Benign Familial Neonatal Seizures

Benign familial neonatal seizures (BFNS), is a rare autosomal-dominant epilepsy of the newborn characterized by recurrent, mainly focal seizures that begin in the first few days of life and remit after a few weeks or months. Seizures start with a tonic posture accompanied by other symptoms such as apnea and other autonomic features. The seizures often progress to focal or bilateral clonic jerks. Generalized seizures have also been reported. Seizures are usually brief, lasting for approximately 1–2 minutes, but may occur as many as 20–30 times a day. The postictal state is short-lived, and the neonates look normal in the interictal period [Ronen et al., 1993]. The interictal electroencephalogram (EEG) pattern either is normal or shows minimal focal or multifocal abnormalities. EEG patterns suggestive of poor prognosis, such as a burst-suppression or inactive EEG, have not been reported. The original description of the disease dates back to the work of Rett and Teubel, who described a four-generation family with nine individuals affected [Rett and Teubel, 1964]. In their original work, the authors emphasized the heritability of the disease and the absence of acquired origin. The term “benign” was added only few years later, to highlight the fact that most affected individuals experienced normal psychomotor development. Nevertheless, 5 percent of BFNS patients present with febrile seizures and 11 percent develop epilepsy later in life [Plouin and Anderson, 2005]. In rare infants, the outcome may be poor [Dedek et al., 2003; Borgatti et al., 2004; Steinlein et al., 2007], and a single family with myokymia later in life has been described [Dedek et al., 2001].

BFNS is the first epilepsy syndrome for which a gene could be localized [Singh et al., 1998; Charlier et al., 1998]. This disorder has been linked to mutations in the neuronal voltage-gated potassium channel subunits, KCNQ2 and KCNQ3 [Singh et al., 1998; Charlier et al., 1998], located on chromosome 20 (20q13.3) and on chromosome 8 (8q24), respectively. KCNQ2 and KCNQ3 subunits mediate the M-current, a neuronal-specific K⁺ current, which plays a key role in controlling neuronal excitability [Biervert et al., 1998; Singh et al., 2003; Wang et al., 1998; Soldovieri et al., 2007]. The majority of cases, 56 percent, are associated with mutations in KCNQ2, with KCNQ3 mutations accounting for only 6.6 percent [Deprez et al., 2009]. More than 60 mutations of KCNQ2 have been reported so far, including missense, frameshift, nonsense, and splice-site mutations. Furthermore, submicroscopic deletions or duplications in KCNQ2 have been shown to account for a significant proportion of unsolved BFNS cases, in which mutations are not detected by direct sequencing [Heron et al., 2007a], and have been found in about one-sixth of BFNS families.

The age dependency of BFNS has been explained by the peculiar spatial and temporal expression pattern of KCNQ2 and KCNQ3 subunits in human development [Kanaumi et al., 2008]. A high degree of expression of KCNQ2 has been identified in the hippocampus, temporal cortex, cerebellar cortex, and medulla oblongata in fetal life. There is a decrease in expression of the gene after birth. The increased expression of KCNQ3 during fetal life compared with infancy has also been confirmed.

Benign Familial Neonatal-Infantile Seizures

Benign familial neonatal-infantile seizures (BFNIS) is an autosomal-dominant benign epilepsy of the early infancy with a high degree of penetrance. Age of onset ranges from 2 days to 6 months in different family members, with a mean age of onset of 3 months. Seizures have a focal onset, often with eye deviation, apnea, cyanosis, and staring. Secondary generalization may follow. Similarly to BFNS, seizures may occur in clusters. The interictal EEG is normal or shows occasional central or posterior focal spikes [Heron et al., 2002; Herlenius et al., 2007]. Seizures abate by 1 year of life, with a low risk of recurrence.

Unlike the previously discussed “benign familial neonatal seizures,” which are associated with potassium channel mutations, this syndrome has been linked to mutations in SCN2A [Heron et al., 2002], the alpha-2 subunit of the neuronal voltage-gated sodium channel [Berkovic et al., 2004a; Helbig et al., 2008]. Eight different mutations in ten families have been described so far [Heron et al., 2002; Berkovic et al., 2004b; Herlenius et al., 2007]. Mutations in SCN2A seem to be specific to this condition, with only very few cases being reported [Helbig et al., 2008]. The SCN2A mutations reported in association with BFNIS are all missense mutations. While the majority are predicted to be gain-of-function, more recently, Misra and colleagues reported three mutations that are a loss-of-function type [Misra et al., 2008]. An increase in the Na⁺ current caused by modifications of voltage-dependent gating, or a reduction in the current caused by reduced channel expression or impaired gating kinetics, has the ultimate effect of increasing neuronal excitability, thus favoring seizure generation or propagation [Avanzini et al., 2007; Misra et al., 2008].

