Duplication of the sodium channel gene cluster on 2q24 in children with early onset epilepsy


Address correspondence to Barbara Goeggel Simonetti, Division of Pediatric Neurology, Department of Pediatrics, Inselspital, University of Berne, CH-3010 Berne, Switzerland. E-mail: barbara.goeggel-simonetti@insel.ch


Purpose:  Sodium channel gene aberrations are associated with a wide range of seizure disorders, particularly Dravet syndrome. They usually consist of missense or truncating gene mutations or deletions. Duplications involving multiple genes encoding for different sodium channels are not widely known. This article summarizes the clinical, radiologic, and genetic features of patients with 2q24 duplication involving the sodium channel gene cluster.

Methods:  A systematic review of the literature and report of two cases.

Key Findings:  Nine individuals with 2q24 duplication involving the sodium channel gene cluster are described (seven female, two male). All presented with severe seizures refractory to anticonvulsant drugs. Seizure onset was in the neonatal period in eight patients with SCN1A-involvement, in infancy in one patient with SCN2A and SCN3A, but no SCN1A involvement. Seizure activity decreased and eventually stopped at 5–20 months of age. Seizures recurred at the age of 3 years in the patient with SCN2A and SCN3A, but no SCN1A involvement. Eight patients had a poor neurodevelopmental outcome despite seizure freedom.

Significance:  This article describes a distinct seizure disorder associated with a duplication of the sodium gene cluster on 2q24 described in otherwise healthy neonates and infants with severe, anticonvulsant refractory seizures and poor developmental outcome despite seizure freedom occurring at the age of 5–20 months.

The neonatal period and infancy carry a high risk for seizures and the development of epilepsies (Hauser, 1994; Adelow et al., 2009). Neonatal seizures are mostly symptomatic, due to disorders like hypoxic ischemic brain injury, hypoglycemia, or electrolyte derangements. Less frequent are seizure syndromes of the neonatal period or epilepsy syndromes with neonatal onset, such as benign familial neonatal epilepsy (BFNE), benign familial neonatal-infantile epilepsy (BFNIE), benign neonatal idiopathic seizures (BNIS), neonatal encephalopathy with suppression bursts (early myoclonic encephalopathy [EME], and early infantile epileptic encephalopathy [EIEE]), or malignant migrating partial seizures in infancy (MMPSI). Although outcome is generally good in BFNEs, BFNIE, and BNIS, it is severe in children with neonatal encephalopathy with suppression bursts or MMPSI. Etiology of the severe encephalopathy syndromes is metabolic, malformative, or often unknown (Yamamoto et al., 2011). Gene mutations were identified in the benign epilepsy syndromes, with neonatal or infantile seizures affecting genes encoding for potassium channels (KCNQ2 and KCNQ3) in BFNEs, or a sodium channel (SCN2A, located on chromosome 2q24.3) in BFNIE (Striano et al., 2006; Herlenius et al., 2007; Deprez et al., 2009). Recently, a new epileptic syndrome was reported described as “familial neonatal seizures with intellectual disability caused by a microduplication of chromosome 2q24.3″ (Heron et al., 2010). An association of chromosome 2 aberrations with epilepsy is well known and not surprising, given the fact that five voltage-gated sodium channel genes (SCN1A, SCN2A, SCN3A, SCN7A, and SCN9A) are located on 2q24 alone. This association is usually based on deletions including the 2q24 region (Grosso et al., 2008). Duplications involving the locus 2q24 have not been widely recognized as a cause for epilepsy. We review the literature on 2q24 duplications and report two new cases with 2q24 duplications involving the voltage-gated sodium channel gene cluster, presenting with severe neonatal seizures but evolving to a relatively good outcome.


A systematic review of the literature was conducted using the search term “2q24” combined with “duplication” in the databases PubMed, EMBASE, and Web of Science for publications in English, German, Italian, French, or Spanish from January 1960 until December 2011. Inclusion criteria were a detailed clinical description of patients with 2q24 duplication, including at least gender, age at onset of seizures, evolution of seizures, and neurodevelopmental outcome. The first and the last author independently appraised the evidence level of the publications.

In addition, the genetic databases OMIM and Genoglyphix (Signature Genomics/Perkin Elmer, Spokane, WA, U.S.A.) were searched for further cases of 2q24 duplications.

