Clinical Science


Pregabalin Drug Interaction Studies: Lack of Effect on the Pharmacokinetics of Carbamazepine, Phenytoin, Lamotrigine, and Valproate in Patients with Partial Epilepsy

Martin J. Brodie, Elaine A. Wilson, David L. Wesche, Christine W. Alvey, Edward J. Randinitis, Edward L. Posvar, Neil J. Hounslow, Nicola J. Bron, G.L. Gibson, and Howard N. Bockbrader

Pregabalin (Lyrica) is a new drug that has been approved in the EU as add-on treatment for partial seizures, and as treatment of peripheral neuropathic pain, in adults. Because many patients with partial seizures require more than one antiepileptic drug (AED) to achieve the best seizure control they can, it is important to determine whether new drugs for treating partial seizures may interact with commonly used AEDs. This paper reports the results of several drug–drug interaction studies of pregabalin and commonly used AEDs. In these studies, pregabalin was given to patients who were each taking one AED, either valproate, phenytoin, lamotrigine, or carbamazepine, to treat their epilepsy. Adding pregabalin to monotherapy with valproate, phenytoin, lamotrigine, or carbamazepine had no effect on the concentrations in the blood of any of these four drugs. Likewise, when concentrations of pregabalin in the blood were measured in these same patients, the concentrations were similar to those seen in the blood of subjects from earlier studies who had received pregabalin by itself. These findings indicate that pregabalin does not interact with any of these four common AEDs. The combinations of pregabalin with either valproate, phenytoin, lamotrigine, or carbamazepine were generally well tolerated by the patients in these studies, and when side effects did occur, they were typically mild to moderate and resolved quickly. Together, these studies demonstrate that pregabalin may be safely added to therapy with valproate, lamotrigine, phenytoin, or carbamazepine without concern for drug–drug interactions.


Ethinyl Estradiol, Not Progestogens, Reduces Lamotrigine Serum Concentrations

Arne Reimers, Grethe Helde, and Eylert Brodtkorb

This work studies the interaction between the antiepileptic medicine lamotrigine (LTG, Lamictal), and hormonal contraception. The participants were females using either no hormonal contraception (18 patients), an ethinyl estradiol (EE)-containing (11 patients) or a progestogen (PG)-only-containing compound (16 patients). Drug fasting blood levels of LTG were analyzed. Comedication with other drugs known to alter LTG blood levels was not allowed. Some patients changed group after the first blood sample and thus served as their own controls. We found that women using EE had only half as much LTG in their blood compared to the women in the control group or the PG-group. There was no difference between women using PG, and controls. Also, there was no difference between women using either oral, topical or parenteral PG. Five women switched from the control-group to the EE-group and experienced a considerable reduction in their blood levels of LTG. A rise of LTG in blood towards control levels was seen in the two women who changed from EE to PG. We conclude that it is the EE-component of oral contraceptives which reduces LTG blood levels. The PG-only compounds did not influence LTG blood levels in this study. These findings should be considered when counseling women with epilepsy in childbearing age.


Characteristics of Male Veterans with Psychogenic Nonepileptic Seizures

Barbara A. Dworetzky, Andreja Strahonja-Packard, Christopher W. Shanahan, Jeanette Paz, Barbara Schauble, and Edward B. Bromfield

In the population of patients referred to epilepsy centers with seizures that are not adequately controlled by medications, 10-58 % do not turn out to have epilepsy, and instead have psychologically based spells referred to as psychogenic nonepileptic seizures (PNES). These patients are usually young women, by a ratio of 3 to 1. Male Veterans with PNES were compared to those with a definite history of epileptic seizures. A small number of patients who may have had both types of seizures were excluded from analysis. All patients were evaluated between 1997 and 2000 and were asked questions about whether they had ever experienced a traumatic brain injury, whether they had taken antiseizure medications (and if so, how many), if they experienced depression, anxiety, or used alcohol or illicit drugs, or had chronic pain or post-traumatic stress disorder. In addition, the medical chart was reviewed to confirm whether they received compensation from the Veteran's Administration for their diagnosis, or had normal or abnormal neuroimaging, neurological examinations, or electroencephalograms (brain waves). Men with PNES were significantly younger, reported more frequent events, and were more likely to report chronic pain, anxiety, and post-traumatic stress disorder than patients with epileptic seizures. They did not differ with respect to whether they reported the loss of bladder control during a seizure, or whether they received disability compensation from the Veteran's Administration.