Neonatal Epileptic Encephalopathies

The International League Against Epilepsy (ILAE) defines two further neonatal epilepsy syndromes, early myoclonic encephalopathy and Ohtahara syndrome (early infantile epileptic encephalopathy). Both syndromes have been defined as epileptic encephalopathies with neonatal onset and a typical suppression-burst EEG pattern [Aicardi and Ohtahara, 2005]. They are associated with severe psychomotor retardation and grave prognosis [Djukic et al., 2006]. A distinguishing feature is the main seizure type, which is myoclonic in the case of early myoclonic encephalopathy, and tonic in the case of early infantile epileptic encephalopathy. Ohtahara syndrome has been linked to a gene encoding syntaxin binding protein 1 (STXBP1), which encodes a vesicle release protein and has been implicated in 35.7 percent of cases [Saitsu et al., 2008; Deprez et al., 2009]. STXBP1 mutations have also been implicated in cases [Saitsu et al., 2008; Deprez et al., 2009] of mental retardation and nonsyndromic epilepsy [Hamdan et al., 2009]. Therefore, the role of STXBP1 in epilepsy is interesting on a couple of levels. First, it extends the genetic epilepsies to include molecules involved in neurotransmission through vesicle release, in addition to the many known ion channel mutations. Second, it redefines a syndrome as “genetic,” which might otherwise have been classified as “structural” or “metabolic.” Thus far, only one other protein involved in synaptic vesicle transmission, Synaspin-1, has been linked to epilepsy and learning difficulties, as reported in one family [Garcia et al., 2004]. Testing for STXBP1 may be useful in nonlesional cases of Ohtahara.

A second gene has recently been implicated in Ohtahara; an ARX protein truncation mutation has been found in one family [Fullston et al., 2010]. ARX is a homeobox gene located on the short arm of the X chromosome, and it has been implicated in a number of different syndromes including West’s syndrome (see below), lissencephaly, and nonsyndromic X-linked mental retardation and epilepsy [Strømme et al., 2002]. Limited human and extensive mouse gene-expression studies show high levels of ARX expression in the fetal brain, particularly in the neuronal precursors of the germinal matrix and the ventricular zone. High levels of expression are also observed in the subventricular zone, the caudate nucleus, putamen, substantia nigra, corpus callosum, amygdala, and hippocampus [Bienvenu et al., 2002; Kitamura et al., 2002; Colombo et al., 2004; Poirier et al., 2004]. This expression during early development and predilection for neuronal tissue suggests that ARX has a pivotal function in neurodevelopment. Recently, homozygous mutations on SLC25A22 have been described in case reports of consanguineous families with neonatal epileptic encephalopathies with suppression-burst [Molinari et al., 2005, 2009].

A final epileptic encephalopathy is characterized by early onset of a severe intractable seizure in girls with normal interictal EEG and severe hypotonia. The syndrome is associated with mutations in X-linked cyclin-dependent kinase-like 5 (CDKL5) [Bahi-Buisson et al., 2008a, b]. Mutations in the CDKL5 gene have also been shown to cause infantile spasms and Rett’s syndrome-like phenotype. Recent studies [Bahi-Buisson et al., 2008a, b] highlighted the key clinical features of this rare epileptic encephalopathy that should help establish a molecular diagnosis. In younger patients under 2 years of age, early-onset epilepsy is probably the most consistent sign in CDKL5 mutations. In all cases, epilepsy starts within 3 months, often in the neonatal period, with very frequent seizures and an interictal EEG that is normal or indicates background slowing. At this age, neurologic examination reveals severe hypotonia and poor eye contact. Subsequently, a large proportion of patients develop epileptic encephalopathy characterized by the occurrence of infantile spasms with hypsarrhythmia, and then multifocal epilepsy. While brain magnetic resonance imaging (MRI) might be normal at onset, it is usually found to be abnormal later in the course of the disease. Although it is nonspecific, most patients exhibit cortical atrophy combined with hyperintensities in the temporal lobe white matter that could be related to abnormal myelination. The evolution of the disorder includes the appearance of hand stereotypies from the age of 18 months to 3 years, hand apraxia, sleep disturbances and deceleration of head growth [Bahi-Buisson et al., 2008a]. All girls with CDKL5 mutations are severely delayed, with limited, if any, autonomy.

