Epilepsy Program

Abstracts presented at the American Epilepsy Society meeting 2012

• Data-Driven Approach to Augment Decision Support for Pediatric Neurology in EMR Era
• Clinician Adherence to Practice Parameters on First Nonfebrile Seizure in Children
• Detection of Epileptogenicity with Non-Linear Analysis of Electroencephalographic Signals
• A Hybrid Wavelet Transform and Machine Learning Algorithm for EEG Spike Detection

DATA-DRIVEN APPROACH TO AUGMENT DECISION SUPPORT FOR PEDIATRIC NEUROLOGY IN EMR ERA

K. Mane, T. Loddenkemper, P. Owen, M. Mikati, I. Fernández, M. Tennison, A. Leviton

UNC-Chapel Hill, Boston Children’s Hospital

RATIONALE:

Epilepsy patient data stored in the electronic medical record (EMR) data has the potential to offer enhanced patient care, if presented in an easy to understand way. Unfortunately, the large volume of data presented in raw format (as numbers) are hard to process and interpret, thus limiting their use. Data displays and computers can aid the physician during everyday practice. Here, we describe how data displays and decision support systems can work in coordination with each other. We show how patient data and comparative population evidence can be visually displayed in a way that enables the physician to make sense of this information in a fast and intuitive manner.

METHODS:

Many variables are recorded in the EMR of children with epilepsy: seizure types and frequencies, medications, outcomes (measured by increase/decrease in seizure frequencies), potential medication side-effects and their severity. Patient enrollment continues at three different pediatric epilepsy centers. Computational approaches can be used to process, aggregate, and build displays of data. Visual representation help the physician better understand patient data, and aggregated evidence from comparative population. Embedded interactions empower the physician to customize the choice of comparables. The display program shows an approach to make use of the existing EMR resources without having the physician search through the patient's file.

RESULTS:

We have now collected pilot data on 20 epilepsy patients. Different data views can be built to reveal patterns in the data. Figure 1 shows dashboard with patient-level data, while Figure 2 highlights an approach to display and use comparative population treatment evidence to assist the physician in decision making. Predictive modeling approaches can be use comparable patient data to identify trend in medication response. Within a single dashboard (Figure 1 or Figure 2), different data views linked together give flexibility for data overlay, and for filter and syncing all data views. Such data displays are likely to help the physician gain a quick understanding of the characteristics of the patient data at the time of visit. Such interactive data displays have the potential to reduce information overload, free up the physician's cognition for higher level data processing, and aid informed decision making.

CONCLUSIONS:

This multicenter pilot study is designed to highlight the role that data displays and a clinical decision support program can play to bridge the complementary skill sets of humans and computers (as an external aid) to rapidly derive useful information from a large dataset. External aid in the form of interactive data displays can be used to augment the information processing abilities of the physician and facilitate decision making.

CLINICIAN ADHERENCE TO PRACTICE PARAMETERS ON FIRST NONFEBRILE SEIZURE IN CHILDREN

J. M. Avallone, A. Patel, N. Baumer, A. Sansevere, L. Doyle, N. Mehta, S. Choi, A. Pinto, T. Loddenkemper

Boston Children's Hospital

RATIONALE:

Monitoring of practice parameter adherence and assessment on the effect on outcome in pediatric patients is limited. This study aims to assess adherence to American Academy of Neurology (AAN) practice parameters in children with first nonfebrile seizure at our center.

METHODS:

We retrospectively reviewed charts of pediatric patients age 1 month - 21 years who presented to our Emergency Department for first time convulsive seizure between January 2010 and December 2011. Patients with ICD 9 codes 345.1, 345.9, and 780.39 were included. Patients who presented with a first time febrile seizure or had a known brain tumor were excluded. Using a standardized assessment tool, we reviewed adherence to the AAN practice parameters related to the recommended ordering of routine electrolytes, toxicology screen, EEG, lumbar puncture, urgent head CT, and brain MRI imaging. When a diagnostic study was ordered that did not adhere to guidelines, we determined if the test was abnormal and if the abnormality could explain the seizure or required urgent management.

