Epilepsy in Patients with GRIN2A Alterations: Genetics, Neurodevelopment, Epileptic Phenotype and Response to Anticonvulsive Drugs

european journal of paediatric neurology xxx (2017) 1e12

Ofﬁcial Journal of the European Paediatric Neurology Society

Original article Epilepsy in patients with GRIN2A alterations: Genetics, neurodevelopment, epileptic phenotype and response to anticonvulsive drugs

* C. von Stu¨lpnagel a,l, , M. Ensslen b, R.S. Møller c,m, D.K. Pal d, S. Masnada e, P. Veggiotti e, E. Piazza e, M. Dreesmann f, T. Hartlieb a, T. Herberhold a, E. Hughes g, M. Koch h, C. Kutzer i, K. Hoertnagel j, J. Nitanda a, M. Pohl k, K. Rostasy h, T.B. Haack n,o,K.Stohr€ a, G. Kluger a,l, I. Borggraefe b a Hospital for Neuropediatrics and Neurological Rehabilitation, Epilepsy Center for Children and Adolescents, Schon€ Klinik Vogtareuth, Vogtareuth, Germany b Department of Pediatric Neurology, Developmental Medicine and Social Pediatrics, Epilepsy Center, University of Munich, Munich, Germany c Danish Epilepsy Centre, Dianalund, Denmark d Department of Basic and Clinical Neurosciences, Institute of Psychiatry, Psychology and Neuroscience, King's College, London, England, United Kingdom e Child Neuropsychiatry Division, Neurological Institute Casimiro Mondino Foundation IRCCS, Pavia, Italy f Sozialpadiatrisches€ Zentrum Potsdam, Ernst von Bergmann Klinik, Potsdam, Germany g King's College Hospital & Evelina Children's Hospital, London, England, United Kingdom h Department of Pediatric Neurology, Developmental Neurology and Social Pediatrics, Vestische Kinder-und Jugendklinik Datteln, University Witten/Herdecke, Datteln, Germany i Kinderzentrum St. Martin, Regensburg, Germany j Cegat GmbH, Laboratory of Medical Genetics, Tuebingen, Germany k Children's Hospital Dritter Orden, Passau, Germany l Paracelsus Medical University Salzburg, Salzburg, Austria m Institute for Regional Health Services, University of Southern Denmark, Odense, Denmark n Institute of Human Genetics, Technische Universitat€ Mu¨nchen, Munich, Germany o Institute of Human Genetics, Helmholtz Zentrum Mu¨nchen, Neuherberg, Germany

Abbreviations: ABPE, Atypical Benign Partial Epilepsy of childhood; ACMG, American College of Medical Genetics; AED, antiepileptic drugs; BECTS, benign focal epilepsy with centrotemporal spikes; CLB, clobazam; CSWS, Continuous Spike Waves during Slow Wave Sleep; EE, epileptic encephalopathy; EEG, electroencephalogramm; ESES, electrical status epilepticus during slow wave sleep; ExAC, Exom Aggregation Consortium; IFE, idiopathic focal epilepsy; LEV, levetiracetam; LKS, Landau Kleffner Syndrome; TPM, topiramate; STM, sultiame; VPA, valproic acid. * Corresponding author. Klinik fu¨ r Neuropadiatrie€ und Neurologische Rehabilitation, Epilepsiezentrum fu¨ r Kinder und Jugendliche, Schon€ Klinik Vogtareuth, Krankenhausstraße 20, D-83569 Vogtareuth, Germany. Fax: þ49 8038 903418. E-mail address: [email protected] (C. von Stu¨ lpnagel). http://dx.doi.org/10.1016/j.ejpn.2017.01.001 1090-3798/© 2017 European Paediatric Neurology Society. Published by Elsevier Ltd. All rights reserved.

article info abstract

Article history: Objective: To delineate the genetic, neurodevelopmental and epileptic spectrum associated Received 30 June 2016 with GRIN2A alterations with emphasis on epilepsy treatment. Received in revised form Methods: Retrospective study of 19 patients (7 females; age: 1e38 years; mean 10.1 years) 8 December 2016 with epilepsy and GRIN2A alteration. Genetic variants were classified according to the Accepted 2 January 2017 guidelines and recommendations of the American College of Medical Genetics (ACMG). Clinical findings including epilepsy classification, treatment, EEG findings, early childhood Keywords: development and neurodevelopmental outcome were collected with an electronic GRIN2A questionnaire. Epilepsy Results: 7 out of 19 patients fulfilled the ACMG-criteria of carrying “pathogenic” or “likely Epileptic encephalopathy pathogenic variants”, in twelve patients the alterations were classified as variants of un- Specialized therapy known significance. The spectrum of pathogenic/likely pathogenic mutations was as fol- lows: nonsense n ¼ 3, missense n ¼ 2, duplications/deletions n ¼ 1 and splice site n ¼ 1. First seizures occurred at a mean age of 2.4 years with heterogeneous seizure types. Pa- tients were treated with a mean of 5.6 AED. 4/5 patients with VPA had an improved seizure frequency (n ¼ 3 with a truncation: n ¼ 1 missense). 3/5 patients with STM reported an improvement of seizures (n ¼ 2 truncation, n ¼ 1 splicing). 3/5 CLB patients showed an improvement (n ¼ 2: truncation; n ¼ 1 splicing). Steroids were reported to have a positive effect on seizure frequency in 3/5 patients (n ¼ 1 each truncation, splicing or deletion). Conclusions: Our data indicate that children with epilepsy due to pathogenic GRIN2A mutations present with different clinical phenotypes and a spectrum of seizure types in the context of a pharmacoresistant epilepsy providing information for clinicians treating children with this form of genetically determined epileptic syndrome. © 2017 European Paediatric Neurology Society. Published by Elsevier Ltd. All rights reserved.