Benign Familial Infantile Seizures

Benign familial infantile seizures (BFIS) syndrome was first described in 1992 by Vigevano and co-workers, who reported on cases with focal seizures in infancy, a family history of convulsions, benign outcome, and autosomal-dominant inheritance [Vigevano et al., 1992]. This syndrome is now included in the most recent classification and terminology proposed by ILAE [Engel, 2001]. Age of onset ranges from 4 to 8 months, with a peak around 6 months. Seizures typically occur in clusters of brief recurrent episodes of up to ten a day, never reaching true status epilepticus. Seizures are often characterized by head and eye deviation that may alternate in laterality for different attacks; even in the same patient, EEG recordings corroborate independent right and left seizure onset, confirming the alternating clinical pattern [Vigevano, 2005].

While the overall prognosis remains excellent in terms of remission from seizures, there is a risk of paroxysmal dyskinesias in later childhood [Lee et al., 1998; Demir et al., 2005]. BFIS is the least well characterized genetically, with multiple loci implicated but few candidate genes. Multiple authors have linked the syndrome to the pericentromeric region of chromosome 16 [Lee et al., 1998; Demir et al., 2005]. There has also been one report linking the syndrome to chromosome 19 [Guipponi et al., 1997], and one linking it to chromosome 1 in a family with hemiplegic migraine and a mutation in the ATP1A2 gene, which encodes a sodium-potassium ATPase transporter [Vanmolkot et al., 2003]. In addition, mutations in SCN2A and KCNQ2 have been described [Striano et al., 2006; Zhou et al., 2006], suggesting a possible overlap of clinical and genetic characteristics of the three benign epilepsy syndromes occurring in the neonatal or neonatal-infantile period.

Dravet’s Syndrome

Dravet’s syndrome, previously known as severe myoclonic epilepsy in infancy (SMEI), was described by Charlotte Dravet in 1978 [Dravet, 1978]. This entity is an epileptic encephalopathy, a condition in which the epileptiform abnormalities themselves are believed to contribute to the progressive disturbance in cerebral function [Engel, 2001]. Onset is at approximately 6 months of age in previously healthy children. Dravet’s syndrome typically presents with prolonged hemiclonic or generalized febrile seizures, sometimes leading to status epilepticus. Myoclonic jerks, atypical absences, and complex partial seizures appear later in the course of the disease. Subsequently, children develop neurological deficits, such as ataxia and corticospinal tract dysfunction [Dravet et al., 1992], as well as psychomotor delay, behavioral disturbances, and significant learning problems [Wolff et al., 2006].

EEG findings may be normal at onset, and then show generalized, focal, and multifocal abnormalities, often associated with a strong photosensitivity. MRI is usually normal or, in a few cases, shows dilatation of the cisterna magna or slight diffuse atrophy [Dravet et al., 2005]. However, most recent neuroimaging techniques can identify hippocampal sclerosis in some patients during the course of the disease [Striano et al., 2007]. The onset during infancy and frequent association with prolonged febrile seizures help to distinguish the syndrome from epilepsy with myoclonic-astatic seizures, also known as Doose’s syndrome.

Major advances have occurred in our understanding of Dravet’s syndrome with elucidation of its association with SCN1A mutations. Since the report of Claes and colleagues in 2001 [Claes et al., 2001], the concept that SCN1A is the major gene responsible for Dravet’s syndrome has been established [Fujiwara et al., 2003; Wallace et al., 2003; Fukuma et al., 2004]. The gene encodes the neuronal sodium channel NaV1.1 alpha subunit, which has four homologous regions, each encoding six transmembrane domains and a region controlling interactions with the permeable ion. Mutations have been found throughout the gene [Gambardella and Marini, 2009], and a website has been created to track the increasing number of mutations at http://www.scn1a.info. There are currently more than 300 known SCN1A mutations, with mutational rates in patients with Dravet’s syndrome of 33–100 percent [Claes et al., 2003; Fukuma et al., 2004; Mulley et al., 2005; Sugawara et al., 2002; Nabbout et al., 2003; Wallace et al., 2003]. Mutations are commonly truncations, although missense mutations can occur [Claes et al., 2003; Kanai et al., 2004; Mulley et al., 2005]. Among SCN1A mutations, those associated with Dravet’s syndrome are quite variable, including frameshift, nonsense, deletion, amplification, and duplication. However, simple missense mutations still account for 40 percent of cases [Mullen and Scheffer, 2009].