RESULTS:

One hundred ten patients met inclusion criteria. Chart review showed that physicians followed AAN guidelines on ordering of toxicology screens (88%; 95% C.I. 82-94%), EEG (96%; 95% C.I. 93-100%), lumbar puncture (96%; 95% C.I. 93- 100%), and non-urgent brain MRI imaging (95%; 95% C.I. 95-99%) well. Although the majority of physicians followed the guidelines regarding routine electrolytes (79%; 95% C.I. 71-87%) and emergent head CT imaging (75%; 95% C.I. 67-84%), adherence to these parameters was lower. There was non-adherence to the parameter on ordering of routine electrolytes in 21% of the patients. Electrolyte results in these patients revealed mild abnormalities in 56% (95% C.I. 34-77%) including hyponatremia, hyperkalemia, and hyperglycemia. Electrolyte findings did not explain the etiology of the seizure or require intervention. There was nonadherence to the parameter on obtaining emergent head CT imaging in 25% of the patients. Of those, 77% had head CT imaging that was not indicated, while emergent imaging was not performed when indicated in 23% of the patients. 65% of the emergent head CTs performed when not indicated were ordered by outside emergency departments prior to transfer to our center. Of the patients who underwent unnecessary head CT imaging, 80% were normal and 20% (95% C.I. 6-44%) showed an abnormal result, which did not require urgent intervention (i.e. chronic changes, PVL).

CONCLUSIONS:

Child neurologists and pediatricians adhere to the majority of AAN practice parameters on children with first nonfebrile seizure. Occasional cases of nonadherence to guidelines including ordering of electrolytes or emergent head CTs within our and referring emergency departments did not alter urgent management supporting available guidelines. As a result of this review, we will implement a hospital wide standardized clinical management and assessment plan for first time nonfebrile seizure in children and continuously assess adherence. Additional efforts to include referring centers into this treatment plan are underway.

DETECTION OF EPILEPTOGENICITY WITH NON-LINEAR ANALYSIS OF ELECTROENCEPHALOGRAPHIC SIGNALS

W. Bosl, I. Sánchez Fernández, T. Loddenkemper

Boston Children's Hospital

RATIONALE:

The diagnosis of epilepsy relies on surrogate markers for epileptogenicity such as the presence of repeated seizures or spikes. The aim of this study is to test the usefulness of a type of non-linear mathematical analysis: recurrence quantitative analysis (RQA) of electroencephalogram (EEG) signals as a direct biomarker of epileptogenicity in the brain.

METHODS:

We enrolled 76 patients that met the following inclusion criteria: 1) age 1 month to 21 years, 2) normal neuropsychological development. 38 patients (cases) had several episodes of electro-clinically documented absence seizures. The remaining 38 patients (controls) met the following criteria: 1) at least one EEG study performed because of the clinical suspicion of seizures, 2) normal EEG study results, 3) an electro-clinical presentation that was not consistent with epilepsy based on a thorough evaluation. We identified and collected sample EEG segments of 10-30 seconds duration each: 1) in the cases, three segments were collected: baseline without spikes, hyperventilation without spikes, and spikes; 2) in controls, two segments were collected: baseline without spikes, and hyperventilation without spikes. RQA was used to compute 19 different nonlinear measures for each of the 19 EEG channels, resulting in 19X19 or 361 features. From these, 20 of the most statistically relevant values were automatically selected and used for classification.

These nonlinear features were used in two different machine learning algorithms to classify the EEG segments as belonging to cases (epileptic patients) or controls (non-epileptic patients). The concordance of this diagnostic classification was compared to the classification made by the clinical epileptologist (gold standard).

RESULTS:

The results of the RQA and of the human epileptologist were highly correlated as shown in table 1. Hyperventilation increases the levels of epileptogenicity of both cases and controls, making them more difficult to differentiate. In order to control for this potential confounder, we performed a subgroup analysis eliminating the tracings during hyperventilation in both cases and controls. The resulting classification of patients into cases and controls was even more correlated with the diagnosis of the human epileptologist (Table 2).

CONCLUSIONS:

Analysis of EEG signals with RQA was highly correlated with the classification of the human epileptologist (gold standard). Non-linear analysis of short EEG signals of patients with epilepsy, even in epilepsy patients with normal EEG on visual inspection, may be a useful method to monitor and diagnose epilepsy.