a re-entry loop (M2) and an intracellular C-terminal domain þ 1. Introduction (CTD). The NTD harbors a Zn2 binding site, which is involved þ in Zn2 mediated inhibition of the receptor. The ligand- Glutamate is the most relevant excitatory neurotransmitter binding domain mediates glycine binding in GluN1 receptor within the central nervous system (CNS) and mediates its subtypes and glutamate binding in GluN2 receptor subtypes, action via three different receptor types, AMPA-, Kainate- and the transmembrane domains M1 to M4 build the ion channel 1 N-Methyl-D Aspartate-(NMDA-)receptors. Disturbance of pore. The intracellular C-terminal domain is the most diver- NMDA receptors can occur by either genetic disruption of gent region among GluN2 subtypes and mediates intracellular receptor-coding genes or antibodies directed against receptor signals. GluN2A is encoded by the gene GRIN2A on human peptides. Both conditions lead to significant human disease chromosome 16.7 GRIN2A consists of 14 exons and encodes such as epilepsy, developmental delay and autoimmune the 1464 amino acid GluN2A subunit of the GluN2 receptor 2e4 encephalitis. NMDA receptors are di-heteromeric iono- complex. tropic complexes consisting of two obligatory GluN1 subunits Mutations within the GRIN2A gene may cause benign focal and two additional GluN2 or other subunits. Four different epilepsy with centrotemporal spikes (BECTS).2 In addition, GluN2 receptor subtypes exist (GluN2A-D). These subtypes GRIN2A gene mutations are more likely to occur in epilepsy determine the functional diversity of the receptor as they subtypes which are believed to be a more severe variant of reveal different expression patterns during brain develop- BECTS as Continuous Spike Waves during Slow Wave Sleep 5,6 ment and maturation. In addition, they show different (CSWS), Landau Kleffner Syndrome (LKS) and Atypical Benign spatial expression in the brain. GluN2A expression in rats is Partial Epilepsy (ABPE) of childhood.8 The latter ones are not detectable at birth, but it is present at P14 (corresponding usually difficult to treat epilepsy syndromes. Attempts have to the first year of life in a human) and is mostly abundant in been made to use the NMDA-receptor inhibitor memantine in the adult rat. Thus, GluN2A is believed to be the most relevant a patient with a mutation leading in vitro to a gain of function GluN2 subunit from childhood to adulthood. GluN2A consist of the channel.9,10 of an extracellular N-terminal (NTD) and a ligand-binding In the present investigation, we studied seven patients domain (LBD), three transmembrane domains (M1, 3 and 4), with childhood-onset epilepsy with pathogenic alterations

within the GRIN2A gene. We report on the phenotypic spec- (Sorting Intolerant From Tolerant) Human Protein” provides trum associated with new mutations and present data on re- information of tolerated versus intolerated mutations for sponses to AEDs and possible correlations between genotype selected amino acid substitutions. SIFT is able to sort func- and epileptic phenotype. tionally neutral from deleterious amino acid changes. Scores <0.05 are likely to predict affected protein function. PROVEAN is capable to provide predictions of the functional consequences 2. Patients and methods for amino acid substitutions, deletions and insertion. A PRO- VEAN score equal to or below a threshold of 2.5 predicts a 2.1. Recruitment “deleterious” effect on the protein. PolyPhen-2 (Polymorphism Phenotyping v2) is a software tool to predict the impact of an 10 centers in Europe (Germany n ¼ 6, England n ¼ 2, Denmark amino acid substitution on the structure and function of a n ¼ 1, and Italy n ¼ 1) contributed information about patients human protein. It uses information about physicochemical with epilepsy and a GRIN2A mutation such as genetic findings, properties. PolyPhen-2 scores of 0,850 are classified to have a neurodevelopmental performance, epileptic phenotype and high probability of a damaging effect on the protein. Mutation treatment response to different AED. Anonymized data were Assessor calculates a functional impact score (FIS) for amino reported using an electronic questionnaire. Inclusion criteria acid residue changes. It uses patterns of evolutionary conser- for the study were patients with an alteration within the vation which are derived from aligned families and sub-families GRIN2A gene and epilepsy. of sequence homologs within and between species. A FIS of 1938 and above is likely to predict a functional change within the 2.2. Descriptive analysis protein. MutationTaster2 comprises information from different biomedical databases as about evolutionary conservation, Assessment of patient characteristics was collected using a splice-site changes, loss of protein features and changes that defined electronic questionnaire containing a battery of might affect the amount of mRNA. different items related to the epileptic and neuro- Mutations were pasted within the input mask using the developmental phenotype in addition to genetic findings. The suggested format for each program, which usually were in line following items were documented: family history of the pa- with the present HVGS nomenclature (www.hvgs.org). tient, early childhood development, intellectual ability, onset of epilepsy, seizures, anticonvulsive treatment, as well as on clinical and diagnostic investigations. Physicians were asked 3. Results to complete information for each drug related to the time of therapy, dosage and effect on seizure frequency (improve- 3.1. Genetic findings ment, no effect or aggravation and if possible the reduction on seizure frequency on a 25% percentage scale). We retrospec- 7 out of 19 patient were classified as “pathogenic” (n ¼ 1) or tively analyzed the efficacy of AED 3-12 months after the “likely pathogenic” (n ¼ 6) according to the ACMG criteria. introduction of each drug (except for steroids and vitamin B6) Detailed information about each item which was sufficient for compared to the baseline period four weeks before starting the category of ACMG classification in any individual case is the new AED. Drug response was defined as a 50% seizure depicted in Table 1 for alterations classified as pathogenic and reduction or more. Seizure aggravation was defined as a more likely pathogenic and in the Supplementary Table 1 for al- than 50% increase in seizure frequency. Ethical approval was terations classified as VUS. Twelve patients carried most likely obtained from the Bavarian State Medical Association variants of unknown significance. A detailed description of (“Bayerische Landesaerztekammer”). the demographic and clinical date of these twelve patients with VUS is presented in Supplementary Tables 2 and 3 2.3. Assessment of genetic variants though they were not acknowledged within the further descriptive analysis regarding neurophysiological-, clinical- The interpretation of the genetic variants in this survey was and AED-response data. performed in accordance to the most recent consensus rec- The distribution of alterations within the GRIN2A poly- ommendations for the interpretation of sequence variants of peptide chain is shown in Fig. 1. Pathogenic or likely patho- the American College of Medical Genetics.11 In brief, informa- genic variants were widely distributed affecting different tion related to segregation, vicinity of the variants to functional regions of the receptor subunit (S1, 2: n ¼ 2, transmembrane relevant domains, absence of the variant in healthy individuals domains M1-M3: n ¼ 3) but spared completely the extracel- (Exome Aggregation Consortium) and computational tools (in lular N-terminal domain and intracellular C-terminal domain. silico analysis) are used among others to classify the variant into The majority of variants of uncertain significance II was the categories “pathogenic”, “likely pathogenic” and “variant of located within the C-terminal domain. uncertain significance”. In silico analysis for prediction of ge- There was no association of type of epilepsy phenotype or netic results regarding computational relevance was obtained developmental variables as age of seizure onset, pharmacor- by using the following algorithms: “SIFT (Sorting Intolerant esistence (failure of at least 2 antiepileptic drugs of adequate From Tolerant) Human Protein”,12 “Protein Variation Effect choice), seizure frequency,19 the presence of an EEG status Analyzer”,(PROVEAN),13 “Polyphen-2”,14 “MutationTaster2”15 pattern (ESES) and speech performance with the type of the and “Mutation Assessor”.16 These programs use different ap- suspected consequence of the mutations (i.e. truncation vs. proaches to determine the pathogenicity of mutations. “SIFT missense mutation).

