Dominant Focal EpilepsiesSarah Weckhuysen, Mary Kurian, Fabienne Picard, and Stephanie Baulac

Published on 21/06/2017 by admin

Filed under Pediatrics

Last modified 21/06/2017

Print this page

rate 1 star rate 2 star rate 3 star rate 4 star rate 5 star
Your rating: none, Average: 2.9 (24 votes)

This article have been viewed 3839 times




The familial focal epilepsies with onset in childhood or adolescence are classified according to the focus of seizure onset in different members of the same family. Seizure onset can be uniform in all affected individuals in a family, as in autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE), autosomal dominant epilepsy with auditory features (ADEAF), and familial mesial temporal lobe epilepsy (FMTLE). Alternatively, different types of focal epilepsy can co-occur in the same family, as in familial focal epilepsy with variable foci (FFEVF). A diagnosis of any of these syndromes can only be made in the presence of a family history of focal epilepsy, as individual affected family members do not differ clinically from patients with sporadic focal epilepsies. During the past few years, a genetic cause has been identified for many of these syndromes, and a thorough familial history taking should be part of every workup of patients with focal epilepsy (Table 16.1).


Clinical Manifestations

Autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE) was the first familial focal epilepsy syndrome described, and the first epilepsy syndrome for which a genetic cause was identified (1,2). Within families, the epilepsy follows autosomal dominant inheritance with a clinical penetrance of about 70%, and the severity of the epilepsy varies greatly among the different affected members of a family (1,3,4). The core seizure phenotype consists of clusters of motor seizures arising from sleep, with a mean age at seizure onset between 8 and 11.5 years (1,3,5). Patients have a mean of eight seizures per night that often cluster over 1 or 2 hours. A nonspecific aura can precede the motor seizure, such as a somatosensory aura with tingling, for example, in the back or in the head. A sensation of being out of breath often accompanies the seizures (6). The ictal motor component includes hyperkinetic (frantic movements of bipedal activity, pelvic thrashing), tonic, or dystonic (sudden, abruptly changing dystonic puppet-like postures) elements. Seizures are of short duration, usually lasting less than 1 minute (1). Awareness is most often retained throughout the seizure. Because of the sometimes bizarre seizure semiology, a misdiagnosis of parasomnias, night terrors, or hysteria is frequent before the diagnosis of epilepsy is finally made (7).

Sleep deprivation (30% of the patients) and stress (30%) are possible provocative factors. Diurnal seizures may be observed in the most severe cases, usually during periods of poor seizure control. Rare secondarily generalized seizures are reported in nearly half of the patients, occurring at the onset of epilepsy or during the course of the disease.

Neurological clinical examination is normal. However, a high incidence of behavioral disorders, such as irritability, aggressiveness, and impulsive behavior, has been reported during adolescence (3), and personality disorders and/or depression can be present during adulthood. Neuropsychological disturbances may occur, but are most often subtle, with features of frontal and sometimes even temporal lobe dysfunction (8). A few reported families present with a more severe phenotype: psychosis and other psychiatric disorders were reported in several individuals from a Norwegian family (9) and in two other families with a severe ADNFLE phenotype; status epilepticus was present in 24% of the patients; behavioral and psychiatric disorders were reported in half of the patients; and intellectual disability was seen in 25% (10). In nearly half of the subjects from four families with ADNFLE, mild intellectual disability was present, showing that cognitive dysfunction is an integral part of the phenotype of ADNFLE (8). As discussed in this chapter, the severity of the epilepsy and associated neuropsychiatric features seem to be at least partly related to the underlying gene mutations.

TABLE 16.1


EEG and Neuroimaging

Interictal EEG is normal, or may show nonspecific focal epileptiform abnormalities in frontotemporal anterior regions, often only visible during sleep recording (1,3,5). Scalp ictal EEG is noncontributive in about half of the patients (4), and movement artifacts often obscure interpretation. Seizures occur during non-rapid eye movement (non-REM) sleep, mainly from stage 2, and start after a transition from sleep to an arousal (6). By using ictal video-EEG recordings and functional neuroimaging, Hayman et al showed different regions of onset within the frontal lobe in different family members, concordant with the differences in ictal symptomatology in individuals from the same family (11).