A HYBRID WAVELET TRANSFORM AND MACHINE LEARNING ALGORITHM FOR EEG SPIKE DETECTION

V. Chavakula, I. Sánchez Fernández, J. Peters, W. Bosl, A. Rotenberg, T. Loddenkemper

Boston Children’s Hospital

RATIONALE:

Manual EEG spike quantification is frequently a subjective and time-consuming process. We have previously developed a double threshold wavelet algorithm for quantification of epileptic spikes, and now report improvements to this automated algorithm via addition of a machine learning component in order to further enhance sensitivity while substantially decreasing the rate of false positives.

METHODS:

We applied our algorithm to EEG recordings from eleven pediatric patients (average age of 8 years) with documented ESES (Electrical Status Epilepticus in Sleep). For each patient, one segment of three to eight minutes duration was first analyzed using a double threshold wavelet method to highlight time points corresponding to potential spike events. For each detected spike, the calculated features were the amplitude, slope, and spike angle in every channel. These features were input to a machine learning algorithm to determine whether the time point was indeed a spike. Machine learning classifier training and testing for each set was performed using a 10-fold cross-validation technique. Results were compared against goldstandard visual review by two board certified clinical neurophysiologists.

RESULTS:

Two different machine learning modalities were tested—Support vector machine (SVM) and Bayesian learning. When SVM was used, the overall sensitivity was 69.6% (+/- 15.3%) and the overall specificity was 95.7% (+/- 10.4%). The kappa statistic was 0.878 and 0.854 respectively for reviewers A and B. When Bayesian learning was utilized, the overall sensitivity was 77.6% (+/- 9.1%) and the overall specificity was 87.6% (+/- 8.3%), and the kappa statistic was 0.780 and 0.771 respectively for reviewers A and B. The detailed results are provided in Table 1, and a graphic summary is shown in Figure 1.

CONCLUSIONS:

Automated spike quantification by wavelet transform is feasible and can be improved by machine learning. Machine learning improved specificity from 64% to as much as 95%, while allowing for maintenance of higher sensitivities. Overall, there was good agreement with the gold-standard of EEG review by human visual inspection. In certain data sets, low sensitivity was noted, which was most likely due to presence of artifact, further highlighted by the fact that these sets displayed the highest amount of disagreement between the two human reviewers. We anticipate that upon further validation, the reported spike detection protocol method may be useful in quantification of epileptiform activity in research and in clinical practice.

Abstracts presented at the American Epilepsy Society meeting 2012

• Outcomes of Vagus Nerve Stimulation (VNS) in Pediatric Epilepsy
• Survey on Terminology on Electrical Status Epilepticus in Sleep and Continuous Spikes and Waves During Sleep
• Malformation of Cortical Development Associated with Early Onset of Medically Interactable Epilepsy in Sturge-Weber Syndrome
• Chromosomal Microarray is High Yield for Identifying Copy Number Variants in Epilepsy Patients

OUTCOMES OF VAGUS NERVE STIMULATION (VNS) IN PEDIATRIC EPILEPSY

C. D. Yu I. Abdelmoumen S. Ramgopal C. Powell K. Remy, M. Libenson, J. Madsen A. Rotenberg, T. Loddenkemper

McGill University, Boston Children's Hospital

RATIONALE:

To examine the efficacy of vagus nerve stimulation (VNS) therapy in children and adolescents with medically refractory epilepsy.

METHODS:

We reviewed the charts of all patients implanted with a VNS at a single center between December 1997 and March 2011. All patients with complete follow-up data at 6 and 12 months were included. Patients were divided into two groups according to the median seizure frequency at baseline. A Wilcoxon Rank sum test was performed to determine if there was significant seizure reduction from baseline to month 12 across various subgroups organized by baseline characteristics.

