Clinical manifestations

Presentation and course

Early in the description of the disease it became evident that in the same pedigree of a GEFS+ family, patients express a highly variable phenotype that is generally mild (128). Although the occurrence of GTCS was originally considered a defining characteristic of GEFS+, these are not the only type of seizures observed to date. Indeed, FS+ and afebrile generalized seizures are sometimes associated with absences or atonic, myoclonic, myoclonic-astatic, or focal seizures (164). Xu and colleagues demonstrated this highly heterogeneous spectrum in their report of 39 families with 196 affected members (162). This variable presentation was subsequently further supported by the study of 303 families in which 22 GEFS+ families were described (49), including phenotypes not generally recognized at the time. In addition, genetic and clinical data suggest that at least some cases of Dravet syndrome (severe myoclonic epilepsy of infancy; SMEI) are closely related to GEFS+.

Neurologic examination is usually normal. However, individual patients may present neurologic symptoms as part of a given epilepsy syndrome (ie, ataxic gait or action myoclonus in Dravet syndrome). The presence of seizures in at least 1 other family member is an important indication that a case is associated with the GEFS+ spectrum; however, de novo pathogenic variants in SCN1A in mild phenotypes within the GEFS+ spectrum have proven that mild GEFS+ is not always inherited (107). Environment, genetic background, and epigenetic modifications may also have an important role in the disease presentation.

Although members of a pedigree may present with well-characterized epilepsy syndromes (myoclonic-astatic epilepsy, Dravet syndrome, cryptogenic temporal epilepsy), others may only have occasional seizures, particularly febrile convulsions. The recognition of GEFS+ is thus frequently difficult in clinical practice. An outline of epileptic syndromes and seizure types that constitute the spectrum of GEFS+ are designated bellow. At least 2 members of a family with phenotypes that fall in the GEFS+ spectrum are required for the diagnosis of GEFS+ (134).

Febrile seizures. Febrile seizures are generalized tonic-clonic seizures occurring during fever higher than 38^o C in children ranging from 3 to 6 months to 6-years-old (128; 108). Febrile seizures are the prominent seizure type of GEFS+, with their incidence ranging between 33% to 44% according to different reviews (25; 164).

Febrile seizures usually last a few minutes and resolve spontaneously. They are considered complex when duration is more than 15 minutes, there is more than 1 episode in 24 hours, and there is the presence of focal seizures. Positive family history for febrile seizures and genetic predisposition are very common in this fever-associated disorder that has an incidence of 3% to 4% in the pediatric population. Outcome is almost always favorable (25).

Febrile seizures plus (FS+). The term febrile seizures plus is defined by the presence of febrile seizures, usually generalized, that occur outside the age limits of between 3 months and 6 years, and/or afebrile generalized tonic-clonic seizures (133; 47). These afebrile generalized tonic-clonic seizures may occur during the classic febrile seizure age range, after febrile seizures remit, or, in rare instances, in late childhood or early adulthood (144; 143; 130).

Febrile seizures plus correspond to a frequent phenotype of GEFS+ syndrome with about 20% of patients fulfilling FS+ criteria; of them, 9% initially presented with simple febrile seizures (164). The prominent feature in this group seems to be the presence of rare generalized tonic-clonic seizures. Oller-Daurella and Oller noted that isolated “grand mal” in childhood frequently started with febrile convulsions and was usually limited to a few seizures (114). Lennox-Buchtal reported that generalized tonic-clonic seizures following febrile convulsions were often nocturnal, infrequent, relatively benign in nature, and with spontaneous remission (82). Nelson and Ellenberg found that 18 (1%) of their patients with afebrile seizures after febrile convulsions had only 1 or a few seizures, with remission before 48 months of age (110; 111).

Extremely prolonged febrile seizures lasting more than 30 minutes could be classified as febrile seizures plus. In fact, the FEBSTAT cohort study of children with febrile status epilepticus revealed a higher incidence of febrile seizures and epilepsy in families of patients with febrile status epilepticus when compared with patients with simple febrile seizures (46.5% vs. 28.3%) (65). A susceptibility for febrile status epilepticus in GEFS+ patients could thus be implied.

Febrile seizures plus are considered a mild subgroup of GEFS+; patients experience a spontaneous remission and a normal neurodevelopment (108).

Febrile seizures (plus) and other seizure types. The first GEFS+ family described by Scheffer and Berkovic in 1997 included few probands that, apart from febrile seizures, suffered from absences, atonic seizures, and 1 proband with myoclonic-astatic epilepsy. Absence seizures in GEFS+ have been reported by other authors as well but their patterns are different from the typical childhood absence epilepsy as seizures are infrequent instead of multiple per day (164).

Myoclonic-astatic epilepsy (or Doose syndrome) is an age-specific epileptic syndrome characterized by generalized myoclonic-atonic seizures that often present with drop attacks in preschool-aged children. Other types of generalized seizures (GTCS, absences, atypical absences) may develop at the course of the disease. A family history of febrile seizures is found positive in more than one fourth of patients (77). A synchronous to seizure onset neurodevelopmental regression results in various outcomes ranging from normal to impaired intelligence (05).

Focal seizures. Focal seizures were responsible for approximately 9% of total seizure manifestation in 409 GEFS+ affected individuals (164). Temporal lobe epilepsy may occur following febrile seizures or FS+, either alone or associated with hippocampal sclerosis. The FEBSTAT prospective study demonstrated that children with febrile status epilepticus have increased risk for developing acute hippocampal injury (141). Frontal lobe epilepsy has been described in GEFS+ families presenting with hemiclonic seizures and orofacial motor seizures (12; 11; 130).

Occipital lobe epilepsy and Panayiotopoulos syndrome have been described in GEFS+ families as phenotypes not generally recognized within the GEFS+ spectrum (49). Four individuals with Panayiotopoulos syndrome in a SCN1A mutation positive GEFS+ family were also reported by Kivity and colleagues (78). The above data led to include epilepsy with centrotemporal spikes in GEFS+ focal epilepsies, but no more cases have been reported to our knowledge in the literature in order to confirm more than a chance association (49).