Magnetic resonance imaging (MRI) studies were classically considered normal in all patients; however, recent reports have described focal cortical dysplasias in families with DEPDC5 mutations (see following text) (12,13). A hypometabolism was observed in a few regions, including the right anterior orbitofrontal cortex, in a fluorodeoxyglucose (FDG)-positron emission tomography (PET) study of several ADNFLE patients (14).

Treatment and Evolution

The seizure semiology in a given individual is very stereotyped throughout life, except for some cases with seizure onset in early childhood, where seizures may evolve from tonic attacks to classical NFLE seizures including dystonic or hyperkinetic components (4). In many patients, seizures persist for years, but then tend to disappear with age, particularly around the fourth or fifth decade, without relapse after cessation of drug therapy (3). Carbamazepine is the most effective antiepileptic medication in this syndrome. It leads to a complete suppression of seizures in about 70% of patients. Interestingly, low dosages of carbamazepine (around 600 mg/day in adults) are usually sufficient, which may indicate pathophysiological mechanisms that are different from those of other epilepsies (15). However, drug resistance to carbamazepine and other antiepileptic drugs is observed in about 30% of patients.

Molecular Genetics

Mutations in genes coding for two different subunits of the neuronal acetylcholine nicotinic receptors (nAChRs), namely CHRNA4 and CHRNB2, have been identified in 10% to 15% of the ADNFLE families (4,1618). A mutation in CHRNA2 was identified in two Italian families with ADNFLE (19,20). The nAChRs are ligand-gated ion channel receptors. Being the first epilepsy genes to be identified, they contributed to the concept of epilepsy being a channelopathy. Nearly all reported nAChR mutations are missense mutations. A study analyzing the clinical features of 19 ADNFLE families with a total of 150 patients showed that some nicotinic receptor mutations, like the Ser284Leu mutation in CHRNA4, appear to be associated with an increased risk for major psychiatric or neuropsychological symptoms such as mental retardation, schizophrenia-like symptoms, or marked specific cognitive deficits, but the risk is low for most other mutations (21).

More recently, in a small percentage of ADNFLE families, mutations have been identified in KCNT1, which encodes a Na+-gated K+ channel (22), and in DEPDC5, a gene encoding DEP (Disheveled, Egl-10 and Pleckstrin) domain containing protein 5 (23,24).

Mutations in KCNT1 were identified in four families with severe forms of ADNFLE. They had an earlier mean age at onset of 6 years, compared to 10 years in the classical form of ADNFLE, and had more frequent intellectual disability, behavioral, or psychiatric problems (22). Remarkably, in the same journal issue, mutations in KCNT1 were described in patients with the seemingly unrelated and much more severe syndrome epilepsy of infancy with migrating focal seizures (EIMFS), previously called malignant migrating partial seizures of infancy (25). Since then, KCNT1 mutations seem to have given rise to a whole spectrum of focal or multifocal epilepsies with different degrees of intellectual impairment (26). Although functional studies support some degree of genotype–phenotype correlation, a few mutations leading to both ADNFLE and EIMFS have been described (26,27). There is preliminary evidence for a positive effect of the antiarrhythmic drug quinidine, a potassium channel blocker, in epilepsies due to KCNT1 mutations (28,29).

Recent studies have further reported DEPDC5 mutations in different familial focal epilepsy syndromes, in particular, familial focal epilepsy with variable foci (FFEVF), but several smaller families with ADNFLE or familial temporal lobe epilepsy (FTLE) have been described too (13,23,24,30). It is possible that the families with ADNFLE or FTLE, in whom DEPDC5 mutations were found, were too small to identify the variable foci within the family, and the subsequent finding of additional affected members might allow us to correct the classification of the epilepsy syndrome. In general, families with ADNFLE and DEPCD5 mutations have a higher rate of drug resistance compared to the classical ADNFLE (24). DEPDC5 is an upstream negative regulator of mammalian target of rapamycin mTOR complex 1 (mTORC1) (12,31). Other proteins of the same mTORC1 system are involved in tuberous sclerosis (32). Interestingly, MRI abnormalities, in particular focal cortical dysplasias, were described in some patients with familial focal epilepsy and DEPDC5 mutations, with a preferential location in the frontal lobe or in the insula (12,13). Some of these drug-resistant patients benefited from surgical treatment, and had a good long-term outcome (13).