& Table 1 e Genetic data on patients with pathogenic and likely pathogenic variants: Predicted by Mutation Taster, #One parent clinically affected (mother had childhood § epilepsy), Splice site defect within intron 7 in the splice site consensus sequence before exon 7 (the acceptor site), Truncations: transcript probably not translated due to nonsense mediated mRNA decay. Truncations Splicing affected Duplication/deletion Missense uoenjunlo adarcnuooyxx(07 1 (2017) xxx neurology paediatric of journal european Patient# 2104 129 138 Phenotype CSWS ABPE LKS CSWS Epileptic Focal epilepsy Epileptic encephalopathy encephalopathy

¨ Mutation c.1586delT het. c.1818G>A het. c.2407G>T het. c.2007þ1G>A duplication exon c.1841A>G het. c.1936A>G het. pae ,e l,Eies nptet with patients in Epilepsy al., et C, lpnagel 4 & 5 het. Protein change p.Phe528Glyfs*22 p.Trp606* p.Glu803* IVS7 n.a. p.Asn614Ser p.Thr646Ala Domain affected M1-4, S2, CTD M2-4, CTD M4, CTD n.a. NTD M2 M3 Segregation Not tested though Pending De novo Father, not Not tested De novo De novo 1 parent affected# affected ExAC-total reports Not reported Not reported Not reported Not reported Not reported Not reported Not reported ExAC-allele frequency Not reported Not reported Not reported Not reported Not reported Not reported Not reported Provean (Score) n.a. Deleterious (12,534) Deleterious (5240) n.a. n.a. Deleterious (4731) Deleterious (4654) SIFT (Score) n.a. n.a. n.a. n.a. n.a. Tolerated (0,07) Damaging (0) PolyPhen-2 (Score) n.a. n.a. n.a. n.a. n.a. Probably damaging (0,997) Probably damaging (0,992) Mutation Assessor (Score) n.a. n.a. n.a. n.a. n.a. Low (1,59) Medium (3,46) MutationTaster2 Disease causing Disease causing Disease causing n.a. n.a. Disease causing Disease causing Summary of criteria PVS1, PM2 PVS1, PM2, PP3 PVS1, PM6, PM2, PP3 PVS1, PM2 PM1, PM2, PM4 PM1, PM2, PM6, PP3 PS2, PM1, PM2, PP3 variant classiﬁcation Interpretation according to Likely pathogenic Likely pathogenic Pathogenic Likely pathogenic Likely pathogenic Likely pathogenic Likely pathogenic GRIN2A the ACMG standards

Abbreviations (only criteria are listed which where encountered within this cohort): PVS1: null variant (nonsense, frameshift, canonical ±1 or 2 splice sites, initiation codon, single or multiexon deletion) in a gene where LOF is a known mechanism of disease. leain:Gntc,neurodevelopment, Genetics, alterations: PS2: De novo (both maternity and paternity conﬁrmed) in a patient with the disease and no family history. PM1: Located in a mutational hot spot and/or critical and well-established functional domain (e.g., active site of an enzyme) without benign variation.

PM2: Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes Project, or Exome Aggregation Consortium”. e PM4: Protein length changes as a result of in-frame deletions/insertions in a nonrepeat region or stop-loss variants. 12 PM6: assumed de novo, but without conﬁrmation of maternity or paternity. PP3: Multiple lines of computational evidence support a deleterious effect on the gene or gene product. european journal of paediatric neurology xxx (2017) 1e12 5

Fig. 1 e Spatial distribution of putative changes within the polypeptide change of the Glun2a protein. Black colour indicates pathogenic or likely pathogenic variants; grey colour indicates variants of uncertain signiﬁcance.

3.2. Patient disposition There were focal motor seizures in two patients and atonic in two other patients. Dialeptic seizures in two patients, one 7 patients (5 males; age: 1.2e15 years; mean 6.9 years; median: together with tonic seizures and dialeptic seizures in another 7 years) with GRIN2A alterations were included. Two patients patient. In the further course of epilepsy the main seizures (#12; #14 (VUS)) have been included in other publications on types were focal seizures (n ¼ 2; one with secondary gener- epilepsy patients.2,17 Family history was positive for epilepsy alization), generalized tonic-clonic seizures (n ¼ 2; one with in 3/7 patients. fever), myoclonic jerks (n ¼ 2) and one patient with eyelid Regarding the early child development the mean age of myoclonias, dialeptic seizures, atonic seizures, tonic seizures sitting unsupported was 8.3 months, while three patients did or Infantile Spasms in one patient each. not learn to sit unsupported nor to walk independently due to In the active phase of the epilepsy, seizures occurred daily severe muscle hypotonia and intellectual disability. Walking in four patients, rarely in two patients and were ongoing in independently was acquired at the age of 11.5 months. First one patient. Febrile seizures occurred in one patient. spoken words were reported at a mean age of 20.8 months The main EEG findings, obtained at a mean age of 4.4 years (range 11 monthse3 years). One child did not acquire any (range 7 monthse9.8 years) were BEPC (benign epileptic po- language at all due to severe intellectual disability and two tentials in childhood) in one patient, while another patient children were less than 18 months old. The current language showed an ESES (electrical status epilepticus during slow ability was evaluated as mildly impaired in one patient, wave sleep) and two patients had both. In two other patients moderately impaired in two and severely impaired in another (all of them younger than 15 months at EEG evaluation) two patients. Two patients did not show active language multifocal sharp waves with secondary generalization were production yet (all bellow 18 months). Intellectual develop- reported and in one patient (5.6 years at EEG evaluation) a ment evaluated at a mean age of 4.7 years (range 1e9 years) multifocal EEG status with emphasize on the left centro- was described as moderately impaired in four and as severely temporo-parietal region was noted. Photosensitivity was not impaired in three patients. Three patients attended A school observed in any patient (see Table 3). for special needs. Three patients were still too young to go to school and in one patient information was missing. In the neurological examination, performed at a mean age 3.4. Response to AEDs of 5.5 years (range 13 monthse11.5 years a wide variety of ¼ neurological findings was reported. A muscular hypotonia Patients were treated with no (n 1) to 8 different antiepi- was noticed in two patients. Further neurological findings leptic drugs (mean: 5.6; median: 5.5); pharmacoresistant epi- reported in the patients were tremor (n ¼ 1), ataxia (n ¼ 2), lepsy (as defined by lack of efficacy of 2 antiepileptic drugs) deficit in gross motor skills (n ¼ 2) or fine motor skills (n ¼ 1), was reported in six patients. The six most frequently used and a severe developmental delay (n ¼ 3) (for details see Table drugs were valproic acid (VPA), sultiame (STM), clobazam 2). (CLB), steroids, levetiracetam (LEV) and topiramate (TPM). Brain MRI performed in all patients (median age 3.8 years, Regarding the long term retention over the whole patient range 3 monthe6.5 years) was either normal (n ¼ 4) or group STM was given in 4/7 and VPA in 3/7 patients after 12 revealed nonspecific features (cerebral hypoplasia and cyst of months and were also reported to have a good anticonvulsive the pituitary gland (n ¼ 1), mild myelination delay (n ¼ 1) or effect with 4/5 VPA patients reporting an improvement and 3/ prominent external CSF (n ¼ 1)). 5 STM patients (see Table 4 for further details). In the analysis of the different genetic subgroups depend- ing on the type of mutation diagnosed the following obser- 3.3. Epilepsy and EEG vations could be noticed: In the patient group with truncation of the GRIN2A gene All patients presented with epileptic seizures. Seizures (n ¼ 3) the most often used AED were VPA, CLB and STM. STM occurred at a mean age of 2.4 years with a wide range between led to a seizure freedom in two patients for 13 months and had 3 months and 5.75 years). Also the semiology of the initial no effect in another patient. VPA was reported to have an seizure type varied much between the different patients. improvement in 3/3 patients, with one reported seizure