The history, clinical, electroencephalographic, and radiographic findings of two patients are described in detail. Written informed consent for publication was obtained from both patients’ parents.

In our patients, array comparative genomic hybridization (CGH) was performed using a 135k Roche-NimbleGen GGX array (Roche-NimbleGen, Madison WI, U.S.A.) The raw data were processed and analyzed with NimbleScan version 2.5.26, SignalMap version as well as the Genoglyphix database version 2.4-6 according to the human genome build 36 (hg18). For confirmation of the duplication, we performed quantitative polymerase chain reaction (qPCR) for three markers within the duplicated region (within the genes GALNT5, SCN1A, XIRP2) as well as Fluorescence in situ hybridization (FISH) analysis (BlueGenome RP11-194P8 for Patient 1, RP11-275I3 for Patient 2).



Patient 1: This patient was found in the genetic database Genoglyphix (Signature Genomics/Perkin Elmer, case 57670) and has not been published elsewhere. The eutrophic baby girl was delivered at term without perinatal complications. Her mother had thyroid hormone substitution, well-controlled gestational diabetes, and premature contractions leading to temporary bed rest; otherwise the pregnancy was uneventful. She was the first child of unrelated parents. Apart from an uncle with learning difficulties, the family history was unremarkable. Seizures were observed on the first day of life, starting with an unusual cry, wide opening of the eyes, flushing, and bulbar and head deviation to the right, lasting for 10–30 s and occurring up to 18 times daily. Interictally, the girl was behaving adequately and showed no pathologic signs. The interictal electroencephalography (EEG) showed excessive focal spikes and sharp waves over the right centrotemporal leads. Ictal EEG depicted right centrotemporal, at times also bicentral, slow, repetitive spike-wave activity, followed by background slowing. Half of the seizures recorded were electroencephalographic only; the other half occurred with the stereotypic clinical seizure described. The results of a metabolic screen including electrolytes and glycemia, as well as cerebral ultrasound and magnetic resonance imaging (MRI) were within the normal range. The seizures continued despite treatment with phenobarbitone and pyridoxine, but finally came to a halt at 8 months of age, after introduction of carbamazepine. Subsequent clinical and EEG controls were normal and the anticonvulsive medication was stopped at the age of 2.5 years. A neuropsychological evaluation at the age of 6.3 years showed a verbal IQ score <70 and a nonverbal IQ of 86. The full scale IQ could not be assessed due to the patient’s attention deficit.

Array CGH revealed a 6.7 Mb duplication of 2q24 with the genomic position chr2: 161,297,937–168,028,317 (hg18), spanning the complete sodium channel gene cluster (SCN1A, SCN2A, SCN3A, SCN7A, and SCN9A). The duplication has been confirmed by FISH analysis (RP11-194P8), and proved to be de novo.

Patient 2 was the first child to unrelated, healthy parents originated from Vietnam, Austria and Croatia. Detailed family history revealed that the mother of Patient 2 presented two early miscarriages, and one of the father’s half sisters had a unilateral agenesis of the kidney. Family history was otherwise unremarkable. Due to an increased nuchal translucency of 6.5 mm, prenatal chromosome analysis had been performed and showed a normal female karyotype 46, XX. In addition, the fetus showed global intrauterine growth retardation (all parameters on the 3rd percentile) and ultrasonographic diagnosis of fetal congenital anomalies of the kidneys and the urinary tract (CAKUT). The mother reported tobacco consumption during pregnancy but no further teratogenic exposure. Delivery was uneventful and on term. At birth, Patient 2 presented with a weight of 2,740 g (P5–P10), length of 45 cm (<P3, −3 standard deviation score [SDS]), and head circumference of 31 cm (<P3, −3.5 SDS). She had no dysmorphic features. The abdominal ultrasound showed agenesis of the right kidney with a compensatory hypertrophic left kidney. The cardiac ultrasound was normal. On the third day of life, repetitive focal tonic and multifocal clonic seizures were observed, starting with central cyanosis and head deviation, followed by multifocal clonic extremity movements, lasting up to 2 min. Interictally, the child was unremarkable. The interictal EEG recorded multifocal sharp waves. The ictal EEG showed a discharge pattern starting with a generalized attenuation of the background activity, followed by sharp and slow waves, secondarily generalizing.

