomy, and 176 cases with lesionectomy and/or epileptogenic foci resection plus bipolar coagulation. Patients with temporal lobe epilepsy (TLE) were 100 cases, and extra-TLE, 42 cases (25 in frontal, 12 in parietal, and five in occipital) and multiple lobe epilepsy, 162. All patients were followed-up by mail or phone annually since 1996. The factors including the sex, age, epileptogenic reasons, epileptic sites, operative fashions, follow-up times were analyzed. Results: The general surgical outcomes: 68.63% (140 of 204) were seizure free (Engel I), and Engel II 14.71% (30 of 204), Engel III 8.82% (18 of 204), Engel IV 7.84% (16 of 204). In male and female patients, the seizure-free ratio was 63.08% and 78.38%, respectively (p < 0.05). In younger patients, the seizure-free percentage was 79.59%, and in older, 58.49%. In 140 cases, abnormal structures were seen in MRI scans, and 91 cases were cured; 64 cases had negative MRI imagines with definite clinical manifestations and positive eletrophysiological examinations, and 49 cases were seizure free. The TLE seizure-free ratio was 73% (73 of 100) and extra-TLE in frontal, parietal, and occipital lobes were 68% (17 of 25), 91.67% (11 of 12), 80% (four of five), respectively, and
Epilepsia, Vol. 44, Suppl. 9, 2003
AES PROCEEDINGS multiple lobes, 56.45% (35 of 62). As for as the follow-up durations: the seizure-free percentage of follow-up for 6–7 years, 5–6 years, 4– 5 years, 3–4 years, 2–3 years, 1–2 years were 63.64% (seven of 11), 100% (all nine), 77.27% (17 of 22), 61.54% (24 of 39), 68.75% (33 of 48), 66.67% (50 of 75), respectively. Six (50%) of 12 cases with purely bipolar coagulation were seizure free, and three (42.86%) of seven cases with corpus callosotomy plus bipolar coagulation and nine cases with hemispherectomy were completely seizure free; 125 (68.31%) of 162 cases with multiple lobe epilepsy were cured. Conclusions: Factors such as sex and age have a little role in determining the prognosis. Patients with secondary epileptogenic reasons have a better outcome than do primary patients, but there is no significant difference. Patients with single cerebral lobe epileptogenic foci have a better result than patients with multiple lobes. Among all patients, hemispherectomy has the best results in controlling the attacks, and combined operations are better than singles.The group of 5- to 6years follow-up durations is better than the others group for having few cases.
1.452 NEUROPSYCHOLOGICAL OUTCOME AFTER SELECTIVE HIPPOCAMPECTOMY WITH TRANSSYLVIAN VERSUS TRANSCORTICAL APPROACH: A RANDOMIZED PROSPECTIVE CLINICAL TRIAL 1 Martin T. Lutz, 1 Christoph Helmstaedter, 2 Hans Clussmann, 2 Johannes Schramm, and 1 Christian E. Elger (1 Department of Epileptology and 2 Department of Neurosurgery, University Clinic, Bonn, Germany) Rationale: Selective amygdalohippocampectomy (SAH) is a surgical treatment option for patients with medically intractable mesial temporal lobe epilepsy. It is performed to limit resection of unaffected tissue while maintaining favorable seizure outcome. In SAH the mesial structures can be approached by different routes, the transsylvian approach (1) and the transcortical approach (2). The former approach bears a greater risk of vascular injury and frontal lobe manipulation, but the advantages or disadvantages with respect to postoperative cognitive outcome are still a matter of debate. Methods: Eighty patients were randomly assigned to one of the two surgical approaches. In 41 patients the transsylvian approach and in 39 patients the transcortical approach was performed. All patients received comprehensive neuropsychological testing of verbal/nonverbal memory, attention, and executive functions before and 7 months after SAH. Results: Seventy-five percent of patients became completely seizure free with no difference depending on the chosen approach. Repeated measurement MANOVA showed that outcomes after both approaches were essentially the same. The only exception was phonemic fluency, which showed a significant improvement after transcortical but not after transsylvian SAH. Conclusions: The results indicate that the surgical approach can be chosen independent of cognitive outcome criteria. Improvement in phonemic fluency after transcortical SAH may reflect selective normalization of cognitive functions after epilepsy surgery. Frontal lobe disturbance during transsylvian SAH might have hindered this process of recovery.
157
Rationale: Holoprosencephaly is usually associated with a diffuse disruption of neuronal proliferation, migration, and cortical organization. Epilepsy surgery is rarely performed. This report evaluates the outcome of hemispherectomy in an infant with holoprosencephaly microsigns. Methods: A 15-month-old girl had global development delay and seizures that started almost immediately after birth. Seizures were described as short cessation of activity followed by blank staring spells, followed by multifocal jerking of arms, legs, and body. She was treated unsuccessfully with phenobarbital, gabapentin, levetiracetam, and lorazepam. Her examination revealed plagiocephaly, right hemiatrophy, and right hemiparesis. Teething and development were delayed. Neuroimaging studies revealed a left hemiphere hypoplasia associated with enlargement of the left ventricle, left periventricular heterotopia, and abnormal gyral patterns on the left. Prolonged video-EEG study supported that all epileptic events were emanating from the left hemisphere. Results: Left hemispherectomy resulted in a complete seizure control. Her first tooth, a single central incisor, appeared at 20 months. Conclusions: This observation illustrates the need to consider epilepsy surgery in patients with refractory partial seizures unresponsive to medical intervention, even when holoprosencephaly microsigns are present.
1.454 SURGICAL OUTCOME AFTER HEMISPHERECTOMY 1,2 Pedro P. Mariani, 1,2 Jose A. Burattini, 1,2 Arthur Cukiert, 1,2 Rodio Brandao, 1,2 Lauro Ceda, 1,2 Leila Frayman, 1,2 Valeria A. Mello, 1,2 Carla Baise, 1,2 Meire Argentoni, 1,2 Cristine M. Baldauf, and 1,2 Cassio R. Forster (1 Neurology and Neurosurgery, Hospital Brigadeiro, and 2 Neurology and Neurosurgery, Clinica de Epilepsia de Sao Paulo, Sao Paulo, SP, Brazil)
1. Wieser HG, Yasargil MG. Selective amygdalohippocampectomy as a Surgical Treatment of Mesiobasal Limbic Epilepsy. Surg Neurol 1982;17:445–7. 2. Olivier A. Transcortical Selective amygdalohippocampectomy in Temporal Lobe Epilepsy. Can J Neurol Sci 2000;27(suppl. 1):S68– 96.
Rationale: Hemispherectomy has been used as treatment modality for refractory epilepsy for many years. More recently, hemispheric epileptic syndromes have been better defined, and the patient population submitted to this type of procedure has become increasingly homogenized. Presently, the main epileptic syndromes submitted to this type of surgery are Rasmussen syndrome, hemiplegic syndromes, hemispheric cortical dysplasia, and Sturge–Weber syndrome. Methods: Thirty-three patients submitted to functional hemispherectomy were studied. Sixteen had Rasmussen syndrome, 13 congenital hemiplegic syndrome, one Sturge–Weber syndrome, and three cortical dysplasia. Surgery consisted of large frontotemporoparietal resection, total callosotomy, and disconnection of the remaining frontal and occipital poles. Thirty patients had contralateral hemiplegia preoperatively. All patients had refractory epilepsy with frequent daily seizures; age ranged from 11 months to 29 years (mean, 8.9 years) and follow-up time from 6 months to 9 years (mean, 3.5 years). Results: Twenty-nine patients have been seizure free after surgery. The other four patients had ≥80% of seizure-frequency improvement. Three patients needed ventriculoperitoneal shunting during late followup. The three patients that were not hemiplegics preoperatively were left with deficit postoperatively (two with Rasmussen syndrome and one with cortical dysplasia); in the others, there was no further neurological deterioration. Thirty-two patients disclosed an aseptic meningitis postoperative syndrome characterized by high fever, meningismus, and preserved neurologic status, with a mean duration of 2 weeks. There was no surgical mortality. Conclusions: Functional hemispherectomy is an effective and safe surgical procedure in selected patient populations. Surgical morbidity and mortality are very low. Special attention should be paid to patients with cortical dysplasia with or without hemimegalencephaly (HME). In those with HME, there should be a compromise between resection and disconnection to avoid unnecessary empty spaces. Patients with cortical dysplasia might bear residual neurologic function within the dysplastic cortex. (Supported by Sao Paulo Secretary of Health.)
1.453 HEMISPHERCTOMY IN AN INFANT WITH HOLOPROSENCEPHALY MICROSIGNS Paul Maertens and Victoria Parada (Neurology, University of South Alabama, Mobile, AL)
1.455 SUPERSELECTIVE, SELECTIVE, AND TAILORED SURGERY FOR TEMPORAL LOBE EPILEPSY: THE ROLE OF SEMIOLOGY, ELECTROCORTICOGRAPHY, AND SODIUM AMOBARBITAL TESTING
REFERENCES
Epilepsia, Vol. 44, Suppl. 9, 2003
158
AES PROCEEDINGS
1 Christopher R. Mascott, 2 Luc Valton, 1 Jean-Christophe Sol, and 3 Tim Frederick (1 Neurosurgery and 2 Neurology, CHU-Toulouse Rangueil, Toulouse, France; and 3 Neurology, Tulane Epilepsy Institute, New Or-
leans, LA) Rationale: In temporal lobe epilepsy, there has been growing evidence that successful surgery for seizure control may involve a tailored approach for each patient. The selection of surgical approach has been made on the basis of extensive invasive recordings by some schools. Other approaches have included extensive awake mapping for speech function and epileptiform activity. We have employed a strategy of only occasional invasive recordings, no awake mapping, electrocorticography, and frequent preoperative sodium amobarbital testing for speech and memory. Methods: We have selected 17 cases where preoperative planning and intraoperative electrocorticography led us to nonstandard tailored selective or supraselective surgical options. The goal was to assess seizure outcome in this patient population with nonstandard surgery. The reasoning underlying a supraselective approach was most often the preservation of verbal memory by subtotal resection of mesiotemporal structures or receptive speech by limiting temporal neocortical resection combined with subpial transection to areas with intraoperative epileptiform activity. For these reasons, the majority of the cases considered here are left temporal cases (14 of 17). A lesion was present in five cases, but none of the cases reported was simple lesionectomy. Results: Surgeries performed included selective hippocampectomy sparing amygdala, amygdalectomy sparing hippocampus, other intentional subtotal mesiotemporal resections, and tailored neocortical resections and/or transactions. All lesions were resected combined with one of the aforementioned strategies. Fifteen of 17 patients were seizure free at 1 year. The two who were not had had prior invasive monitoring, clearly implicating the subsequently operated on temporal lobe in seizure generation but developed different frontal lobe type seizures after surgery. One of the two has been seizure free for >2 years after additional frontal surgery. The other is also a candidate for frontal lobe surgery. Conclusions: We assessed seizure outcome in a small series of patients who had nonstandard temporal lobe surgery. Seizure outcomes would suggest that if an epileptogenic network is sufficiently disrupted, results can be excellent, even with highly selective surgical strategies. It remains crucial to continue to refine strategies that will permit selecting the most minimal procedure that will effectively disrupt a particular epileptogenic network.
1.456 EARLY AND IMMEDIATE POSTOPERATIVE PSYCHIATRIC FINDINGS IN PATIENTS SUBMITTED TO CORTICOAMYGDALOHIPPOCAMPECTOMY FOR REFRACTORY EPILEPSY 1,2 Valeria A. Mello, 1,2 Carla Baise, 1,2 Cristine M. Baldauf, 1,2 Meire Argentoni, 1,2 Cassio R. Forster, 1,2 Leila Frayman, 1,2 Arthur Cukiert, 1,2 Jose A. Burattini, 1,2 Pedro P. Mariani, 1,2 Joaquim O. Vieira, 1,2 Rodio Brandao, and 1,2 Lauro Ceda (1 Neurology and Neurosurgery, Hospital Brigadeiro, and 2 Neurology and Neurosurgery, Clinica de Epilepsia de Sao Paulo, Sao Paulo, SP, Brazil) Rationale: Corticoamygdalohippocampectomy (CAH) is the commonest procedure carried out for the treatment of refractory epilepsy.
Important limbic relays are targeted by the procedure, and psychiatric alterations might be expected after it. We have studied the early and immediate postoperative findings in patients submitted to CAH. Methods: Twenty-nine patients submitted to CAH received psychiatric evaluation 1 week before surgery and 1 week and 1 month after surgery. Psychiatric evaluation included a diagnostic interview (SCID), rating for the Hamilton 21 Depression Scale (HAM21), Brief Psychiatric Evaluation Protocol (BPRS), and Young Mania Scale (EAM-m). Thirteen patients were submitted to left and 16 to right CAH. Results: Overall, nine patients had preoperative psychiatric disease. Seven patients presented significant psychiatric alterations during the studied period (six within the first week and one after 1 month, postoperatively), and three of these had previous psychiatric disease. Two of them have been submitted to left and five to right CAH. One patient developed depression 1 week after left CAH; three patients developed mania 1 week after CAH (two right and one left CAH); one patient developed mania and depression after right CAH; one patient developed psychotic symptoms 1 week after surgery, and one patient developed mania 1 month after surgery. All patients received treatment targeted at the specific psychiatric disturbances. All except one were off medications 1 month after surgery. Conclusions: Almost a quarter of the patients submitted to CAH developed psychiatric alterations during the early and immediate postoperative periods. Right-side CAH seems to be related to a higher psychiatric morbidity during the studied period. These patients often need brief medical therapy for the treatment of the psychiatric disorder, especially patients with hypomania/mania. A larger series might consolidate the statistical analysis. (Supported by Sao Paulo Secretary of Health.)
1.457 TRANSCALLOSAL RESECTION OF HYPOTHALAMIC HAMARTOMA IN THE TREATMENT OF REFRACTORY EPILEPSY 1 Yu-tze Ng, 1 John F. Kerrigan, 1 Harold L. Rekate, 2 Jeffrey V. Rosenfeld, 1 Erin C. Prenger, 1 Margaret R. Varland, and 1 Robert F. Spetzler (1 Neuroscience, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Center, Phoenix, AZ; and 2 Neurosurgery, Alfred Hospital, Melbourne, Victoria, Australia) Rationale: Hypothalamic hamartomas (HHs) are rare developmental malformations of the tuber cinereum that typically result in gelastic seizures followed by the development of other seizure types including tonic, tonic–clonic, and complex partial. The seizures are extremely refractory to antiepileptic drugs. Previously, the resection of these HHs was thought to be either not feasible and/or would not affect the seizures of these patients. We decribe the successful resection of HHs in seven patients with resultant dramatic cessation of their seizures in four (Table 1). Methods: Surgical: The procedure was performed with some modifications using the transcallosal–intraforniceal approach originally described by Rosenfeld et al. (Neurosurgery 2001;48:108–18). Using frameless stereotaxis, the callosotomy is made in the midline ∼3 cm in length. The approach proceeds between the two leaves of the septum pellucidum without retractors placed on the columns of the fornices. At that point, the surgeon has a view of the floor of the third ventricle (hypothalalmus), and the hamartoma is resected. The MRI brain scans show pre- and postoperative images, respectively, in one of the patients.
TABLE 1. The clinical characteristics of the patients Patient 1 2 3 4 5 6 7
Age (yr)
Age at sz onset (yr)
Preop number of sz types
Preop total sz frequency (per day)
% HH resected
Postop number of sz types
Postop total sz frequency (per wk)
8 4 9 4 3 9 18
0.1 0.5 2 0.3 0 0.3 0.1
1 1 4 2 2 4 4
2–3 4–8 11–16 3–6 15–30 16–23 12–44
T >75% 95% T T T T
– – 2 – – 1 2
– – 7 – – 1 3
Sz, seizure; HH, hypothalamic hamartoma; T, total.