Photic- and Pattern-induced Seizures: Expert Consensus of the Epilepsy Foundation of America Working Group

Graham Harding, Arnold J. Wilkins, Giuseppe Erba, Gregory L. Barkley, and Robert S. Fisher

In August 2004, the Epilepsy Foundation of America convened a workshop to begin to develop an expert consensus on photosensitive seizures. Literature and data were reviewed, and consensus derived from discussion. A flash is a potential hazard if it has luminance at least 20 cd/m2, occurs at a frequency of least 3 Hz, and occupies a solid visual angle of at least 0.006 steradians (about 10% of the central visual field or 25% of screen area at typical viewing distances). A transition to or from saturated red also is considered a risk. A pattern with the potential for provoking seizures contains clearly discernible stripes, numbering more than five light–dark pairs of stripes in any orientation. When the light–dark stripes of any pattern collectively subtend at the eye from the minimum expected viewing distance a solid angle of more than 0.006 steradians, the luminance of the lightest stripe is greater than 50 cd(m2, and the pattern is presented for 0.5 seconds or longer, then the pattern should display no more than five light–dark pairs of stripes, if the stripes change direction, oscillate, flash or reverse in contrast; If the pattern is unchanging or smoothly drifting in one direction, no more than eight stripes. These principles are easier to apply in the case of fixed media, for example, a prerecorded TV show, which can be analyzed frame-by-frame, as compared to interactive media. A consensus view of stimuli likely to provoke visually-evoked seizures can be developed.


Photic- and Pattern-induced Seizures: A Review for the Epilepsy Foundation of America Working Group

Robert S. Fisher, Graham Harding, Giuseppe Erba, Gregory L. Barkley, and Arnold Wilkins

This report summarizes background material presented to a consensus conference on visually-provoked seizures, convened by the Epilepsy Foundation of America. A comprehensive review of literature was performed. Photosensitivity, an abnormal EEG response to light or pattern stimulation, occurs in approximately 0.3-3% of the population. The estimated prevalence of seizures from light stimuli is approximately 1 per 10,000, or 1 per 4,000 individuals 5-24 years old. People with epilepsy have a 2-14% chance of having seizures precipitated by light or pattern. In the Pokemon cartoon incident in Japan, 685 children visited a hospital in reaction to red-blue flashes on broadcast television. Only 24% who had a seizure during the cartoon had previously experienced a seizure. Photic or pattern stimulation can provoke seizures in predisposed individuals, but such stimulation is not known to increase the chance of subsequent epilepsy. Intensities of 0.2-1.5 million candlepower are in range to trigger seizures. Frequencies of 15-25 Hz are most provocative, but the range is 1-65 Hz. Light–dark borders can induce pattern-sensitive seizures, and red color also is a factor. Seizures can be provoked by certain TV shows, movie screen images, videogames, natural stimuli (e.g., sun on water), public displays, and many other sources. Recommendations on reducing risk of seizures have been developed by agencies in the UK, Japan, and the International Telecommunications Union, affiliated with the UN. The Epilepsy Foundation of America has developed a consensus of medical experts and scientists on this subject, reported in an accompanying work.


Clinical Features of Patients with Posterior Cortex Epilepsies and Predictors of Surgical Outcome

Charles L. Dalmagro, Marino M. Bianchin, Tonicarlo R. Velasco, Veriano Alexandre Jr., Roger Walz, Vera C. Terra-Bustamante, Luciana M. Inuzuka, Lauro Wichert-Ana, David Araújo Jr., Luciano N. Serafini, Carlos G. Carlotti Jr., João A. Assirati Jr., Hélio R. Machado, Antonio C. Santos, and Américo C. Sakamoto

Epilepsies originating from the posterior brain areas such as occipital, parietal, or occipital border of the temporal lobe, or from any combination of these regions, are known as posterior cortex epilepsies (PCEs). It is important to know that these epilepsies can be surgically treated. In this study we analyzed the clinical characteristics of all PCE patients referred for surgery at our center from 1994 to 2003, and searched for predictors of surgical outcome. A total of 81 patients were retrospectively analyzed. They were divided in surgical and nonsurgical groups, and their main characteristics were compared. Detailed analysis of all variables of the surgical cases were performed in the search for predictors of seizure outcome. In this study the main risk factors for PCEs included gliosis (34.56%), malformation of cortical development (33.33%), tumors (8.64%), brain trauma (3.70%), Sturge-Weber disease (4.93%), vascular malformations (3.70%), family history of epilepsy (3.70%), history of CNS infections (2.46%), and low IQ (2.46%). Out of the 81 patients, 44 were submitted to surgery at the time of the completion of this study. Surgical treatment was highly effective in improving seizures (p(0.001) when compared to previous pharmacological treatment alone. Twenty-eight patients (65.11%) became seizure free after surgery versus none in the nonsurgical group. Regarding outcome predictors, patients with shorter duration of epilepsy and those without neurological abnormalities on clinical exam had higher chances of favorable evolution.