An extensive metabolic work-up revealed no pathologic findings. Apart from a mild corpus callosum hypoplasia, cerebral MRI including H-spectroscopy, showed no further anomalies. The seizures continued despite administration of multiple anticonvulsive drugs (see Table 1), and ceased at 2 weeks of age with a combination therapy of phenobarbitone and oxcarbazepine. There was a seizure-free interval up to the age of 2 months, after which recurrent seizures of the same semiology occurred, clustering up to several times daily, despite anticonvulsant plasma levels in the therapeutic range. An afebrile urinary tract infection was diagnosed and successfully treated with antibiotics. The seizures stopped after valproic acid had been added to phenobarbitone and oxcarbazepine. Subsequently, the girl remained seizure free despite the fact that her parents had spontaneously withdrawn the anticonvulsant medication when she was 1.5 years of age. At last follow-up at the age of 2 years, the child presented with a normal developmental quotient despite a mild dystonic movement disorder. The EEG was normal.

Table 1.   Seizure and EEG features in patients with 2q24 duplications
Patient no. (author of publication)Age at epilepsy onsetSeizure semiologySeizure durationSeizure frequencyResponse to AED (+ pos. effect, − no effect)EEG featuresOutcome
  1. ADH, attention deficit and hyperactivity; CZP, clonazepam; CBZ, carbamazepine; FA, folinic acid; GMF, gross motor function; GTCS, generalized tonic–clonic seizure; LEV, levetiracetam; LZP, lorazepam; MDZ, midazolam; ID, intellectual disability; NS, not stated; nvIQ, nonverbal IQ; OXC, oxcarbazepine; PB, phenobarbitone; PHT, phenytoin; PLP, pyridoxal phosphate; PN, pyridoxine; PR, pharmacoresistant epilepsy (no response to >2 anticonvulsive drugs); PSW, polyspike waves; SSW, sharp and slow waves; SW, sharp waves; TPM, topiramate; VGB, vigabatrin; VPA, valproic acid; vIQ, verbal IQ.

1 (proband, III: 2 in: Heron et al., 2010)Neonate (18 days)Generalized tonic with apnea, At 4 months: tonic and clonic seizures with cyanosis, eye deviation, later also absence and myoclonic seizuresNSInitially few times per week. 4 months: 15×/day− PB, CZP, PHT, PN, LTG
Neonate: independent bilateral midtemporal and central SSW complexes
4 months: diffuse background slowing, frequent independent bilateral central and temporal discharges
17 months: very frequent high amplitude spike, SW and slow wave activity, approaching hypsarrhythmia
Generalized paroxysms of SSW
18 months: marked reduction in epileptiform activity, better organization of background
Seizures ceased by age 20 months
IQ < 40, autistic features, behavioral aggression
2 (proband’s sister, III: 4 in: Heron et al., 2010)Neonate (2 days)Generalized clonic, myoclonic or tonic with eye deviation and apneaNSNSPB, PHTAt 4 months: SSW in right temporal regionSeizures stopped by 5 months
Mild motor developmental delay, FSIQ 50–60
3 (proband’s sister, III: 5 in: Heron et al., 2010)Neonate (3 days)Generalized clonic, myoclonic or tonic with eye deviation and apneaNSNSPB, PHTAt 3 weeks: normalSeizures stopped by 5 months
Normal early milestones, FSIQ 50–60
4 (proband’s mother, II: 2 in: Heron et al., 2010)Neonate (2 days), possibly in uteroMyoclonic, generalized seizuresNSNSPB, PHTNSSeizures stopped at 2 weeks
Borderline intellect
5 (Raymond et al., 2011)Neonate (2 weeks)Focal tonic, autonomic, second
20–60 sInitially 2–3, then 20–30/day− PN, LEV, TPM
+ PB
Ictal: Diffuse voltage attenuation followed by bilateral frontocentral rhythmic medium voltage α/β activity, then high voltage rhythmic sharp activitySeizure reduction at age 5 months
Moderate delay in GMF at 6 months
6 (Okumura et al., 2011)Neonate (3 days)Focal motor (not further specified)10–20 s5–20/day− MDZ, LEV, PB
+ PB (20 mg/kg/day), VPA
Interictal: abnormal background activity with spiky transientsSeizure free since age 5 months
Severe global developmental delay
7 (Patient 1)Neonate (1 day)Focal tonic, autonomic20–30 s12–18/day− PB, PN
Interictal: focal spikes and SW (right centrotemporal)
Ictal: right centrotemporal and bicentral SSW, followed by background slowing
Seizure free since age 8 months, normal EEG without AED at 2.5 years
Age 6.4 years: vIQ <70, nvIQ 86, FSIQ-test not completed. ADH
8 (Patient 2)Neonate (3 days)Multifocal clonic30 s–2 min5–20/day− PB, PHT, LZP, VPA, LEV, PN, FA, PLP
+ OXC comb. with PB, VPA
Interictal: multifocal (frontoparietal and right frontotemporal) SW
Ictal: 4 Hz PSW and SW activity, frontal onset, secondarily generalizing, 90 s duration, clinically multifocal cloni
Seizures stopped at age 6 months, off AED since age 9 months
Normal development and EEG at age 2 years
9 (Vecchi et al., 2011)3 monthsFocal, second. GTCS
Age 2 years: atonic seizures
ClustersNS+ CBZ
+ VPA for atonic seizures
Multifocal PSWSeizure free up to age 2 years (CBZ stopped at 13 months), then recurrence of epilepsy with atonic seizures
At 7 years seizure free, ID