RESULTS:

252 patients were identified. Complete follow-up data was available for 69 patients (29 girls, mean implantation age 11.1 years, SD 5.3). Mean age of seizure onset was 28.6 months (SD 34.3). Ten patients had generalized epilepsy, 30 had focal epilepsy, and 29 had both generalized and focal features. Forty-seven and 44 subjects had reductions in seizure frequency at months 6 and 12, respectively. The overall median percentage change (seizures/month) from baseline was -50% [25%, 75% Quartiles: (-73%, 0%)] at 6 months (p=0.005) and -40% [25%, 75% Quartiles: (-75%, 25%)] at 12 months (p=0.058). 4 patients were seizure-free at 12 month follow-up. Patients with higher baseline seizure frequency (>45/month) experienced a greater percentage reduction in seizures, with a median reduction of 61% [25%, 75% Quartiles: (-94%, -50%)] (p<0.001) and 69% [25%, 75% Quartiles: (-96%, -30%)] (p<0.001) at 6 and 12 months, respectively. Patients with a lower baseline seizure frequency (≤45/month) experienced a median seizure reduction of 25% [25%, 75% Quartiles: (-50%, 50%)] at 6 months (p=0.94) and 0% [25%, 75% Quartiles: (-50%, 100%)] at 12 months (p=0.39). Of the excluded patients, 6 had treatment discontinued during the 12 month follow-up period. Reasons for discontinuation included infection (n=3), nausea (n=1), skin breakdown (n=1), and perceived lack of efficacy (n=1). Sixteen of the 69 patients experienced adverse effects which did not warrant discontinuation of therapy, including voice changes, coughing, throat tickle, difficulty sleeping, and pain over the VNS area. Outcomes were compared across multiple subgroups including gender, age at VNS implantation, seizure semiology, previous history of SE, number of medications at baseline, presence or absence of MRI lesion, and duration of epilepsy prior to implantation. There were no significant differences in outcome in any of these subgroups (p>0.05). In the low baseline group, however, there was a tendency for patients who were younger at the time of VNS implantation to experience moderately better outcomes (p=0.06).

CONCLUSIONS:

VNS therapy resulted in a reduction in seizure frequency at both 6 and 12-month follow-up periods. These data suggest that VNS is a viable treatment for patients with medically refractory epilepsy, but may provide more consistent seizure reduction in patients with higher baseline seizure frequencies.

SURVEY ON TERMINOLOGY ON ELECTRICAL STATUS EPILEPTICUS IN SLEEP AND CONTINUOUS SPIKES AND WAVES DURING SLEEP

J. M. Peters, I. Sánchez Fernández, K. E. Chapman, A. T. Berg, T. Loddenkemper

Boston Children's Hospital

RATIONALE:

The terms "Electrical status epilepticus in sleep (ESES)" and "Continuous spikes and waves during sleep (CSWS)" are variably used in the literature. The aim of this study was to collect the current views of clinical epileptologists regarding the meaning and use of these terms.

METHODS:

We administered an online survey asking the members of the American Epilepsy Society about their views on terminology and conceptualization of ESES and CSWS. The survey was administered through an online survey website (www.surveymonkey.com) and distributed through the "Professional Connection" website of the American Epilepsy Society. All responses were automatically registered by the online survey website.

RESULTS:

We received 78 responses (72 complete, 6 had some missing data). The demographic features of the respondents are summarized in Table 1. The terms ESES and CSWS were considered synonymous by 39 (54%) respondents, not synonymous by 28 (39%), 5 (7%) respondents stated that they did not know the answer, and 6 respondents skipped this question. The minimum EEG cut-off value of epileptiform activity during sleep for ESES was considered: at least 85% by 37 (52%) respondents, at least 50% by 27 (38%) and any amount of spike-waves by 1 (1%); 6 (9%) were not sure about the cut-off value, and 6 respondents skipped this question. The cut-off value was considered mandatory by 32 (44%), typical but not required by 32 (44%), 8 (11%) did not know the answer and 6 respondents skipped this question. The diagnosis of CSWS was performed based on both clinical and EEG assessment together by 47 (65%) respondents, and based on EEG assessment alone by 21 (29%). The segment used to quantify the epileptiform activity was: all phases of sleep during the whole night by 17 (26%), all non-REM phases of sleep during the whole night by 32 (49%), comparison of epileptiform activity during wakefulness and sleep by 10 (15%), and variable segments for every patient by 3 (5%); 8 respondents used other segments, and 13 respondents skipped this question. Regarding the extent of discharges, 11 (15%) respondents only quantified bilateral and synchronous spike-waves, 22 (31%) quantified bilateral spike-waves even if not synchronous, and 30 (42%) quantified unilateral spike-waves as equivalent to bilateral spike-waves; 9 (13%) respondents did not know the answer and 6 skipped the question. The quantification of epileptiform activity was performed as the percentage of one-second bins occupied by spike-waves by 42 (64%) respondents and as the total number of spike-waves per unit of time in 24 (36%); 4 respondents used other methods and 12 skipped this question. The understanding of the nature and prognosis of CSWS is graphically represented in Figure 1.