Please cite this article in press as: von Stu¨ lpnagel C, et al., Epilepsy in patients with GRIN2A alterations: Genetics, neurodevelopment, epileptic phenotype and response to anticonvulsive drugs, European Journal of Paediatric Neurology (2017), http://dx.doi.org/10.1016/ j.ejpn.2017.01.001 6 plpi hntp n epnet niovliedus uoenJunlo adarcNuooy(07,http://dx.doi.org/10.1016/ (2017), Neurology Paediatric of Journal European drugs, anticonvulsive j.ejpn.2017.01.001 to Stu response von and as: phenotype press epileptic in article this cite Please Table 2 e Clinical data from the 7 patients with GRIN2A alterations: CC: head circumference; f: female; n.a.: not applicable; n.d.: no data; m: male; mo: months; max.: maximum; P: percentile; r.: right; y: years; mean age eval.: mean age at evaluation. Clinical data from the 7 patients with GRIN2A alterations (1) Patient ID Truncation Splicing affected Duplication/Deletion 2 10 4 12 9 Age at the time of the study (years) 15 9.8 7 11.58 1.42 Sex m m m f m Ancestry Turkish European European European European Protein change of GRIN2A p.Phe528Glyfs*22 p.Trp606* p.Glu803* IVS7 n.a. Level of intellectual disability/age at evaluation Moderate/6 y Moderate/9 y Moderate/6 y Moderate/n.d. Severe/17 mo 1 (2017) xxx neurology paediatric of journal european Developmental stages Age of sitting/walking 6 mo/11 mo 9 mo/11 mo 9 mo/11.5 mo 9 mo/14 mo Not yet/not yet Age of ﬁrst words/ﬁrst sentences 3 y/3 y 11mo/28 mo. 18 mo/loss of 18 mo/n.d. Not yet ¨

pae ,e l,Eies nptet with patients in Epilepsy al., et C, lpnagel speech at 2.5 y Current language ability Severely impaired Mildly impaired Moderately Moderately n/a impaired impaired Regressive episode during With worsening of EEG regression n. d. Yes, speech no n/a the development/age in language and attention Clinical Examination Age at examination 7 y 9.8 y 5.5 y 11.5 y. 17 mo Height in cm (P.)/weight in kg 28.1 kg (75 P.); 132 cm (50 P.); Height P 3e10; 124 cm (97. P.); 158 cm (84. P.) 79 cm (25.P.), 8.6 kg (<3.P), (P.)/head circumference in cm (P.) Head: 54 cm (84 P.) Weight P10-25 27.5 kg (97. P); 40.8 kg (44. P.). 42.5 cm (<3. P.) head 55.5 cm (>97 P.) Neurologic examination Ataxia; muscular hypotonia Tremor, difficulty in Difficulty in gross Difficulty in Severe developmental gross motor skills motor skills fine motor skills delay (impaired fixation, no head control, no active movement) MRI Age at examination 6 y 6 y 5y 6.5 y 3 mo Result normal normal normal normal Delayed myelinisation, otherwise normal examination GRIN2A Clinical data from the 7 patients with GRIN2A alterations (2) Patient ID Missense Summary leain:Gntc,neurodevelopment, Genetics, alterations: 13 8

Age at the time of the study (years) 2.6 1.16 Mean 6.94y

¼ ¼ e

Sex mf f2m 5 12 Ancestry European European Protein change of GRIN2A p.Asn614Ser p.Thr646Ala Level of intellectual disability/age at evaluation Severe/n.d. Severe/1 y Moderate n ¼ 4; severe n ¼ 3; mean age eval. 4.7 y

Developmental stages Age of sitting/walking Not acquired Not yet/not yet Mean: 8.3 mo/11.9 mo; 3 not acquired Age of first words/first sentences Not acquired Not yet Mean 20.8 mo/32 mo; 3 not acquired (yet), 1 loss of speech at 2.5 y Current language ability Severely impaired n/a Mildly: 1; moderately: 2; severe: 2þ2 no language yet Regressive episode during no Loss of fixation and Regressive episodes: n ¼ 3 the development/Age beginning of strabism 2 mo european journal of paediatric neurology xxx (2017) 1e12 7

reduction of 50e75%. 2/3 patients who were taking CLB had an improvement, while one patient reported no effect. Steroids, LEV, TPM and ESM received 2/3 patients each. Of the two pa- 1,

¼ tients receiving LEV no one had an anticonvulsive effect. No therapeutic effect was also noticed in the two patients on 2;

¼ TPM. Steroids led to a slight seizure improvement of 25%, and 3

¼ had no anticonvulsive effect in the other patient. Of the two patients with ESM one reported an improvement of 50%, while 1; 2;

¼ one stopped the medication because of cognitive impairment ¼ which became aggravated. One patient each became seizure free on PER or OXC. Due to the even smaller number of patients with a

2; tremor: n missense mutation of the GRIN2A gene (n ¼ 2) it was not ¼ possible to ﬁnd one AED which was given more often than another. One patient each received VPA, CLB, Steroids, LEV, normal

TPM, OXC, LTG or VGB. In VPA and LTG an improvement was difficulties in gross motor skills:difficulties n in coordination/fine motor: n severe developmental delay: n 4 Muscular hypotonia: n ataxia: n reported, LTG led to seizure freedom and was given for 27 months. The other AED did not show a good therapeutic effect in these patients. The patient with a duplication of the GRIN2A gene received STM, Steroids, TPM, PB and Vit. B 6 respectively. While STM and Vit B6 were reported to have no therapeutic effect, Ste- roids, TPM or PB improved the seizure situation by 50e75% each.