Epilepsia, Vol. 44, Suppl. 9, 2003
AES PROCEEDINGS
159
Melinda Nolan, Ryoichi Sakuta, Nathaniel Chuang, Hiroshi Otsubo, James Rutka, O. Carter Snead, Cynthia Hawkins, and Shelly Weiss (Brain and Behaviour Institute, Hospital for Sick Children, Toronto, ON, Canada) Rationale: Dysembryoblastic neuroepithelial tumors (DNTs) are a pathologically distinct group of tumors commonly associated with medically intractable chronic epilepsy and a favorable prognosis after surgical resection. We describe the clinical, MRI, and pathological characteristics and outcome in children after surgical resection of pathologically confirmed DNTs to ascertain prognostic features for seizure-free outcome. Methods: Neurology, neurosurgery, and pathology databases from 1993 to 2002 were used to retrospectively identify children with confirmed DNTs with seizures and a minimum of 12-month follow-up. Clinical, radiologic, and pathological features were examined. Results: Twenty-six children, with mean seizure duration 2.6 years (range, 0–10.4) before surgery, were identified. Mean age at surgery was 10.0 years (4.0–18.0). DNTs were found in the temporal lobe (39%), frontal lobe (31%), parietal lobe (23%), and occipital lobe (8%). Preoperatively, the majority of DNTs (77%) had a typical MRI appearance, with a well-defined, cortically based, multicystic mass. Enhancement was identified in 50%. Residual DNT was evident in 15 of the 22 children where pre- and postoperative MRI was available. Three children demonstrated recurrence of tumor. Pathologically the majority of DNTs had complex glioneuronal features, and some degree of cortical dysplasia was found in surrounding tissue of all tumors where sufficient tissue was available for examination (18 of 26). Seizure outcome was good in 85% of children at 12 months (Engel class I); however, at longer follow-up (mean, 4.3; range, 1.0–14.4 years), only 65% remained seizure free. Seven children had further surgical resection because of continued medically intractable seizures. Risk factors for seizure recurrence (age at onset, duration of seizures preoperatively, MRI appearance of DNT, location of tumor, residual spikes on electrocorticography, residual tumor on MRI postoperatively, pathologic classification of DNT, degree of associated cortical dysplasia) were examined with respect to seizure outcome at 12 months and at long-term follow-up. Only the presence of residual tumor was a significant risk factor for seizure recurrence at long-term follow-up (p = 0.015). Although not statistically significant, all four children with continuing seizures at 12 months had residual DNTs on MRI. Conclusions: Children with DNTs represent a group with surgically treatable epilepsy; however, outcome is not always favorable. Residual tumor is a risk factor for continued or recurrent seizures, and some children may require multiple surgical resections. Cortical dysplasia commonly coexists with DNT. Recurrent tumor can occur despite the lowgrade nature of this neoplasm. Although the majority of children remain seizure free after surgical excision of DNT without recurrence, a considerable number fail to respond or have recurrent seizures after a brief seizure-free period.
(Figs. 1 and 2). Neurologic: The baseline seizure types and frequencies were recorded pre- and postoperatively. Other changes and/or complications were also recorded. Results: The seven patients were followed up for an average of 8.9 weeks. Their average age was 7.9 years and they consisted of five males. Preoperatively, all patients had multiple daily seizures. Four patients are currently seizure free, and three patients have had >90% reduction in seizure frequency. Most patients had a mild, transient, postoperative hyponatremia; one patient had hyperphagia, and another decreased shortterm memory; both are improving. Conclusions: Transcallosal resection of HH is safe and the best treatment for refractory (to medical therapies), mixed seizures that typifies these patients. 1.458 DYSEMBRYOBLASTIC NEUROEPITHELIAL TUMORS IN CHILDHOOD: A STUDY OF LONG-TERM OUTCOME AND PROGNOSTIC FEATURES
1.459 HOW MANY PALLIATIVE SURGICAL PROCEDURES FOR INTRACTABLE EPILEPSY? 1,2 Frank J. Ritter, 1,2 Michael D. Frost, 1,2 Willie T. Anderson, 3 Mary E. Dunn, 1,2 Patricia E. Penovich, and 1,2 John R. Gates (1 Minnesota Epilepsy Group, PA of United Hospital and Children’s Hospital and Clinics-St Paul, St. Paul, 2 Department of Neurology, University of Minnesota, Minneapolis, and 3 Neurological Associates, Ltd, St. Paul, MN) Rationale: Rarely are seizures completely controlled by either corpus callosotomy (CC) or vagal nerve stimulation (VNS). If a patient has had one of these procedures, helpful or not, will the other procedure be of benefit? We reviewed our experience with patients who had both procedures to answer the following questions: Were seizures decreased by either palliative procedure, or by both? Were the number of antiepileptic medications (AEDs) taken by the patient decreased after either procedure? To which therapeutic intervention was seizure-frequency reduction attributed? Was there a subjective improvement in the patient’s quality of life (QOL)? Methods: Records of surgical patients were searched to find those who had both CC and VNS. The medical records were reviewed for demographics, seizure type(s), medication(s), and seizure frequency before and after surgeries. A seizure-frequency decrease of ≥80% was Epilepsia, Vol. 44, Suppl. 9, 2003
160
AES PROCEEDINGS
considered a positive outcome after CC. A decrease of seizures of ≥50% after VNS was considered positive. Results: Twenty patients (13 male/seven female) had both CC & VNS surgeries; 18 of 20 (90%) had Lennox–Gastaut syndrome (LGS) with multiple seizure types, and two had partial secondarily generalized seizures. Before their last surgery, an average 11 AEDs had failed and in 50%, the ketogenic diet failed. In 14 of 20 (70%) a partial (four) or complete (10) CC occurred bofore VNS. Seven of 14 had a positive response to CC before VNS. In six of 14 (42%) the VNS reduced seizures by >50%. Four of six VNS responders had a positive response to previous CC, two did not. After the VNS, four patients had an increase in the number of AEDs, one decrease and nine unchanged. Six of 20 (30%) had VNS before CC. All had LGS, and none had a positive response to VNS. All six had a positive response to CC, two partial CC, four complete CC, three had number of AEDs increased, two decreased and one, no change. Fourteen of 20 patients (or parents/care giver) reported an improvement in QOL after the second surgical procedure. In two patients the improvement in seizure control and QOL was attributed solely to a change in AEDs. The rest attributed these improvements to both surgeries or the most recent procedure. Conclusions: In this refractory group of patients, 90% with LGS, >50% appeared to benefit from a second palliative surgical procedure. Only rarely were AEDs reduced, but a subjective improvement was noted in QOL.
1.460 IQ AND MEMORY SUBTESTS ON PRESURGICAL NEUROPSYCHOLOGICAL EVALUATION DO NOT PREDICT SEIZURE OUTCOME IN MESIAL TEMPORAL LOBE EPILEPSY SURGERY Sara R. Rosset, Karinne O. Rezek, Erica R. Coimbra, Roger Walz, Vera C. Terra-Bustamante, Veriano Alexandre, Jr., Tonicarlo R. Velasco, Joao A. Assirati, Jr., Carlos G. Carlotti, Jr., Antonio C. Santos, and Americo C. Sakamoto (Neurology, Psychiatry and Clinical Psychology, University of S˜ao Paulo, Ribeir˜ao Preto, S˜ao Paulo, Brazil) Rationale: Mesial temporal lobe epilepsy related to hippocampal sclerosis (MTLE-HS) is the most common surgically remediable epileptic syndrome. Anterior and mesial temporal lobectomy is a safe and effective treatment for medically intractable MTLE-HS. The literature regarding prognostic factors for surgical outcome of MTLE-HS is often contradictory. We investigated the predictive value of the presurgical neuropsychological performance in the postoperative outcome of MLTE-HS patients. Methods: We included 191 consecutive patients surgically treated for MTLE-HS between 1995 and 2000, 93 males, 98 females, mean age of 34.9 years. According to the surgical outcome, patients were classified in two groups: (a) seizure-free group (Engel class I); and (b) non–seizurefree group (Engel II to IV). The clinical, demographic, neuroimaging, neurophysiologic, and cognitive performance (IQ, Visual Reproduction I, Visual Reproduction II–30 min, Logical Memory I, and Logical Memory II–30 min) were analyzed. Results: There were no differences among the three tests and postsurgical seizure outcome (p > 0.5). Univariate analysis indicated that patients with lower RVLII disclosed higher risk to persist with seizures (OR, 2.63; CI 95%, 1.195–5.81; p = 0.14). After the multiple logistic regression analysis adjusted for clinical, neuroimaging, and neurophysiologic imbalances, this lower performance in the RVLII did not show significant association with postsurgical seizure outcome (OR, 2.69; CI 95%, 0.78–9.29; p = 0.12). Conclusions: Cognitive performance in the presurgical neuropsychological tests did not show association with postsurgical seizure outcome in a consecutive group of MTLE-HS. (Supported by Apoio FAPESP, FAEPA, CNPq.)
1.461 NEUROPATHOLOGIC FINDINGS IN DYSEMBRYOPLASTIC NEUROEPITHELIAL TUMOR (DNT): CORRELATION WITH ADJACENT CORTICAL DYSPLASIA 1,5 Ryoichi Sakuta, 2 Hitoshi Otsubo, 2 Melinda A. Nolan, 2 Shelly K. Weiss, 1 Cynthia Hawkins, 3 James T. Rutka, 4 Nathaniel A. Chuang,
Epilepsia, Vol. 44, Suppl. 9, 2003
4 Sylvester H. Chuang, and 2 O. Carter Snead (1 Division of Pathology, DPLM, 2 Divisions of Neurology, 3 Neurosurgery, and 4 Diagnostic Imaging, The Hospital for Sick Children, University of Toronto, Toronto, ON, Canada; and 5 Department of Pediatrics, Koshigaya Hospital, Dokkyo University School of Medicine, Koshigaya, Saitama, Japan)
Rationale: The present study intends to identify factors of pathologic features that predict postoperative clinical outcome in children with seizures secondary to dysembryoplastic neuroepithelial tumor (DNT). Methods: We retrospectively reviewed pathologies of patients who underwent epilepsy surgery between 1993 and 2002 at The Hospital for Sick Children, Toronto. We reviewed clinical, radiologic, and neuropathologic features. Results: There were 13 males and 13 females. Mean age at onset of seizures was 7 years (range, 2–13 years). The mean age at surgery was 10 years (4–18 years). Fifteen patients experienced complex partial seizures; seven patients, complex partial seizures with secondarily generalized seizures; three patients, simple partial seizures; and one patient with tuberous sclerosis who had infantile spasms developed partial seizures. Nine tumors were located in the temporal lobe, nine in frontal lobe, six in parietal lobe, one in occipital lobe, and one in frontoparietal lobe. Histopathologic characteristics of DNTs include the specific glioneuronal elements in all patients. The main pathologic type of DNT was complex in 18 patients (69%), and DNT accompanied with adjacent cortical dysplasia in 17 patients (65%). In the 17 patients with DNT with adjacent cortical dysplasia, seven were Palmini’s classification grade 1, eight were grade 2, and two were grade 3. Postsurgical follow-up ranged from 1 to 14 years (mean, 4 years): 17 patients were seizure free and nine have had seizures including eight patients with CD. There was a statistical siginificance between seven of 17 patients (41%) with DNT and CD requiring further surgeries due to recurrent intractable seizures, and none of nine patients without CD requiring second surgery. There was no correlation between existence and grades of CD and postsurgical seizure outcomes. Conclusions: Children with DNT and CD often had recurrent seizures and required further epilepsy surgery. Adjacent CD may play a role of epileptogenesis in DNT, thus the appropriate resection of DNT and adjacent CD should be considered.
1.462 TEMPORAL LOBE EPILEPSY SURGERY: ANALYSIS OF FAILURES AND THE ROLE OF REOPERATION Vicenta Salanova, Omkar Markand, and Robert Worth (Neurology, Indiana University, Indianapolis, IN) Rationale: Temporal lobe epilepsy surgery (TLES) benefits >80% of patients, and normalizes the late mortality rate in patients who become seizure free (Salanova et al. Epilepsia 2002). Few centers have analyzed patients who continue to have seizures. Our purpose was to analyze the reason for surgical failures and the role of reoperation and to compare these patients with seizure-free patients, to see if there were significant differences between the groups. Methods: From 1984 to 2002, 262 patients with TLE underwent surgery after a comprehensive presurgical evaluation. There was no surgical mortality. Patients were followed up at 3, 6, and 12 months and yearly therafter; 65% became seizure free (class I), 19% had rare seizures (class II), and 16% continued to have seizures (class III, IV). Patients with persistent seizures (class III, IV) underwent reevaluation to include prolonged video-EEG recordings and MRIs. The clinical history, seizure type, postsurgical MRI, pathology, types of resections, outcome, the reasons for failure, and the late mortality rate were analyzed and compared with those patients who became seizure free. Results: Analysis of failures (41 pts, class III, IV): Mean age at seizure onset was 11 years, mean age at surgery, 26.6 years; and mean duration of epilepsy, 15.2 years; 12% had febrile seizures (FSs), 29% head trauma, and 7% encephalitis. 56% had epigastric or experiential auras, 52.5% had abnormal imaging, 34% bitemporal interictal epileptiform discharges, and 20% posterior temporal localization; 84% had a pathological diagnosis, and 18% had dual pathology. Postsurgical MRI showed residual posterior mesial temporal structures (PMTS) in 26 of 30 (86.6%), residual PMTS and posterior temporal lesions (PTL) in two (6.6%), and PTLs in two (6.6%). The recorded temporal lobe seizures were localized to the
AES PROCEEDINGS side of the previous resection; 21 patients underwent reoperation (11 on the left); 14 had resection of the posterior MTS, five of the posterior MTS and basal posterior temporal cortex, and two of the posterior MTS and PTLs. There was no surgical mortality, but 57% became seizure free, and 24% had rare seizures. We compared these patients with those seizure free and found that 45% of seizure-free patients had FS, 12% had head trauma, and 70% had abnormal imaging. Fewer seizure-free patients had invasive recordings, and the late mortality rate was lower than in those patients with persistent seizures. Conclusions: When compared with seizure-free patients, patients who failed temporal lobe epilepsy surgery were less likely to have a history of FS and abnormal imaging, and more likely to have a history of head trauma, encephalitis, and posterior temporal localization, suggesting larger epileptogenic zones; 57% became seizure free after resection of the posterior MTS and posterior basal temporal cortex. Patients who fail surgery should be reevaluated for reoperation. 1.463 ATTENTION IN CHILDREN WITH SEVERE DRUGRESISTANT EPILEPSY 1,2 Rocio Sanchez-Carpintero, 3 Elizabeth B. Isaacs, 4 Ingram Wright, 1 William F.J. Harkness, and 1 Brian G.R. Neville (1 Neurosciences Unit, Institute of Child Health and Great Ormond Street Hospital, London, United Kingdom; 2 Pediatric Neurology Department, Clinica Universitaria de Navarra, Pamplona, Navarra, Spain; and 3 MRC Childhood Nutrition Research Centre, and 4 Cognitive Neurosciences Unit, Institute of Child Health, London, United Kingdom) Rationale: Children with severe drug-resistant epilepsy (SDRE) are at risk of cognitive impairments. Attention may play a role in those impairments, as it has been found defective both in children with benign rolandic epilepsy and complex partial seizures. Our aim is to explore selective, sustained, and divided attention in two groups of children with SDRE, treated either medically or surgically. Attention in children with SDRE is compared with attention in the general population, and relations between attention and epilepsy variables studied. The effect of surgical control of epilepsy on attention is explored. Methods: The 28 children with SDRE were tested in three specific aspects of attention with standardised measures at baseline and at 7 months’ follow-up. Ethical approval was obtained. Test of Everyday Attention for Children (TeaCh) and Faces was used to explore the different aspects of attention. Fourteen of the subjects had a substantial seizure reduction by epilepsy surgery immediately after baseline assessment. Fourteen did not have surgery, acting as controls. The pooled attention of the 28 patients at baseline was tested against the normal population. The relations between attention and seizure frequency, age at first seizure, number of antiepileptic drugs, past history of convulsive status epilepticus, right hemisphere EEG involvement, and frontal lobe involvement were studied by using correlations. Repeated measures ANOVA was used to compare performance on attention at baseline and at follow-up in both surgical and control groups. Results: Children with SDRE showed deficits in selective (TeaCh) [mean, 6.1 (3.0); p < 0.001] and divided attention (TeaCh) [mean, 5.5 (4.1); p < 0.001] in comparison to the normal population. IQ was significantly correlated with attention, but no other variables were related to attention measures. At follow-up, children who had epilepsy surgery tended to improve in Faces 3 min [F(1, 25) = 4.0; p = 0.057], and improved in Faces 6 min [F(1, 24) = 5.4; p = 0.03] in comparison to controls. In these two measures the pooled 28 subjects did not show impairment when compared with the general population. Conclusions: Children with SDRE have impaired selective and divided attention in comparison to the normal population. This is probably related to the cause of the epilepsy rather than to the effect of seizures alone. The attention impairments are partly dependent on IQ. It is difficult to separate the primary pathology and the effects of early-onset epilepsy as causes of attentional difficulties, but seizure relief by surgery does not invariably resolve these problems. Nevertheless, children improve in the attentional tasks they were good at before, suggesting that learning processes might be facilitated by epilepsy control in tasks where there is not previous impairment. [Supported by a grant (Ex 2001 33419023) from the Ministry of Education, University General Directorate, Spanish Government.]