Array CGH performed on DNA extracted from lymphocytes revealed a 10.1 Mb duplication on 2q24 (chr2: 157,787,313–167,928,769) spanning the complete sodium channel gene cluster as well. The duplication has been confirmed by qPCR for all three markers and proved to be de novo. FISH analysis did not reveal evidence for a translocation of the duplicated fragment.

Review of the literature

The systematic literature search revealed 27 publications (PubMed, 6; EMBASE, 12; Web of Science, 9 publications). After excluding duplicates, 14 remained. The 14 full-text articles were screened for the inclusion criteria. Four articles describing seven patients from four families were included in this study (Heron et al., 2010; Okumura et al., 2011; Raymond et al., 2011; Vecchi et al., 2011). All publications were case reports.

Including the two patients presented in this article, nine cases of 2q24 duplications with early onset epilepsy are described in the medical literature. The seizure characteristics and EEG features are summarized in Table 1, and the clinical and radiographic findings in Table 2. Seven of the nine patients are female. Ethnicity was Caucasian and Asian in one child (patient 2), Caucasian in six children, and not stated in two. Family history was positive for cognitive problems in five patients, among them the four patients belonging to the same family described by Heron et al. (2010). Apart from the family described by Heron et al. (2010), two patients had a positive family history for epilepsy, one of them for febrile seizures, as well (Raymond et al., 2011; Vecchi et al., 2011). None of the eight patients’ parents, in whom according information was given, were consanguineous. Figure 1 depicts the genetic aberrations detected in patients with 2q24 duplication. The duplication was found to be de novo in three patients (Okumura et al., 2011), including the two novel mutations presented herein, and in the mother of the index patient in the family described by Heron et al. (2010). A mosaic of the duplication was detected in the father of the child presented by Raymond et al. (2011). No information was provided on the genetic findings in the parents of the child without involvement of the SCN1A gene (Vecchi et al., 2011).

Table 2.   Clinical and radiographic patient characteristics
Patient no. (author, year of publication)SCN genes involvedGenderEthnicityConsanguinityFamily history for epilepsyFurther personal medical historyCerebral imaging
  1. CAKUT, congenital anomalies of the kidneys and urinary tract; F, female; M, male; IUGR, intrauterine growth retardation; NS, not stated; wGA, weeks of gestational age.