Mutations are usually de novo, although cases of germline mosaicism have been reported and should be taken into account for genetic counseling [Gambardella and Marini, 2009]. Patients for whom standard polymerase chain reaction (PCR) fails to detect an SCN1A mutation may undergo additional testing with multiplex ligation-dependent probe amplification, which detects microchromosomal copy number variations, such as deletions or duplications, in an additional 10–25 percent of PCR-negative cases [Madia et al., 2006; Mulley et al., 2006; Suls et al., 2006; Marini et al., 2009]. In terms of treatment for Dravet’s syndrome, topiramate, stiripentol, valproate, clonazepam, and levetiracetam, as well as the ketogenic diet, have been effective. Lamotrigine and carbamazepine may induce seizure aggravation [Guerrini et al., 1998; Wakai et al., 1996].

The spectrum of SCN1A-related epilepsies has been extended in the past few years, and several subgroups related to Dravet’s syndrome have now been reported, such as the syndrome of borderline SMEI (SMEB), characterized by a lack of myoclonic seizures and generalized spike-wave discharges, and the intractable childhood epilepsy with generalized tonic-clonic seizures (ICE-GTC), in which tonic-clonic seizures predominate [Osaka et al., 2007; Fujiwara, 2006; Rhodes et al., 2005]. SCN1A has also been implicated in cases of familial migraine [Cestele et al., 2008; Vahedi et al., 2009] and in a variety of other epilepsy syndromes. While no association has been found with simple febrile seizures [Petrovski et al., 2009] in the 5 percent of Dravet cases found to have familial mutations, family members often have other phenotypes, such as generalized epilepsy with febrile seizures plus (GEFS+) [Wallace et al., 2001b; Sijben et al., 2009]. In a recent comprehensive study, Harkin and colleagues have analyzed a cohort of 188 patients with various epileptic encephalopathies and found SCN1A mutations in 48 percent of patients, mainly novel; 96 percent were de novo [Harkin et al., 2007]. No patients with West’s syndrome or progressive myoclonic epilepsy had an SCN1A mutation. However, mutations were not restricted to those with typical epileptic encephalopathies. Interestingly, a few patients with cryptogenic generalized or focal epilepsy, myoclonic-astatic epilepsy, and Lennox–Gastaut syndrome also carried mutations.

Because SCN1A is implicated in a variety of other epilepsy syndromes, some of which are more benign, there is new potential for confusion regarding prognosis for affected patients, and a new need for clinicians to understand the subtleties of genetic testing for SCN1A mutations. While many different SCN1A mutations have been identified in patients with Dravet’s syndrome, most of which are unique to individuals, several recurrent mutations have also been found [Mulley et al., 2005]. Mutations are spread throughout the gene and, therefore, have different predicted functional effects on the protein [Kanai et al., 2004]. However, it is apparent that wherever and whatever the functional effects of these mutations are, they all lead to a similar seizure phenotype. Collating the available functional data on such mutations does result in an obvious explanation of the shared epilepsy phenotype; however, mathematical modeling has predicted an increased excitability via augmented action potential firing [Spampanato et al., 2004]. Mutations in SCN1A are collectively the most frequent genetic cause of epilepsy. Hence, further efforts to clarify the precise pathophysiology are likely to be important to the fundamental understanding of epileptogenesis.