161
1.464 POSTOPERATIVE INDEPENDENT LIVING, WORKING, AND DRIVING AFTER A TEMPORAL LOBECTOMY FOR MEDICALLY REFRACTORY EPILEPSY 1 Howard W. Schacht, 1,2 John R. Gates, 1,2 Patricia E. Penovich, and 1,2 Deanna L. Dickens (1 Minnesota Epilepsy Group, P.A. of United Hospital and Children’s Hospitals and Clinics-St. Paul, St. Paul, and 2 Department of Neurology, University of Minnesota, Minneapolis, MN) Rationale: The purpose of the surgical intervention for epilepsy treatment for medically refractory temporal lobe epilepsy is to render the patients seizure free, with the clear implication that they can now drive, as well as continue as or become productive members of society by being gainfully employed and living independently. This study was to obtain a long-term postoperative assessment of success in the areas of driving, employment, and independent living. This study involved patients who had received temporal lobectomies at the Minnesota Epilepsy Group between 1986 and 2001 and still actively being followed. Methods: The study consisted of 44 patients, 21 males, 23 females, all of whom had received a temporal lobectomy between 1986 and 2001. Patients were sequentially selected who have met this fairly simple criterion of a limited temporal lobectomy of the right or left (X, right; Y, left) within that time. Demographic, seizures, and medical data were retrospectfully collected from both groups. Patients were classified as being seizure free, which could include simple partial seizures, or having persistent complex partial or tonic–clonic seizures. Results: We documented the number of patients driving, living independently, and working before surgery and compared that with their seizure outcome and their conditions after the procedure. We concluded that the patients who worked and lived independently before surgery continued working and living independently. Those who did not work or live independently before surgery were able to achieve this state with a few exceptions described later. Before surgery, 13 patients were driving; after surgery, 24 were driving. Of the remaining 20 patients, five were not driving because of seizure frequency, eight because of cognitive impairment, one because of poor vision, four because of the expense or desire to drive, and two were unknown. Thirty-one patients were working before surgery and 34 after surgery. Of the remaining 10 patients, only two were not working due to seizure activity, one due to memory impairment, one due to inability to handle pressure, five with mental impairment, and one unknown. Conclusions: Temporal lobectomy does result in long-term improvement in independent living, gainful employment, and driving. These results appear to have been sustained for 3–17 years.
1.465 TEMPORAL LOBE EPILEPSY SURGERY WITH LIMITED RESOURCES IN TWO CENTERS OF ARGENTINA 1 Walter H. Silva, 4 Roberto Giobellina, 1 Damian Consalvo, 1 Patricia Solis, 1 Pablo Salgado, 1 Brenda Giagante, 1 Silvia Oddo, 4 Viviana Abatedaga, 1 Luciana D’Alessio, 1 Estela Centurion, 1 Patricia Saidon, 3 Alejandra Rabadan, 2 Ricardo Vazquez, 2 Eduardo Seoane, and 1 Silvia Kochen (1 Epilepsy Center, Department of Neurology, Hospital Ramos Mejia.CONICET-CEFYBO-UBA, 2 Department of Neurosurgery, Hospital Ramos Mejia, and 3 Department of Neurosurgery, Italiano Hospital, Buenos Aires, and 4 Epilepsy Center Argentina, Lennox Fundation, Cordoba, Cordoba, Argentina) Rationale: Surgical treatment seems to be the best option for patients with refractory symptomatic temporal lobe epilepsy (TLE). The aim of this study was to analyze the surgical results of two epilepsy centers of Argentine operated between October 1996 to March 2002. Methods: We selected 42 patients who were operated on because of a diagnosis of medically intractable TLE and who had ≥1 year of postsurgical follow-up. All the patients were evaluated using a multidisciplinary approach that include a complete medical and neurologic history, outpatient EEG, MRI of the brain, and video-EEG. Neuropsychological tests were performed in 37 (88%) and intracarotid amobarbital test in only one patient. Deep electrodes were implanted in two patients. Seizure outcome was assessed using Engel’s classification. Results: There were 24 females and 18 males; mean age, 35 years. In 40 patients, an anterior temporal lobectomy (ATL) was performed,
Epilepsia, Vol. 44, Suppl. 9, 2003
162
AES PROCEEDINGS
and in two patients, a lesionectomy was done. Twenty-six patients were operated on on the right side and 16 on the left side. The histopathologic findings showed a low-grade tumor in six patients, hippocampal sclerosis in 30, dual pathology in two, cavernous angiomas in one patient, and neuronal migration disorder in two. The mean follow-up period was 2.5 years (range, 1–4 years). Thirty-two patients (78%) were in class I, four patients were in class II, four patients were in class III, and two patients were in class IV. Conclusions: Our results showed a postoperative outcome comparable with series of developed countries. The number of patients that were operated on during the period of analysis was significantly lower than the potential candidates. This finding is of great significance for the creation of epilepsy surgery programs in developing countries
1.466 NEUROPSYCHOLOGICAL OUTCOME TWO TO FOUR YEARS AFTER EPILEPSY SURGERY IN CHILDREN AND ADOLESCENTS 1,2 Mary Lou Smith, 2 Irene M. Elliott, and 3 Lucyna Lach (1 Psychology, University of Toronto at Mississauga, Mississauga, and 2 Neurology, Hospital for Sick Children, Toronto, ON, and 3 Social Work, McGill University, Montreal, QC, Canada) Rationale: The purpose of this study was to investigate the neuropsychological outcomes of epilepsy surgery in children and adolescents, using a controlled, prospective design. We have previously reported that children who undergo epilepsy surgery do not differ from controls with intractable epilepsy 1 year after surgery. In this phase of the research we extended the follow-up period to 2–4 years after surgery to address the question of whether an advantage in neuropsychological function may emerge over time. Methods: Two groups of children with medically refractory epilepsy were studied at three points in time. The surgical group (n = 23) underwent a baseline assessment before surgery and at 1 and 2–4 years postoperatively. A comparison group (n = 15) was studied at comparable points in time. At baseline, the groups were comparable in age, age at seizure onset, seizure frequency, number of AEDs, sex, handedness, and IQ. Intelligence, verbal memory, visual memory, visual sustained attention, and academic skills were examined by using objective, standardized tasks. Results: Multivariate analyses of variance (with Group and Time as independent variables) were conducted on four categories of neuropsychological function: intelligence, memory, attention, and academic skills. A significant effect was found only for the academic area, and was an effect of time [F(2, 210) = 6.62; p < 0.002]; over time, the standard scores on the academic measures declined for both groups. Within the surgical group, there were no differences in test performance relating to seizure outcome. Conclusions: In this longitudinal investigation, changes in neuropsychological performance were not found to be associated with either surgical status or seizure outcome. Academic scores declined in both the surgical and the comparison group; findings suggest a slowed rate of acquisition of skills over time relative to the normative population. Benefits of surgery and seizure control are not apparent in these aspects of neuropsychological function within the first 2–4 years after surgery. (Supported by The Ontario Mental Health Foundation.)
1.467 CHANGE IN SEIZURE TYPE AFTER ANTERIOR TEMPORAL LOBECTOMY FOR REFRACTORY EPILEPSY Michael R. Sperling, Cornelius Robens, and Joseph I. Tracy (Neurology, Thomas Jefferson University, Philadelphia, PA) Rationale: It has been reported that secondarily generalized tonic– clonic seizures (GTCSs) become more frequent after anterior temporal lobectomy (ATL). We tried to replicate these findings in a larger patient sample, further investigate postoperative changes in seizure severity, and evaluate risk factors for postoperative occurrence of GTCSs.
Epilepsia, Vol. 44, Suppl. 9, 2003
Methods: All clinical data were maintained prospectively in an electronic database. Seizure type and frequency were registered for patients before and after ATL for refractory epilepsy. The following variables were assessed to determine if they influenced likelihood of occurrence of GTCSs after ATL: age at time of surgery, gender, age at seizure onset, history of febrile seizures, latency in years between risk factor for epilepsy and onset of seizures, contralateral memory function as measured by the Wada test, full-scale IQ, preop seizure frequency, preop seizure type, and presence or absence of preoperative GTCSs. Nonparametric analyses and parametric statistics were used as appropriate. This project was approved by the Jefferson IRB. Results: Of 464 patients who underwent ATL between 1986 and 2001, 146 patients had one or more postoperative complex partial seizures (CPSs) or GTCSs. Among the 146 patients who experienced seizure recurrence, the overall prevalence of GTCSs did not change: 68 patients (46.58%) had preop GTCSs, whereas 73 patients (50%) had postop GTCS (p = 0.74). However, seizure type did change in individual patients; only 86 patients (58.90%) had identical seizure types before and after surgery. Of 78 patients with only CPSs before surgery, 27 (34.62%) developed GTCSs after surgery. Of 68 patients who had GTCSs before surgery, 22 (32.35%) had only CPSs after surgery. GTCSs comprised a greater proportion of the total number of seizures before ATL (18.08%), than after surgery (35.89%) (dependent samples t test, p < 0.01). Preop seizure type was the best predictor of postop seizure type, and the presence of preop GTCSs and a greater preop GTCS frequency were the only predictors of postop GTCS occurrence; no other risk factors examined influenced postop GTCS occurrence. Conclusions: Seizure severity may change after ATL, both for the better and the worse in individual patients. Although overall GTCS prevalence does not increase, there is an increase in the proportion of GTCS after surgery. This could reflect greater efficacy in reducing CPSs than GTCSs by ATL, but also suggests that intrinsic mechanisms for regulating seizure propagation are altered by surgery. Both local and distant circuits regulating excitability may be affected. (Supported in part by the NIH.)
1.468 NONLESIONAL ORBITOFRONTAL EPILEPSY: CLINICAL FEATURES AND RESULTS OF RESECTIVE EPILEPSY SURGERY 1 Thaddeus S. Walczak, 1 James R. White, 1 Teresa A. Tran, 1 Ilo E. Leppik, 2 Robert E. Maxwell, 1 Jeanne L. Beattie, and 1 Robert J. Gumnit (1 Neurology, MINCEP Epilepsy Care, and 2 Neurosurgery, University of Minnesota, Minneapolis, MN) Rationale: There is almost no information regarding clinical and seizure features or results of resective epilepsy surgery in nonlesional orbitofrontal epilepsy. Methods: Case series collected from 95 patients undergoing intracranial EEG recording before resective epilepsy surgery. Patients included if (a) high-quality MRI showed no lesions; (b) intracranial electrode array sampled temporal neocortical (TN), temporal mesiobasal (MB), orbitofrontal (OF), and lateral frontal regions; and (c) ictal EEG onset preceded clinical onset and involved OF cortex in most seizures. Clinical features and surgical outcome determined from medical record review and phone follow-up. Intracranial EEG reviewed. Results: Six patients (three males) identified. Mean age at seizure onset was 9 years. Two had nonspecific auras. All had complex partial seizures; in four automatisms suggested frontal lobe involvement, in two temporal lobe involvement. None had tonic–clonic seizures. Four had at least weekly seizures, and two had daily seizures at presentation. With scalp interictal EEG, four of six had inferior frontal-anterior temporal, five of six sphenoidal, and one of six midtemporal discharges. Intracarotid amytal test found evidence of hippocampal dysfunction in four of six. Thirty-five seizures were recorded with intracranial EEG (4–12/case). Sixteen emerged from OF regions alone (four of six cases), nine from OF and TN regions simultaneously (four of six cases), four from OF and MB regions simultaneously (two of six cases), five from TN alone (three of six cases), and one from MB regions. EEG onset preceded clinical onset by 2–154 s (median, 7 s) in 33 of 35 seizures. In seizures with OF onset, propagation to TN occurred in 1–23 s (median, 5 s),
AES PROCEEDINGS always before spread to MB regions. One patient could not have surgery because seizure-onset area overlapped with language cortex. One patient underwent OF resection, four had multilobar resections including OF, TN, and MB regions based on extent of seizure-onset area. Four have been seizure free since surgery, one has rare seizures (mean duration of follow-up, 20 months; range, 2–57 months). The patient with persistent seizures had OF resection alone. No complications or neurologic sequellae were seen. Conclusions: Patients with nonlesional OF epilepsy often have seizure-onset areas involving both OF and TN regions either individually or simultaneously. Seizure-onset areas and other features indicate that the epileptogenic lesion involves both OF and TN areas. Nontheless, aggressive intracranial recording can define seizure-onset areas adequately. Multilobar resection is safe and very effective. (Supported by MINCEP Epilepsy Care.)
1.469 SURGICAL OUTCOME AND PROGNOSTIC FACTORS IN MEDICALLY INTRACTABLE NEOCORTICAL EPILEPSY 1 Chang-Ho Yun, 2 Sang Kun Lee, 3 Chun-Ki Chung, and 2 Kwang-Ki Kim (1 Neurology, College of Medicine, Inha University, Inchon, and 2 Neurology and 3 Neurosurgery, College of Medicine, Seoul National University, Seoul, Republic of Korea) Rationale: Majority of previous reports on the surgical outcome and its predictors in medically intractable epilepsy had focused on medial temporal lobe epilepsy. We performed this study to evaluate the postoperative outcome and the prognostic significances of clinical and demographic factors, results of presurgical evaluations, and pathologic findings in medically intractable neocortical epilepsy (NE). Methods: Among the patients operatively managed between 1995 and 2001, 193 patients with NE (126 male, 67 female; age at surgery, 26.7 ± 7.8 years) and sufficient postoperative follow-up (>1 year) were included. The primary outcome variable was the status of patients in the last postoperative year, classified as either seizure free or not free. Demographic, clinical, electroencephalographic, MRI, PET, SPECT, and pathologic data were analyzed. Results: One hundred eleven (57.5%) patients were free of seizure. Multifocal ictal onset (p = 0.001), extratemporal ictal focus (p = 0.003), absence of epileptogenic lesion on MRI (p = 0.003), nonlocalized ictalonset pattern on scalp EEG (p = 0.02), nonlocalizable finding in PET (p = 0.003), and pathologic diagnosis as malformations of cortical development predict poor surgical outcome with univariate analysis. By using multivariate analysis, the findings of MRI, PET, and ictal scalp EEG were the significant predictors of surgical outcome. Conclusions: Moderate proportions of surgically remedied NE patients remained seizure free. Presence of epileptogenic lesion on MRI and localizing patterns in PET and ictal scalp EEG are factors associated with good postsurgical outcome.