Finally, a few mutations have also been identified in the corticotropin-releasing hormone (CRH) gene (33,34). However, functional studies have shown contradictory results, consisting of both increased and diminished levels of CRH expression, and more evidence is needed to securely implicate the gene in the etiology of ADNFLE.


ADNFLE Caused by nAChR Mutations

The nAChRs are ligand-gated ion channels permeable for cations (Na+, K+, Ca2+). The endogenous ligand is acetylcholine (ACh), whereas nicotine is an exogenous ligand. The nAChRs are pentameric receptors, and they occur in different functional subunit combinations. The most widely expressed nAChR in the human brain is the α4β2 nAChR. Most of the α4β2 nAChRs are presynaptic and have a neuromodulatory role: They enhance the release of the neurotransmitter present in the neurons on which they are located (GABA, glutamate, dopamine, norepinephrine, serotonin, or ACh). Other nAChRs are postsynaptic and mediate fast excitatory synaptic transmission. The nAChRs are involved in brain development and plasticity and participate in many different brain functions, including modulation of sensory inputs, sleep, mood, attention, reward, and learning (35). As for their role in ADNFLE, electrophysiological studies of ADNFLE mutations tend to show a gain of function of the mutant receptor (36), and mouse models point toward alterations in the control of GABAergic transmission in the prefrontal cortex. The precise mechanisms through which mutations in a ubiquitously expressed gene lead to a frontal lobe nocturnal epilepsy are, however, still little understood (16).

ADNFLE Caused by KCNT1 Mutations

The KCNT1 or slack channel, is a sodium-gated potassium channel; its presence in the central nervous system is ubiquitous, but its location in the cortex is most prominent in the frontal lobe, according to a study performed in rodents (37). In vitro functional studies showed an increased current of the mutant KCNT1 channels (27). This gain of function was generally greater for mutations identified in EIMFS than in ADNFLE. Many mutations are located in the large cytoplasmic C-terminal domain of KCNT1, which interacts with a large network of proteins, among which is the FMRP protein. Loss of FMRP is the cause of Fragile X mental retardation, and the disturbance of the interaction between KCNT1 and FMRP might possibly play a role in the frequent occurrence of cognitive impairment in KCNT1-related pathology (38).


See paragraph on FFEVF.


Temporal lobe epilepsy is the most common focal epilepsy type in adolescents and adults. Many patients have underlying congenital or acquired brain lesions, and the entity is, therefore, often considered a typical lesional epilepsy. Nevertheless, through the years several familial cases have been described, and a genetic cause has been identified for a substantial subset of them. Based on seizure semiology, two types of familial temporal lobe epilepsies can be distinguished: a lateral form with typically auditory or dysphasic features (autosomal dominant epilepsy with auditory features, ADEAF) and a mesial form with prominent psychic and vegetative symptomatology (familial mesial temporal lobe epilepsy, FMTLE). Whereas the mesial form of familial TLE encompasses a heterogeneous group of epilepsies, familial lateral temporal lobe epilepsy is clinically and genetically a much more homogeneous entity.


Clinical Manifestations

Autosomal dominant epilepsy with auditory features (ADEAF, previously called familial lateral temporal lobe epilepsy, or autosomal dominant partial epilepsy with auditory features, or ADPEAF) is a familial focal epilepsy syndrome characterized by seizures with auditory symptoms and/or receptive aphasia. The diagnosis of epilepsy might be overlooked in some family members due to the sometimes subtle symptomatology. Careful clinical history taking is therefore crucial.