75 P.), 9 kg (2525 P.), P.) The patient with a mutation affecting the splicing of the e e GRIN2A gene was treated with VPA, CLB, LEV, STM and Ste- roids respectively. The best anticonvulsive effect was reported of CLB, which led to seizure freedom, while STM was reported Prominent external CSF spaces, normal myelination 77 cm (50 45 cm (10 axial hypotonia to reduce the seizure frequency by 50%. LEV and Steroids showed an improvement of the seizure situation of 25% each and VPA was reported to have provoked a seizure aggravation. (For further details please see Table 4).

4. Discussion

Epilepsy-aphasia syndromes were previously a group of rare, severe epileptic encephalopathies of often unknown etiology. Recently, the genetic basis of at least a part of these syndromes was unraveled by finding GRIN2A mutations in four families,18 of which one had been formerly reported by others pituitary gland and a displaced pituitary 92 cm (40. P.); 12CC kg non (9.P.); available as a new syndrome consisting of autosomal dominant rolandic epilepsy, mental retardation, and speech dyspraxia.19 Mutations of GRIN2A were further identified as an important risk factor for idiopathic focal epilepsy (IFE) by detecting het- erozygous mutations in GRIN2A in a significant proportion of individuals with IFE.2 In the latter patient cohort, mutations occurred significantly more frequently in severely affected individuals with CSWS. In another investigation it was demonstrated that about 20% of patients with Landau- eKleffner-syndrome, CSWS or electroclinically atypical Result Cerebral hypoplasia, cyst of the Height in cm (P.)/weight in(P.)/head kg circumference in cm (P.) Neurologic examination Ataxia, severe developmental delay Severe developmental delay; rolandic epilepsy, which are often associated with speech impairment, had a de novo or inherited mutation in the GRIN2A gene as genetic etiology.8 In contrast to the above-mentioned studies our study group consisted of a more heterogeneous epileptic phenotype of 7 patients with epilepsy and GRIN2A mutations. This was reflected by the different epilepsy syndromes, which had been initially classified in our cohort spanning from such severe MRI Age at examination 2.5 y 4 mo Mean: 3.8 y Clinical Examination Age at examination 2.5 y 13 mo Mean 5.5 y forms like epileptic encephalopathies in early infancy (n ¼ 2),

Please cite this article in press as: von Stu¨ lpnagel C, et al., Epilepsy in patients with GRIN2A alterations: Genetics, neurodevelopment, epileptic phenotype and response to anticonvulsive drugs, European Journal of Paediatric Neurology (2017), http://dx.doi.org/10.1016/ j.ejpn.2017.01.001 8 plpi hntp n epnet niovliedus uoenJunlo adarcNuooy(07,http://dx.doi.org/10.1016/ (2017), Neurology Paediatric of Journal European drugs, anticonvulsive j.ejpn.2017.01.001 to Stu response von and as: phenotype press epileptic in article this cite Please Table 3 e Epilepsy features in our 7 GRIN2A patients. foc.: focal; GTCS: generalized toniceclonic seizures; ABPE: atypical benign partial epilepsy; CSWS: continuous spike wave during slow wave sleep syndrome; EE: epileptic encephalopathy; ESES: electrical status epilepticus during slow wave sleep; LKS: Landau Kleffner syndrome; m: months; AED: Anti-epileptic drugs: CBZ: carbamazepine; CLB: clobazam; ESM: ethosuximide; LEV: levetiracetam; LTG: lamotrigin; OXC: oxcarbamazepine; PB: phenobarbital; PER: perampanel; PHT: phenytoin; STM: sultiame; TPM: topiramate; VGB: vigabatrin; Vit B6: Vitamin B6; VPA: valproic acid; EEG: electroencephalogram; BEPCs: benign epileptic potentials of childhood; ESES: electrical status epilepticus during slow wave sleep. Epilepsy features in our 7 GRIN2A patients (1) Patient ID Truncation Splicing affected Duplication/Deletion 2 10 4 12 9

Protein change of GRIN2A p.Phe528Glyfs*22 p.Trp606* p.Glu803* IVS7 n.a.

Age at seizure onset 5.75 y 3y 3.5 y 3 y 3 m 1 (2017) xxx neurology paediatric of journal european (m:months or y:years) Seizure type at onset Atonic seizures Focal seizures Atonic seizures Absences Tonic, dialeptic Seizure types during Focal atonic seizure Focal, sec generalized Eyelid myoclonia; generalized e Myoclonic; GTCS ¨

pae ,e l,Eies nptet with patients in Epilepsy al., et C, lpnagel disease course of the right arm; clonic seizures dialeptic; atonic Initially classiﬁed Epilepsy syndrome CSWS ABPE LKS CSWS EE Febrile seizures No No No No No Status epilepticus Yes (ESES) No No Yes (ESES) Yes Frequency of seizures Rare Ongoing Daily Daily Daily Lifetime/current anti-epileptic treatment n ¼ 8: VPA, CLB, ESM, n ¼ 7: VPA, CLB, LEV, n ¼ 5: VPA, CLB, ESM, n ¼ 5: VPA, CLB, n ¼ 6: STM, Steroids, LEV, STM, Steroids, TPM; STM, TPM, OXC, PER STM, Steroids; LEV, STM, Steroids TPM, PB, Vit B6; VPA PERþSTM VPAþSTMþCLB LEVþSTM PB, TPM Pharmoresistance Yes Yes Yes Yes Yes EEG Age at examination 7 y 1.4 y 5.6 y 5 y 9.8 y BEPCs Yes No No Yes Yes ESES Yes No Yes Yes No Other abnormalities Multifocal sharp waves/spike waves EEG status multifocal with ee with secondary generalization emphasize on left centro-temporo-parietal GRIN2A Epilepsy features in our 7 GRIN2A patients (2) Patient ID Missense Summary leain:Gntc,neurodevelopment, Genetics, alterations: 13 8