December 8, 2003 Platform Session A: Pediatrics 3:30 p.m.–5:30 p.m. A.01 NLSTEPSS: A POPULATION-BASED STUDY ON CONVULSIVE STATUS EPILEPTICUS IN CHILDHOOD 1,2 Richard F.M. Chin, 1 Brian G.R. Neville, 2 Helen Bedford, 2 Angie Wade, 2 Catherine Peckham, and 1,3 Rod C. Scott (1 Neurosciences Unit; 2 The Centre for Paediatric Epidemiology; and 3 Radiology and Physics Unit, Institute of Child Health, London, United Kingdom) Rationale: Status epilepticus (SE) is the most common neurologic emergency in childhood. Population-based estimates of the incidence, aetiology, treatment, and natural history of SE are required for appropriate allocation of resources for management of SE. There are few population-based studies on SE, and all such studies have been primarily or exclusively based on adult populations. Thus, a population-based study that primarily addresses SE in a paediatric population is required.
163
The North London convulsive STatus EPilepticus in childhood Surveillance Study (NLSTEPSS), an ongoing study, is such a population-based study. Methods: In NLSTEPSS, all children aged 29 days to 15 years with convulsive status epilepticus (CSE) within North London are identified by using a multitiered notification system involving a research collaborative network of paediatricians from all hospitals in North London. Details of the event are obtained within 2 weeks of notification. Incident cases are defined as lifetime first episodes of CSE. Occurrence is defined as the total number of cases/100,000 children/year. The childhood population of North London, derived from Census 2001, was the denominator in incidence and occurrence estimations. SE is classified as prolonged febrile convulsion (PFC), acute symptomatic, remote symptomatic, idiopathic/cryptogenic, or unclassified. PFC is defined as SE during a febrile (>38◦ C) illness in a previously neurologically normal child, aged between 6 months and 5 years, in the absence of CNS infection. In contrast, SE is classified as acute symptomatic if it occurs in a previously neurologically normal child within a week of a defined acute neurologic insult. Results: The 148 (median age, 3.0 years; range, 0.08–15.9 years; 72 boys) episodes have been identified. The occurrence is 33/100,000 children/year [95% confidence interval (CI), 27–38/100,000 children/year]. Relative risk (RR) of occurrence was 2.7 for nonwhite compared with white children. The 79 (67%; median age, 2.0 years; range, 0.08–15.9 years; 44 boys) were incident (incidence, 18/100,000 children/year; 95% CI, 14–22/100,000 children/year). Of children, 78% were younger than 5 years, and 49% were younger than 1 year; 27 (35%) had PFC, 10 (13%) had acute symptomatic, 29 (36%) remote symptomatic, eight (10%) idiopathic/cryptogenic, and five (6%) were unclassified SE. During hospitalisation, four (5%) died. Three had meningitis, and one had glutaric aciduria type I. Intensive care was required in 36 (46%) cases. Conclusions: The incidence of CSE in childhood is ∼18/100,000 children/year and is most common in the younger age range. Nonwhite children are at greater risk of CSE than are white children. The reason for this is unclear, but data on the effect of socioeconomic status are being analysed. PFC is a significant cause for CSE in childhood, accounting for a third of incident cases. Fatality during hospitalisation is low but is most likely to be associated with an acute symptomatic cause. (Supported by The Wellcome Trust.)
A.02 IS INTRAOPERATIVE ELECTROCORTICOGRAPHY RELIABLE IN CHILDREN WITH INTRACTABLE NEOCORTICAL EPILEPSY? Eishi Asano, Krisztina Benedek, Aashit Shah, Csaba Juhasz, Otto Muzik, Diane C. Chugani, Sandeep Sood, and Harry T. Chugani (Departments of Pediatrics, Neurology, Radiology, and Neurosurgery, Children’s Hospital of Michigan, Wayne State University, Detroit, MI) Rationale: To study the relation between spike frequencies during electrocorticography (ECoG) under general anesthesia and prolonged subdural EEG recording in children with intractable neocortical epilepsy. Methods: Thirteen children (aged 1.9–15.5 years) underwent a 10min intraoperative ECoG and prolonged subdural EEG recording, using 64- to 120-channel subdural electrodes. During ECoG, isoflurane was maintained at 0.5–1.0% to minimize the effect on spiking. With a spikedetection program, the spike frequency during ECoG as well as during prolonged subdural EEG was determined for each channel. The spike frequency for ECoG was compared with prolonged subdural EEG for each patient (Wilcoxon signed-ranks test). The spatial pattern of spike frequency during ECoG was compared with prolonged subdural EEG (Spearman’s rank correlation). The relation between the most frequently spiking electrode on ECoG and the ictal-onset zone was analyzed in cases in which at least three seizures were captured during prolonged subdural EEG. Results: In 10 of 13 patients, the spike frequency was lower during ECoG than during prolonged subdural EEG (mean z = −6.1; p < 0.001). In one patient, there was no significant difference in the spike frequency between ECoG and prolonged subdural EEG. In the other two patients, there were very few spikes during both ECoG and prolonged subdural EEG, and appropriate comparison of spike frequency was not possible. A
Epilepsia, Vol. 44, Suppl. 9, 2003
164
AES PROCEEDINGS
positive correlation was seen between the spike frequency patterns during ECoG and prolonged subdural EEG (mean rho = 0.60; p < 0.001) in nine cases, all of which exhibited frequent spikes (>3/min) at least in one electrode during ECoG. In contrast, there was no correlation between the spike-frequency pattern during ECoG and prolonged subdural EEG (mean rho = 0.08) in the other four cases, all of which showed rare spikes (3/min) had the most frequently spiking electrode a sa part of ictal onset, whereas all four patients with rare spikes (3/min) are seen in at least one electrode during ECoG. Rare spikes (48 h in 15 (71%) of 21, and six infants continued to have ENSs.
Conclusions: ENSs, an important early marker of acute encephalopathy after newborn heart surgery, were common in a large, contemporary cohort of infants. The burden of ENSs varied widely, but many experienced numerous seizures. Seizures are a candidate outcome end point in future neuroprotection trials in this patient population (Fig. 1). (Supported by The Fannie E. Rippel Foundation and the American Heart Association.)
AES PROCEEDINGS A.05 PRENATAL CONTRIBUTIONS TO NEONATAL SEIZURES: IMPLICATIONS FOR EPILEPSY RISK 1,2 Mark S. Scher (1 Pediatric Neurology, Rainbow Babies and Children’s Hospital, and 2 Pediatrics and Neurology, Case Western Reserve University, Cleveland, OH) Rationale: Neonatal seizures (NSs) are surrogate markers that carry a high risk for neurologic morbidity, including epilepsy, and occur either as part of an evolving newborn encephalopathy (NE) from birth or as a new clinical sign (non-NE) after an asymptomatic immediate postpartum period. Methods: Two-hundred thirty-six neonates [120 preterm, mean gestational age (GA), 31.5 weeks] had EEG-confirmed seizures with or without clinical accompaniments over a 13-year period (1983–1996). Results: One hundred seven (45%) infants had seizures without NE compared with a cohort of 129 who expressed abnormal tone and arousal from birth consistent with NE, only five of whom had intrapartum asphyxia that explained NSs, with NE based on fetal surveillance and neonatal assessments. Prenatal conditions, particularly during the third trimester, contributed to NSs with or without NE, which included maternal diseases such as hypertensive disorders of pregnancy, vasculopathies, and genitourinary infections; placental diseases such as fetal vascular thrombosis, chorioamnionitis/funisitis, villus dysmaturity and infarction; and fetal conditions such as multiple gestation, intrauterine growth restriction, hydrops fetalis, and fetal stroke. Neonatal illnesses also contributed to NSs with or without NE including CNS infections, acute intracranial hemorrhage, generalized sepsis, and persistent pulmonary hypertension of the newborn. Only 22 (9%) newborns (15 preterm) had congenital brain malformations and/or inborn errors of metabolism that were ascertained to be the principal etiologies for NSs. Seventy-two (31%) infants died (51 preterm), and 37% of survivors had recurrent seizures during childhood (age 1–13 years). Conclusions: A recent multidisciplinary task force between obstetrics and pediatrics recommended that antepartum factors should be considered more often than intrapartum factors to explain the association between NE and later cerebral palsy. Maternal, placental, and fetal conditions, particularly during the third trimester, should also be considered by epileptologists for patients with a history of NSs with or without NE, because they are at a higher risk for particularly localization-related epilepsy during childhood. (Supported by NS34508.)
A.06 HIPPOCAMPAL ABNORMALITIES AFTER PROLONGED FEBRILE CONVULSION ARE REVERSIBLE: A LONGITUDINAL MRI STUDY 1,2 Rod C. Scott, 2 David G. Gadian, 1 Brian fnm>G. Neville, and 2 Alan Connelly (1 Neurosciences Unit, and 2 Radiology and Physics Unit, Institute of Child Health, London, United Kingdom) Rationale: Prolonged febrile convulsion (PFC) is the most common form of status epilepticus in childhood and may have a causative relation with mesial temporal sclerosis (MTS), the most common structural lesion identified in patients requiring surgery in the treatment of epilepsy. Our recently published data on children investigated within 5 days of a PFC is suggestive of hippocampal oedema [large hippocampal volume (HCV) and prolonged T2 relaxation time (T2 )] that is resolving over a 5-day period, suggesting that PFC can injure the hippocampus. However, those data do not rule out the possibility of a preexisting lesion such as hippocampal dysgenesis. Thus, the aims of the current follow-up study are to determine whether HCV and T2 have changed 4–8 months after a PFC, to determine whether follow-up HCV and T2 are different in patients and control subjects, and to determine whether HCV or T2 side-to-side asymmetry has increased. Methods: Patients were requested to have follow-up magnetic resonance (MR) investigations 4–8 months after their initial scan. They were assessed for evidence of seizures, developmental delay, or behavioural abnormalities. Subsequently, MR investigations including a FLASH 3-D dataset for HCV measurement and T2 mapping were obtained. Pairedsample t tests and general linear modeling, after adjustment for age and intracranial volume, were used for statistical analysis. Hippocampal
165
asymmetry was characterised by an asymmetry index (AI). HCV, T2 , or AI that falls outside the 95% prediction limits for control subjects is considered abnormal. Results: The 14 patients (14–31 months) have had follow-up investigations. The second scan was carried out a mean of 5.5 months after the immediate scan. The 14 age-matched controls had HCV, and 15 age-matched controls had T2 . Of the 14 patients, four have had further seizures. Two had short febrile convulsions, one had prolonged febrile convulsions (PFCs), and one had nonfebrile seizures. There was a significant reduction in HCV (p = 0.002) and T2 (p = 0.018) between the first and second investigations. There is now no difference in HCV (p = 0.91) or T2 (p = 0.99) in patients when compared with controls. There is an increase in HCV asymmetry in patients at follow-up (p = 0.044). Five of 14 patients had an abnormal AI for HCV. Three had one hippocampus outside the lower 95% prediction limit for control subjects. Conclusions: A bilateral reduction in HCV and T2 between the first and second scans confirms that the original findings are temporary and are more consistent with hippocampal oedema than with a preexisting abnormality. The change in hippocampal symmetry in the patient group is consistent with injury and neuronal loss at the time of PFCs, especially in the three individuals who now have a single small hippocampus. As there is no T2 abnormality, the hippocampi do not meet the criteria for MTS. There may be a lag period of several years between a PFC and the onset of epilepsy, and therefore some of these patients may be in the process of developing MTS. (Supported by The Wellcome Trust.)
A.07 EVIDENCE FOR PARIETOFRONTAL DYSCONNECTION IN EPILEPTIC SYNDROMES WITH CONTINUOUS SPIKES AND WAVES DURING SLOW SLEEP 1 Xavier De Tiege, 1 Serge Goldman, 1 Steven Laureys, 2 Denis Verheulpen, 3 Catherine Chiron, 3 Olivier Dulac, and 2 Patrick Van Bogaert (1 PET/Biomedical Cyclotron Unit, and 2 Department of Pediatric Neurology, ULB-Hopital Erasme, Brussels, Belgium; and 3 Department of Pediatric Neurology, Hopital Saint Vincent de Paul, Paris, France) Rationale: Syndromes with continuous spikes and waves during slow sleep (CSWS) are age-related epileptic syndromes that associate cognitive dysfunctions and specific EEG abnormalities. Previous positron emission tomography (PET) studies using [18 F]-fluorodeoxyglucose (FDG) have shown that some affected patients have focal hypermetabolism. We hypothesized that particular metabolic patterns could characterize this subgroup of patients. Methods: Among children with epilepsy investigated with FDG-PET in our center, 11 patients, aged 3 to 10 years, had the EEG pattern of CSWS and hypermetabolism on PET. These patients were considered as one group, which was analyzed by using a voxel-based method, Statistical Parametric Mapping 99, with a control group of young adults. In a first step, a simple comparison between the two groups was performed. In a second step, we searched for disease-induced changes in the contribution of a brain area to the level of metabolic activity in another brain area by using “pathophysiological interactions” here defined as an application of the psychophysiological interactions to a pathological condition. Results: When compared with the control group, the patient group showed the coexistence of highly significant hypermetabolic areas in the right parietal lobe and hypometabolic areas in the frontal lobes (p < 0.05). Epilepsy-induced changes in the parietofrontal metabolic relations were then searched for. When considering the peak voxel values in the right parietal hypermetabolic areas, significant pathophysiological interactions with the metabolism in the right and the left frontal lobes were found (p < 0.05). Regression plots also indicated that hypermetabolic areas in the right parietal lobe were associated with a loss in functional connectivity with the right frontal lobe (Fig. 1) and inhibition in the left frontal lobe (Fig. 2). Conclusions: This study shows evidence for parietofrontal dysconnection in epileptic syndromes with CSWS. This altered modulation seems to be the consequence of a loss of functional connectivity or an inhibition mechanism. The presence of frontal hypometabolism is in agreement with the neuropsychological deficits frequently observed in epilepsy patients with CSWS (Figs. 1 and 2).
Epilepsia, Vol. 44, Suppl. 9, 2003
166
AES PROCEEDINGS onset GM and MSs. Diagnoses of seizures and epileptic syndromes was based on the International League Against Epilepsy (ILAE) classifications and were validated independently by at least two epileptologists. An extended pedigree was constructed for each family, and all available affected relatives were interviewed, examined, and underwent EEG studies. Results: Twenty-two probands were female and 10 were male subjects (2.2:1 ratio), with the age during enrollment averaging 24.2 years (range, 11–52 years). Pyknoleptic absence seizures (pASs) started at 6.5 years (range, 2–10 years). By enrollment, pASs had persisted an average 15.6 years (range, 5–47 years). MSs started between 8 and 24 years and had persisted for 2–30 years by enrollment. All probands had at least one GM seizure that appeared from 6 to 22 years and had been recurring for 30 years. Forty-six percent of probands showed 3-Hz single spike and slow wave complexes or 3- to 5-Hz single spike–wave complexes. Fifty-four percent had single spike–wave complexes mixed with polyspike–wave complexes. Five percent showed fast polyspike and slow wave complexes as the sole EEG trait. Neuroimaging and neurologic examinations remained normal during follow-up. In 82%, monotherapy with valproate achieved seizure control; the rest needed polytherapy. In 75%, another family member was affected with epilepsy. Among 99 nonproband family members, 57 were females vs. 42 males. There was almost as much paternal as maternal transmission. More relatives (43.6%) had absences alone or in combination with MSs or GM. Conclusions: CAE persisting with MSs and GM seizures of adolescent onset is a female-preponderance syndrome characterized in probands and affected family members by pyknoleptic absences and EEG patterns of both 3-Hz spike–wave complexes and 3- to 7-Hz polyspike–wave complexes. Maternal transmission was equal to paternal transmission. These results indicate a persisting CAE syndrome distinct from remitting CAE and classic JME. (Supported by NINDS grant No. 5RO1NS042376-03.)