The spectrum of GEFS+ has been enriched by epilepsy with auditory features. Bisulli and colleagues described a mother and a daughter with prominent auditory octal manifestations that belonged to a family where 8 members had a GEFS+ typical phenotype (16). A missense pathogenic mutation in the SCN1A gene was detected. In another GEFS+ four-generation family with SCN1A mutation and 9 affected subjects, the electroclinical features indicated a temporal-parietal-occipital carrefour epilepsy (125).

Dravet syndrome. Dravet syndrome is associated with the most severe form of GEFS+ phenotype. This distinctive genetic syndrome is associated with different types of drug-resistant seizures that appear during the first year of life. Dravet syndrome patients demonstrate a susceptibility to febrile seizures and status epilepticus; cognitive and motor decline appear soon after seizure onset in a previously normal developing child. Dravet syndrome represents the most typical single-gene (SCN1A) associated epilepsy, with an incidence estimated at 1 in 15,500 (149).

Veggiotti and colleagues first reported the cases of 2 brothers diagnosed with Dravet syndrome who belonged to a family of GEFS+ and suggested that Dravet syndrome might be the most severe end of the GEFS+ spectrum (153). Singh and colleagues found a high incidence of a family history of febrile seizures in the families of 12 patients with Dravet syndrome and reached a similar conclusion (144). To illustrate the phenotypic variability, Grant and Vasquez reported on a 10-year-old boy, with normal early development, who experienced febrile seizures at 2 years of age and later experienced multiple types of seizures (afebrile myoclonic seizures, spasms, and absences) (56). A family history of simple febrile seizures was noted. EEG and MRI were all normal at onset, and video-EEG performed at the age of 4 years and 8 months demonstrated interictal and ictal abnormalities. The child also presented behavioral difficulties and, during the second decade, presented with difficult-to-control afebrile generalized seizures. The patient was considered to have a moderate phenotype within the GEFS+ spectrum relative to myoclonic-astatic epilepsy or benign myoclonic epilepsy of infancy.

Although the SCN1A gene has been correlated with less severe GEFS+ phenotypes, data suggest that patients with SCN1A de novo or inherited heterozygous mutations present in around 80% with severe Dravet syndrome clinical spectrum (149).

A prediction model for early diagnosis of SCN1A-related epilepsies has been proposed by Brunklaus and colleagues in 2022 (22). This model will tend to accurately discriminate patients that will develop Dravet syndrome versus other GEFS+ phenotypes. This prediction model could allow an objective estimation at disease onset of which a child will develop Dravet syndrome versus GEFS+, assisting clinicians with prognostic counseling and helping to take decisions on early use of precision therapies (22).

Prognosis and complications

Although the overall phenotype of GEFS+ families tends to be mild (several members will only have febrile seizures, FS+, or rare GTCS that will remit until early adolescence), prognosis can only be discussed for individuals. Therefore, the syndromic diagnosis for each affected member must be considered separately.

Regarding the overall spectrum of clinical features associated with GEFS+, the most severe end of the spectrum is represented by Dravet syndrome with its poor prognosis and the least severe end by febrile non-recurrent convulsions. Patients with Dravet syndrome appear to be susceptible to frequent seizures or status epilepticus (24).

Genetic epilepsy with febrile seizures plus may be complicated by sudden unexpected death in epilepsy (SUDEP). Myers and colleagues described 2 GEFS+ families in which at least 1 individual had suffered a SUDEP, according to the Epilepsy Pharmacogenomics Research Database. The first individual was a 22-month-old girl who was carrying a pathogenic SCN1B variant and presented with a FS+ phenotype. In the second family, the elder of 2 brothers with atypical multifocal SCN1A-associated Dravet syndrome died suddenly at the age of 7 years after status epilepticus (109). Status epilepticus and SUDEP have a high incidence of Dravet syndrome; more data are needed to confirm their relationship with milder GEFS+1 phenotype, given the association of the SCN1B mutation with Brugada and long QT syndrome.

Data regarding outcomes and comorbidities in SCN1A-related GEFS+ have become available (34). In a Dutch cohort of 164 patients, walking disabilities and severe behavioral problems affect 71% and 43% of patients with Dravet syndrome respectively, whereas only mild learning problems and psychological/behavioral problems are reported for 27% and 38% respectively non-Dravet syndrome patients. Eighty-five percent/majority of patients with non-Dravet syndrome become seizure-free at 10 years follow-up.

Clinical vignette

A 7-year-old male patient was referred for consultation after a generalized tonic-clonic seizure of short duration. The boy had experienced febrile seizures twice during his second year of life on the occasion of viral upper respiratory infections; no investigation was proposed. He had an otherwise unremarkable perinatal and medical history. Neurodevelopment was appropriate for age and neurologic examination was strictly normal at referral. A few weeks after the appointment, he had another episode of febrile seizures. Standard scalp EEG showed generalized spike-wave activity. A brain MRI was normal. He had an older sister with a history of 2 febrile convulsions by the age of 3 years, followed by epileptic episodes of head-nodding and rare falls (orienting to a syndromic diagnosis of myoclonic-atonic epilepsy). Both brother and sister responded relatively well to valproate, but the sister had to remain under treatment until the age of 11 years. Seizure freedom and normal psychomotor development at the age of 12 years was achieved for both siblings.

Extensive history taking revealed that the father had also experienced febrile seizures until the age of 8 years and that 1 of his uncles was considered to be epileptic at a young age, but detailed data were not available.

Different individuals of this family have presented with heterogenous clinical manifestations (FS+ in the father, GTCS, and other seizure types in the others), orientating the diagnosis to GEFS+ and the performance of genetic testing.