Localizing Value of Ictal–Interictal SPECT Analyzed by SPM (ISAS)

Kelly A. McNally, A. LeBron Paige, George Varghese, Heping Zhang, Edward J. Novotny, Jr., Susan S. Spencer, George Zubal, and Hal Blumenfeld

Epilepsy is a devastating illness with a major economic and psychosocial impact. While medications are often beneficial, millions suffer from epileptic seizures that are not stopped by medications. Hope for a cure can be offered to these patients through selective surgery to remove the region of seizure onset. However, prior to surgery, the region of seizure onset must be identified with a high level of certainty. We have developed a new method of analyzing noninvasive brain images to identify the region of seizure onset. The method, called ISAS (which stands for Ictal–interictal SPECT Analyzed by SPM) uses a statistical approach comparing blood flow changes in the brain of epilepsy patients to normal individuals. Full technical details of the methods used for this study are provided at so that other centers can easily implement ISAS. The method emphasizes the need to identify a single unambiguous region of seizure onset. The analysis is based on objective measurement of blood flow changes during seizures, and eliminates much of the bias introduced by subjective readings of brain images by human eye. We tested ISAS in a group of patients and found that when ISAS detected a single unambiguous region, it was always the correct location of seizure onset. This was confirmed when epilepsy surgery at these locations cured the patients of their seizures. We are hopeful that noninvasive brain imaging and improved analysis through ISAS can offer a cure for many patients with epilepsy.


Diffusion Tensor Imaging in Late Posttraumatic Epilepsy

Rakesh K. Gupta, Sona Saksena, Atul Agarwal, Khader M. Hasan, Mazhar Husain, Vikas Gupta, and Ponnada A. Narayana

Epilepsy is common sequelae of traumatic brain injury (TBI). We present our experience with Diffusion tensor imaging (DTI), a relatively newer MRI technique in the differentiation of patients with chronic TBI with epilepsy from without epilepsy. The results of this study are based on the quantitative DTI data from twenty three chronic TBI patients who underwent DTI scans (with late post traumatic epilepsy n/14, without epilepsy n/9). Eleven healthy age matched controls were also scanned using the same protocol. Region-of-interest (ROI) analysis was performed within and beyond T2/fluid attenuating inversion recovery (FLAIR) visualized abnormality and the corresponding contralateral normal appearing region for determining the DTI derived metrics, fractional anisotropy (FA), mean diffusivity (MD). A significant reduction in regional mean FA ratio along with significant increase in regional mean MD ratio was observed in TBI patients as compared to controls. The mean regional FA ratio was significantly lower in TBI patients with epilepsy than in those without epilepsy. The tissue volume with low FA value was also found to be higher in TBI patients with epilepsy than without. Our results indicate that epileptogenesis in TBI is associated with more severe microstructural brain parenchymal damage as suggested by DTI. These findings may provide new insights into the pathophysiology of epilepsy and in planning specific surgical strategies related to post traumatic epilepsy (PTE).


Predicting Posttraumatic Epilepsy with MRI: Prospective Longitudinal Morphologic Study in Adults

Anna Messori, Gabriele Polonara, Flavia Carle, Rosaria Gesuita, and Ugo Salvolini