December 8, 2003 Platform Session B: Imaging 3:30 p.m.–5:30 p.m. B.01 VOXEL-BASED MORPHOMETRY OF GREY AND WHITE MATTER IN PATIENTS WITH TEMPORAL LOBE EPILEPSY Neda Bernasconi, Simon Duschesne, Andrew Janke, Frederick Andermann, and Andrea Bernasconi (Neurology and Neurosurgery, Montreal Neurological Institute, Montreal, QC, Canada) A.08 CHILDHOOD ABSENCE EPILEPSY PERSISTING WITH ADOLESCENT GRAND MAL AND MYOCLONIC SEIZURES: CLINICAL AND EEG DESCRIPTION OF 32 FAMILIES 1,2 Marco T. Medina, 1,2 Reyna Duron, 3 Maria E. Alonso, 1 DongSheng Bai, 4 Sonia Khan, 1 Gregorio Pineda, 1,5 Julia N. Bailey, 6 Ignacio P. Castroviejo, 3 Astrid Rasmussen, 3 J. Ramos-Peek, 3 Sergio Cordova, 3 Francisco Rubio-Donnadieu, 3 Adriana Ochoa, 3 A. Jara-Prada, 1 Katerina Tanya B. Perez-Gosiengfiao, and 1 Antonio V. DelgadoEscueta (1 Comprehensive Epilepsy Center, University of California at Los Angeles, Los Angeles, CA; 2 Neurology Training Program, National Autonomous University of Honduras, Tegucigalpa, Honduras; 3 National Institute of Neurology and Neurosurgery, Mexico, DF, Mexico; 4 Neurosciences Dept., Riyadh Armed Forces Hospital, Saudi Arabia; 5 Neuropsychiatry Institute, UCLA, Los Angeles, CA; and 6 Pediatric Neurology Dept., University Hospital La Paz, Madrid, Spain) Rationale: Childhood absence epilepsy (CAE) with or without grand mal accounts for 5–15% of all epilepsies. Four subsyndromes of CAE have been described by our group, as well as by other authors. One of them is CAE that persists with adolescent or adult myoclonic seizures (MSs) and grand mal (GM) seizures. Methods: We prospectively studied 17 families of European ethnic origin and 15 families of mixed European and American Indian origin recruited from 1978 to 2002 (32 probands and 99 affected nonproband family members). Families were ascertained through a proband with persisting CAE starting between 2 and 10 years old plus adolescent-
Epilepsia, Vol. 44, Suppl. 9, 2003
Rationale: In temporal lobe epilepsy (TLE), morphometric magnetic resonance imaging (MRI) studies have shown extrahippocampal structural damage. These studies are labor intensive, focus on few areas, and have low intra- and interrater reliability. Voxel-based morphometry (VBM) is an automated technique allowing identification of differences in grey matter (GM) and white matter (WM) with no a priori region of interest, allowing whole-brain analysis between groups. The purpose of this study was to determine whole-brain GM and WM abnormalities in TLE by using VBM. Methods: We studied 87 patients with intractable TLE (mean age, 35 years) and 51 neurologically normal controls (mean age, 33 years). Seizure focus was defined as right (n = 40) or left (n = 47) if >70% of seizures were recorded from one side. Volumetric MRI showed ipsilateral hippocampal atrophy in all patients. MRI 3-D images were acquired on a 1.5-T scanner by using a T1 -fast field-echo sequence (slice thickness, 1mm). Image processing included (a) automated correction for intensity nonuniformity; (b) normalization of grey-level intensities; (c) linear registration of images to a standardized stereotaxic space; (d) classification of brain tissue into GM, WM, and CSF; (e) blurring of GM and WM binary masks with an isotropic gaussian kernel of 5 mm FWHM to generate 3-D maps of GM and WM “density.” Statistical maps of differences between patients’ and controls’ densities were obtained by using a general linear model. Significance was set at p < 0.05, corrected for multiple comparisons. Results: GM reduction: Compared with normal controls, patients with left and right TLE had diffuse GM reduction in the hippocampus
AES PROCEEDINGS ipsilateral to the seizure focus. In patients with left TLE, GM reduction was also present in the contralateral hippocampal tail. Both patient groups had ipsilateral GM reduction in various frontal lobe areas (prefrontal, dorsolateral prefrontal, orbitofrontal, central, and cingulate). In patients with left TLE, ipsilateral reduction was also present in the insular and superior temporal cortices. Contralateral GM decrease was present in frontal and occipital areas in left-TLE patients and in the superior temporal gyrus in right-TLE patients. In addition, both groups had bilateral thalamic GM reduction. WM reduction: patients with left and right TLE showed a diffuse reduction of temporal lobe WM (temporopolar, entorhinal, superior temporal, temporal stem, and fusiform) ipsilateral to the seizure focus, and decreased WM in the body of corpus callosum. Right-TLE patients had ipsilateral WM reduction in the frontal and postcentral areas. Conclusions: GM pathology in TLE extends beyond the hippocampus, involving mainly the frontal lobe and the thalamus. Ipsilateral WM reduction is present in the anterior temporal lobe. This pattern of abnormalities is compatible with preferential degeneration of reciprocal parahippocampal–frontal and parahippocampal–thalamic pathways connecting areas responsible for attention and memory processing.
B.02 ONTOGENETIC EFFECT OF A GABAA -Receptor Subunit Mutation (γ2 -R43Q) in the Living Human Brain: A Structural MRI Study David Reutens, Marco Fedi, Amanda Wood, Carla Marini, and Samuel Berkovic (Medicine, The University of Melbourne, Heidelberg, Victoria, Australia) Rationale: We report the first study in the living human brain of the structural consequences of a missense mutation of the γ 2 subunit (R43Q) of the γ -aminobutyric acid A (GABAA ) receptor associated with inherited generalized epilepsy (IGE) and febrile convulsions. Previous studies in individuals affected by this mutation have demonstrated enhanced cortical excitability associated with widespread reduction in benzodiazepine-receptor binding. The GABAA receptor is known to play an important role during brain development by affecting neuronal differentiation and migration. Visual inspection of magnetic resonance imaging (MRI) scans in subjects with the mutation suggested a consistent difference in callosal morphology. Here we sought to examine this structural abnormality systematically. Methods: Subjects comprised 10 subjects affected by the GABAA γ 2 subunit mutation (3 men; mean age, 43 years) and 47 controls (23 men; mean age, 34 years). All subjects underwent high-resolution T1 weighted MR images. The images were placed in a standard stereotaxic coordinate space and, by using a semi-automated algorithm, the corpus callosum was segmented, and callosal thickness was measured at 39 equally spaced nodes along its length. The statistical significance of differences in regional callosal thickness between groups was assessed by using a nonparametric permutation test. Results: The posterior midbody of the callosum was significantly thinner in individuals with the GABAA mutation than in controls. Gray
167
matter and white matter volumes also were measured in subjects affected by the mutation and a subset of age-matched normal subjects. No significant difference in volume was observed. Conclusions: These findings provide compelling in vivo evidence of a role for the GABAA receptor in brain development and, in particular, the development of the corpus callosum. Previous studies have demonstrated subtle abnormalities in brain architecture in patients with IGE. It is possible that the morphologic consequences of the R43Q mutation are part of the link between genotype and clinical phenotype. B.03 HIGH-RESOLUTION DIFFUSION TENSOR IMAGING IN PATIENTS WITH CORTICAL DYSPLASIA: A STUDY AT 3 TESLA 1 M. Halko, 1 S. Knake, 1 P.E. Grant, 1 D.H. Salat, 1 D.S. Tuch, 1 S.S. Stufflebeam, 1 H. Shriaishi, 2 E.B. Bromfield, 3 D.L. Schomer, and 1 E. Halgren (1 MGH/MIT/HMS Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital/Harvard Medical School, Charlestown; and 2 Brigham and Women’s Hospital and 3 Beth Israel Deaconess Medical Center, Boston, MA) Rationale: Cortical dysplasias (CDs) are a common cause of epilepsy associated with pathologic cortical and white matter organization. The ability to localize focal CD considerably modifies the postsurgical outcome. The aim of this study was to investigate whether diffusion tensor imaging (DTI) is able to detect subtle microstructural changes in the white matter (wm) organization associated with CD. Methods: Five patients (mean age, 25.2 years) with temporal lobe epilepsy (TLE) and CD in the affected temporal lobe and three patients (mean age, 25.3 years) with focal, nonlesional, extratemporal epilepsies (ETEs) were investigated. In three of the TLE patients, the diagnosis of CD was confirmed histologically. CDs have been diagnosed by using an eight-channel-array surface coil MRI. DTI was collected by using a multishot high-resolution echo-planar imaging technique (2 mm slices), gap, 0 mm; TR, 12,000 ms; TE, 69 ms; 128 × 128 acquisition matrix; b-value, 700 s/mm2 ) with a 3-T whole-body scanner (Siemens Trio; Siemens, Erlangen, Germany). Diffusion-weighted images were acquired during an 8-min scan with 60 averages. Whole-brain fractional anisotropy (FA) maps were coregistered to a low-b image. Mean FA was calculated within a 5-mm-diameter spherical region of interest (ROI) deep in the central wm of the temporal lobe in all patients. The fractional change ratio was defined by subtracting the FA from the pathologic side from the one of the healthy side and dividing it by the FA of the normal side. Results: CD showed significantly widespread increased FA values despite smaller visible detectable lesions in the ROI of the affected side (p = 0.014). Table 1 gives an overview of the results. The mean difference was 0.15. There were no hemispheric differences in ETE (p = 0.4). Conclusions: DTI seems to be a promising, noninvasive tool to detect even subtle structural changes of wm organization. The method needs further validation but seems to be a promising new tool, especially for detecting CD in patients with nonlesional focal epilepsies, and might be integrated in the presurgical workup in the future. (Supported by MIND Institute.)
TABLE 1. Overview of the results Patient no. TLE 1 2 3 4 5 ETE 6 7 8
Age (yr)
Diagnosis
Diagnosis
Healthy side
Focus side
Fractional change ratio
32 27 17 27 23
rt TLE lt TLE rt TLE rt TLE rt TLE
Dysplasia Dysplasia Dysplasia Dysplasia Dysplasia
0.386 (lh) 0.353 (rh) 0.323 (lh) 0.163 (lh) 0.170 (lh)
0.494 (rh) 0.459 (lh) 0.410 (rh) 0.486 (rh) 0.283 (rh)
0.28 0.30 0.27 0.51 0.66
26 23 27
ETE ETE ETE
nl nl nl
0.343 (rh) 0.509 (lh) 0.412 (rh)
0.312 (lh) 0.553 (rh) 0.378 (lh)
0.09 0.08 0.08
TLE, temporal lobe epilepsy; FA, fractional anisotropy; ETE, extratemporal epilepsy (all patients had frontal lobe seizures and frontal lobe CD); nl, nonlesional; rh, right hemisphere; lh,left hemisphere.
Epilepsia, Vol. 44, Suppl. 9, 2003
168
AES PROCEEDINGS
B.04 POSTICTAL DIFFUSION TENSOR IMAGING 1,2 Beate Diehl, 1 Mark R. Symms, 1 Philip A. Boulby, 1 Tuuli Salmenpera, 4 Claudia A.M. Wheeler-Kingshott, 3,4 Gareth J. Barker, and 1 John S. Duncan (1 MRI Unit, Department of Clinical and Experimental Epilepsy, Institute of Neurology, University College London, London, United Kingdom; 2 Dept. of Neurology, The Cleveland Clinic Foundation, Cleveland, OH; 3 Neuroimaging Research Group, Institute of Psychiatry, London, and 4 Department of Neuroinflamation, NMR Research Unit, Institute of Neurology, London, United Kingdom) Rationale: To evaluate postictal diffusion tensor imaging (DTI) as a tool to delineate the zone of seizure onset. Methods: DTI and statistical parametric mapping (SPM) were used to examine objectively the diffusion properties of 20 patients with intractable epilepsy (18 with focal epilepsy, two with generalized epilepsy) both postictally and interictally and to compare them with 27 normal controls who were scanned twice. Scans were obtained as soon as possible after a seizure and again after a seizure-free interval >24 h in 10 patients and 2–18 h in the other patients. Three sets of statistical tests were performed on each patient’s fractional anisotropy and mean diffusivity scans: interictal versus controls, postictal versus controls, and a third “difference analysis” to test for significant changes in diffusion between postictal and interictal scans in comparison with the differences noted in the two sets of control scans. Results: Compared with the control group, 13 of the 20 patients (72% of the patients with focal epilepsy) had increases in mean diffusivity in the interictal scan compared with a single set of control scans. No decreases in mean diffusivity compared with the controls were detected. The differential analysis detected relative decreases in mean diffusivity postictally in nine patients (50% of the patients with focal epilepsy). These changes were focal in seven patients. In six of those, the side of the epileptic focus was known, and colocalization was present in three. No changes in anisotropy were noted between the post- and interictal states. Conclusions: This study showed that diffusivity is frequently abnormally increased in patients with focal epilepsy and that postictally, a significant proportion of patients show a relative decrease in diffusivity, probably reflecting cellular swelling in the area of seizure onset and possibly also in areas of spread. Postictal diffusion changes, however, appear complex and dynamic, and timing after the seizure may be critical. Anisotropy measures appear to be less sensitive to those changes. Further systematic studies are needed, and correlation with outcome after epilepsy surgery will determine the role of the postictal diffusivity measures in the presurgical evaluation of epilepsy patients. (Supported by Action Medical Research, U.K.)
B.05 FUNCTIONAL REORGANIZATION OF THE FACE MOTOR AREA IN PATIENTS WITH EPILEPTOGENIC LESIONS OF THE CENTRAL REGION Alexandre Carpentier, Stephane Clemenceau, Stephane Lehericy, Remy van Effenterre, and Michel Baulac (Departments of Neurosurgery, Neuroradiology, and Epilepsy, Hopital de la Pitie-Salpetriere, Universite Paris VI, Paris, France) Rationale: Slowly evolving brain damage such as low-grade tumor may result in plastic changes of cortical motor networks. The nature of these plastic changes and the factors related to these changes are largely unknown in patients with brain tumor, although their knowledge may have important functional value. Functional magnetic resonance imaging (fMRI) is a powerful tool in the evaluation of motor cortex functionality in the preoperative planning of surgical procedures in and around the rolandic region. We used fMRI to determine the factors related to reorganization of motor circuits in the presence of brain epileptogenic lesions. Methods: Ten normal subjects and 76 patients (45 low-grade astrocytomas, 20 high-grade gliomas, 11 congenital vascular malformations) were studied by using fMRI with a face motor paradigm. Activation maps were constructed for each subject, and changes in position/amplitude/surface of the motor activation on the lesion side were compared with the activation pattern obtained in the hemisphere con-
Epilepsia, Vol. 44, Suppl. 9, 2003
tralateral to the lesion. Histology, lesion location, primary face motor area infiltration, mass effect, epilepsy, and facial palsy were studied in terms of impact on cortical motor fMRI interhemispheric asymmetry index by using a multivariate statistical analysis. In the motor cortex, a laterality index (LI) was calculated as follows: LI = [(L − R)/(L + R)], with L and R being the numbers of activated voxels in the left and right hemispheres, respectively. Results: Asymmetric pixel count in motor cortex activation was observed in patients with face area infiltration (p < 0.0001), mass effect (p < 0.002), and high epileptic seizures frequency (p = 0.0035). Histology, age, handedness, lesion side, epilepsy duration, and age at epilepsy onset had no impact on cortical activation. Only three patients had preoperative facial palsy. Postoperative facial palsy occurred in six patients and resolved completely within 15 (±30) days. Conclusions: Three main factors were associated with reorganization of the primary face motor area: motor cortex infiltration, mass effect and epileptic seizure frequency. Postoperative findings show that functional motor recovery occurs even in patients with direct lesion of the primary motor face area. This study supports the specificity of epilepsy as a cortical plasticity inductor. (Supported by Assistance Publique, Hopitaux de Paris, INSERM EMI 0224 and EPI007).