Biological basis

Etiology and pathogenesis

GEFS+ is a familiar disorder with great genetic and phenotypic variability. The various types of seizures or epilepsy syndromes are transmitted by autosomal dominant inheritance with incomplete penetrance, but recessive inheritance, de novo, and mosaic cases have also been identified (21; 107). However, the genetic cause is until today identified in a minority of families. In a study of 60 GEFS+ families with 409 affected individuals, only one third of the tested families (50/163) had a pathogenic variant in a GEFS+ known gene (164). Complex inheritance seems to be more common than monogenic inheritance (164). The majority of genes identified so far include sodium channels and GABA-A receptors, suggesting that this “genetic syndrome” belongs to the “channelopathies.” A subsequent effort to correlate single gene mutations with mechanisms of epileptogenesis suggests that mutations found in Dravet syndrome and GEFS+ syndromes cause phasic-GABApathy rather than NMDA-pathy or tonic GABA-pathy (53). However, the fact that other GEFS+ loci (8p23-p21, GEFS+6) do not appear to contain ion channel genes or neurotransmitter receptor genes led Baulac and colleagues to even conclude that “identification of the responsible gene might bring to light a mechanism involved in epileptogenesis” (10). The possible pathophysiological role of identified genes associated with GEFS+ is discussed in detail in the following section.

SCN1B gene (GEFS+ type 1). A locus for the GEFS+ spectrum was mapped to chromosome 19q13 (158), where the voltage-gated sodium channel beta1 subunit gene (SCN1B) was subsequently identified. A missense mutation of this gene (C121W) was found in 2 unrelated families (158; 157). In vitro functional studies (voltage-clamp recording in Xenopus laevis oocytes) showed changes consistent with loss of function. In particular, the mutant protein failed to accelerate recovery from inactivation of the associated alpha2 subunit (158); however, these findings were not confirmed in a mammalian cell system (97). Co-expression of mutant and wild-type beta subunits with the alpha subunits, caused an intermediate inactivation rate (102) arising from binding competition between the inactive mutant subunit and the alpha subunit; thus, accounting for dominant inheritance. In heterozygotes, association between inactive beta subunits and alpha subunits would generate a persistent sodium current rendering neurons hyperexcitable and apt to initiate firing under small depolarizations (98). Whole-cell voltage clamp experiments were performed at 22°C and 34°C using Chinese hamster ovary cells, and temperature-dependent changes were found to be consistent with increased neuronal excitability of GEFS+ patients harboring the C121W mutation (48). Further in vivo rodent models led to the hypothesis that SCN1B-C121W confers a deleterious gain-of-function mutation in GEFS+ patients (79).

In a Belgian family, a second mutation was identified in the SCN1B gene (p.170_E74del) and was identified as a deletion of 5 amino acids in the extracellular immunoglobulin-like fold of SCN1B (06). The functional consequences of the p.170_E74del mutation remain to be explored. Two novel mutations (R85C and R85H) in the extracellular immunoglobulin-like domain of the sodium channel beta1 subunit have also been identified in individuals from 2 families with GEFS+ (131). They severely reduce beta subunit-mediated modulation of sodium channel function, and this could increase neuronal excitability and underlie GEFS+ pathogenesis (161). Brackenbury and colleagues suggested that SCN1B is critical for neuronal proliferation, migration, and pathfinding during the critical postnatal period of brain development (19). They described an abnormal neuronal patterning, which occurred during early postnatal brain development in SCN1B-null mice and preceded hyperexcitability. The β1, β2, and β4 subunits of voltage-gated sodium channels reportedly function as cell adhesion molecules. A crystallographic analysis established that the β1 gene mutations associated with GEFS+ impaired the β1-mediated cell-cell adhesion, and this should underlie the GEFS+ pathogenesis (140).

The pathophysiological mechanism of the heterozygous SCN1B mutation p.Asp25Asn (D25N; c.73G> A) of the β1 subunit reported in a patient with GEFS+ was investigated by Baroni and colleagues (09). Studies in human embryonic kidney 293 (HEK) cells imply that D25N mutation of the β1 subunit causes a maturation (glycosylation) defect of the protein and disables its interaction with the α subunit. The authors concluded with the possibility of an emerging mutation-specific GEFS+ treatment based on protein maturation (09).

Missense novel biallelic variants of the SCN1B gene in homozygous condition have been reported in 2 siblings with febrile plus seizures, 1 of them mimicking Dravet syndrome (32). This c.265C>T variant predicting p.Arg89Cys is located in the extracellular immunoglobulin loop domain of the protein, which mediates interaction of the beta‐1 subunit with cellular adhesion molecules (32). It was thus considered pathogenic, expanding the spectrum of recessive inheritance in GEFS+. At the same time, an analysis of 22 Dravet syndrome patients without SCN1A variants has identified only 2 exon SCN1B variants (c.351C> T, p.G117G, and c.467C> T, p.T156M) in very low frequencies, thus doubting the causative connection of heterozygous SCN1B mutations with Dravet syndrome (55).

SCN1A gene (GEFS+2). It is estimated that the most frequent mutations found in GEFS+ families are within the SCN1A gene, accounting for 5.6% of the cases in large series, 10% to 11.5% in other reports, and 20% in more recent estimations (94; 162; 132). Two GEFS+ families with linkage to chromosome 2q24-33 were reported to harbor missense mutations within the sodium channel alpha subunit gene SCN1A (51). Both mutations affect highly conserved residues that encode for the putative voltage sensor of the transmembrane region of the channel. Functional studies showed that these mutations had different functional consequences. One mutation (R1648H) accelerated the recovery from inactivation, with consequent neuronal hyperexcitability. A computational model for SCN1A was further formulated and used to elucidate molecular mechanisms exhibited by this mutant (R1648H) (74). Mice heterozygous for the same R1648H mutation (RH mice) have demonstrated an increased susceptibility to cocaine-induced seizures (123). The same mouse model experienced chronic alterations in sleep regulation (115). To understand how other GEFS+ causing SCN1A mutations affect neural circuits and behaviors, Das and colleagues established a second knock-in mouse model using CRISPR/Cas9 editing (33). The K1270T mice represent an important tool for identifying similarities and differences in mechanism of action with R1648H and other mutations (33). The T875M mutation increases the slow inactivation mode, which reduces the accessibility of the channel protein for opening (51). However, electrophysiological studies of human SCN1A with the mutations T875M, R1648H or W1204R in a mammalian expression system (50), suggest these mutations confer a gain of function that may be responsible for symptoms associated with GEFS+ (88). Moreover, the mutations I1656M and R1657C were consistently associated with a loss of function (87) and mutations V1353L and A1685V with a complete abrogation of sodium channel activity. For 14 affected individuals of an Italian family, a missense mutation in the SCN1A gene was associated with a phenotype of only febrile seizures; 3 individuals later developed temporal lobe epilepsy (30). A novel SCN1A missense mutation in exon 21 (p.K1372E) was identified in 17 individuals in an Ashkenazi Jewish family with important phenotypic heterogeneity, from unaffected carriers to Dravet syndrome (54). Another example of remarkable clinical heterogeneity among family members was given by Passamonti and colleagues in a 3-generation Italian family carrying the c.2946+5G>A splicing mutation (116). Moretti and colleagues described 2 additional families in which affected individuals had biallelic SCN1A variants (103). Both patients (one per family, from related parents in each case) had fever-sensitive epilepsy beginning in the first months of life, followed by afebrile seizures and without severe cognitive impairment. Other authors have reported 2 siblings from a consanguineous pedigree with epilepsy phenotype compatible with GEFS+ associated with a homozygous likely pathogenic SCN1A variant (92). These 2 patients were compared to 10 previously published with epilepsy and variants in compound heterozygosity or homozygosity in the SCN1A gene. A case of 2 missense mutations in cis (p.[Arg1525Gln; Thr297Ile]) in all affected individuals of an Italian family showing GEFS+ and idiopathic generalized epilepsy features is also reported (15). Functional studies of the novel p.Arg1525Gln mutation in association with a previously known SCN1A cis mutation revealed significant shift in the activation curve towards more positive potentials of Nav1.1 channels (15).