Posttraumatic epilepsy (PTE) is a well-recognized complication of head injury, and there is open discussion on its pathogenesis. Focal brain lesions documented in the acute stage represent accepted risk factors, but the relation between PTE and posttraumatic brain abnormalities as revealed by magnetic resonance imaging (MRI) has not been established. We prospectively evaluated morphological risk factors for PTE by using a panel of MRI sequences in serial assessments up to 2 years after trauma. MRI abnormalities indicative of gliosis and hemosiderin, respectively, were assessed in the images of the 135 adults who completed the study; clinical follow-up was 5-10 years. Statistical analysis of the results showed that sequelae of focal brain lesions that required surgical treatment (subdural hematomas-contusions) were a significant PTE risk factor, as were some, but not any sequelae of nonsurgical hemorrhagic contusions. With the combined use of different MRI sequences, we identified three patterns of morphological appearance of these, and those showing hemosiderin dregs completely surrounded by gliosis since early follow-up MRI (50% of the total number of lesions) were not, at difference with those showing gliosis incompletely surrounding hemosiderin dregs and time-related changes from incomplete to complete gliosis wall around hemosiderin. Although our results require confirmation from further MRI studies because of several limitations, it would appear that follow-up MRI examination in the early chronic stage can differentiate among low, intermediate and high-risk sequelae of head injury, and that follow-up MRI studies can yield new evidence to the still open debate on posttraumatic epileptogenesis.


Cerebral Damage in Epilepsy: A Population-based Longitudinal Quantitative MRI Study

Rebecca S. N. Liu, Louis Lemieux, Gail S. Bell, Sanjay M. Sisodiya, Philippa A. Bartlett, Simon D. Shorvon, Josemir W. A. S. Sander, and John S. Duncan

There have been concerns that epileptic seizures may cause brain damage. Studies in animals have shown that even a small number of relatively mild seizures may damage brain structures such as the hippocampus, which is important for memory function. Only recently have researchers performed repeated magnetic resonance imaging (MRI) brain scans on individuals with epilepsy to determine whether seizures cause brain damage in humans, shown as loss in brain volume. In the largest MRI study investigating whether seizures damage the brain, we performed brain scans on 190 patients of different ages, 3.5 years apart. We measured the whole brain and specific structures within the brain, and compared changes in their brains, with those seen in healthy control subjects of similar ages. Previous brain injuries, such as convulsions related to high temperatures in early childhood were associated with lower brain volumes at the start of the study. The overall rates of brain volume loss were similar in individuals with and without epilepsy, and amongst patients with different kinds of epilepsy. The main factor influencing the rate of volume loss was the age of the patient and not the number of seizures. There was significant shrinkage of a part of the brain in 17% of those with epilepsy, compared with 7% of those without epilepsy. We concluded that epilepsy is associated with brain shrinkage in only a minority of those with the condition. More studies are needed to identify who is at particular risk of this occurring, and how this may be prevented.


Unusual Findings in Brain Biopsies of Two Patients with Acute Magnetic Resonance Imaging Lesions Associated with Focal Seizures

Shearwood McClelland III, Jenny M. Libien, Steven S. Chin, David J. Adams, Stanley R. Resor, Jr., Stephen Chan, and Robert R. Goodman

Patients with focal seizures often have MRI abnormalities in the brain region of their presumed seizure focus. Neoplasms, ischemic infarctions, inflammatory processes, and other specific pathologic entities have been diagnosed by biopsies of such MRI abnormalities. Two patients with this presentation had brain lesion biopsies with a leading presumptive diagnosis of glial neoplasm but were found to have indistinct histopathology. Each patient presented with focal seizures (right parietal region, right hippocampus) corresponding with focally increased T2 signal on MRI. In both patients, the preoperative clinical suspicion was for neoplastic or inflammatory processes. Several weeks after seizure onset, craniotomy in patient 1 and stereotactic needle biopsy in patient 2 revealed mild gliosis with reactive vascular changes and perivascular hemosiderin deposition, not diagnostic of but compatible with venous congestion (or possibly venous thrombosis). Postoperatively, patient 1 had brief sensory seizures that stopped 5 months after surgery, while patient 2 did not develop subsequent seizures. Both patients had normalization of their MRI (except for postoperative changes) and have remained seizure-free. We describe two patients who had brain biopsies of striking focal increased T2 signal MRI abnormalities associated with seizures. Pathologic findings contradicted our preoperative suspicions (neoplasm or inflammatory process), compatible with (but not conclusive for) subacute venous congestion/thrombosis. These findings indicate that patients with seizures may have an associated discrete intra-axial MRI lesion that is not neoplastic. To our knowledge, this is the first report of focal seizure-associated MRI lesions with biopsy findings compatible with venous congestion/thrombosis.