1 (proband, III: 2 in: Heron et al., 2010)3′ end of SCN1A, SCN2A, SCN3AMCaucasianNoneYNeonatal hemolytic anemia; birth 36 wGANormal cerebral MRI at 4 months
2 (proband’s sister, III: 4 in: Heron et al., 2010)3′ end of SCN1A, SCN2A, SCN3AFCaucasianNoneYBirth 36 wGANormal cerebral MRI
3 (proband’s sister, III: 5 in: Heron et al., 2010)3′ end of SCN1A, SCN2A, SCN3AFCaucasianNoneYNeonatal hemolytic anemia, birth 36 wGANormal cerebral MRI
4 (proband’s mother, II: 2 in: Heron et al., 2010)3′ end of SCN1A, SCN2A, SCN3AFCaucasianNoneYNihilNS
5 (Raymond et al., 2011) SCN1A, SCN2A, SCN3A FNSNonePositive (paternal grandmother)Birth 35 5/7 wGANS
6 (Okumura et al., 2011) SCN1A, SCN2A, SCN3A, SCN7A, SCN9A FNSNoneNegativeNihilMRI normal
7 (Patient 1) SCN1A, SCN2A, SCN3A FCaucasianNoneUncle with learning difficultiesWhite spellsNormal MRI
8 (Patient 2) SCN1A, SCN2A, SCN3A FCaucasian (3/4) and Asian (1/4)NoneNegativeGlobal IUGR, CAKUTMRI: callosal hypoplasia, otherwise normal
9 (Vecchi et al., 2011) SCN2A, SCN3A MCaucasianNonePositive for epilepsy and febrile seizuresNihilMRI normal at 3 months
Figure 1.

Extent of 2q24 duplication in patients with early onset epilepsy. The sizes of the six different duplications on 2q24 vary between 1.5 and 12 Mb. Apart from the last patient (Vecchi et al., 2011) all duplications affect SCN1A. In the first two patients (Heron et al., 2010; Raymond et al., 2011), SCN2A and SCN3A are duplicated as well. In the following three cases (patients 1, 2, and Okumura et al., 2011), the respective duplication covers the complete SCN cluster on 2q24.

All patients with an involvement of the SCN1A gene showed a seizure onset in the neonatal period. Seizure semiology was described in detail in seven patients (including the two presented by us), three of them starting with a tonic deviation of the head, two had facial flushing, one had cyanosis, two had an unusual cry, followed by clonic movements in two, multifocal in one, generalized in the other (Heron et al., 2010; Raymond et al., 2011). The members of the family described by Heron showed tonic seizures with eye deviation and apnea, such as generalized clonic, myoclonic and, in the index patient later in his life, absence seizures. Seizures were refractory to multiple anticonvulsant agents in five patients, including the index patient of the family described by Heron et al. (2010), whereas the mother and two sisters of this index patient were seizure free on phenytoin combined with phenobarbitone. Seven of the eight patients with involvement of the SCN1A gene were seizure free by 20 months of age, the latest, even without anticonvulsant medication (including the two patients presented by us). The eighth patient was only 6 months old at last follow-up, when he had a significant reduction in seizure frequency. The patient without SCN1A involvement had a seizure-free interval in the second year of life, when atonic seizures occurred (Vecchi et al., 2011). At last follow-up at the age of 7 years, he was again seizure free. Outcome was guarded in the majority of patients, with severe (6 of 9) or moderate (2 of 9) developmental delay or cognitive problems. Patient 2 presented here was the only child with normal development despite minor neurologic findings (mild dystonic movement disorder). Follow-up duration varied from 6 months up to adulthood.


The main finding of this review is the distinguished clinical evolution in patients with 2q24 duplications involving the SCN gene cluster with severe, anticonvulsant-refractory seizures of neonatal onset, occurring several times daily in otherwise healthy newborns. Seizure semiology was heterogeneous in the neonatal phase and in infancy, comprising focal or generalized tonic seizures with deviation of head or eyes, apnea, and cyanosis or facial flushing, such as multifocal clonic or myoclonic seizures. Effective anticonvulsant drugs were valproic acid, high-dose phenobarbitone, carbamazepine, oxcarbazepine, vigabatrin, clonazepam, and phenytoin. However, valproic acid, phenobarbitone, phenytoin, and the benzodiazepines midazolam and lorazepam were ineffective in some patients. The positive response to phenytoin, carbamazepine, and oxcarbazepine is surprising, taking into account that these agents are contraindicated due to their capacity to exacerbate epileptic activity in patients with Dravet syndrome, which is based on a SCN1A mutation leading to haploinsufficiency in the majority of cases (Marini et al., 2011). The different gene dose in patients with duplications versus patients with haploinsufficiency may explain the diverging response to anticonvulsant drugs acting on the sodium channel, such as phenytoin, carbamazepine, and oxcarbazepine. The present data are based on a small number of patients with 2q24 duplication and does not allow drawing a conclusion on the efficacy of the different treatment regimen used in these patients.

