Generalized Epileptic Disorders: An Update

Generalized epileptic disorders are characterized by seizures in which there are symptoms of bilateral cerebral involvement, and an electroencephalogram in which synchronous abnormal activity is recorded in both hemispheres at seizure onset (1). Generalized epilepsies result from alterations in the function of the whole brain or of major brain systems, and are often the expression of inborn errors in which the molecular etiology can be identified through genetic approaches. Thus, insights gained from genetic studies of the generalized epilepsies are beginning to answer the age-old question: "What is the cause of epilepsy?" Indeed, concepts of pathogenesis developed in generalized epilepsies may provide clues to the etiologies of the more common focal epileptic disorders. Having passed what is truly a milestone event in epilepsy research—the cloning of the first human idiopathic generalized epilepsy genes (see 2,3)—it seemed an opportune time to review the scope of research progress on the generalized epilepsies. We begin with a brief summary of the clinical features of the generalized epilepsies and then describe new developments in the genetics of these disorders. Current understanding of the cellular and network mechanisms leading to some forms of generalized seizures also are considered. Finally, drug treatment is discussed in the context of the pathophysiologic mechanisms. This review originated at Focus on Epilepsy V, an international symposium held in Quebec City in May 1999. A list of the faculty that participated in this conference is given later (4,5). Generalized epilepsy syndromes are often classified as idiopathic or symptomatic. Idiopathic epilepsies have been described as disorders "not preceded or occasioned by another" (i.e., with no underlying cause other than a possible hereditary predisposition). Idiopathic epilepsies are categorized according to the age at onset, clinical and EEG characteristics, and genetic etiology (4). According to the International League Against Epilepsy (ILAE) definition, idiopathic generalized epilepsies include the following forms: (a) benign neonatal (familial) convulsions, (b) benign myoclonic epilepsy in infancy, (c) childhood (pyknolepsy) and juvenile absence epilepsies, (d) juvenile myoclonic epilepsy, (e) epilepsy with grand mal seizures on awakening, and (f) seizures precipitated by specific modes of activation. However, even this rather extensive list does not encompass all the situations that can be encountered in clinical practice. For instance, some idiopathic epilepsies with generalized tonic–clonic convulsions, such as sleep grand mal or grand mal at random, do not yet have well-defined syndromic definitions. Typical absence seizures characterized by a short lapse of consciousness with no (or minimal) motor and autonomic manifestations are the most innocuous type of generalized seizure. At the other extreme are generalized tonic–clonic seizures with dramatic motor manifestations and impairments in autonomic function that may be life threatening. However, idiopathic generalized tonic–clonic seizures are highly sensitive to antiepileptic drugs (AEDs), and thus the clinical outcome is usually favorable. Conversely, seizures characterized by less dramatic motor manifestations, such as atonic or akinetic seizures, are often resistant to drug therapy and can severely impair the patient's quality of life. In symptomatic generalized epilepsies, seizures occur in conjunction with a spectrum of other neurologic and systemic symptoms, and are usually associated with structural brain lesions or alterations in brain metabolism. In those cases in which the genetic defects have been identified, the mutations perturb brain development, energy production, or other important aspects of brain function. However, the specific ways in which these defects result in seizures has not yet been elucidated. In contrast to the idiopathic epilepsies, symptomatic epilepsies are usually less sensitive to AEDs and are often associated with severe neurologic or psychological impairment. Some forms (e.g., West and Lennox–Gastaut syndromes, epilepsy with myoclonic–astatic seizures or myoclonic absences) can be either cryptogenic or symptomatic of different inborn disorders, whereas early myoclonic encephalopathy and infantile epileptic encephalopathy with suppression burst are always symptomatic of metabolic or malformative encephalopathies. Although the true generalized nature of the discharges accompanying many of the symptomatic generalized epilepsies is debated, the syndromes are nevertheless currently classified among the generalized epilepsies. The ILAE classification has a number of drawbacks. First, its clinical value is uneven: some diagnoses, such as juvenile myoclonic epilepsy, identify a definite type of epilepsy, whereas others, such as myoclonic–astatic epilepsy, correspond to a composite spectrum of clinical forms whose boundaries are not universally agreed on. Second, the definitions of idiopathic and cryptogenic forms with their different implications, but