Intellectual Prognosis of Status Epilepticus in Adult Epilepsy Patients: Analysis with Wechsler Adult Intelligence Scale–Revised

Naoto Adachi, Kousuke Kanemoto, Reimi Muramatsu, Masaaki Kato, Nozomi Akanuma, Masumi Ito, Jun Kawasaki, and Teiichi Onuma

Various systemic and brain condition can cause prolonged seizures, termed status epilepticus (SE). After recovering from an episode of SE, some patients suffer from cognitive dysfunction as well as other serious neurological symptoms. However, it has not been well studied whether and to what extent SE could lead to an intellectual deterioration in epilepsy patients who have already been treated with antiepileptic drugs (AED). To investigate this issue, we tested intellectual function before and after an episode of SE in 15 adult epilepsy patients. As an index of intellectual function, we used full scale IQ of the Wechsler Adult Intelligence Scale-Revised. The change of FIQ before and after the SE was compared with the differences between the two test trials performed in an equivalent interval in 40 clinically matched patients without SE. We found no significant changes, either increase or decrease, in intellectual ability as a consequences of SE. Furthermore, there were no significant associations between FIQ changes and SE-related variables, i.e., age at the episode of SE, duration of SE, type of SE. Our findings suggest that in epilepsy patients, SE per se does not often result in long-lasting intellectual dysfunction.


Characterizing Magnetoencephalographic Spike Sources in Children with Tuberous Sclerosis Complex

Koji Iida, Hiroshi Otsubo, Ismail S. Mohamed, Chiyuki Okuda, Ayako Ochi, Shelly K. Weiss, Sylvester H. Chuang, and O. Carter Snead III

In tuberous sclerosis complex (TSC), tubers and sclerotic patches on the brain cause seizures. Magnetoencephalography (MEG) measures magnetic fields caused by electrical activity in cells and locates centers of seizure activity as spike sources (SSs) for surgical planning. We evaluated data from seven children with TSC, characterized the MEG SS results, and correlated them with prolonged video-electroencephalography (VEEG) and MRI. We classified MEG results according to whether the SS were in clusters or were scattered, and whether the SSs were anatomically related to tubers. Two patients had a single, unilateral SS cluster with other scattered SSs. The clustered SSs corresponded to prominent tubers on MRI and active seizure areas on VEEG. Two patients had bilateral clustered SSs, with other scattered SSs. The cluster locations partly overlapped multiple prominent tubers. VEEG discharges recorded between seizures were bilateral or diffuse. Three patients had only bilateral scattered SSs. These scattered SSs partly overlapped seizure-onset regions and between-seizure-onset regions on VEEG. In one patient with equally bilateral scattered SSs, the SSs overlapped a prominent tuber and seizure-/between-seizure-onset zones in the right frontal brain. The significance of the study is that the characteristics of noninvasively determined MEG SSs can help define the primary epileptogenic zone (single cluster, scattered SS, in unilateral hemisphere) and reduce the extent of invasive VEEG needed for surgical planning. When MEG SSs are focal, intracranial VEEG grids need to cover only the focal area; when MEGSSs are scattered, however, VEEG must cover at least the entire area of MEG SSs.


Hemispherectomy for Catastrophic Epilepsy in Infants

Jorge A. González-Martínez, Ajay Gupta, Prakash Kotagal, Deepak Lachhwani, Elaine Wyllie, Hans O. Lüders, and William E. Bingaman

The removal of one of the cerebral hemispheres (also called hemispherectomy) is performed successfully to treat medically intractable hemispheric epilepsy in adolescent and older children, providing remarkable results in terms of seizure control and quality of life. Nevertheless, experience and surgical indications in extremely young children are still unknown. In fact, the technical challenges of this drastic procedure, together with a reluctance to refer very young patients for “elective” surgery, often delays surgical treatment despite the need for earlier intervention. Here, we report our single-surgeon experience with 18 children younger than two years-old who had hemispherectomy procedures to treat hemispheric life threatening epilepsy. In the end of the follow-up period, complications occurred in 16.7% with no death. Twelve out of 22 procedures (54.5%) resulted in incomplete disconnection (connected epileptic tissue left behind). Type of surgical procedure and bilateral electroencephalogram (EEG) abnormalities were not associated with persistent seizures after surgery. Incomplete disconnection was the only variable statistically associated with persistent seizures after surgery. In conclusion, hemispherectomy for seizure control provides excellent and dramatic results with a satisfactory complication rate. Our results support the concept that early surgery should be indicated in highly selected patients with severe, medically intractable and hemispheric epilepsy. Safety factors as an expert team in pediatric intensive care unit, neuro-anesthesia and an experience pediatric epilepsy neurosurgeon are mandatory.