B.06 ABERRANT BIOENERGETIC NETWORK RELATES TO HIPPOCAMPAL HISTOPATHOLOGY IN MTLE 1 Jullie W. Pan, 2 Jung H. Kim, 3 Aaron Cohen-Gadol, 3 Dennis D. Spencer, and 4 Hoby P. Hetherington (1 Neurology, Albert Einstein Coll. Medicine, Bronx, NY; 2 Pathology and 3 Neurosurgery, Yale University Sch. Medicine, New Haven, CT; and 4 Radiology, Albert Einstein Coll. Medicine, Bronx, NY) Rationale: Data from mesial temporal lobe epilepsy (mTLE) patients have shown abnormalities in fluorodeoxyglucose (FDG) uptake in the subcortical nuclei, which have been correlated to dentate and hilar cell loss. Whereas the cerebral metabolic rate of glucose (CMRglc) is a sensitive evaluation for rate of substrate use, the tissue’s current bioenergetic state can be directly assessed with 31 P spectroscopy. Previously bioenergetic abnormalities have been seen with 31 P spectroscopic imaging in the hippocampus and used in the lateralization of epileptogenic regions. In this study, we correlated regional 31 P spectroscopic imaging data for bioenergetic state in relation to postoperative histopathology in 14 unilateral mTLE patients. Methods: Three-dimensional 31 P MR spectroscopic imaging was performed at 4 Tesla in patients and controls (n = 14). Regions studied included the bilateral hippocampi (amygdala, pes hippocampus, hippocampal body), thalamus, and striatum. Bioenergetic impairment was determined by the phosphocreatine-to-adenosine triphosphate ratio (PCr/ATP) and evaluated in relation to clinical and neuropathologic evaluation of hippocampal glial fibrillary acidic protein (GFAP) staining, gliosis, and neuronal loss. Results: Cross-hemispherically, the ipsilateral thalamic PCr/ATP correlated significantly with the contralateral hippocampal PCr/ATP (R = 0.55; p < 0.02). The ipsilateral thalamus also significantly correlated with dentate GFAP staining (R = −0.60; p < 0.02; Fig. 1), whereas the bilateral striate PCr/ATP related significantly to CA1 GFAP staining (R = −0.57; p < 0.03 ipsi; R = -0.60; p < 0.02 contra; data not shown). Correlation of the striatal PCr/ATP to dentate GFAP staining revealed a logarithmic dependence, which was significant (R = 0.70; p < 0.005; Figs. 1 and 2).
Conclusions: GFAP expression is known to be induced with repetitive seizures and is elevated in chronic epilepsy. The bioenergetic data
AES PROCEEDINGS also likely reflect acute seizure occurrence and interictal spike spread. Based on the neuropathology in mTLE, where dentate mossy fiber axons → CA3 → CA1, the imaging correlations with GFAP staining suggest that bioenergetically, the ipsilateral thalamus functions upstream with the dentate, whereas the bilateral striate functions downstream with CA1. This multistep connectivity may also contribute to the nonlinear dependence of dentate GFAP staining seen with striate energetics. These data support a view wherein a network of bioenergetic impairment is seen in unilateral TLE, involving the hippocampus, thalamus, and striatum. In particular, the extent and distribution of energetic insufficiency appears to depend on the varying components of damage in the ipsilateral hippocampus (Fig. 1). (Supported by NIH NINDS P01 39092, R01 40550, Charles A. Dana Foundation.)
B.07 LOCALIZATION OF THE EPILEPTOGENIC ZONE BY TEMPORAL CLUSTERING ANALYSIS IN THE RESTING fMRI 1 Victoria L. Morgan, 2 Amir Arain, and 2 Bassel Abou-Khalil (1 Radiology and 2 Neurology, Vanderbilt University, Nashville, TN) Rationale: Accurate localization of the epileptogenic zone before epilepsy surgery is often not possible with existing noninvasive methods. Novel noninvasive methods of localization are needed for such situations. We investigated the value of resting functional magnetic resonance imaging (fMRI) with temporal clustering analysis for correct localization of the epileptogenic zone. Methods: We investigated two patient groups: six patients with temporal lobe epilepsy (TLE), who subsequently became seizure free or almost seizure free after temporal resection, and three patients with refractory suspected extratemporal epilepsy, who also had a complete presurgical evaluation. fMRI scanning was performed with baseline antiepileptic drugs (AEDs). Images were acquired over several minutes with the patients awake but with eyes closed. This resulted in a 4-D dataset of 200 image volumes over time. A histogram of the number of voxels in the volume whose signal intensity reached a maximum at each time point was created (temporal clustering technique). Points in the histogram >100 voxels reaching maximum were considered relevant to interictal or subclinical ictal activation of the epileptogenic zone. These times were used to create an activation map by using SPM99 software. For patient set 1, the hemisphere of the hippocampus with highest peak activation (T value) was compared with the side of resection. For patient set 2, the activation maps were compared with results of standard presurgical assessments. Results: The histograms for each of the TLE patients demonstrated at least one peak. In all patients, the hippocampus on the side of resection showed the highest correlation with the peaks of the histogram (p = 0.005). The extratemporal patients demonstrated activations in regions consistent with the clinical semiology and the overall conclusions of the presurgical evaluations. All patients had extratemporal frontal seizure semiology and poorly localizing or misleading ictal and interictal EEG data. The first two patients had nocturnal sleep-related seizures with tonic posturing and no alteration of awareness. The fMRI analysis indicated frontal regions of activation consistent with seizure semiology and not demonstrated by other tests. The third patient had nocturnal hypermotor seizures. MRI was normal, and EEG, positron emission tomography (PET), and ictal single-photon emission computed tomography (SPECT) data favored right temporal localization. The fMRI analysis demonstrated several extratemporal areas of activation that invasive recordings confirmed to be an epileptic network activated with every seizure. These regions were also not demonstrated by other noninvasive tests. Conclusions: Resting fMRI with temporal clustering analysis demonstrates excellent potential for localization of the epileptogenic zone. It can be used to evaluate the whole brain with relatively high, threedimensional spatial resolution in 0.5 Hz). We developed a genuine DC-EEG technique for bedside recordings to assess the utility of slow ictal EEG responses (DC shifts) in determining the localization of seizure origin. Methods: More than 40 patients with focal epilepsy were examined by using DC-EEG at bedside for up to several days during video-EEG recording for preoperative evaluation. Ictal slow EEG signals were analyzed both visually and by using 3D-source localization tools. Results: We observed scalp negative DC shifts with amplitudes typically of up to several hundreds of microvolts at the vertex with a mastoid reference during seizures. 3D source localization demonstrated a focal onset of the DC shifts. Slow EEG signals enabled the identification of the site of seizure origin, even in cases (e.g., mesial temporal lobe onset) in which standard scalp EEG yielded equivocal results. Conclusions: Full-bandwidth EEG recording with DC-EEG is feasible in a clinical environment (bedside recordings). Localization of the slow EEG components can provide valuable information about the site of onset of focal seizures, which is of particular relevance for presurgical evaluation (Fig. 1). (Supported by The Academy of Finland, Finnish Cultural Foundation, Arvo and Lea Ylpp¨o Foundation, and the Regional Epilepsy Center, University of Washington.)
C.04 MEG, EEG, AND MRI CORRELATIONS IN EXTRATEMPORAL EPILEPSY 1,2 Marta Garc´ıa-Fern´ andez, 1,3 Felipe Quesney, 1 Carlos Amo, 1 Fernando Maest´u, 1 Alberto Fern´andez, 4 Jaime Campos-Castell´o, and 1 Tom´as Ortiz (1 Centro MEG, Universidad Complutense de Madrid; and 2 Clinical Neurophysiology Department, Hospital Cl´ınico Universitario San Carlos, Madrid, Spain; 3 Montreal Neurological Institute, McGill University, Montreal, QC, Canada; and 4 Neuropediatrics Department, Hospital Cl´ınico San Carlos, Madrid, Spain ) Rationale: A bilateral representation of the epileptic disturbance, a widespread unilateral distribution of the interictal spiking, or its absence, represent significant limiting factors of EEG localization in extratemporal epilepsy (ETE). We studied the localizing effectiveness of magnetoencephalography (MEG) in ETE and compared it with the EEG and magnetic resonance imaging (MRI) results. Methods: The 22 patients (nine females; mean age, 24 years) with ETE underwent EEG, MEG (148 channels), and 3D-MRI (1.5 and 3 Tesla). The anatomic distribution of the epileptogenic area/lesion was frontal (F), 14; central (C), three; parietal (P), four; and insular (I), one patients. Ten patients had MRI lesions (seven F, two C, and one P). Results: Seven (32%) of 22 patients had bilateral epileptiform EEG discharges (six F and one P). In five of them (71.4%), MEG provided a correct lateralization. Four (18%) of 22 patients (three F and one P), showed widespread unilateral EEG spiking. MEG accuratly localized the epileptic activity in all of them. Five (23%) had normal EEGs. MEG demonstrated localized/lateralized epileptiform activity in all of them. In three (13.5%), MEG improved the EEG localization or corrected the EEG lateralization in regard to the MRI lesion. In the remaining three (13.5%), the MEG and EEG findings colocalized in the same brain region. MEG and neuroimaging findings were concordant in all 10 patients with MRI lesions. Conclusions: Whereas EEG achieved proper localization of the epileptogenic area in seven (31.8%) of 22 patients with ETE, MEG reliably localized the epileptogenic area in 20 (90%). Concordance of MEG–MRI findings was seen in all patients with structural lesions. Our
REFERENCES 1. Vanhatalo S, Holmes MD, Tallgren P, et al. Very slow EEG responses lateralize temporal lobe seizures: an evaluation of noninvasive DCEEG. Neurology 2003;60:1098–102. 2. Lagerlund TD, Gross RA. DC-EEG recording: a paradigm shift in seizure localization? Neurology 2003;60:1062–3.
C.06 A NEW METHOD FOR DYNAMIC INFORMATION ANALYSIS IN EPILEPSY 1 Leon D. Iasemidis, 1 Awadhesh Prasad, 1 Mallika Mukherjee, 1 Levi Good, and 1,2 Jie Wu (1 The Harrington Department of Bioengineering, Arizona State University, and 2 Neurology, Barrow Neurological Institute, Phoenix, AZ) Rationale: Modern methods for dynamic analysis of signals recorded from the brain (like EEG, field potentials, spike trains) involve the notion of a multidimensional state space and measure of the complexity
Epilepsia, Vol. 44, Suppl. 9, 2003
172
AES PROCEEDINGS
and stability of the embedded system in the state space. So far, the foundation blocks of this approach had been the information content (e.g., Renyi information) and the rate of change of information content (e.g., Lyapunov exponents, Kolmogorov entropy). However, still missing are reliable measures of the direction and magnitude of information flow that could characterize the interactions and quantify the coupling of different parts of a system in the state space. Such measures will be of paramount importance in science in general and in epilepsy in particular. Results from the application of such a measure to epilepsy will be presented. Methods: We introduce a new method to analyze the interactions between brain signals in the spatiotemporal domain. The thus-defined cross short-term maximum Lyapunov exponent (CSTLA,B max ) measures the rate of information flow (bits/s) from brain site A to B. We have applied this measure to (a) data from dynamic models of coupled chaotic oscillators, (b) long-term multielectrode EEG recordings from patients with focal epilepsy, and (c) intracellular and extracellular cell membrane potential recordings in “seizures” produced by hyperhermia in immature rat hippocampal slices. Results: In all simulation data, under different spatial configurations and types of oscillators, the CSTLmax correctly estimated the direction of flow of information. Its application to the human EEG data showed the following novel results: (a) correct (p < 0.01) lateralization and localization of the epileptogenic focus in both patients analyzed (our two most difficult cases) by identification of the driver electrode sites; (b) elevated, intermittent, dynamic driving of extrafocal sites by the focus in the preictal periods, and of the focus by the extrafocal sites in the postictal periods. Its application to the data from hippocampal slices showed prevalence of unidirectional information flow between extra- and intracellular spaces before the accompanying spreading depression (SD) at low temperatures, and balanced bidirectional flow after SD (high temperatures). This last observation is in accordance with the physiologic basis of the phenomenon (Wu J, Fisher RS. J Neurophysiol 2000; 84:1355–60). Conclusions: The newly introduced measure of cross Lyapunov exponents provides new insights into the epileptogenic process (see also Iasemidis LD. IEEE Trans Biomed Eng 2003;50:549–58). Results from the simulation (computer), clinical (in vivo), and experimental (in vitro) data used in this study support this conjecture. The results from the new measure’s application to hippocampal-slice data suggest that CSTLmax can also be a useful tool in studies of the basic mechanisms of epilepsy. (Supported by NIH, Whitaker Foundation.) C.07 ON THE PREDICTABILITY OF SEIZURES 1,2 Florian Mormann, 1,3 Thomas Kreuz, 3 Ralph G. Andrzejak, 1,2 Christoph Rieke, 3 Kraskov Alexander, 1 Christian E. Elger, and 1 Klaus Lehnertz (1 Department of Epileptology, and 2 Helmholtz-Institute for Radiation and Nuclear Physics, University of Bonn, Bonn; and 3 John von Neumann Institute for Computing, Research Center J¨ulich, J¨ulich, Germany) Rationale: An important issue in epileptology is whether epileptic seizures can be anticipated before their occurrence. Of particular interest is the question whether information extracted from the EEG of epilepsy patients can be used for the prediction of seizures. Several studies have claimed evidence for the existence of a preseizure state that can be detected by using linear and nonlinear EEG analysis methods. Most of these studies, however, were performed on short, selected recordings, and little experience exists with continuous long-term recordings over several days. In this study, we evaluated the predictive performance of a variety of measures derived from the theory of dynamic systems. Methods: We compare different linear and nonlinear measures comprising both univariate (derived from a single EEG signal) and bivariate approaches (characterizing the synchronization between two EEG signals recorded simultaneously from different locations of the brain) in terms of their ability to distinguish between the seizure-free interval and the preseizure period. We analyzed intracranial continuous multiday EEG recordings from five patients with temporal mesial lobe epilepsy undergoing presurgical diagnostics covering >300 h of multichannel EEG. We use receiver-operating characteristics (ROC curves) to quan-
Epilepsia, Vol. 44, Suppl. 9, 2003
tify the degree to which the amplitude distribution of the time profiles calculated from the seizure-free interval can be distinguished from the amplitude distribution from the preseizure period for a particular measure. To assess the statistical significance of the obtained performance values, we use a recently proposed method termed seizure time surrogates. Results: Our analysis shows a similar performance of linear and nonlinear measures. We do, however, find a distinct difference between bivariate and univariate approaches, with a higher performance of the bivariate measures. Validation analysis using seizure time surrogates shows only performance values for the bivariate measures to be statistically significant. Conclusions: The predictive performance of univariate measures reported in earlier publications could not be confirmed. Results for the bivariate measures, conversely, provide statistically significant evidence for the existence of a preseizure state that is reflected by changes in synchronization as characterized by these measures. (Supported by the Deutsche Forschungsgemeinschaft).
C.08 PUT YOUR SEIZURE-PREDICTION STATISTICS TO THE TEST: THE METHOD OF SEIZURE TIME SURROGATES 1 Ralph G. Andrzejak, 1,2 Thomas Kreuz, 2 Florian Mormann, 1 Alexander Kraskov, 3 Christoph Rieke, 2 Christian E. Elger, and 2,3 Klaus Lehnertz (1 John von Neumann Institute for Computing, Research Center J¨ulich, J¨ulich; and 2 Department of Epileptology and 3 Helmholtz-Institute for Radiation and Nuclear Physics, University of Bonn, Bonn, Germany) Rationale: A rapidly growing number of studies deals with the prediction of epileptic seizures. For this purpose, various techniques derived from linear and nonlinear time series analysis have been applied to the EEGs of epilepsy patients. In none of these works, however, is the performance of the seizure-prediction statistics tested against a null hypothesis, an otherwise ubiquitous concept in science. In consequence, the evaluation of the reported performance values is problematic. Here we propose the technique of seizure time surrogates based on a Monte Carlo simulation to remedy this deficit. Our approach allows formal testing of the null hypothesis of the nonexistence of the preictal state. Methods: To illustrate the technique of seizure time surrogates, we analyzed the spatiotemporal distribution of a nonlinear measure (ξ ) that was calculated from the intracranial EEG of patients with mesial temporal lobe epilepsy. Seizure time surrogates were constructed by replacing the original seizure times with times randomly chosen from the interictal intervals. The total number of seizures and the distribution of intervals between consecutive seizures were imposed as constraints on the seizure time surrogates. From the time profiles of ξ , a simple evaluation statistic was calculated for the original seizure times, as well as for an ensemble of 19 seizure time surrogates. The measure ξ together with the applied evaluation statistics are referred to as seizure-prediction statistics, the performance of which is denoted by R. Provided that a preictal state exists and that our prediction statistic is be able to detect it, highest values of R should be obtained for the original seizure times. Results: We obtained a rather high value of R for the original seizure times. At the first glance, this finding could be regarded as evidence for the existence of a preictal state and the capability of the seizure-prediction statistic to detect it. However, the value of R obtained for the original seizure times was well within the distribution of R values obtained for the seizure time surrogates. Hence, our technique allowed successful disclosure of the insignificance of a high value of the performance of a seizure-prediction statistic. Conclusions: The aim of the present study was not to prove or disprove the existence of a preictal state. Neither was our aim to propose a certain seizure-prediction statistic. Rather, the aim was to introduce the technique of seizure time surrogates that allows validation of the performance of any given seizure-prediction statistics. We have provided an example in which this technique helped to disclose tantalizing results of a seizure-prediction statistic as misleading. In future applications, we expect seizure time surrogates to be a powerful tool to differentiate statistics unsuited for a detection of the preictal state from more promising approaches.