Mutational analysis of SCN1A in patients with Dravet syndrome has revealed a high frequency (70%) of mutations, which are shown to be dominant and occur de novo as sporadic cases. In a series of 333 patients with Dravet syndrome, all possible types of mutations were identified along the length of the SCN1A gene; missense mutations were the most common (41). SCN1A truncation mutations thus far appear to be exclusive to Dravet syndrome; however, one report identified a novel de novo splice-site mutation within the SCN1A gene, which led to a protein truncation in a patient with focal epilepsy and FS+ (80). A subsequent mutation was identified in a patient with a milder FS+ phenotype (70). The preferred mechanism to explain the effect of the wide spectrum of mutations is haploinsufficiency, ie, total loss of function of the mutated allele (41), although there are attempts to elucidate mechanisms concerning how missense mutations might influence clinical phenotype in Dravet syndrome and GEFS+. Volkers and colleagues performed a biophysical analysis of 3 SCN1A missense mutations (R865G, R946C, and R946H), which they detected in 6 patients with Dravet syndrome, and compared the functionality of the R865G mutation with that of a R859H mutation detected in a GEFS+ patient. The 2 mutations reside in the same voltage sensor domain of Na(v) 1.1. The 2 Dravet syndrome mutations, R946C and R946H, were nonfunctional. However, the novel voltage sensor mutants, R859H (GEFS+) and R865G (Dravet syndrome), produced sodium current densities similar to those in wild-type channels. Their results suggest that the R859H mutation causes GEFS+ as a result of different biophysical defects in Na(v) 1.1 gating (154). Sugiura and colleagues reported 2 missense mutants at the same residue provoking different degrees of loss of function between GEFS+ and Dravet syndrome (147).

Approximately 10% of patients with Dravet syndrome have SCN1A mutations that are inherited. Furthermore, Depienne and colleagues reported that inherited SCN1A mutations were highly associated with mosaicism, affecting 13 of 19 families (40). The use of real-time PCR to selectively amplify and quantify mutant alleles present at low levels in the transmitting parent is an approach that is likely to enhance the management of families who may suffer from Dravet syndrome and have "major consequences for genetic counseling of asymptomatic parents with only one affected child" (37). Two SCN1A mosaic mutations were identified in 2 unrelated families with FS+ and partial epilepsy (139), indicating that SCN1A mosaicism may also occur in patients other than those with Dravet syndrome, who exhibit a milder form of epilepsy. Compared to the more extensive truncating and splice-site mutations of SCN1A found in Dravet syndrome, these mutations were missense mutations. The authors highlighted that for such families with mild cases of febrile seizures, even with asymptomatic parents, the possible high rate of recurrence associated with SCN1A mosaicism should be considered during genetic counseling. Myers and colleagues reported 7 patients with GEFS+ phenotypes with de novo SCN1A pathogenic variants, including a pair of monozygotic twins, that caused milder phenotypes (107). Four pathogenic variants were located in the pore-forming regions of SCN1A, one in the N-terminal and one in the voltage sensor region. They suggest that this genetic syndrome is not necessarily inherited or familial because it can present in sporadic individuals, and as such, family history is not essential for its diagnosis (107).

The voltage-gated sodium channel genes SCN1A and SCN1B were shown to be up-regulated in the hippocampus of spontaneously epileptic rats (58) although it is still unclear whether these changes play a causal role in epilepsy in these rats or whether they are secondary to seizures. Dutton and colleagues found that preferential inactivation of one SCN1A allele in parvalbumin interneurons of the neocortex and hippocampus in P22 mice resulted in reduced seizure thresholds, whereas similar inactivation of SCN1A from excitatory neurons did not significantly affect seizure thresholds (46).

A practical reference guide for the genetic testing of SCN1A in epilepsy has been published (67). More than 1200 mutations to date have been identified in the SCN1A gene, and the list is continuously growing, with targeted population-based sequencing of the gene in patients with Dravet syndrome and GEFS+ (119; 152).

GABRG2 gene (GEFS+ 3). The GABA-A receptor gamma2 subunit gene, GABRG2 (5q34), is also reported to be involved in the pathogenesis of GEFS+, confirming locus heterogeneity for this spectrum of epilepsy phenotypes (13; 155). Eight mutations have been identified so far, of which 6 were identified in GEFS+ families: R43Q, R139G, K289M, P83S, the c.1329delC frameshift mutation, and p.R136* (13; 155a; 08; 81; 150; 73). The mutation p.R136* was discovered in a family with 5 epileptic members, 3 of them with history of febrile seizures. The mutation segregated with the febrile seizure component of this family and was also present in an asymptomatic infant (73). A mutation (Q359X) was also identified in a family where the proband had Dravet syndrome (60). An attempt to elucidate the pathophysiological role of these mutations has been reported (76; 66; 18; 68; 69; 73). Claes and colleagues found different mutations of GABRG2 in 7 children with Dravet syndrome (29). These were de novo mutations observed in sporadic cases. It is possible that other mutations of the same gene may be transmitted and are responsible for cases of GEFS+ with other types of seizure (143), the variable picture resulting from different types of mutation or influences from other genes.