The Accuracy of Outcome Prediction Models for Childhood-onset Epilepsy

Miranda Geelhoed, Anne Olde Boerrigter, Peter Camfield, Ada T. Geerts, Willem Arts, Bruce Smith, and Carol Camfield

On the day of diagnosis it would be highly desirable to know which children with epilepsy will eventually become seizure free, be able to stop daily medication treatment and have remission from (outgrow) their epilepsy. Two large studies (Nova Scotia and the Netherlands) identified children when their epilepsy started and then followed them for at least 5 years to see which children had remission. We combined these two studies to allow consideration of 1055 children with epilepsy and used powerful statistical methods (classification tree models and stepwise logistic regression) to see if clinical aspects of the child's epilepsy at diagnosis and results of electroencephalographic (EEG) testing could be used to predict which children would have remission from their epilepsy and which would not. Overall 59% of the children were in remission at the end of follow up (seizure-free and no longer receiving daily medication). The statistical models allowed us to predict the outcome correctly for 70%. Factors that helped to predict the outcome were the type of epilepsy (syndrome), age of first seizure (12 years of age, number of seizures before treatment, abnormalities in neurological function and intelligence. However even in this large group of children our prediction schemes were incorrect in 30%-some were expected to remit and did not while others had a surprise remission when it was not anticipated. Based on currently available clinical and EEG variables, predicting the outcome of childhood epilepsy is difficult and appears to be incorrect in about one of every three patients.


Safety and Tolerability of the Ketogenic Diet in Pediatric Epilepsy: Effects of Valproate Combination Therapy

David A. Lyczkowski, Heidi H. Pfeifer, Soumit Ghosh, and Elizabeth A. Thiele

The ketogenic diet (KGD) is a high-fat, low-carbohydrate, restricted-protein diet that can effectively treat intractable epilepsy. Valproate (VPA) is a widely-used antiepileptic drug. Like other treatments, both the KGD and VPA are associated with side effects. Previous limited studies have suggested that the combination of the KGD and VPA might possibly increase the risk of serious side effects including pancreatitis, kidney dysfunction, and liver toxicity. Because of the potential for such side effects, some have avoided using these two therapies in combination. We studied 71 children with epilepsy treated with the KGD, of whom 24 were also taking VPA. In general, side-effect profiles and efficacy of the KGD were not affected by VPA. The patients on VPA and KGD cotherapy did not show signs of pancreatitis, kidney dysfunction, or serious liver toxicity. In some patients, the combination of VPA and KGD provided optimal seizure control. Cotherapy with the KGD and VPA should not be excluded because of concerns about safety and tolerability.


Ketogenic Diet in Patients with Dravet Syndrome

Roberto Horacio Caraballo, Ricardo Oscar Cerósimo, Diego Sakr, Araceli Cresta, Nidia Escobal, and Natalio Fejerman

The ketogenic diet (KD) has been used as a therapeutic alternative to antiepileptic drugs (AEDs) for refractory epilepsy. The diet consists of an intake of three or four times as much fat as carbohydrates and protein combined. In this retrospective study, we evaluate the efficacy and tolerability of the KD in patients with Dravet Syndrome (DS), a severe myoclonic epilepsy occurring in infants. Between March 1, 1990 and August 31, 2004, 52 patients who met diagnostic criteria of DS were enrolled in our study. Twenty of them were placed on the KD using the Hopkins protocol and followed for a minimum of one year. Three of the 20 original children stayed on the diet for 12 months, four children for 2 years, four children for 3 years and two children for 4 years. One year after initiating the diet, 13 of the initial patients (65%) remained on the diet. Two patients (15%) were seizure free and eight children (61.7%) had a 75-99% decrease in seizures and the remaining three children (23%) had a 50% to 74% decrease in seizures. Thus, 1 year after starting the diet, 10 children (77%) had achieved a (75% decrease in their seizures. Four patients have been off the diet for more than two years; one of them is seizure free, two have sporadic seizures and one, who abandoned the diet after two years of adhering to it, relapsed. No differences in seizure control when compared to age, sex, or seizure type were found. Considering the severity and intractability of seizures in patients with Dravet syndrome, the fact that 11 of the 13 children who remained on the diet had a significant reduction in number of seizures shows that the KD is at present an interesting therapeutic alternative. Even in patients in whom seizure reduction was not dramatic, quality of life improved and in all of them the number of AEDs was lowered to one or two. We consider that children with Dravet syndrome should be offered the KD immediately after failing three adequate trials of AEDs.