ill-defined boundaries, are not fully satisfactory. Third, it is becoming apparent that certain idiopathic generalized epilepsy syndromes, which appear clinically to be pure epilepsy phenotypes, are in fact genetically heterogeneous. Despite these reservations, the 1989 ILAE classification provides the clinician with a useful diagnostic framework and a guide to natural history, symptoms, and prognosis; in addition, it represents a starting point for genetic studies. The past decade has witnessed the genetic identification of several symptomatic generalized epilepsies, and recently, of a few rare idiopathic epilepsies. Examples of symptomatic generalized epilepsy syndromes with an established genetic defect include (a) myoclonic epilepsy and ragged red fiber disease (MERRF), in which alterations in mitochondrial oxidative phosphorylation are caused by mutations in the mitochondrial gene that codes for tRNA(Lys) (5); (b) Miller–Dieker lissencephaly, in which cortical maldevelopment is associated with a defect in the LIS-1 gene (6) encoding platelet-activating factor acetylhydrolase (7); and (c) progressive myoclonus epilepsy of the Unverricht–Lundborg type (EPM1), in which there is a mutation in the gene for cystatin B, a protein inhibitor of cysteine proteases (8–10). Recently the role of the cystatin B gene defect in the production of myoclonic seizures was confirmed by studies in which the cystatin B gene was disrupted in mice with a gene-targeting strategy (11). Cystatin B–deficient mice exhibited myoclonic seizures (but not tonic–clonic seizures as in human EPM1) and ataxia associated with apoptotic degeneration of cerebellar granule cells. However, cerebellar degeneration is unlikely to account for the myoclonic seizures because other mouse mutants with cerebellar degeneration lack such seizures. The genetic defects in several other progressive myoclonus epilepsies including Lafora disease (EPM2) and certain forms of neuronal ceroid lipofuscinoses (Batten disease) also have been identified. These myoclonus epilepsies are associated with intracellular inclusion bodies in a wide variety of cell types. In EPM2, there is a defect in a gene encoding the protein laforin, which has a consensus amino acid sequence indicative of a protein tyrosine phosphatase (12,13). Mutational inactivation of such an enzyme is presumed to affect the control of glycogen metabolism. At least eight genes underlie the neuronal ceroid lipofuscinoses (CLN1-8), of which four have been characterized (CLN1, CLN2, CLN3, CLN5). It is now generally believed that the neuronal ceroid lipofuscinoses represent a new class of lysosomal storage disorders (14–16). An additional symptomatic generalized epilepsy syndrome, the Angelman syndrome, results from a complex defect affecting imprinted genes on maternal chromosome 15q11-q13. Angelman syndrome can be produced by gene deletions, paternal uniparental disomy, or mutations disrupting the imprinting mechanism (17). Abrogation of the UBE3A gene, which codes for E6-AP ubiquitin protein ligase (also known as ubiquitin ligase 3A), an enzyme involved in protein degradation and processing, is believed to be the cause of Angelman syndrome. However, recent evidence suggests that defects in associated γ-aminobutyric acid subtype A (GABAA)-receptor genes may play an important modifier role (18). Although the most common form of Angelman syndrome (type 1A) is produced by an ∼4-Mb deletion in maternal 15q11-13, a small number (4–6%) of Angelman syndrome patients (type IV) have a point mutation affecting the UBE3A gene. In extended families with such mutations, heterozygotes are normal if the mutant chromosome is inherited from the father, but are affected if it is inherited from the mother. This occurs because the UBE3A gene is maternally imprinted in a brain-specific fashion; that is, in certain neuronal types including Purkinje cells, hippocampal neurons, and olfactory mitral cells, only the copy of the gene inherited from the mother is expressed. Such imprinting does not occur in most other neuronal types or in somatic tissues. Type IV Angelman syndrome patients exhibit many of the neurobehavioral features associated with the type 1A form of the disorder, but they may be less severe. In particular, individuals with UBE3A mutations may have fewer apparent EEG abnormalities and few if any seizures (19). Interestingly, there is a cluster of GABAA-receptor genes within the region deleted in type IA Angelman syndrome. When one of these genes GABRB3 (encoding the GABAA receptor β3 subunit) is deleted in the mouse, the animals have seizures and other features of the Angelman syndrome (20). Deficiency of GABRB3 may function as a modifier of the epilepsy phenotype, accounting for the greater severity in the deletion cases. Thus, type IA Angelman syndrome may represent the first form of human epilepsy in which a genetic defect affecting brain GABAergic function plays a clearly identified pathogenetic role. Major progress has recently been made in understanding the pathogenesis of certain idiopathic generalized epilepsies. In two such