AES PROCEEDINGS December 8, 2003 Platform Session D: Translational Research 3:30 p.m.–5:30 p.m.
D.01 SEIZURE INHIBITION AND DELAYED EPILEPTOGENESIS BY IN VIVO RAAV-MEDIATED NEUROPEPTIDE Y GENE TRANSFER 1 Cristina Richichi, 2 Deborah E. Lin, 1 Daniela Stefanin, 1 Daniele Colella, 1 Giuliano Grignaschi, 3 Gunther Sperk, 2 Matthew J. During, and 1 Annamaria Vezzani (1 Neuroscience, Mario Negri Inst for Pharmacol Res, Milano, Italy; 2 Molecular Medicine, University of Auckland, Auckland, New Zealand; and 3 Pharmacology, Univ of Innsbruck, Innsbruck, Austria) Rationale: We investigated the effect of long-term neuropeptide Y (NPY) gene transduction in the rat hippocampus on epileptogenesis and limbic seizures by using a recombinant adeno-associated viral vector (rAAV). Methods: rAAV carrying the NPY gene under the control of neuronal enolase promoter (2 µl rAAV-NSE-NPY, 4.2 × 108 ) was injected in both septal and temporal aspects of the rat hippocampus bilaterally. Controls were rats injected with equivalent volumes of empty vector. Two different rAAV-NSE-NPY vectors were used with serotype 2 or serotype 1 and 2 capsid proteins, respectively. Four to 8 weeks after gene delivery, rats were implanted with hippocampal and cortical electrodes for EEG analysis of seizures. One week after surgery, rAAV-NSE-NPY rats and their controls (n = 6–8 each group) were injected intrahippocampally (ih, 40 ng) or intracerebroventricularly (icv, 250 ng) with kainic acid or electrically stimulated in the ventral hippocampus according to a wellestablished fast kindling protocol. NPY gene transduction was assessed by in situ hybridization analysis of mRNA and by peptide immunocytochemistry. Results: The expression of the transgene significantly increased 5–7 days after ih administration of the vector; gene transduction was maximal at 4 weeks and lasted for ∼6 months. With serotype 2 rAAV-NSE-NPY, the peptide was specifically increased in hilar interneurons and their axonal projections for ∼1.5 mm from the injection site. In serotype 2 rAAV-NSE-NPY–injected rats, the number of EEG seizures and their duration were reduced by 50% on average and the time to seizure onset was delayed by twofold after ih or icv kainate. With serotype 1 and 2 rAAV-NSE-NPY, the peptide was increased also in the terminal field of mossy fibers for ∼3 mm from the injection site. These rats had a 75% reduction of EEG seizures and a twofold delay in their onset after ih or icv kainate. In serotype 1 and 2 rAAV-NSE-NPY–injected rats, the threshold for afterdischarge induction was increased by 40% in the stimulated hippocampus, and the number of stimuli to reach stages 3 and 4–5 of kindling was increased by twofold. Conclusions: These data show that (a) rAAV mediates neuron-specific long-term transduction of NPY gene in the rat hippocampus; and (b) NPY overexpression in hilar interneurons and in mossy fibers potently inhibits seizures and delays epileptogenesis. Thus, long-term increase of an inhibitory peptide in crucial brain sites may represent a novel strategy for antiepileptic treatment. [Supported by HFSP RG0045-2000 (A.V.,G.S.); Fondazione Cariplo (A.V.); and New Zealand Health Res Council (M.D.).]
D.02 IDENTIFICATION OF THE SECOND LAFORA PROGRESSIVE MYOCLONUS EPILEPSY GENE 1,2 Elayne M. Chan, 1 Edwin J. Young, 1 Iulia Munteanu-Oprea, 1 Leonarda Ianzano, 1,2 Stephen W. Scherer, and 1,3 Berge A. Minassian (1 Program in Genetics and Genomic Biology, The Hospital for Sick Children, 2 Molecular and Medical Genetics, The University of Toronto, and 3 Division of Neurology, Department of Pediatrics, The Hospital for Sick Children, Toronto, ON, Canada) Rationale: Lafora disease (LD) is a progressive myoclonus epilepsy with onset in teenage years. It is an autosomal recessive disorder and
173
one of the most severe forms of juvenile epilepsy. Symptoms include myoclonus, ictal visual hallucinations, and other seizures, paralleled by a severe cognitive decline and death within 10 years of onset. Mutations in the EPM2A gene are present in ∼70% of LD families, and the gene in the remaining families was so far unknown. The EPM2Aencoded protein, named laforin, contains a carbohydrate-binding motif and a dual-specificity phosphatase domain. Recently we identified the first interactor of laforin, a novel protein of unknown function named EPM2AIP1. Despite these findings, the pathway involving laforin and EPM2AIP1 and their involvement in the onset of LD is still unclear. Methods: One approach we took to identify protein components that are involved in the same pathway was to search for the second LD gene by genetic mapping studies. A genome-wide linkage scan was performed by using four families of a French-Canadian “isolate” that do not segregate any EPM2A mutations. Results: Two-point lod scores at one marker gave a lod score of 2.84 at a recombination fraction of θ = 0. Additional markers in the region were genotyped, and all affected individuals were homozygous across a 2.2-Mb region for a seven-marker haplotype that was not present in 100 ethnically matched controls. Additional families were included in the study and genotyped at markers in the critical interval. One family further refined the region to a 0.84-Mb region based on loss of homozygosity at four of the seven markers that define the critical interval. Four genes in this region were examined for sequence changes. Mutations in one gene were identified in 21 families of various origins, including the FrenchCanadian cluster. The mutations include six missense changes and five deletions. Conclusions: We will describe the identification of the second Lafora disease gene and the characterization of its protein product. [We thank the many clinicians for providing patient DNA, especially Eva Andermann, Antonio V. Delgado-Escueta, and Guy A. Rouleau. This work was funded by the Canadian Institute of Health Research (CIHR).]
D.03 TLE ASSOCIATED WITH MUTATIONS OF THE GENE ENCODING THE REGULATORY SUBUNIT OF THE SODIUM CHANNEL IN GEFS+ 1 Ingrid E. Scheffer, 2 Robyn H. Wallace, 2 Leanne Dibbens, 1 Samantha Turner, 1 Bronwyn Grinton, 1 Rita Singh, 1 Judith Adams, 1 Mary Connellan, 2 John C. Mulley, and 1 Samuel F. Berkovic (1 Epilepsy Research Institute, University of Melbourne, Heidelberg, Victoria; and 2 Centre for Medical Genetics, Women’s and Children’s Hospital, North Adelaide, South Australia, Australia) Rationale: Generalized epilepsy with febrile seizures plus (GEFS+) is a familial epilepsy syndrome with heterogeneous phenotypes including febrile seizures (FSs), febrile seizures plus (FS+), and mild and severe forms of generalized epilepsy including myoclonic–astatic Epilepsy (MAE) and severe myoclonic epilepsy of infancy (SMEI). Less commonly focal epilepsy may occur, including temporal lobe epilepsy (TLE). GEFS+ is associated with mutations in the sodium channel–regulatory β1 subunit gene, SCN1B, and the pore-forming α1 and α2 subunit genes, SCN1A and SCN2A, as well as the γ -aminobutyric acid (GABA)A receptor γ 2 subunit gene GABRG2. Although SCN1B was the first gene associated with GEFS+, only one SCN1B mutation, C121W, has been reported in two Australian families. This mutation is localized in the extracellular domain of the protein, a region important in channel gating kinetics. In contrast, many SCN1A mutations occur in patients with GEFS+ and SMEI. We report here two GEFS+ families with novel mutations of SCN1B and analyze the phenotypes occurring in the six known families with SCN1B defects with particular emphasis on the TLE phenotype. Methods: Six families with GEFS+ underwent detailed epilepsy characterization in all available affected family members by using a validated seizure questionnaire. All previous electroclinical data were obtained. Seizure types were classified according to the international classification. SCN1B mutational analysis was performed. Results: There were 47 affected individuals in six families with proven mutations of SCN1B. Novel mutations of the same amino acid occurred in two families; one individual had FS, seven have FS+. The C121W mutation was found in four families (39 individuals), including the two
Epilepsia, Vol. 44, Suppl. 9, 2003
174
AES PROCEEDINGS
previously reported. The phenotypes of the C121W mutation included FS in 16 individuals, FS+ in 12, MAE in two, FS+ with absences in one, and three were unclassified. TLE occurred in five individuals with the C121W mutation; two had preceding FS, and two, FS+. TLE started at a mean age of 12 years; median, 5.5 years. Two patients had hippocampal sclerosis (HS); one had recurrent febrile status epilepticus and bilateral HS; the other did not have prolonged seizures and had right HS. There were 12 unaffected carriers with the C121W mutation and one with another mutation. Conclusions: Our series of six families illustrates the phenotypic spectrum of SCN1B defects including five individuals with TLE. TLE was not always associated with FS or HS. The intriguing role of SCN1B in TLE and HS is more complex than the accepted paradigm of a prolonged seizure resulting in HS and TLE. SCN1B joins the select group of genes so far recognized as contributing to TLE and HS. The emerging importance of GEFS+ mutations in SCN1B is emphasized by the finding of novel mutations in the extracellular loop, as well as its role in increasing numbers of GEFS+ families. (Supported by NH&MRC; Bionomics Limited.)
D.04 IDENTIFYING FACTORS THAT REGULATE EPILEPSY GENES WITH A TRANCRIPTOME-QTL METHOD BASED ON MICROARRAY DATA FROM BXD MICE 1 Robyn H. Wallace, 2 Yan Cui, and 1 Robert W. Williams (1 Anatomy & Neurobiology, and 2 Molecular Sciences, Center of Genomics & Bioinformatics, University of Tennessee, Memphis, TN) Rationale: Several ion channel mutations have been identified in rare families with mendelian inheritance of epilepsy. Large numbers of epilepsy patients have since been screened for mutations in ion channel genes, with little success. The cause of epilepsy in the majority of patients remains unknown. This study aims to identify regulators of neuronal ion channel gene expression and to test the hypothesis that epilepsy can be caused by mutations in these transcriptional regulators. Methods: Quantitative trait locus (QTL) mapping has the potential to be applied to the analysis of the variation of gene expression. This is achieved by treating each gene on a microarray chip as a quantitative trait, and associating levels of gene expression with genotyping data. C57BL/6 and DBA/2 mice are known to differ in their susceptibility to seizures; therefore some of the differences in gene expression between these two strains will be related to seizures. Recombinant inbred (RI) strains derived from a C57BL/6J female and a DBA/2J male are members of the BXD set. By studying alterations in transcription in the BXD mapping panel, we were able systematically to map genes that modulate ion channel gene expression. This method is called “transciptome-QTL” mapping. All microarray and QTL mapping results are publicly available at WebQTL (webqtl.roswellpark.org). Results: We used transcriptome-QTL mapping to identify chromosomal regions that control the expression of selected sodium channels, potassium channels, γ -aminobutyric acid (GABA) receptors, and acetylcholine receptors. Transcription factors were selected as obvious candidates for the regulation of gene expression; however, the QTLs detected may represent genes that are several steps upstream in the regulatory pathway. It was of interest that the LIM homeobox proteins (Lhx) featured predominantly. We first checked the abundance of Lhx genes to determine if they were so common that one would be expected on every chromosome, and found that there are only eight known Lhx genes. Lhx1, Lhx2, Lhx3, and Lhx5 are all very close to the QTLs that control variation in expression levels of Gabra1, Kcnq2, Chrnb2, and Scn1b. Lhx4 and Lhx 9 are within 20 cM of Gabrg2. Conclusions: Homeobox genes have previously been associated with epilepsy, and it has been suggested that molecules controlling cell-fate decisions also are operative during seizure-induced neurogenesis and plasticity. Lhx proteins play an essential role in cell differentiation and development of both neuronal and nonneuronal tissue. Therefore, the variation in ion channel expression may be due to alterations in cell types or cell numbers. It also is possible that different members of the Lhx gene family directly control the expression of different ion channel proteins. Misexpression of a Lhx gene in epidermal cells of ascidian larva led to ectopic expression of a voltage-gated sodium channel that is normally
Epilepsia, Vol. 44, Suppl. 9, 2003
expressed in the CNS; therefore Lhx proteins may control membrane excitability by regulating ion channel gene expression.
D.05 DISCRETE POPULATIONS OF NESTIN-EXPRESSING CELLS IN HUMAN EPILEPTIC HIPPOCAMPUS: IMPLICATIONS FOR CELL-REPLACEMENT STRATEGIES Alexander A. Sosunov, Xiaoping Wu, Peter D. Crino, Robert R. Goodman, and Guy M. McKhann II (Neurological Surgery, Columbia University, New York, NY; and Neurology, University of Pennsylvania, Philadelphia, PA) Rationale: Epilepsy surgical tissue is a valuable source of progenitor cells. Based on developmental neurobiology and in vitro mammalian work, many nestin-expressing cells in the CNS are stem or precursor cells for neurons and/or glial cells. However, human nestin+ cells have been immunophenotyped to a limited degree. We sought to characterize the antigenic and structural properties of nestin+ cells in human temporal lobe epilepsy (TLE) tissue. Methods: Immunohistochemical and electron microscopic (EM) investigation of nestin+ cells in human hippocampus from surgical patients with intractable TLE: 31 hippocampi were studied, 21 with hippocampal sclerosis (HS) and 10 without sclerosis, based on magnetic resonance imaging (MRI) and neuropathologic criteria. Results: Four types of nestin+ cells were identified: (a) dadial glial cells, observed mainly in young patients, have characteristic structural features and are glial fibrillary acidic protein (GFAP)+, vimentin+ (Fig. 1a), S100β+ (mainly in nucleus), and glutamine synthatase- (GS); (b) fibrous-like astrocytes have large, long processes and are usually found in str. lacunosum-moleculare or near areas of sclerosis. These cells are strongly vimentin+; (c) unusual nestin+ cells with small perikaryon and curved processes that are tightly apposed to neuronal cell bodies. These cells are located in hilus, CA3, and regions of CA1, in areas where reactive astrocytes also are found. Detected in eight hippocampi (four cases with HS, four without HS), these cells are GFAP-, GS-, CD68(microglial marker), and vimentin+, S100β+ (Fig. 1b), and NG2+; and (d) bipolar cells found in the hilus (Fig. 1c) and surrounding white matter, often in contact with blood vessels. They are GFAP-, S100β-, GS-, NG2-, and vimentin+. By EM, the third and fourth cell types have features of undifferentiated cells: condensed nuclear chromatin and few cytoplasmic specialized organelles.
FIG. 1. A: Radial glia cell: nestin+ (red), vimentin+ (blue), GFAP+ (green). B: Nestin+/S100β+ (red) in perineurinal position in hilus. C: Bipolar cell (peroxidase method) (Supported by Klingenstein Foundation, Parents Against Childhood Epilepsy, NIH R21 NS 42334.)