GABRD gene (GEFS+5). GABRD, which encodes the delta subunit of GABA-A receptor, is a putative susceptibility gene for GEFS+. Mutations in this gene were identified in a family of siblings with FS+ and febrile seizures (42). In vitro studies demonstrated that expression of the GABRD mutation, E177A, in human embryonic kidney cells produced "markedly reduced current when exposed to a saturating concentration of GABA” (130).

GABRB3 gene. GABRB3 encodes for the beta3 subunit of the GABAA receptor, GABRB3 mutations cause reduced receptor function, predicting impairment of GABA-mediated inhibition, and it has been associated with severe epilepsies. In a large cohort of 416 patients with various epilepsies, 22 patients were found to have presumed pathogenic variants in GABRB3. Eighteen were missense mutations, 3 were truncating, and 1 was a partial duplication of exons 1 through 9. Fourteen were de novo mutations; 3 mutations segregated in the family in a dominant fashion, 1 was inherited from the asymptomatic mother who was mosaic for the mutation, and 2 were inherited from an unaffected parent (101). The phenotype described is quite wide, including FS, early-onset absence epilepsy, myoclonic atonic epilepsy, unclassified focal epilepsies, Dravet syndrome like, West syndrome, Lennox-Gastaut syndrome, and other epileptic encephalopathies. The discrepancies found between phenotype and genotypes lead the authors to hypothesize that other factors besides the mutation influence the phenotype, for example, overall genetic background, a complex combination of different genetic variations or more specific genetic factors with larger detrimental or protective effects (101).

Other genes and genetic aspects to consider.

SCN9A gene (GEFS+7). Singh and colleagues identified mutations in the SCN9A gene in patients with Dravet syndrome as well as simple febrile seizures, self-limited afebrile seizures, and temporal lobe epilepsy (142). Another study reported that SCN9A susceptibility variants may contribute to complex inheritance for cases of Dravet syndrome without SCN1A mutations (106). Subsequent reports further support the association of monogenic SCN9A mutations and GEFS+ spectrum (164; 03; 165).

STX1B gene (GEFS+9) and synaptopathies. Schubert and colleagues reported the identification of mutations in STX1B encoding syntaxin-1B6 that are associated with both febrile seizures and epilepsy (135). First the authors presented data from 2 large pedigrees previously reported with linkage to chromosome 16p11.2 and 16p12-q12 (83; 159). Then 3 more mutations were identified in 449 familial or sporadic cases, and analyses of zebrafish larvae showed seizure-like behavior and epileptiform discharges that were highly sensitive to increased temperature. Two missense mutations resulted in nonconservative amino acid substitutions within the highly conserved SNARE motif region of STX1B. The authors propose this class of mutations as “synaptopathies.”

An article reported 3 affected members of a GEFS+ Chinese family carrying a novel heterozygous mutation of c.705 locus of STX1B (151). The same team attempted a literature review with a total of 10 mutations of the STX1B gene (half of them missense mutations) in 36 patients with GEFS+ phenotype, further supporting the association of the locus with this genetic syndrome (151). The convergence of mutations in genes encoding the synaptic vesicle release machinery suggests that this is a significant pathogenic pathway and that other associated genes may also contribute to epilepsy.

HCN1 (GEFS+10). Hyperpolarization-activated cyclic nucleotide-gated channels control neuronal excitability and few mutations reported so far in their encoding genes have been linked to epileptic encephalopathies. In a cohort of 33 patients with novel pathogenic or likely pathogenic variants, 36% of the patients had the first during a febrile illness whereas the GEFS+ spectrum was in total amongst the most predominant phenotypes (95). The authors also conducted further genotype and functional analysis suggesting that more severe phenotypes were related with variants within or close to transmembrane domains of the channel; different variants result to a wide biochemical alteration ranging from complete loss-of-function to important changes in activation kinetics and/or voltage dependence (95). The role of HCN1 mutations in GEFS+ pathogenesis needs further investigation.

SCN2A gene. A missense mutation encoding the alpha-subunit of neuronal voltage-gated Na(+) channel type II (Na(v)1.2) in a patient with febrile seizures associated with afebrile seizures was first reported by Sugawara and colleagues in 2001 (146). Experimental models showed that R188W mutation-related neuronal hyperexcitability was correlated with slower inactivation of the mutant channel than wild-type resulting in augmented Na (+) influx, whereas the Na(+) channel conductance was not affected (146).

Possible causative role of the SCN2A gene in GEFS+ is further supported by the newly reported heterozygous mutation c.1399G>A on exon11 of SCN2A (Nav1.2) in a twin family with GEFS+ (86). A study among Egyptian children established an association between SCN2A c. 56 G/A genetic polymorphism (SCN2A rs17183814) and febrile seizure plus (14). Moreover, this study showed that carriers of this SCN2A polymorphism tended to respond poorly to antiepileptic drugs.

Copy number variations. Hartmann and colleagues performed a genome-wide screening for copy number variations (CNVs) in 36 patients with SCN1A-negative fever-associated syndromic epilepsies (61). Phenotypes included Dravet syndrome (n=23; 64%), GEFS+/FS+ (n=11; 31%), and unclassified fever-associated epilepsies (n=2; 6%). They identified 13 rare CNVs in 8 out of 36 (22%) individuals. These included known pathogenic CNVs in 4 out of 36 (11%) of patients: a 1q21.1 duplication in a proband with Dravet syndrome, a 14q23.3 deletion in a proband with FS+, and 2 deletions at 16p11.2 and 1q44 in 2 individuals with fever-associated epilepsy with concomitant autism and/or intellectual disability. Three additional CNVs were classified as likely pathogenic and 6 CNVs were of unknown significance. They suggested that fever-associated epilepsy syndromes may be a recurrent clinical presentation of known microdeletion syndromes.