Another difference from Dravet syndrome is the fact that none of the patients had seizures related to fever. Finally, epilepsy onset is usually in infancy in patients with Dravet syndrome, whereas the patients with 2q24 duplications involving the SCN1A gene presented herein had seizures starting in the neonatal period. Partial duplications of SCN1A have been described as a rare cause of Dravet syndrome (Marini et al., 2009), which may be explained by disruption of the gene consequently leading to haploinsufficiency.

The patient without SCN1A-gene involvement showed a different clinical pattern with seizure onset at the age of 3 months only, evolving to epilepsy with atonic seizures after a seizure-free interval during the second year of life. As some other patients and despite the fact that he carried a mutation for two sodium channels (SCN2A and SCN3A), as well, his seizures stopped with carbamazepine (Vecchi et al., 2011).

SCN3A is so far not associated with an OMIM-annotated phenotype. There is one report of a patient with focal seizures and a mutation in SCN3A (Holland et al., 2008). SCN2A mutations, deletions, and partial duplications (all leading to haploinsufficiency) are associated with benign familial infantile seizures, and rarely cause an early infantile epileptic encephalopathy. The genomic region on 2q24 harbors several genes of unknown function and, to the best of our knowledge, there is no further apparent epilepsy-causing gene. However, some genes may have an influence on seizure activity, as demonstrated for Slc4a10-deficient mice by Jacobs et al. (2008).

Although the number of patients is not sufficient to prove a significant genotype–phenotype correlation, SCN1A seems to be one of the major causative genes influencing the phenotype caused by a duplication of the region.

None of the patients showed obvious dysmorphic features. Brain MRI showed mild callosal hypoplasia in one, and no pathologic findings in six of the seven patients in whom this information was provided. Unfortunately, up to now, none of the patients had repeat imaging after the age of 2 years, when progression of myelination allows a better yield in detecting structural anomalies.

The association of congenital renal anomalies with duplication of the 2q24 region, as found in one of our patients, has not been described so far, and there is currently no evidence on genes contained in the 2q24 region being major candidates for renal abnormalities. Even though one of our patients has a unilateral renal agenesis with a positive family history (father’s half sister), the patient’s father is not a carrier of the duplication. Moreover, the aunt with the unilateral renal agenesis concerned had neither a history of intellectual impairment nor a seizure disorder. Therefore, we suppose a different etiology of the renal agenesis. Outcome was guarded in the majority of patients (8 of 9), despite seizure freedom. Seizures stopped during the first 2 years of life in most patients (7 of 9). However, the origin of intellectual disability may be multifactorial in the cases published by Heron et al., taking into account that the mother, who had a microduplication of chromosome 2q24, showed neonatal seizures but still developed to borderline intellect, whereas her husband, the father of the affected three children, who had been unavailable for genetic testing, was reported to be intellectually impaired and to have had seizures during childhood. Moreover, one of the siblings without 2q24 microduplication showed intellectual disability without a seizure disorder (Heron et al., 2010).

Four patients had a parent with the same duplication; three belonged to the same family described by Heron et al. (2010), with their mother, who had neonatal seizures but no significant intellectual disability, carrying the 2q24 microduplication. As hypothesized earlier, the intellectual disability in the three siblings (one boy, two girls) affected by neonatal seizures and intellectual impairment, and one of their brothers with intellectual impairment but no seizures, might have been inherited by their father who was not available for genetic testing. Another patient’s father had a mosaic with a healthy phenotype. This father’s own mother, thus the paternal grandmother of the patient, had seizures of unknown origin, not otherwise described, which are most probably unrelated to the epilepsy syndrome in the index patient (Raymond et al., 2011).

In conclusion, this review suggests that 2q24 duplication including the sodium channel gene cluster is associated with severe, anticonvulsant refractory epilepsy in otherwise healthy infants, starting during the neonatal period and carrying a poor neurodevelopmental outcome, despite eventually achieved seizure freedom between the ages of 5–20 months in most patients.


We kindly thank Robert N. Walker, MD, for sharing information on the first patient presented in this article.


None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.