disorders—benign familial neonatal convulsion (BFNC) and generalized epilepsy with febrile seizures plus (GEFS+)—defects have been identified in genes coding for subunits of voltage-gated ion channels. Thus, at least some idiopathic generalized epilepsies appear to be members of a growing class of channelopathies that include a broad range of paroxysmal disorders, including certain cardiac arrhythmias, episodic ataxias, periodic paralyses, congenital myotonias, and migraine (2,21). It has long been recognized that genetic factors play a role in the idiopathic generalized epilepsies (22). However, for the most common forms, including childhood absence epilepsy (incidence, one in 1,000), juvenile myoclonic epilepsy (incidence, one in 2,000), and juvenile absence epilepsy (incidence, one in 3,000), linkage studies have so far failed to reveal replicable susceptibility loci. Identification of the genetic basis of these disorders has been challenging because inheritance occurs in a complex (nonmendelian) fashion, and the syndromes may be genetically heterogeneous. It also remains to be determined whether major gene effects or multiple minor susceptibility alleles contribute to the genetic variation (23). In the case of the far less common BFNCs (incidence, one in 100,000), inheritance usually occurs in an autosomal dominant fashion with high (85%), but not complete penetrance (24). This has allowed the genetic basis of this idiopathic generalized epilepsy syndrome to be identified with a positional cloning strategy. BFNC occurs in two genetic forms: (a) EBN1, whose gene has been localized to chromosome 20q13.3; and (b) EBN2, a rarer form to date conclusively identified in only one pedigree, in which the gene is on chromosome 8q24 (25,26). The genes affected in EBN1 and EBN2 have recently been identified and shown to encode the two novel homologous voltage-gated K+ channels KCNQ2 and KCNQ3. Evidence indicates that KCNQ2 and KCNQ3 coassemble to form heteromeric K+ channels. BFNC mutations in either KCNQ2 or KCNQ3 result in a small (∼20–30%) reduction in the current amplitude, which is believed to be sufficient to cause the disorder. The dominant inheritance seems to result from haploinsufficiency rather than from a dominant negative effect produced by the mutant channel subunits (as is the case in the long QT syndrome) (27). A characteristic feature of BFNC is that seizures are expressed mainly during the first several weeks or months of life. In the majority of affected individuals, long-term development is normal, and only ∼16% of patients have epileptic seizures later in life (28,29). It may be postulated that the expression of KCNQ2/KCNQ3 channels is at a critical level in infancy but that there is redundancy in these or related K+ channels later in life, accounting for the disappearance of symptoms in most affected individuals. Alternatively, changes in other brain excitability mechanisms could be responsible for the age-dependent resolution of the seizure phenotype. Evidence has been presented that KCNQ2/KCNQ3 heteromers represent a member of the M-current family of K+ channels (30,31). The M currents are low-threshold, slowly activating K+ currents that exert inhibitory control over neuronal excitability (32). Opening of M-current channels is suppressed by neurotransmitters acting on G protein–linked receptors, including muscarinic cholinergic receptors. Inhibition of M currents result in membrane depolarization and an increase in neuronal input resistance, making the cell more likely to fire actions potentials. Thus, it is not surprising that a defect in M current in BFNC would be associated with enhanced seizure susceptibility. Recently voltage-activated Na+ channels have been implicated in seizure susceptibility in the complex syndrome referred to as "generalized epilepsy with febrile seizures plus" (GEFS+), a phenotypically diverse disorder in which there are febrile seizures in childhood and afebrile seizures later in life. (The syndrome is distinct from common benign febrile seizures.) Two genetically distinct forms of GEFS+ have been described, the first associated with mutations in the auxiliary β1 Na+-channel subunit SCN1B (GEFS+1) (33), and the second with the principal α Na+ channel subunit SCN1A (GEFS+2) (34). Both of these disorders are inherited in an autosomal dominant fashion. Functional studies of expressed Na+ channels composed of mutant subunits are beginning to provide insight into how the genetic defects in these generalized epilepsy syndromes lead to seizure susceptibility. The β1 protein implicated in GEFS+1 regulates the kinetic properties of the main Na+ channel α subunit, causing speeding of inactivation and acceleration in recovery from inactivation. In GEFS+1, there is a C→G transversion within the coding region of the SCN1B gene. The mutation changes a conserved cysteine residue to tryptophan, thus disrupting a putative disulfide bridge, which maintains an extracellular immunoglobulin-like fold of the β1 subunit. The immunoglobulin motif normally allows the β1 to associate with the α subunit, and this interaction