CONCLUSIONS:

Our data show that the terms ESES and CSWS are used differently by epileptologists. Variability may impact clinical care, communication and clinical research. A common terminology and diagnostic criteria are urgently needed.

MALFORMATION OF CORTICAL DEVELOPMENT ASSOCIATED WITH EARLY ONSET OF MEDICALLY INTRACTABLE EPILEPSY IN STURGE-WEBER SYNDROME

A. L. Pinto, L. Chen, S. Prabhu, H. Lidov, A. Poduri, M. Sahin, M. Takeoka

Boston Children's Hospital

RATIONALE:

Severe Sturge-Weber Syndrome (SWS) cases may present in early life with intractable seizures; these cases may be associated with developmental disorganization of the cortex, but the frequency and significance of this association remains unclear. In this study we compared the clinical features, neuroimaging and neuropathological findings in such severe cases of SWS who required surgical treatment for intractable epilepsy.

METHODS:

We retrospectively reviewed the clinical features, preoperative magnetic resonance imaging (MRI) studies, and pathological findings of all SWS patients who underwent excisional surgery for intractable epilepsy at Boston Children's Hospital between 1994 and 2011.

RESULTS:

Eleven patients (male/female = 4/7) with SWS were identified who underwent surgery for intractable epilepsy (mean age 13 +/- 6.2 months), including hemispherectomy (n = 10) and focal cortical resection (n = 1). Mean age at seizure onset was 15 +/- 11 weeks of age. In 50% (n=6) of patients, patients had 2 different types of seizures; 2 patients had 3 different types of seizures. Focal clonic seizures were the most common type and occurred in 9 patients; apnea as second type of seizure occurred in 4 patients. Brain MRIs were reviewed in 6 of 11 patients and had findings consistent with Sturge-Weber syndrome, in all cases. Five cases showed further evidence of accelerated myelination. Areas of poor differentiation of white and gray matter were seen in some patients on MRI, while histopathological examination revealed findings consistent with cortical dysplasia.

CONCLUSIONS:

In our patients, cortical malformations were consistently identified in children with severe SWS who underwent early epilepsy surgery. Clinical presentation in this subgroup is severe, with early onset of catastrophic epilepsy, medically intractable seizures, and associated cognitive and motor deficits. Changes consistent with cortical malformation may be associated with the more severe early onset intractable epilepsy rather than vascular malformation changes, considering that such dysplastic changes may be more epileptogenic; further studies will be necessary to confirm such trend, and also to assess whether such dysplastic changes are developmental, or secondary to vascular and perfusion anomalies in such patients.

CHROMOSOMAL MICROARRAY IS HIGH YIELD FOR IDENTIFYING COPY NUMBER VARIANTS IN EPILEPSY PATIENTS

H. Olson, J. M. Avallone, Y. Shen, B. Wu, A. Poduri

Boston Children's Hospital, Harvard Medical School

RATIONALE:

The objective was evaluate the yield of clinical chromosomal microarray (CMA) testing in a retrospective series of patients with epilepsy, and to identify novel or rarely reported overlapping copy number variants (CNVs) in this population.

METHODS:

We identified 977 patients with an ICD9 code for epilepsy or seizures who had a clinical CMA performed between October 2006 and February 2011 in the Children's Hospital Boston system. We retrospectively reviewed medical records to identify if the patients met clinical criteria for epilepsy and define abnormalities on CMA.