Fortin and colleagues reported an overrepresentation of CNVs in their small cohort of chromosomal microarray analysis in 12 GEFS+ families (52). Four of the twelve families had at least 1 copy number CNV identified (15q11.2 deletion, 19p13.3 deletion affecting CACNA1A, 10q11.22 duplication, and 22q11.2 deletion).

PRRT2 gene. The proline-rich transmembrane protein 2 (PRRT2) gene was identified to be related to paroxysmal kinesigenic dyskinesia, and infantile convulsions with paroxysmal kinesigenic dyskinesia, but He and colleagues screened PRRT2 exons in a cohort of 136 epileptic patients with febrile seizures including FS+, GEFS+, and Dravet syndrome (63). PRRT2 genetic mutations were identified in 25 out of 136 (18.4%). Five loss-of-function and coding missense mutations were identified.

CHRNA4 gene. CHRNA4 encodes for the alpha4 subunit of the neuronal nicotinic acetylcholine receptor, and it was identified in 3 individuals with episodic dyskinesia and febrile seizures that were PRRT2-negative (71).

An important genetic study by Heron and colleagues identified 8 previously unreported missence variants in SLC32A1 gene (64). For the scope of the study, whole genome sequencing data from 1165 epilepsy patients from the Epi4K dataset and 1329 Australian patients from the Epi25 dataset were interrogated. Two variants cosegregated with the phenotype in 2 large GEFS+ families with 8 and 10 affected subjects. Six further variants were identified in smaller families with GEFS+ or idiopathic generalized epilepsy (64). These variants are predicted to alter GABA transport into synaptic vesicles, leading to altered neuronal inhibition.

Other genes associated with a small subset of patients with Dravet syndrome include SCN1B (117), SCN2A (137), and PCDH19 (38; 93). The PCDH19 gene encodes protocadherin 19 and is believed to play a major role in epileptic encephalopathies (previously implicated in epilepsy and mental retardation limited to females) (43). In a study by Depienne and colleagues, 150 unrelated patients (113 females) with febrile and afebrile seizures were screened for PCDH19 mutations (39). Mutations were identified in 18 patients with highly variable clinical features but almost always included a high sensitivity to fever and clusters of brief seizures. The familial condition was suggestive of GEFS+ in a family, with 3 affected females presenting with partial cryptogenic epilepsy. Thus, this X-linked gene appears to mainly affect females, and the presence of an affected mosaic male indicates that cellular interference is at least 1 pathogenic mechanism (38). Another example of maternal transmission in a GEFS+ family was a report of a balanced translocation between chromosome X and 14 in which the breakpoint on the X chromosome disrupted a gene that encodes an auxiliary protein of voltage-gated Na channels, fibroblast growth factor 13 (FGF13) (122). Rigbye and colleagues further investigated whether the mutation of FGF13 would explain other cases of GEFS+ compatible with X-linked inheritance (124). They screened the coding and splice site regions of the FGF13 gene in a sample of 45 unrelated probands in which GEFS+ segregated in an X-linked pattern. The authors subsequently identified a de novo FGF13 missense variant in an additional patient with febrile seizures and facial edema. Their data suggest FGF13 is not a common cause of GEFS+.

Despite all the advances in GEFS+ genetics, reported mutation frequencies of known GEFS+ related genes vary between different series. For example, of the 3 genes SCN1A, SCN1B, and GABRG2, only a single mutation in SCN1A was identified in a study of 19 GEFS+ Scandinavian families (136), and no mutations were identified in any of the genes in a study of 7 GEFS+ Italian families (17) and 2 GEFS+ Tunisian families (105). In the latter study, an insertion of a T nucleotide within intron 12 of the SCN1A gene was found in 2 probands and both parents, but it remains unclear whether this is a pathogenic mutation or a polymorphism. In a study of another 8 GEFS+ Italian families with mutation screenings of the same genes, mutations were found in 2 families in the SCN1A gene (121). In a study of 23 GEFS+ Chinese families, SCN1A and GABRG2 mutations were both observed in 2 of the patients, but no SCN1B mutations were identified (148). A study concerning another 3 GEFS+ Chinese families did not find any mutations in the 4 genes, SCN1A, SCN1B, GABRG2, and SCN2A, but results showed a 6 cM candidate interval at 5q33-34 in 1 family (85). Sequencing candidate genes GABRG2 and GABRA1 in this region did not identify a causative mutation. Advances in genetics (next-generation/whole exome sequencing) will provide further insights in the genetic landscape of GEFS+, as implied by the discovery of an H258R mutation in the KCNAB3 gene in a Chinese family with GEFS+ (44).

Genetic heterogeneity and modifier genes. The mechanism that generates the phenotypic variability of GEFS+ in families is unclear, and the sequence of events that leads to febrile seizures and later development of various forms of epilepsy in some family members remains unknown. The fact that GEFS+ patients may harbor the same mutations (eg, SCN1A mutations) yet have different phenotypes indicates that additional genetic or environmental factors influence clinical presentation.

As stated by Scheffer and Berkovic, the most likely explanation for the phenotypic variability in GEFS+ is an effect of other modifier genes (128; 99). A study to investigate modifier genes in patients with Dravet syndrome who already harbor SCN1A mutations identified mutations in the CACNB4 gene, which encodes for the calcium channel beta-4 subunit (113). Hawkins and colleagues demonstrated that variants in SCN2A, KCNQ2, and SCN8A influence the phenotype of mice carrying the SCN1A R1648H mutation and suggested that ion channel variants may contribute to the clinical variation seen in patients with GEFS+ (62). A few years later, Makinson and colleagues showed that mice with SCN8A reduced expression had seizure resistance, reduced susceptibility to epileptiform activity, reduced hippocampal network excitability, and extended life (90). To further investigate the complex interference of different sodium voted gate channels in epilepsy, Makinson and colleagues introduced the SCN1A-R1648H mutation, identified in a family with GEFS+, into the corresponding position (R1627H) of the mouse SCN8A gene (89). Heterozygous R1627H mice exhibited increased resistance to some forms of pharmacologically and electrically induced seizures, and the mutant SCN8A allele ameliorated the phenotype of SCN1A-R1648H mutants. Paradoxically, at the homozygous level, R1627H mice did not display increased seizure resistance and were susceptible to audiogenic seizures. The authors also observed increased hippocampal pyramidal cell excitability in heterozygous and homozygous SCN8A-R1627H mutants, and decreased interneuron excitability in heterozygous SCN8A-R1627H mutants. These results demonstrate that the effects of a mutation may not be the same in 2 different voltage gated sodium channels and that targeting the SCN8A gene could on the one hand confer seizure protection but on the other hand increase seizure susceptibility (89).