Protein change of GRIN2A p.Asn614Ser p.Thr646Ala e

Age at seizure onset 5 m 8 m Mean 2.4 y 12 (m:months or y:years) Seizure type at onset Focal motor seizure Tonic seizures 2 atonic; 2 focal; 1 tonic; 1 tonic þ dialeptic; 1 absences Seizure types Febrile GTCS Tonic; infantile spasms and myoclonic 2 foc (1 sec. generalized).; 2 GTCS during disease course (1 febrile); 3 myoclonic (1 eyelid myoclonia); 1 dialeptic/absences; 1 atonic; 1 tonic; 1 Infantile Spasms Epilepsy syndrome Focal epilepsy EE 2 CSWS; 2 EE; 1 focal epilepsy; 1 ABPE 1 LKS Febrile seizures Yes No 6 no; 1 yes Status epilepticus No No 4 no; 3 yes (2 CSWS) Frequency of seizures Rare Daily 4 daily; 2 rare; 1 ongoing european journal of paediatric neurology xxx (2017) 1e12 9

to the severe forms of the spectrum of IFE like CSWS (n ¼ 2) ; and Landau Kleffner syndrome (n ¼ 1) and further to the milder forms such as atypical benign partial epilepsy (n ¼ 1) ¼ ; ESM: 2 and focal epilepsy (n 1). Therefore our cohort consisted of two different age groups of seizure onset, which was in three

patients within the ﬁrst year of life, represented as tonic and

; TPM: 4 dialeptic seizures, in one patient also as focal seizures and in the other four patients within the expected age range for ; Vit B6: 1

idiopathic focal epilepsy between 3 and 5.75 years. Together ; LEV: 4

with the occurrence of epilepsy, the patients showed impaired

, PB: 1 language ability. 3/3 patients with an epilepsy onset before nine months of age even did not acquire any language due to severe intellectual disability in the course of an epileptic en- ; PER: 1 ; Steroids: 5 cephalopathy, while in 2/4 patients of the older age group a regression of development and mainly language was reported ; LTG: 1 ; CLB: 5 by the parents or caregivers. So besides the heterogeneous study population all the patients had epilepsy followed by 5.6 AED

no no no speech impairment and at least a mild intellectual disability, ¼ ; VPA: 5 ; VGB: 1 suggesting that GRIN2A alteration might be a cause of EE, as previously reported.18 By deﬁnition, EE is an epilepsy disorder yes; 1 yes; 4 yes; 4

in which the “epileptic activity itself may contribute to severe e OXC: 2 Median n STM: 5 cognitive and behavioral impairments above and beyond what might be expected from the underlying pathology alone”.20 Nevertheless, this definition remains controversial, as it has been shown eg in patients with Dravet syndrome that neither seizure burden nor EEG changes correlate with cognitive outcome.21 These findings underscore in contrast to the actual definition of EE that the genetic alteration is capable for causing both seizures and cognitive decline independently form each other. This concept would imply that reduction of seizure frequency with an AED would automatically result in a cognitive improvement. To our knowledge our study is the first to report on the effect of commonly used AED treatment in patients with LEV

7: VPA, CLB, LEV, Steroids, þ GRIN2A mutations. The European expert opinion from 2007 ¼ revealed VPA as treatment of ﬁrst choice for the classical Multifocal sharp waves/spike waves with secondary generalization n TPM, OXC, VGB; VPA benign childhood epilepsy with centro-temporal spikes.22 In concordance to this VPA was one of the most used AED in our study group with 5/7 patients receiving VPA. It showed an improvement of the seizure situation in 4/7 patients In the subgroup truncation 3/3 patients received VPA and all reported an improvement, as did the 1/2 patient in the group with missense mutations.

1: LTG Another frequently used AED in our study was STM in 5/7

¼ patients, having an improvement on seizure situation in 3/5 e n LTG patients, with two patients becoming seizure free for 12 months (both patients with a truncation of the GRIN2A gene) and one patient with the mutation affecting the splicing showing a seizure reduction of 50% for a period of 48 months. STM is widely used for BECTS in Germany and has been shown to be highly efﬁcient in achieving seizure freedom and EEG normalization in randomized controlled trials.23,24 Seizure freedom to STM in BECTS was also seen in a randomized head to head trial with levetiracetam in about 90% ESESOther abnormalities no no 3 BEPCs noof no subjects treated with STM.25 Caution with 3 respect to high STM dosages of 8 mg/kg and more was raised in a case series in patients with BECTS showing signiﬁcant impaired cognitive performance.26

anti-epileptic treatment Especially in CSWS according to Caraballo et al. the AED PharmoresistanceEEG Age at examination 7 mo No 1.4 y Yes Mean 4.4 y 6 Lifetime/current drugs commonly used were also CLB, STM and ESM alone or in

** § Table 4 e Overview of the used AED in the 7 GRIN2A patients: n.d. ¼ no data; Dupl/Delet ¼ Duplication/Deletion; Therapy ongoing,#Number of cycles, mg/body surface AED: Anti-epileptic drugs: CLB: clobazam; ESM: ethosuximide; LEV: levetiracetam; LTG: lamotrigin; OXC: oxcarbamazepine; PB: phenobarbital; PER: perampanel; PHT: phenytoin; STM: sultiame; TPM: topiramate; VGB: vigabatrin; Vit B6: Vitamin B6; VPA: valproic acid. AED Mutations Patients Mean duration Mean dose Improvement No effect Aggravation After 3 After 6 After 12 (n¼) (months) mg/kg (n¼) (n¼) (n¼) mo. (n¼) mo. (n¼) mo. (n¼) VPA Truncation 3/3 59 35.8 3 0 0 3 3 3 ** Missense 1/2 1 n. d. 1 0 0 1 (2017) xxx neurology paediatric of journal european Splicing 1/1 n.d. 25 0 1 1 Total 5/6 30 30.4 4 1 1 3 3 3 CLB Truncation 3/3 16 0.5 2 1 0 2 2 1 ¨