Generalized Epilepsy with Febrile Seizures Plus

Generalized epilepsy with febrile seizures plus (GEFS+), mentioned above, was first described by Sheffer and Berkovic, and is associated with SCN1A mutations in roughly 10 percent of cases [Scheffer and Berkovic, 1997]. Mutations are generally missense, and the syndrome is characterized by an autosomal mode of inheritance, a broad spectrum of phenotypes, and a penetrance of 60 percent. Cases are mostly considered to have generalized epilepsy. The most common phenotype is characterized by the association of febrile seizures (FS), or febrile seizures extending beyond 6 years of age (febrile seizures plus, FS+), with afebrile generalized tonic-clonic seizures. This condition is a heritable syndrome, so clinical characteristics may be present in multiple family members. Less common phenotypes show the association of FS+ with absences, or myoclonic or atonic seizures, while the most severe phenotype includes myoclonic-astatic epilepsy. Partial seizures have also been reported, though rarely [Baulac et al., 1999].

West’s Syndrome

The final syndrome to consider that has infantile onset and a genetic etiology, in a minority of cases, is West’s syndrome. While the majority of cases of infantile spasms have no genetic association, mutations in ARX, Aristaless-related homeobox gene, have been associated with cases of X-linked recessive infantile spasms or “ISSX” [Fullston et al., 2010]. Genetic testing may therefore be appropriate in cases in which multiple males are affected by infantile spasms and mental retardation; dystonia is also a feature [Poirier et al., 2008; Shinozaki et al., 2009]. A second, dominant form of ISSX has been associated with a balanced X-autosomal translocation that disrupts CDKL5 (see Neonatal Epileptic Encephalopathies). A final syndrome of epilepsy and mental retardation limited to females has also been described [Deprez et al., 2009; Hynes et al., 2009; Depienne et al., 2009; Dibbens et al., 2008]. This syndrome is often associated with autistic, obsessive, or aggressive traits, and has been linked to various mutation of PCDH19 (protocadherin 19). (For a broader review of infantile spasms, please refer to Chapter 56.)

Early- and Late-Onset Childhood Occipital Epilepsy

Early-onset benign childhood occipital epilepsy was first described by Panayiotopoulos in 1989 as an epilepsy syndrome characterized by the “ictal triad of nocturnal seizures, tonic deviation of the eyes and vomiting” [Panayiotopoulos, 1989]. The age of onset of Panayiotopoulos syndrome (PS) is typically between 3 and 5 years of age; it affects boys and girls equally and is characterized by a low seizure burden. However, seizures can be prolonged, are marked by autonomic features, and often involve alteration in consciousness. Seizures may march to involve clonic movement of the head and upper extremities or secondary convulsion. A minority of patients may develop seizures during the day. The syndrome is benign, with remission before age 12.

The EEG reveals normal background activity, but frequent surface-negative spike and slow-wave complexes over the occipital region, occurring singularly or repetitively with a frequency of 2–4 Hz. Discharges may be apparent bilaterally with a persistent voltage asymmetry or a shifting predominance, or they may be unilateral. The spike-wave complexes attenuate with eye-opening and fixation, and are induced in darkness or with eye closure. Generalized discharges may also be present, and were noted in 50 percent of the original series. Other focal discharges also occur. The occipital spike wave remains present despite treatment with antiepileptic medications, but tends to decrease with age [Capovilla et al., 2009].

The Gastaut type, also known as childhood epilepsy with occipital paroxysms (CEOP), was first described by Camfield in 1978 [Camfield et al., 1978] and elaborated by Gastaut in 1982 [Gastaut, 1982]. CEOP has a later onset, presenting between 3 and 16 years, and seizures are characterized by visual symptoms rather than gaze deviation. Visual symptoms may include amaurosis, phosphenes, illusions, or hallucinations, and may progress to involve automatisms or hemiclonic seizures. Events are followed by migraine and classically have a diurnal pattern. EEG abnormalities similarly involve occipital spike wave activated by eye closure. However, the prognosis is somewhat poorer in terms of chance for remission.

The Panayiotopoulos and Gastaut forms of CEOP are probably better viewed as a clinical spectrum. One-third of patients with CEOP appear to have a mixed syndrome, with features of both Panayiotopoulos and Gastaut forms [Taylor et al., 2008]. Similarly, while there is evidence for a genetic component to CEOP, with some reports of concordant monozygotic or dizygotic twins, the concordance does not appear to be higher in monozygotic twins, indicating that other epigenetic or environmental factors likely also play a role [Taylor et al., 2008]. Additionally, while a family history of epilepsy can be established in roughly 36 percent of patients with either form of childhood occipital epilepsy, affected family members have been found to have a variety of generalized or focal epilepsy syndromes rather than CEOP alone.

Benign Epilepsy with Centrotemporal Spikes