RESULTS:

On review of records, 4 patients were excluded because CMA results were not in the chart and 167 patients were excluded because they did not have clinical epilepsy. Some excluded patients had febrile seizures only or evaluation for spells which were not thought to be seizures by the treating neurologist. Of the remaining 806 patients, the CMA showed at least one CNV in 324 patients (40%) and was normal in 482 (60%). The total number of CNVs was 437 (1-4 per patient) of which there were 183 deletions (42%) and 254 (58%) duplications. Thirty-five were confirmed as de novo, 183 were inherited, and inheritance was unknown in 219. Size ranged between 18 kb and 19 Mb, except one 142 Mb duplication and 3 chromosomal trisomies (X, 21, and 9). One hundred forty-seven CNVs were over 500 kb in size (33%). This series includes at least 5 deletions and 1 duplication involving known epilepsy related genes (KCNQ2, SCN1A, PLCB1, and CACNB4), and 24 CNVs in known hotspots for CNVs in idiopathic epilepsy (15q11.2, 15q13.3, 15q11-q13, 16p11.2, 16p13.11). It also includes CNVs associated with known syndromes that are associated with epilepsy such as Potocki-Lupski syndrome, Mowat-Wilson syndrome, Angelman syndrome, Williams-Beuren region reciprocal syndrome, MECP2 duplication syndrome, Phelan-McDermid syndrome, 22q11 duplication syndrome, 1p36 deletion syndrome, and 6q deletion syndrome. In addition, there are numerous CNVs that occur multiple times with overlap. These may have novel associations with epilepsy, but require further investigation and comparison with controls. For example there are three overlapping deletions involving Xp22.31 (859kb, 1.6 Mb, and 9.6 Mb in size), one de novo and two of unknown inheritance. There are two case reports in the literature of patients with deletions in this region and associated epilepsy and intellectual disability, one with cortical heterotopias. There are 5 overlapping duplications in the same region, ranging from 311 kb to 1.6 Mb in size.

CONCLUSIONS:

The yield of CMA testing in epilepsy patients is high, and not infrequently identifies known abnormalities associated with epilepsy. Further phenotypic evaluation of rare or novel repetitive CNVs is warranted and may identify new genotypephenotype correlations or strengthen rarely reported genotype-phenotype associations. For example, data support the association of CNVs at Xp22.31 with epilepsy.

Abstracts presented at the American Epilepsy Society meeting 2012

• Concordance of Epilepsy in Patients with Polymicrogyria Enrolled in the Epilepsy Phenome/Genome Project
• Somatic and Germline Mutations in AKT3 Cause Epileptic Cortical Malformations

CONCORDANCE OF EPILEPSY IN PATIENTS WITH POLYMICROGYRIA ENROLLED IN THE EPILEPSY PHENOME/GENOME PROJECT

C. Shain, S. Ramgopal, I. Parulkar, A. Poduri, .. EPGP Investigators

Boston Children's Hospital

RATIONALE:

Polymicrogyria (PMG) is a heterogeneous and highly epileptogenic malformation of cortical development. The potential role of epilepsy surgery in patients with PMG has been increasingly recognized. In the PMG cohort of patients from the Epilepsy Phenome/Genome Project (EPGP), we sought to evaluate whether the predicted region of seizure onset based on seizure semiology and EEG data corresponded to the location of the PMG seen on MRI. In addition, we assessed the outcomes of epilepsy surgery in the subgroup of patients for whom this was performed.

METHODS:

Participants were recruited through EPGP, a multi-center collaborative including 26 sites from the United States, Argentina, Australia, and New Zealand. Detailed phenotypic data were reviewed for all patients with epilepsy and PMG. Epilepsy lateralization was assessed using EEG and seizure semiology, and then compared to MRI to establish concordance. In patients who underwent epilepsy surgery, we evaluated outcomes at last follow-up.

RESULTS:

We identified 90 patients (43 female) with PMG and epilepsy. The cohort included pediatric and adult patients with a median age of 11 years (<1-55). Most patients presented with localization-related epilepsy (64.4%); complex partial seizures were the most common seizure type (50%). Half of the patients had bilateral PMG on MRI (50%). Of the 45 unilateral cases, 17 (37.8%) were located in the left hemisphere and 28 (62.2%) were located in the right hemisphere. When comparing PMG and epilepsy lateralization, 70.6% of the 17 patients with left-sided malformations had concordant data suggesting they might be potential candidates for surgery. From this group, one participant underwent epilepsy surgery. Of the 28 patients with right-sided malformations, 75% had right-sided seizure onset with 4 undergoing epilepsy surgery. One patient with bilateral/multifocal epilepsy also underwent epilepsy surgery. Among the patients with bilateral PMG, 15.6% had seizure onset on the right, 15.6% on the left, 35.6% bilateral/multifocal, and 11.1%generalized with 3 patients from this group undergoing epilepsy surgery (2 right, 1 left, see figure). In total, 9 patients in our series underwent epilepsy surgery. They presented with a range of seizure semiologies and did not have strictly focal or unilateral malformations. After surgery, 7 patients had major improvement (Engel score I), 1 patient had overall worsening of seizures (Engel score IV), and 1 patient was seizure-free but had insufficient follow-up for Engel scoring.