pae ,e l,Eies nptet with patients in Epilepsy al., et C, lpnagel Missense 1/2 3 0.5 0 0 0 0 0 0 Splicing 1/1 60 0.5 1 0 0 1 1 1 Total 5/6 26.3 0.5 3 1 0 3 3 2 STM Truncation 3/3 36 8.8 2 1 0 3 3 3 Splicing 1/1 48 6.9 1 0 0 1 1 1 Dupl./Delet. 1/1 6 12 0 1 0 1 1 0 Total 5/5 30 9.2 3 2 0 5 5 4 x Steroids Truncation 2/3 1# 20 110eee x Missense 1/2 1# 20 010eee x Splicing 1/1 2# 20 100eee x Dupl./Delet. 1/1 1# 20 100eee x Total 5/7 e 20 320eee LEV Truncation 2/3 2.5 54 0 2 0 1 0 0 ** Missense 1/2 7 670 1 0 110 Splicing 1/1 50 30 1 0 0 1 1 1 Total 4/6 19.8 50.3 1 3 0 3 2 1 GRIN2A TPM Truncation 2/3 9 4.6 0 2 0 2 1 1 Missense 1/2 1 4.5 0 1 0 0 0 0 Dupl./Delet. 1/1 12 13 1 0 0 1 1 1 leain:Gntc,neurodevelopment, Genetics, alterations: Total 4/7 11 7.4 1 3 0 3 2 2 ESM Truncation 2/3 5.5 33.5 1 0 1 2 1 0

Total 2/3 19.5 33.5 1 0 1 2 1 0 e OXC Truncation 1/3 25 56 1 0 0 1 1 1 12 Missense 1/2 2 30 0 1 0 0 0 0 Total 2/5 13.5 43 1 1 0 1 1 1 LTG Missense 1/2 27 4.2 1 0 0 1 1 1 VGB Missense 1/2 2 55 0 0 0 0 0 0 ** PER Truncation 1/3 7 8 1 0 0 110 PB Dupl./Delet. 1/1 19 6 1 0 0 1 1 1 Vit B6 Dupl./Delet. 1/1 1 30 0 1 0 eee european journal of paediatric neurology xxx (2017) 1e12 11

combination and for refractory cases high dose steroids.27 selection bias due to recruitment of all patients from tertiary Both, improving the EEG by reducing IEDs and enhancing epilepsy centers, and therefore presenting with seizures and a language skills have been reported in CSWS.28 In our patients lack of a standardized protocol. In addition, although software CLB and high dose steroids were also commonly used in 5/7 tools used in this study are widely distributed to help to patients each. CLB had a positive effect in 3/5 patients, of interpreting genetic results, they are not able to serve to pre- whom 2 had an ESES (n ¼ 1 truncation; n ¼ 1 splicing affected) dict the functional relevance of gene mutation alone. This is and did not show aggravation, and also steroids had a positive especially true in patients with inherited mutations with un- effect in 3/5 patients (n ¼ 1 each of patient with truncation; affected parents. Although one could speculate on incomplete splicing affected or duplication/deletion), of whom two had an penetrance of the mutation, this has not yet been shown in ESES. larger series. Consequently, a significant amount of GRIN2A LEV has also been reported to be effective in atypical mutations in this cohort were classified as variants of uncer- rolandic epilepsy and other variants.29 It was given in our tain significance. Nevertheless, by providing a detailed cohort in 4/6 patients, of whom one reported a positive effect description of the epileptic phenotype and genotype in each (one patient with affected splicing mutation of the GRIN2A patient (see Supplementary Tables), researches and epi- gene), while the other three did not see an effect at all. leptologists might identify patients with the same gene al- TPM was investigated in some trials with BECTS and led to terations. In addition, another important tool to investigate seizure freedom in a majority of patients.30,31 In our study the functional relevance of a gene mutation remains using group TPM received 4/7 patients, of whom only one reported in vitro approaches to test the functional properties i.e. with an improvement in seizure situation; three did not notice an respect to ligand affinity, ion current changes and down- effect. These observations are in discordance to the others stream signaling all of which were not addressed by this who noticed a positive therapeutic efficacy in idiopathic focal study. This is especially true when new therapeutic ap- epilepsy of TPM.31 proaches as memantine are raised. Altogether the GRIN2A patients responded well to the commonly used drugs in idiopathic focal epilepsy, above all to VPA, STM CLB and Steroids. 5. Conclusions Besides those patients with GRIN2A-associated epilepsy responding to common AEDs, a significant number continue Our data indicate that children with epilepsy and GRIN2A al- having seizures and encephalopathic EEG patterns. Thus, a terations present with different clinical phenotypes and a more individualized therapy targeting GluN2 directly might be spectrum of seizure types in the context of a pharmacor- an option in these patients. To date, there is one report on a esistant epilepsy providing also the first observation on the use clinical case investigating the effect of memantine in GRIN2A of standard AED for clinicians treating children with this form associated epilepsy.10 Before clinical application, the muta- of genetically determined epileptic syndrome. Given the low tion of the patient was tested in vitro for response to mem- rate of definitively pathogenic or likely pathogenic mutations in antine, in detail, retained sensitivity of NMDAR was shown this cohort of epileptic patients carrying GRIN2A mutations and this observation was translated into clinical application to (seven out of 19) it should be emphasized that genetic results the patient. The affected patient revealed reduced seizure should be interpreted with caution and rigorously adapted to burden with memantine. However, caution is warranted as the standards of the American College of Human Genetics. the NMDAR antagonist memantine might only work in patients with mutations leading to a gain of function of the receptor as described in the patients above and it might be better Conflict of interest to perform individual functional analysis before introduction 32 of memantine in a patient with GRIN2A-associated epilepsy. None. In previous reports in patients with GRIN2A-associated epilepsy, de novo mutations were clustered around transmembrane domains and the functional domains S1 and S2 Appendix ASupplementary data suggesting that these regions exhibit relevant functional 32 properties of the receptor. This was also true in the two Supplementary data related to this article can be found at subjects carrying likely pathogenic missense variants in our http://dx.doi.org/10.1016/j.ejpn.2017.01.001. cohort. Variants of unknown significance most likely clustered around the C-terminal domain suggesting a less significant functional importance of this domain as suggest before. references No missense mutation was observed in this cohort within the NTD despite of the likely functional significance of this domain. The NTD domain harbours the zinc-binding motif 1. Paoletti P, Bellone C, Zhou Q. NMDA receptor subunit and mutations within this region were shown to diminish diversity: impact on receptor properties, synaptic plasticity e zinc-mediated receptor inhibition.2 The lack of occurrence of and disease. Nat Rev Neurosci 2013;14:383 400. 2. Lemke JR, Lal D, Reinthaler EM, et al. Mutations in GRIN2A missense variants in this domain in our cohort is most likely cause idiopathic focal epilepsy with rolandic spikes. Nat Genet due to the small sample size. 2013;45:1067e72. There are several limitations of our study. These are 3. Leypoldt F, Armangue T, Dalmau J. Autoimmune related to its retrospective character, small sample size, encephalopathies. Ann N Y Acad Sci 2015;1338:94e114.