Seizure onset is mostly in adolescence, but onset at any age between 4 years and 50 years has been described (39,40). Affected individuals have focal seizures with or without disturbance of consciousness. Most patients also have secondarily generalized tonic–clonic seizures, but these are rare, and often nocturnal (41). Auras most frequently consist of simple auditory symptoms as humming, buzzing, or ringing. More complex symptoms, such as distortion of sounds, volume changes, or auditory hallucinations, are also seen. A subset of patients has focal seizures consisting of ictal receptive aphasia, during which they lose the ability to understand spoken language despite conserved consciousness (40,42,43). Reflex seizures due to sudden noise, like a ringing telephone, have been described in several families (41). The characteristic seizure semiology of the syndrome is thought to reflect seizure onset in the lateral temporal lobe; hence also the name “autosomal dominant lateral temporal lobe epilepsy (ADLTLE),” which can be found throughout the literature. Although the presence of auditory or aphasic seizures is mandatory for the diagnosis of this syndrome, other seizure types with sensory (eg, visual, olfactory, cephalic) auras and motor or autonomic symptoms can co-occur (42,44,45). They are nevertheless much rarer.

Seizures are usually infrequent and well controlled with any antiepileptic drug used for focal seizures. Rare families, including family members with therapy-resistant seizures, have been described, though, and two family members of a large Norwegian family with predominantly aphasic seizures died suddenly during sleep (40,46). Affected individuals do not show other abnormalities besides seizures, and neurological exam is normal.

EEG, Neuroimaging, and Other Investigations

Interictal EEG is most often normal; however, focal epileptic discharges in the temporal lobe are present in one-third of patients (40,42,47). Ictal EEGs have been described in a few patients, and showed ictal onset in the mid-temporal, anterior temporal, or fronto-temporal region (40,46,47). Standard MRI of the brain is normal. One study showed aspecific malformations of the temporal lobe, such as volumetric enlargement, but this has not been confirmed in further studies (48,49). However, several functional studies showed evidence for subtle abnormalities of the lateral temporal lobe network and of language processing: on functional MRI with an auditory description decision task, individuals with ADEAF had significantly less activation than controls, and magnetoencephalography with auditory stimuli showed a significantly delayed peak 2 auditory evoked field latency in affected individuals (49). Furthermore, in the Norwegian family with predominant aphasic seizures, a strong asymmetry of long-latency auditory evoked potentials with reduced left N1-P2 amplitudes was shown (50), and abnormal phonologic processing was demonstrated in a Sardinian family by means of a fused dichotic listening task (51).

Molecular Genetics

Following the description of three families with linkage to chromosome 10q22, two groups independently described mutations in the gene LGI1 (leucin rich, glioma inactivated 1) as a cause of ADEAF (45,52). LGI1 is a secreted protein, and it was the first non-ion channel gene linked to epilepsy. Subsequent studies have shown LGI1 mutations to be present in 33% to 50% of familial cases, and 2% of sporadic cases with lateral temporal lobe epilepsy (41,42,53). Penetrance is reduced, and is thought to be around 67% (54). There seems to be no major phenotypic differences between families with or without mutations, although families with mutations tend to have more family members with auditory symptoms and fewer with autonomic symptoms (42). Truncating mutations, rare whole gene deletions, and missense mutations have all been described, suggesting a loss of function mechanism for LGI1-related epilepsy. In line with this mechanism, most examined missense mutations have been shown to lead to a loss of secretion of the protein, with accumulation in the endoplasmatic reticulum and Golgi apparatus before their premature degradation (5557).

Interestingly, LGI1 is also a target for autoantibodies in a form of acquired autoimmune limbic encephalitis, an adult-onset neurological disorder including cognitive impairment and epilepsy, with typically faciobrachial dystonic seizures (58).

For years, LGI1 was the only known gene linked to ADEAF. More recently, however, heterozygous missense mutations in the gene RELN (reelin), encoding another secreted protein, have been described in 7 of 40 (17.5%) Italian ADEAF families (59). Many clinically unaffected mutation carriers were present, suggesting a relatively low penetrance rate of 60%. Reelin is also secreted by the liver, and serum levels of RELN were shown to be reduced in patients with RELN mutations. Recessive loss-of-function mutations in this gene were already known to be a cause of lissencephaly with cerebellar hypoplasia (LCH), a severe disorder with cerebral migration defects, intellectual disability, and seizures (60). Epilepsy has not been reported in the heterozygous mutation carriers in the few LCH families described to date, but this can be compatible with the significantly reduced penetrance seen in ADEAF families.