CONCLUSIONS:

The majority of the EPGP cohort with PMG and epilepsy were found to have concordance between laterality of PMG and predicted region of seizure onset based on seizure semiology and EEG data. Our data from the subset of patients who underwent epilepsy surgery suggest that surgery can be successful for carefully selected patients with PMG. Future phenotypic analysis with the addition of genotypic data may provide further insight into the prediction of surgical candidacy among patients with PMG and epilepsy.

SOMATIC AND GERMLINE MUTATIONS IN AKT3 CAUSE EPILEPTIC CORTICAL MALFORMATIONS

A. Poduri, G. D. Evrony, X. Cai, P. C. Elhosary, R. Beroukhim, M. K. Lehtinen, L. B. Hills, E. L. Heinzen, A. Hill, R. S. Hill, B. J. Barry, B. F. Bourgeois, J. J. Riviello, E. Vining, B. Carson, A. J. Barkovich, P. M. Black0, K. L. Ligon, C. A. Walsh

Boston Children's Hospital, Harvard Medical School

RATIONALE:

Hemimegalencephaly (HMG) is a congenital brain malformation characterized by an enlarged, malformed cerebral hemisphere that often requires hemispherectomy for treatment of refractory epilepsy. We recently identified somatic mosaic mutations involving the gene encoding the serine threonine kinase AKT3: an activating point mutation (c.49G→A, creating E17K) and two trisomies of chromosome 1q containing AKT3 (Poduri et al., Neuron, 2012). Here we further characterize the E17K mutation using single cell sequencing of neurons and glia, and we report a novel germline truncating variant in AKT3 in a patient with HMG.

METHODS:

We have identified 29 cases of HMG from which DNA from brain, leukocytes, and/or buccal cells was available. All research was conducted in accordance with the IRB of Boston Children's Hospital. We used copy number analysis of SNP data, quantitative PCR, and candidate gene sequencing. We further studied the AKT3 E17K mosaic mutation case using single cell sorting (NeuN+ vs NeuN-) and single cell sequencing to establish the relative proportions of neurons and glia that were mutationpositive.

RESULTS:

(1) Having previously identified the AKT3 mosaic E17K mutation in DNA from brain tissue (and not leukocytes) in an individual with HMG, we sequenced AKT3 in our other cases. We have identified a novel AKT3 variant, W330X, in DNA obtained from a buccal cell sample in one case. This truncation does not disrupt a critical activation site (T305) but is predicted to result in the loss of the C-terminal regulatory domain, which could potentially cause activation of AKT3. Functional studies of the W330X variant are underway. (2) We further characterized the AKT3 E17K somatic mosaic mutation using single cell sorting and sequencing. Sequencing of neurons (NeuN+) and glial cells (NeuN-) confirmed the mutation is present in both cell types, at a rate of 39}7% (SE) in single NeuN positive cells (neurons) and 27}8% in single NeuN negative cells (consisting of glial and other non-neuronal cells), suggesting a de novo mutation in a progenitor cell that generates neuronal and non-neuronal cells.

CONCLUSIONS:

We now provide a genetic explanation for four cases of HMG involving activation of AKT3 in brain, either by duplication or activating point mutation leading to hemispheric overgrowth. Somatic mutations arising de novo in neuronal-glial progenitors in the developing brain can be responsible for HMG. Together, our data suggest that AKT3 is an important regulator of brain development and that mutations of this gene, whether germline or post-zygotic and affecting a subset of neurons and glia in one hemisphere, lead to HMG. These mutations in HMG present two models, somatic and germline, for the genetics of epileptic cortical malformations.