is presumably perturbed by the mutation. The mutant β1 subunit no longer speeds the Na+-channel kinetics but instead slightly prolongs inactivation and slows recovery from inactivation. Precisely how these alterations in gating result in febrile seizures is not well understood. One idea is that the mutant β1 subunit could result in greater persistent Na+ current, which would lead to enhanced neuronal bursting, as has been shown in neocortical pyramidal neurons (35). If this effect becomes more significant at higher temperatures, it could explain the temperature dependence of the seizures. To date, mutations in two highly conserved residues in the S4 voltage-sensing region of SCN1A have been associated with GEFS+2. Based on analogies with the effects of induced mutations in a related Na+ channel, it has been suggested that the mutations in GEFS+2 could also modify Na+-channel inactivation, leading to enhanced bursting. An important question raised by the identification of defects in K+ and Na+ channels in rare monogenic idiopathic generalized epilepsies is whether genetic variation in the same genes represents a contributory factor for the more common idiopathic generalized epilepsy syndromes. A recent analysis of the gene structure of the KCNQ2 gene revealed that the gene consists of at least 18 exons that can occur in alternatively spliced forms (29). In addition, a single nucleotide polymorphism resulting in an amino acid substitution T752N was identified. The polymorphism is not a functionally important mutation that contributes to the BFNC phenotype. Nevertheless, such allelic variation could theoretically confer susceptibility to epilepsy, either by itself or in combination with other susceptibility loci. However, the results to date do not implicate this particular polymorphism as a source of susceptibility for idiopathic generalized epilepsy (36). Apart from KCNQ2 and KCNQ3, an enormous diversity of K+ channels contributes to the regulation of brain excitability (37). Pharmacologic inhibition of K+ channels is a potent convulsant stimulus (38). In addition, mice with induced deletions of various K+-channel genes exhibit seizures (39,40). K+-channel genes therefore represent significant candidates in the search for epilepsy susceptibility genes. Indeed, defects in the KCNA1 (Kv1.1) voltage-dependent K+-channel gene in episodic ataxia type 1 are associated with a 10-fold increase in the frequency of epilepsy (41). Similarly, the identification of Na+-channel defects in GEFS+ confirms the importance of the Na+-channel genes as candidates. In retrospect, it seems logical that defects in K+ and Na+ channels, the key mediators of excitability in the nervous system, would be critical to the pathogenesis of some forms of generalized epilepsy. Future genetic studies should logically focus on the role of K+ and Na+ channels as well as other ion channel types in susceptibility to idiopathic generalized epilepsies. The 3-Hz spike-and-wave discharge associated with a generalized absence attack represents the most stereotyped seizure pattern in clinical electroencephalography. These seizures, first described by Gibbs et al. (42), occur in individuals with a normal interictal EEG and originate bilaterally, do not evolve in time, and are not associated with postictal depression (Fig. 1A). Generalized spike-and-wave–like discharges were later reproduced experimentally by Jasper and Droogleever-Fortuyn (43) during low-frequency stimulation of midline and intralaminar thalamic nuclei in the cat, suggesting that they may have a subcortical origin ("centrencephalic hypothesis"). However, some years later, this view was challenged by the discovery that bilateral application of convulsants to the frontal cerebral cortex of cats and monkeys could produce a similar pattern of generalized, 3-Hz spike-and-wave activity (44,45). At approximately the same time, studies of the photosensitive Senegalese baboon Papio papio, which exhibits generalized seizures, demonstrated a cortical origin of the electrographic activity recorded in this genetic model of primary generalized epilepsy (46,47). This evidence, along with an extensive review of clinical data, led Gloor (48,49) to propose the term "generalized corticoreticular epilepsy" to reconcile the centrencephalic and cortical mechanisms. This concept assigned essential roles to both thalamus and cortex in the elaboration of generalized 3-Hz spike-and-wave discharges. Electrographic characteristics of generalized spike-and-wave discharges recorded in a patient (A), in a cat injected with i.m. penicillin injection (B), and in a WAG/Rij epileptic rat (C). Note the similarities that include generalization at onset and throughout the discharge, a nonprogressive feature of the electrographic pattern, and the absence of postictal depression. L, left; R, right; F, frontal; C, central; P, posterior; O, occipital; Cx, cortex (A, from ref. 117; B,118; C,56, with permission). The validity of the corticoreticular hypothesis has been tested over the years in several models of absence attacks with generalized spike-and-wave