Mechanistic insight in GEFS+ can be achieved by in vitro and in vivo experimental models.

Murine models have demonstrated that a loss of function of Na(V)1.1 channel (SCN1A) causes refractory seizures and ataxia (163; 112) and have furthermore demonstrated the significance of the genetic background in determining phenotype as heterozygote mice present different characteristics according to different strains (163). This was confirmed in Dravet syndrome mice in pure 129/SvJ genetic background that have fewer seizures with a higher threshold for thermally induced seizures and normal cognitive function compared to mice with an identical mutation but in a C57BL/6J background. It seems that the genetic background is enough to drastically alter the epileptic phenotype, as clinically evidenced in GEFS+ and Dravet syndrome (126).

Further in vitro studies of the murine models have demonstrated the importance of the GABAergic interneurons (163; 75). Pathophysiologic mechanisms concerning mutations of the GABRG2 gene are beginning to be elucidated. Bouthour and colleagues showed that on raising temperature, both the number of GABA(A) receptor clusters and the frequency of miniature inhibitory postsynaptic currents decreased in neurons expressing the K289M mutation (18). Raising temperature also increased the membrane diffusion of synaptic GABA(A) receptors. The authors concluded that the K289M mutation confers an enhanced sensitivity to neuronal activity and that hyperthermia may then trigger the escape of receptors from synapses and, thereby, further reduce the efficacy of GABAergic inhibition. For the missense mutation R43Q of the same gene, a lower temperature threshold for thermal seizures was displayed in mice (66). For another mutation, Q351X, mutant γ2(Q351X) subunits were shown to be degraded more slowly than wild-type gamma2 subunits and formed aggregates within rat cortical neurons (76). Another group showed a significant change in the expression of several epilepsy-related genes in a GABRG2 gene knockout cell-line (HT22^GABRG2KO) model of HT-22 mouse hippocampal neuronal cells (84). In fact, GABRA1and CACNA1A genes showed a temperature-induced decrease in expression, whereas an important alteration in the MAPK and PI3K-Akt signaling pathways and an upregulation of the matrix metalloproteinases family was also observed after GABRG2 knockout, resulting in a multifactorial impairment of GABAergic pathway signaling.

Data on the SCN1A gene using novel technologies [ie, induced pluripotent stem cell (iPSC) technology] further support NaV1.1 loss-of-function and the hypoexcitability of GABAergic neurons as the main mechanisms related to seizures, at the same time exploring the gene’s mechanistic complexity related to technical issues and to pathophysiologic remodeling (91). However, the genotypic-phenotypic correlation remains challenging as only a few SCN1A variants have been functionally assessed. Brunklaus and colleagues showed that electrophysiological data from mammalian expression systems successfully predicted disease severities. Milder phenotypes retained a degree of channel function measured as residual whole-cell current, whereas the Dravet syndrome phenotype often showed no whole-cell current (23). Xie and colleagues generated 2 pairs (GEFS+ patient line from 2 different individuals with the K1270T SCN1A mutation and their controls) of isogenic human iPSC lines by CRISPR-Cas9 gene editing and compared their electrophysiological properties in inhibitory and excitatory iPSC-derived neurons (160). They found that the K1270T mutation causes cell type-specific alterations in sodium current density and evoked firing but also that there are interactions between the mutation and the genetic background. These studies provide encouraging results for potential SCN1A-related disease biomarkers and patient-specific antiseizure therapies.

GEFS+ and inflammation. Choi and colleagues investigated the possible implication of inflammation in the pathogenetic mechanism of febrile seizures and GEFS+ (28). In their study, postictal serum cytokine levels and genetic variants in the cytokine genes interleukin (IL)-1β, IL-6, and high mobility group box-1 of 100 Korean children diagnosed with febrile seizures or GEFS+ were analyzed and compared to controls. Genetic variants located in IL-1β-31 and IL-1β-511 promotor regions were correlated with higher postictal IL-1β levels in febrile seizures, implementing that these variants may represent a host genetic factor for provoking febrile seizures.

Epidemiology

The prevalence of febrile seizures is between 2% and 5% in North America and Europe but around 7% to 14% of the population in Japan and the Pacific Islands. For children with febrile seizures, afebrile seizures occur in 2% to 7% of the population, which is 2 to 10 times the prevalence of that in the general population (05). Retrospectively, the percentage of febrile seizures in temporal epilepsy is estimated at approximately 25%, but does not reach the frequency reported in patients with Dravet syndrome that according to different reviews may even reach 100%. Incidence and prevalence of GEFS+ are not known.

Three quarters of individuals in GEFS+ families had epilepsy onset under 10 years, and equal sex ration have been reported (49). In a prospective population-based national cohort only 2 of 263 epilepsy patients were diagnosed with GEFS+ (149). Therefore, the disorder is either very rare or very difficult to diagnose because of its heterogeneity.

Prevention

Preventative measures are not available. No data are available to suggest that early treatment of febrile convulsions diminishes the risk of FS+ or epilepsy phenotypes associated with FS+ and other seizure disorders.

Differential diagnosis

GEFS+ is a term used to refer to a number of different epileptic seizures and epilepsy syndromes present in the same family. The presence of FS+ in 1 family member may be considered as a distinguishing feature of GEFS+. When dealing with an individual patient with no family history of seizures, a diagnosis other than common febrile seizures is rather impossible.