Conclusions: There are four populations of immunohistochemically and anatomically discrete nestin+ cells in human hippocampus from TLE patients. The neurogenic and gliogenic potential of each of these populations remains under investigation and will ultimately determine their usefulness in cell-replacement strategies. Bipolar nestin+ cells may represent a white-matter pool of neuronal/glial precursors, less differentiated than the multipotential neural progenitors isolated from adult human white matter with A2B5 surface antigen sorting (Nunes et al. Nat Med 2003;9:439–47) (Fig. 1).
AES PROCEEDINGS D.06 INTERICTAL NEUROMETABOLITE LEVELS IN THE ANTERIOR VERSUS POSTERIOR HIPPOCAMUS IN TEMPORAL LOBE EPILEPSY: A MICRODIALYSIS STUDY 1,2 Idil Cavus, 3 Willard S. Kasoff, 4 Robert S. Sherwin, 1 John Krystal, and 2 Dennis D. Spencer (1 Psychiatry and 2 Neurosurgery, Yale University, New Haven, CT; 3 Albert Einstein College of Medicine, Bronx, NY; and 4 Medicine, Yale University, New Haven, CT) Rationale: The damage to the anterior mesial temporal structures in mesial temporal lobe epilepsy (mTLE) is thought to be more extensive. Our group previously reported that anterior hippocampus is more bioenergetically impaired compared with the posterior hippocampus [Firlik et al. Epilepsia 2001;42(suppl 7):67] and that the epileptogenic hippocampus is marked by increased extracellular glutamate (Glu) [Cavus et al. Epilepsia 2002;43(suppl 7):247]. Therefore, with microdialysis, we measured the levels of several neurometabolytes in the anterior (pes) and posterior (body) of the epileptogenic hippocampus in neurosurgical patients with mTLE. Methods: Microdialysis studies were performed in seven patients with mTLE who were being evaluated for temporal lobectomy. Microdialysis probes attached to depth electrodes were implanted in the ipsilateral pes and body of the hippocampus. The interictal basal levels of glutamate (Glu), glutamine (Gln), γ -aminobutyric acid (GABA), glucose (Glc), and lactate (Lac) were measured by using the zero-flow method 5–60 h before any seizure activity and high-performance liquid chromatography (HPLC). Data were log-transformed and analyzed by using paired t tests and and correlation (z score) statistics. Results: In five patients, the epileptogenic focus was localized to the hippocampal pes, and in two patients, to both the pes and the body. The Glu levels were significantly elevated in the pes in all patients (10.4 ± 11.6 vs. 4.8 ± 5.6 µM; p < 0.003). Although Gln levels were similar in the pes and the body (730.4 ± 539.3 vs. 873.8 ± 765.8 µM), the Gln/Glu ratio was significantly decreased in the pes (108.9 ± 67.6 vs. 283.5 ± 213.9; p < 0.002). The lower Gln/Glu ratio indicates impairment in the glutamate–glutamine cycling in the pes. There were no significant differences between the pes and the body for all other measurements (GABA, 782.6 ± 870.3 vs. 461.5 nM; Glc, 2.3 ± 1.93 mM vs. 2.3 1.7; Lac, 6.0 ± 1.4 vs. 5.7 ± 4.0 mM; p > 0.05). However, for all neurometabolytes except Lac, there were highly significant correlations between their levels in the pes and the body (p < 0.02): 70–90% of the variability in the body was explained by the variability in the pes. Conclusions: The extracellular Glu levels were more elevated and the glutamate–glutamine cycling was more impaired in the pes versus the body of the epileptogenic hippocampus. The relative elevation of Glu in the pes may be a consequence of impaired Glu reuptake or reverse transporter release due to impaired energy supply. This is consistent with the increased impairment in the bioenergetic status in the anterior versus posterior hippocampus reported previously from our group. (Supported by NIH-PO1 NS 39092-01 and BIRCWH 1K12DA14038-01 for I. Cavus.)
D.07 COMBINED BOLD, FMRI, AND EEG RECORDINGS DURING SPONTANEOUS SPIKE–WAVE SEIZURES AND BICUCULLINE-INDUCED GENERALIZED TONIC–CLONIC SEIZURES IN WAG/RIJ RATS 1 Hrachya Nersesyan, 1 Sohil Patel, 1 Maritza Rivera, 2 Douglas L. Rothman, 2 Fahmeed Hyder, and 1,3 Hal Blumenfeld (1 Neurology, 2 Diagnostic Radiology, and 3 Neurobiology, Yale University School of Medicine, New Haven, CT) Rationale: Functional magnetic resonance imaging (fMRI) using blood oxygen level–dependent (BOLD) contrast has become a popular tool to investigate brain function. It is known that the BOLD signal depends on cerebral blood flow (CBF), oxidative metabolism, and blood volume. However, the relation between neuronal activity and the BOLD signal is not fully understood, particularly in abnormal conditions such as epileptic seizures. Our previous studies have shown that during somatosensory stimulation, there is a direct relation between the rate of neuronal spiking and oxidative metabolism derived from the BOLD signal. In addition, as we report in another presentation at this meeting, increases in neuronal firing during spike–wave discharges (SWDs) and
175
generalized tonic–clonic seizures in WAG/Rij rats are accompanied by commensurate increases in CBF. Here we sought to determine whether BOLD signals change during SWD and generalized tonic–clonic seizures are similar to or differ from the observed increases in neuronal firing and CBF. Methods: In vivo BOLD fMRI measurements were performed in a 7-T horizontal bore spectrometer in WAG/Rij rats during spontaneous SWDs and bicuculline-induced generalized tonic–clonic seizures under fentanyl/haloperidol anesthesia and neuromuscular blockade. EEG was recorded simultaneously with carbon filament radio-translucent electrodes to determine the timing of seizures. BOLD fMRI signals were then analyzed by comparing images acquired during seizures with baseline images, and superimposed on high-resolution anatomic images in the coronal plane. Results: Comparison of ictal and interictal epochs revealed bilateral increases in BOLD fMRI signals in cortical and subcortical structures during both SWDs and bicuculline-induced generalized tonic–clonic seizures. However, unlike generalized tonic–clonic seizures, SWDinduced increases in BOLD signal were less intense and more localized, mainly involving primary somatosensory cortex and corresponding thalamic nuclei. Interestingly, the BOLD fMRI signal changes during generalized tonic–clonic seizures were also not uniform throughout the brain, but had very intense areas of activation interrupted by relatively “silent” symmetric cortical areas. Conclusions: BOLD fMRI signals increase during both SWDs and generalized tonic–clonic seizures in parallel with increases in neuronal spiking and regional CBF. During both SWDs and generalized tonic– clonic seizures, the whole brain is not involved uniformly; rather, selective networks, particularly in the anterior brain regions, are most intensely involved. (Supported by NIH NS02060, NIH MH067528, NIH NS037203, and the Patterson Trust.)
D.08 EPILEPTOGENESIS INDUCED BY TRAUMATIC BRAIN INJURY IN RATS 1 Jari P.T. Nissinen, 1 Irina Kharatishvili, 2 Tracy K. McIntosh, and 1 Asla Pitk¨ anen (1 Department of Neurobiology, A.I.Virtanen Institute/University of Kuopio, Kuopio, Finland; and 2 Head Injury Center, Department of Neurosurgery, University of Pennsylvania, Philadelphia, PA) Rationale: Various brain insults, such as status epilepticus, encephalitis, and head trauma, are associated with an elevated risk of epilepsy later in life. Up to 50% of patients with penetrating brain injury can develop epilepsy. To investigate molecular and cellular mechanisms associated with increased epileptogenesis after head trauma, our first challenge was to develop an animal model of head trauma–induced epileptogenesis. Methods: Head trauma was conducted by fluid percussion injury in adult Spraque–Dawley rats (n = 26). Five weeks after head trauma, three cortical and one hippocampal electrodes were implanted for videoEEG recordings. To assess the occurrence of epileptic seizures, animals were monitored with continuous (24 h/day) video-EEG monitoring for 7 days at different time points (8–10, 19–20, 27–29, and 36–38 weeks) after trauma. Behavioral seizure severity was scored by modified Racine scoring scale. After video-EEG recordings, animals were perfused for histology. Results: Seventeen surviving animals were followed up in the long term. Seven (41%) of 17 animals with head trauma developed epilepsy. The mean seizure frequency was 0.30 ± 0.33 sz/day (range, 0.04–1.0; median, 0.18). The mean seizure duration was 102 ± 41 s (range, 61– 134 s; median, 86), and the mean behavioral seizure severity, 3.2 ± 1.7 (range, 1.0–5.0; median, 4.0). Histologic analysis showed that mossy fiber sprouting was denser in epileptic animals compared with nonepileptic traumatized or sham-operated control animals. Further, sprouting was most intense in the septal hippocampus ipsilaterally. Conclusions: This study shows for the first time that spontaneous seizures develop in a subpopulation of rats as a sequel of cortical fluid– percusssion injury. This new model of human posttraumatic epilepsy offers a tool to understand molecular mechanisms of epileptogenesis and can be used to test novel antiepileptogenic compounds in preclinical design. (Supported by Finnish Cultural Foundation, Sigrid Juselius Foundation, Vaajasalo Foundation.)
Epilepsia, Vol. 44, Suppl. 9, 2003
176
AES PROCEEDINGS
December 9, 2003 Plenary Session 1: Rhythms of the Brain: Neuronal Oscillations and Epilepsy 8:30 a.m.–11:00 a.m.
PL1.01 PLENARY SESSION: RHYTHMS OF THE BRAIN: NEURONAL OSCILLATIONS AND EPILEPSY 1 Paul A. Rutecki, 2 John Jefferys, 3 Charles L. Wilson, 4 John Huguenard, and 5 Mark Quigg (1 University of Wisconsin, Madison, WI; 2 University of Birmingham, Birmingham, England, United Kingdom; 3 UCLA, Los Angeles, and 4 Stanford University, Stanford, CA; and 5 University of Virginia, Charlottesville, VA) The brain and its neural networks are characterized by oscillatory behavior that underlies normal function and seizures. The mechanisms of normal oscillatory activity may relate to the generation of abnormal synchronization and seizure generation. In this symposium, different time domains of oscillations will be considered in relation to normal physiology and seizure generation. The first two presentations by Drs. Jefferys and Wilson will address fast oscillations (>30 Hz) in hippocampal neural networks under normal conditions, in animal models of epilepsy, and in people with intractable complex partial seizures. Have conventional techniques filtered out the important features of EEG activity? Dr. Huguenard will discuss the thalamocortical networks that underlie normal sleep oscillatory activity and can synchronize the cortex to produce generalized seizures. Attention will be paid to γ -aminobutyric acid (GABA)-mediated inhibition and known molecular abnormalities that lead to human epilepsy. How does thalamocortical circuitry become unstable to produce seizures? Besides oscillations that occur on the millisecond and second time frames, the brain is also affected by circadian rhythms. These longer oscillations may influence seizure susceptibility, and, conversely, seizures may disturb circadian or other rhythms. Dr. Quigg will address oscillations that contribute to human circadian and ultradian rhythms and how these processes may dictate when seizures occur. Effects of seizures on normal oscillatory activity will be considered. At the end of the symposium, the attendee will have reviewed new information regarding oscillatory behavior of neuronal networks and how neuronal oscillations underlie both normal brain function and epilepsy.
December 9, 2003 Plenary Session 2: Focal Lesions and Epilepsy: From Sequence to Cells to Surgery 8:30 a.m.–11:00 a.m.
PL2.01 PLENARY SESSION: FOCAL LESIONS AND EPILEPSY: FROM SEQUENCE TO CELLS TO SURGERY Peter B. Crino (Neurology, University of Pennsylvania, Philadelphia, PA) Focal brain lesions such as developmental and vascular malformations, tumors, and localized infections are common causes of intractable epilepsy in children and adults. Many lesions such as focal cortical dysplasias occur sporadically, whereas others, such as tuberous sclerosis complex, are autosomal disorders resulting from mutations in identified genes. Vascular malformations can be sporadic or be associated with single gene defects. Tumors such as astrocytomas and gangliogliomas likely reflect multigenic or multifactorial etiologies. Cerebral cysticercosis results from infection by Taenia solium. Despite distinct etiologies, these diverse disorders share selected histopathologic similarities that may provide clues to epileptogenesis in these lesions. Focal lesions are often amenable to surgical resection and improved seizure control. As a result of epilepsy surgery, there has been an increase in the availability of resected brain tissue samples from patients with focal epilepsy. Direct
Epilepsia, Vol. 44, Suppl. 9, 2003
analysis of tissue specimens by using cDNA arrays, protein quantification, pharmacology, and gene sequencing permits new strategies to study the pathogenesis of each lesion subtype. Over the past 10-year period, major advances have been made in defining the cellular constituents, embryonic lineage, and molecular pharmacologic profile of cells in focal brain lesions. These findings have permitted a more complete analysis of epilepsy phenotypes in these patients. Several findings that link these lesion subtypes include the presence of inflammatory cells, excessive glial proliferation, vascular endothelial changes, neuronal cell death, and altered expression of selected genes encoding neurotransmitter-receptor subunits. In addition, direct pharmacologic and electrophysiologic assays have shown that changes in neurotransmitter-receptor expression can result in enhanced excitability in several lesion types. Despite distinct etiologies, common cellular and molecular features are shared by developmental, vascular, neoplastic, and infectious brain lesions that may help identify pivotal mechanisms leading to epileptogenesis in these disorders. [Supported by Tuberous Sclerosis Alliance, National Institutes of Health (R01NS04502, R21 NS42334, R21-NS39938), Parents Against Childhood Epilepsy (PACE), and the Esther A. and Joseph Klingenstein Fund.]
December 9, 2003 Poster Session 2 11:00 a.m.–5:00 p.m. Status Epilepticus
2.001 CREUTZFELD/JACOB DISEASE WITH NONCONVULSIVE STATUS EPILEPTICUS AND RECEPTIVE APHASIA David A. Cohen, Ekrem Kutluay, Jonathan C. Edwards, and Ahmad Beydoun (Neurology, University of Michigan, Ann Arbor, MI) Rationale: To describe a rare manifestation of sporadic Creutzfeld– Jacob Disease (CJD) with nonconvulsive status epilepticus and receptive aphasia. CJD typically presents as a rapidly progressive dementia often accompanied with myoclonus. Typical EEG findings consist of generalized and symmetric periodic sharp waves recurring at 0.5–2.0 Hz. Isolated cases of aphasia have been described in CJD, but aphasic status epilepticus has not. Methods: The patient was a 26-year-old man whose initial symptoms consisted of subtle cognitive changes manifested by making spelling errors in e-mails to his friend. Over the subsequent 3 weeks, he had a rapid deterioriation manifested by forgetfulness, confusion, and difficulty performing simple tasks and was brought to the hospital by his parents. On admission, his examination was significant for a Wernicke-type aphasia with an otherwise normal neurologic examination. Results: EEG was diagnostic of partial nonconvulsive status epilepticus of left hemispheric onset. Several anticonvulsants were tried including oxcarbazepine, levetiracetam, phenobarbital, topiramate, phenytoin, and valproic acid. The status was refractory to medical treatment and the patient required midazolam (MDL) coma. After MDL taper, the EEG was characterized by PLEDs over the left hemisphere with occasional electrographic seizure activity. At times the left hemispheric PLEDs evolved to BIPEDs, maximal over the left hemisphere. A brain biopsy was consistent with a spongiform encephalopathy. Quinacrine, an experimental therapy for CJD, was administered but had no effect on the EEG or the clinical course. The patient was discharged to hospice care and died 2.5 months after his initial presentation. Autopsy revealed sporadic CJD, MMII subtype. Conclusions: CJD can present with a variety of signs and symptoms. Our patient is unique in that he presented with nonconvulsive status epilepticus and a Wernicke’s aphasia. Both of these entities are exceedingly rare presentations of CJD, and never before has a patient been described with both of these findings. In patients with unexplained status epilepticus or progressive aphasia, CJD should be considered in the differential diagnosis.