discharges (50–56). These studies have demonstrated a similarity between the synchronized activity occurring in thalamocortical during sleep and the spike-and-wave discharges of absence epilepsy in the In addition, it has been shown that which is for the generalized nature of the discharge over broad of the two is by and It should be that synchronous spike-and-wave discharges in animals exhibit that are higher than the in Nevertheless, as in the spike-and-wave discharges in models exhibit generalization at seizure onset and throughout the discharge, a nonprogressive electrographic pattern, and absence of postictal depression (Fig. the of spike-and-wave discharges in these animals consists of by minor motor manifestations as and as is the case in with absence of the early evidence thalamocortical to spike-and-wave discharges was in studies in cats injected with penicillin In this in model of absence seizures, depression of thalamic activity by of both and spike-and-wave discharges (Fig. In addition, the thalamus is of with the (e.g., spike-and-wave discharges are not (Fig. Finally, spike-and-wave discharges in the cat brain are in both thalamic nuclei and by cortical excitability is by depression Thus, the and of thalamus and cortex is for generalized spike-and-wave discharges in the penicillin have been from studies with various rat that exhibit seizures of the thalamic activity induced by a of is by disappearance of generalized spike-and-wave discharges induced by i.m. injection of activity is most at the of the cortical and thalamic spike-and-wave and with of normal thalamic EEG from the thalamus of a cat, and i.m. injection of Note that spike-and-wave discharges are not in the posterior; R, right; L, left; P, (A, from ref. with permission). extracellular and intracellular have insight into the cellular mechanisms underlying the spike-and-wave discharge in models of absence seizures. the of the spike-and-wave there is a increase in in and whereas during the the is to (Fig. and A similar pattern of burst activity has been in of the thalamus and during spike-and-wave discharges in genetic models (Fig. from neocortical neurons have also demonstrated that the of the spike-and-wave complex in the penicillin model represents an induced by thalamocortical (Fig. In the is associated with to A similar sequence of intracellular has been recorded in cortical and thalamic neurons in genetic absence epilepsy of during spike-and-wave discharges (Fig. It is, whether the of the that with the of the spike-and-wave complex occurs through inhibitory mechanisms. for a cortical and a thalamocortical during spike-and-wave discharges recorded in a cat injected with i.m. of the EEG and EEG and of single activity by the negative of the cortical EEG Note that both cortical and thalamic with between increase in during the and a reduction of this of to during the of the spike-and-wave discharges recorded in a genetic absence epilepsy of rat from in the thalamus and in the cortex and from the the of and activity at an time (A, from ref. B, EEG and intracellular recorded during a spike-and-wave discharge in a cat injected i.m. with Note that the is associated with a depolarization that whereas the of the spike-and-wave complex is associated with a of the are in the intracellular at membrane from a cortical during the of a spike-and-wave discharge recorded in the EEG from a genetic absence epilepsy of (A, from ref. with permission). The most important source only one in of inhibition on thalamocortical is the thalamic which is composed of GABAergic thalamic neurons are with a current, which can The activity produced by such can sleep In the model of absence seizures, lesions of thalamic spike-and-wave discharges neurons from may have enhanced current The along the cell and inhibitory in the membrane of the neurons the a current is that of In addition, a may with the current in the of Thus, spike-and-wave discharges are by of in thalamic principal neurons, which cortical cells, which in neurons within the is through mechanisms that may burst in thalamocortical to characteristic of the spike-and-wave discharges during an absence seizure. This view is by evidence in brain from mice in which the β3 subunit of the GABAA receptor was deleted In the the β3 subunit is mainly to the a in the β3 inhibition is in these neurons (but not in thalamocortical This results in a dramatic increase of the of thalamocortical neurons to the inhibition of cells. evidence in of a role of mechanisms in the of spike-and-wave discharges from with a drug that and EEG alterations similar to those during absence attacks receptors, causing membrane which to of channels The result is, activity and burst in thalamocortical The of thalamic and cortical during generalized spike-and-wave also has been in In these the burst of thalamic neurons as well as the properties of have been shown to be critical in spike-and-wave discharges similar have been to thalamocortical spike-and-wave discharges with different (i.e., in in These have shown that the

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.