At the initial evaluation, differential diagnosis should be done (as in all epileptic patients) from symptomatic seizures following acute brain injury, CNS infection, neoplasia, etc. Subsequently, treatable inherited metabolic epilepsies (pyridoxine-dependent epilepsy, Glut-1 transporter deficiency, etc.) should be differentiated based on clinical and EEG phenotype (05).

Weber and colleagues reported a high rate of febrile seizures in a 4-generation family with benign familial infantile seizures (and linkage to the major benign familial infantile seizures locus on chromosome 16) (159); this is an association previously not described and demonstrates overlap between GEFS+ and benign familial infantile seizures. Moreover, mutations in the SCN2A gene are also associated with benign familial neonatal infantile seizures (BFNIS), GEFS+, Dravet syndrome, and some intractable childhood epilepsies (138).

Similarly, in individual patients, differential diagnosis from Dravet syndrome may be difficult. The syndrome usually manifests with unilateral febrile convulsions. The accompanying symptoms (frequent generalized, unilateral or focal seizures precipitated by fever, myoclonic phenomena, EEG abnormalities) are usually not present at onset. Dravet syndrome can be considered as part of the GEFS+ spectrum when other family members fulfill the general criteria described above.

Associated or underlying disorders

In one prospective case-control study by Sager and colleagues, patients with GEFS+ have demonstrated a significantly high impairment in both articulation and auditory discrimination of phonemes compared with healthy individuals (127). The authors conclude that early diagnosis and treatment of this condition could prevent potential problems as the development of dyslexia in the future. However, in this study, the individual genetic causes of GEFS+ patients have not been mentioned.

Diagnostic workup

Diagnosis of GEFS+ remains exclusively clinical, and it is not possible with either biological or genetic markers. For a family to be considered as presenting clinical characteristics within the GEFS+ spectrum, the presence of several types of seizures (including febrile seizures, FS+, absence, or myoclonic, partial, or astatic seizures) or epilepsy syndromes among more than 1 family member is a prerequisite. Penetrance ranges between 62% and 76%.

Detailed history taking, performed in a methodical and comprehensive manner, is an absolute requirement, and the epilepsy phenotype of all affected members must be described. A routine procedure that aims to diagnose seizure type and epilepsy syndrome of a given individual must be followed. No specific EEG pattern is characteristic of GEFS+, and consequently, all affected members should be subjected to a careful neurologic examination and have a high-quality EEG recording and eventually an MRI.

SCN1A genetic testing should be considered in GEFS+ when affected individuals fulfil the generally proposed criteria of: (1) seizures with fever, warmth, or vaccination; (2) prolonged or hemiconvulsive seizures; (3) photoparoxysmal epileptic response; (4) worsening under treatment with sodium channel inhibitors. The above testing is advised as part of an epilepsy multigene panel (100).

Management

GEFS+ is not an epilepsy syndrome under the classification of the ILAE, and no global management has been suggested. Individual patients should be treated according to their specific type of epilepsy or type of seizures.

Thus, for patients who have simple or complex febrile seizures, a treatment may not be necessary as the severity of the adverse effects of antiepileptic treatment could outweigh the benefits they may produce (96). However, careful instructions for rescue medication use should be carefully given. According to the FEBSTAT study only 23% of febrile seizures patients with a subsequent febrile seizure lasting more than 10 minutes were administered rectal diazepam (65).

Patients presenting with FS+, an antiepileptic drug, usually with a wide spectrum of applications, is administered--the choice being totally empirical.

For Dravet syndrome, for example, lamotrigine and carbamazepine should probably be avoided, at least at the early stages of the disorder as they could exacerbate seizures (57; 20; 72). The use of stiripentol (30-100 mg/kg/24 h) in the management of Dravet syndrome (27; 26), especially in combination with valproate and clobazam, is well established. Topiramate is also valuable in this case and more effective in association with valproate and clobazam (31). Fenfluramine has also proven effective in these cases. Cannabidiol at doses of 10 mg/day has demonstrated efficacy in Dravet syndrome patients in a randomized, double-blinded control trial.

Associated seizures will necessitate treatment on the basis of general clinical experience to avoid inappropriate drug choices for specific seizure types (118; 04; 05).

Vagus nerve stimulation was used in a girl with GEFS+ who presented with GTCS and focal seizures with loss of consciousness with over 75% and 80% reduction respectively 4 years after vagus nerve stimulation initiation, but more studies are needed to confirm its effectiveness and tolerance (59). Effectiveness beyond the Dravet syndrome spectrum needs also to be established for ketogenic diet in GEFS+.

Surgery should be considered as a possible treatment in cases with focal seizures. On the one hand, there are reports of individuals from GEFS+ families with temporal lobe epilepsy that become seizure-free after temporal lobectomy (131). On the other hand, Skjei and colleagues reported on 6 patients who underwent epilepsy surgery for intractable focal seizures (complete or partial frontal lobectomies or resection on frontoparietal region) (145). All patients were eventually found to have SCN1A mutations, and seizures continued despite surgery. Surgical histopathology showed evidence of subtle cortical dysplasia in 4 of 6 patients. More data are thus needed on epilepsy surgery’s efficacy in GEFS+.

An emergency seizure plan written and extensively explained to the family should be provided to all GEFS+ families. Protective helmets for patients experiencing frequent drop attacks should be prescribed. Neurologic and neuropsychological surveillance is essential for optimal outcome. Occupational therapy should be offered to appropriate cases. Vaccination management depends on personal, family, and genetic background. Individualized approaches should be considered (36).

Special considerations

Pregnancy

General recommendations for patients with epilepsy apply to affected members of a GEFS+ family. Women of fertile age should be warned of fetal exposure to valproate and other drugs and for possible interactions with contraception.

The pattern of inheritance is mainly autosomal dominant, but it could also be recessive, or a de novo mutation could be detected so genetic counseling is complex and difficult, and there is extreme variability of clinical phenotypes (21; 107). Because SCN1A has a complex inheritance pattern, SCN1A testing has been recommended in all individuals with febrile plus seizures or Dravet syndrome and in familial cases consistent with GEFS+ (107).