4. Turner SJ, Mayes AK, Verhoeven A, Mandelstam SA, 19. Scheffer IE, Jones L, Pozzebon M, Howell RA, Saling MM, Morgan AT, Scheffer IE. GRIN2A: an aptly named gene for Berkovic SF. Autosomal dominant rolandic epilepsy and speech dysfunction. Neurology 2015;84:586e93. speech dyspraxia: a new syndrome with anticipation. Ann 5. Monyer H, Burnashev N, Laurie DJ, Sakmann B, Seeburg PH. Neurol 1995;38:633e42. Developmental and regional expression in the rat brain and 20. Berg AT, Berkovic SF, Brodie MJ, et al. Revised terminology functional properties of four NMDA receptors. Neuron and concepts for organization of seizures and epilepsies: 1994;12:529e40. report of the ILAE Commission on Classiﬁcation and 6. Paoletti P. Molecular basis of NMDA receptor functional Terminology, 2005e2009. Epilepsia 2010;51:676e85. diversity. Eur J Neurosci 2011;33:1351e65. 21. Nabbout R, Chemaly N, Chipaux M, et al. Encephalopathy in 7. Kalsi G, Whiting P, Bourdelles BL, Callen D, Barnard EA, children with Dravet syndrome is not a pure consequence of Gurling H. Localization of the human NMDAR2D receptor epilepsy. Orphanet J Rare Dis 2013;8:176. subunit gene (GRIN2D) to 19q13.1-qter, the NMDAR2A subunit 22. Wheless JW, Clarke DF, Arzimanoglou A, Carpenter D. gene to 16p13.2 (GRIN2A), and the NMDAR2C subunit gene Treatment of pediatric epilepsy: European expert opinion. (GRIN2C) to 17q24-q25 using somatic cell hybrid and radiation Epileptic Disord 2007;2007(9):353e412. hybrid mapping panels. Genomics 1998;47:423e5. 23. Bast T, Volp A, Wolf C, Rating D, Sulthiame Study G. The 8. Lesca G, Rudolf G, Bruneau N, et al. GRIN2A mutations in inﬂuence of sulthiame on EEG in children with benign acquired epileptic aphasia and related childhood focal childhood epilepsy with centrotemporal spikes (BECTS). epilepsies and encephalopathies with speech and language Epilepsia 2003;44:215e20. dysfunction. Nat Genet 2013;45:1061e6. 24. Rating D, Wolf C, Bast T. Sulthiame as monotherapy in 9. Marwick K, Skehel P, Hardingham G, Wyllie D. Effect of a children with benign childhood epilepsy with centrotemporal GRIN2A de novo mutation associated with epilepsy and spikes: a 6-month randomized, double-blind, placebo- intellectual disability on NMDA receptor currents and Mg(2þ) controlled study. Sulthiame Study Group. Epilepsia block in cultured primary cortical neurons. Lancet 2000;41:1284e8. 2015;385(Suppl 1):S65. 25. Borggraefe I, Bonfert M, Bast T, et al. Levetiracetam vs. 10. Pierson TM, Yuan H, Marsh ED, et al. Mutation and early- sulthiame in benign epilepsy with centrotemporal spikes in onset epileptic encephalopathy: personalized therapy with childhood: a double-blinded, randomized, controlled trial memantine. Ann Clin Transl Neurol 2014;1:190e8. (German HEAD Study). Eur J Paediatr Neurol 2013;17:507e14. 11. Richards S, Aziz N, Bale S, et al. Standards and guidelines for 26. Wirrell E, Sherman EM, Vanmastrigt R, Hamiwka L. the interpretation of sequence variants: a joint consensus Deterioration in cognitive function in children with benign recommendation of the American College of medical genetics epilepsy of childhood with central temporal spikes treated and genomics and the association for molecular pathology. with sulthiame. J Child Neurol 2008;23:14e21. Genet Med 2015;17:405e24. 27. Caraballo RH, Veggiotti P, Kaltenmeier MC, et al. 12. Kumar P, Henikoff S, Ng PC. Predicting the effects of coding Encephalopathy with status epilepticus during sleep or non-synonymous variants on protein function using the SIFT continuous spikes and waves during slow sleep syndrome: a algorithm. Nat Protoc 2009;4:1073e81. multicenter, long-term follow-up study of 117 patients. 13. Choi Y, Chan AP. PROVEAN web server: a tool to predict the Epilepsy Res 2013;105:164e73. functional effect of amino acid substitutions and indels. 28. Wirrell E, Ho AW, Hamiwka L. Sulthiame therapy for Bioinformatics 2015;31:2745e7. continuous spike and wave in slow-wave sleep. Pediatr Neurol 14. Adzhubei IA, Schmidt S, Peshkin L, et al. A method and server 2006;35:204e8. for predicting damaging missense mutations. Nat Methods 29. von Stulpnagel C, Kluger G, Leiz S, Holthausen H. 2010;7:248e9. Levetiracetam as add-on therapy in different subgroups of 15. Schwarz JM, Cooper DN, Schuelke M, Seelow D. “benign” idiopathic focal epilepsies in childhood. Epilepsy MutationTaster2: mutation prediction for the deep- Behav 2010;17:193e8. sequencing age. Nat Methods 2014;11:361e2. 30. Kang HC, Eun BL, Wu Lee C, et al. The effects on cognitive 16. Reva B, Antipin Y, Sander C. Predicting the functional impact function and behavioral problems of topiramate compared to of protein mutations: application to cancer genomics. Nucleic carbamazepine as monotherapy for children with benign Acids Res 2011;39:e118. rolandic epilepsy. Epilepsia 2007;48:1716e23. 17. Moller RS, Larsen LH, Johannesen KM, et al. Gene panel 31. Kremenchugskaia MR, Globa OV, Kuzenkova LM. The use of testing in epileptic encephalopathies and familial epilepsies. topiramate in the treatment of focal epilepsy in children. Zh Mol Syndromol 2016;7:210e9. Nevrol Psikhiatr Im S S Korsakova 2013;113:33e8. 18. Carvill GL, Regan BM, Yendle SC, et al. GRIN2A mutations 32. Strehlow V, Heyne HO, Lemke JR. The spectrum of GRIN2A cause epilepsy-aphasia spectrum disorders. Nat Genet associated epilepsy. Epileotologie 2015;32. 2013;45:1073e6.