ADEAF Due to LGI1 Mutations

LGI1 is a primarily neuron-expressed protein, consisting of a N terminal signal peptide defining it as a secreted protein, a leucine-rich repeat domain, and another C-terminal repeat region called the epitempin repeat (61,62). The protein derives its name from the fact that it was initially found to be downregulated in high-grade gliomas, and tumor expression studies suggest a role for LGI1 in reducing tumor invasion (63). Families with LGI1 mutations, however, do not have an excess of gliomas or other brain tumors. Little was known about the function of the protein at the time mutations in families with ADEAF were identified, but cellular and animal model studies have greatly contributed to our understanding of the function of the protein since then. First of all, copurification studies have identified the main binding partners of LGI1 as the presynaptic voltage-gated potassium channel Kv.1.1, the presynaptic transmembrane proteins ADAM11 and ADAM23, and the postsynaptic protein ADAM22 (6466). Most likely, all together they form a trans-synaptic multiprotein complex (67). The nature of LGI1’s binding partners supports a role of LGI1 in regulation of synaptic transmission. The Kv1.1 channel is predominantly localized to the axons and nerve terminals, and controls the neuronal firing rate and neurotransmission through its hyperpolarizing action. LGI1 has been shown to slow the Kvβ1 subunit-mediated inactivation of the Kv1.1 subunit (64). More extensively studied is LGI1’s interaction with ADAM22, a postsynaptic protein that itself binds to the postsynaptic scaffolding protein PSD-95. Through its interaction with this complex, LGI1 may increase the number of AMPA receptors, thus increasing glutamatergic neurotransmission (65). Besides its role in synaptic transmission, LGI1 also seems to be involved in neurodevelopment, as it was shown to stimulate axonal outgrowth and to play a role in maturation and pruning of glutamatergic neurons (6870).

Several rodent models have further confirmed the role of LGI1 haploinsufficiency in the pathogenesis of epilepsy. Homozygous knock-out animals have early-onset frequent spontaneous seizures and die after 2 to 3 weeks. Heterozygous knock-out animals, which more closely recapitulate the human genotype, have a normal life span and do not have spontaneous seizures, but show an increased susceptibility to sound-induced and PTZ-induced seizures (65,7173). A conditional knock-out mouse model underlined the key role of glutamatergic neurotransmission in LGI1-related epilepsy, by showing that deletion of LGI1 in cortical glutamatergic neurons alone is sufficient to reproduce the epileptic phenotype, while deletion in GABAergic interneurons did not change the seizure threshold (74). In summary, while it is clear that LGI1 loss of function leads to seizures, the precise molecular processes through which this occurs remain unclear. More specifically, its differential role during different stages of life and its predilection for lateral temporal lobe pathology remain topics for further study.

ADEAF Due to RELN Mutations

Reelin is a very large (3460 amino acids) secreted glycoprotein, mainly expressed in the brain and liver. The protein gets its name from the identification of homozygous deletions of its encoding gene in the reeler mouse, which has tremor, ataxia, and neuronal migration defects (75). Prenatally, reelin is secreted by the Cajal Retzius cells in the marginal zone of the developmental cortex and by granule cells in the cerebellum. In the postnatal brain it is mainly expressed in GABAergic interneurons (76). Reelin is known to have an important role in the correct cortical lamination during embryogenesis and in postnatal modulation of synaptic plasticity (77,78). It is thought to act by binding to apolipoprotein E receptor 2 (ApoE2) and very low-density lipoprotein receptor (VLDLR), which, in turn, leads to phosphorylation of the intracellular adaptor protein disabled homolog 1 (DAB1) (79). Interestingly, after the recent discovery of mutations in RELN in families with ADEAF, colocalization of reelin and LGI1 was shown in neurons of the hippocampus and temporal lobe cortex of immature and mature rats (59). These findings suggest a functional interplay between both proteins, and the understanding of this common pathway might help to elucidate the molecular biology underlying ADEAF.


Buy Membership for Pediatrics Category to continue reading. Learn more here