Genetic epilepsy with febrile seizures plus, previously known as generalized epilepsy with the febrile seizures plus (FS+), was first discovered by Scheffer and Berkovic in an Australian family in 1997 (24). Since it was found that the phenotype of focal epilepsy can occur in the GEFS+ family, it was renamed genetic epilepsy with FS+. GEFS+ is an EE with a milder phenotype than DS; it is also related to the multiple gene mutations, including SCN1A, SCN2A, SCN1B, GABRD, SCN9A, STX1B, and Fgf13 (25, 26). We have previously found in animal models that GABRG2 mutations can also cause GEFS+ (27). Various clinical phenotypes can occur in the GEFS+ family, ranging from the most common febrile seizures (FS) and FS+ to the severe EE known as DS. In 2000, Escayg et al. first found mutations in the SCN1A gene (Thr875Met and Arg1648His) in GEFS+ families (28) (Figure 2, Table 1). Aside from DS, SCN1A gene mutations are the most common pathogenic genes for GEFS+. In fact, GEFS+ and DS are different manifestations of epilepsy caused by SCN1A mutations.

Doose syndrome, also known as epilepsy with myoclonic atonic seizure (EMAS), was previously called myoclonic astatic epilepsy (MAE), a rare childhood EE (55). First reported by Doose in 1970, the International League Against Epilepsy (ILAE) in 2010 changed its name to epilepsy with myoclonus-atonic seizures based on the characteristics of epileptic seizures (56, 57). Usually, Doose syndrome develops from seven months to 6 years, and the peak age of onset is 2 to 4 years. Children usually have normal mental and motor development before the onset. Most children start with a generalized tonic–clonic seizure (GTCS). The initial seizures can be very frequent, followed by a variety of generalized seizures, including myoclonic seizures, dystonic seizures, myoclonic–dystonic seizures, and atypical absence; a small number of children may have tonic seizures in the later stages (55).

Doose syndrome is associated with mutations in a variety of epilepsy genes, including SCN1A, SCN1B, CACNA1H, SLC2A1, GABRG2, CHD2, SLC6A1, STX1B, GABRB3, SYNGAP1, and WDR45 (33). In 2005, Ebach et al. reported three cases of EMAS with SCN1A gene mutations (29). In 2007, Harkin et al. found one case of Doose syndrome due to SCN1A mutation in 188 patients with EE (30). Interestingly, Dimova et al. also found a case of EMAS caused by SCN1A gene mutation in a GEFS+ family. The patient started with a febrile seizure at the age of three, after which subsequent multiple myoclonic and myoclonic–astatic seizures appeared (31). Recently, Hinokuma et al. found one microdeletion at 2q24.2 involving SCN1A in 29 patients with Doose syndrome (33) (Figure 2, Table 1). All of the foregoing extends the phenotype of the SCN1A gene mutation to Doose syndrome.

Epilepsy of infancy with migrating focal seizures, previously known as infantile migratory partial epilepsy (MPSI) or infantile malignant migratory partial seizure (MMPSI), is a rare and early-onset developmental EE inherited in an autosomal dominant pattern. It is characterized by onset within 6 months of birth and mainly manifests in the form of frequent, migratory, and varying types of focal seizures. Epileptic seizures are related to the multifocal EEG discharge. Similar to DS, this disease is often associated with severe cognitive impairment and motor impairment. However, unlike DS, the most common causative gene is KCNT1 mutation.

Freilich et al. first identified the SCN1A mutation in a female infant diagnosed with MPSI. The female infant, who was delivered to term, developed epilepsy at 10 weeks after birth, accompanied by progressive hemiplegia, apnea, and progression of multifocal migratory partial epileptic seizures, leading to a recurrence of epileptic status and death at 9 months (34). In the same year, another case of SCN1A mutation was found in another patient with MPSI (35). In 2012, Shein et al. reported a case of SCN1A mutation-induced MPSI with good therapeutic effect assisted by hypothermia (36). In 2015, Lim et al. reported three cases with SCN1A mutation (MPSI) (38). In the same year, Zhang and colleagues found 46 cases of genetic mutations in 253 children with unexplained epilepsy and intellectual/developmental disabilities, of which only one was an SCN1A mutation causing malignant migrating partial seizures of infancy (37). In 2016, Shang et al. conducted genetic testing on nine cases of EIMFs and found that two (22.2%) carried an SCN1A mutation (39). Recently, Fang et al. also found one SCN1A mutation patient in five patients with EIMFS (41) (Figure 2, Table 1). SCN1A is currently considered to be the third most common type of genetic variation in EIMFS (58).

West syndrome, also known as infantile spasms (IS), is a refractory classic EE characterized by repetitive epileptic spasms (ES) and hypsarrhythmia (44). The etiology of the West syndrome is complex and varied. Genetic studies of individuals with unexplained IS have identified over 37 genes as pathogenic (59). However, SCN1A was not reported in a recent review of West syndrome, indicating its rarity in this disease (42, 59). Hattori et al. reported a case of partial epileptic seizures at four months and a West syndrome infant at 8 months with characteristic facial appearance, big toe abnormalities, and developmental delay. Chromosome and gene sequencing revealed the deletion of the SCN1A gene and 2q31.1 region [arr 2q24.3q31.3 (166,303,447–180,982.972) × 1 (build19)] (43). In 2003, Wallace et al. found one case of SCN1A mutation in 23 patients with West syndrome (42). Ilona et al. found only one SCN1A mutation in 45 patients clinically diagnosed with West syndrome by genetic testing (44). Most recently, Na et al. performed targeted gene panel sequencing for 150 early onset DEE infants aged ≤ 3 months and only one patient with SCN1A mutation was found. These findings indicate that the phenotypic heterogeneity of SCN1A mutation has extended to West syndrome (Figure 2, Table 1).

Lennox–Gastaut syndrome is a childhood EE whose main clinical features include multiple types of drug-resistant seizures, intellectual disability, and abnormal EEG with diffuse spines slow complex wave or paroxysmal fast activity. The etiology of LGS is also complex and varied; about 75% of cases have obvious causes such as cortical malformations, posthypoxic ischemic results, postmeningitis/encephalitis, or metabolic encephalopathy, while about 25% are cryptogenic (60). LGS is associated with a variety of genetic mutations, including ion channel genes (SCN2A, KCNT1, GABRA1, SCN8A, and GABRB3), transcription regulation genes (CHD2), neurocutaneous syndrome-related genes (TSC1 and TSC2), metabolic genes (Alg13), and others (45, 61, 62). However, SCN1A mutations rarely occur in LGS (30, 47). Harkin et al. found an SCN1A mutation in one out of 188 epileptic encephalopathy patients diagnosed with LGS (30). In 2009, Selmer and colleagues examined the SCN1A gene in 22 adult patients with LGS and found a mutation in one of them (47) (Figure 2, Table 1). In summary, SCN1A is rare, but it can still occur in LGS.

Rett syndrome is a rare single-gene disease that is more prevalent in females. RTT patients usually have an early stagnation period of onset 6–18 months after birth, and then enter a rapid regression period of development. The typical phenotype includes intractable epileptic seizures and severe mental retardation, particularly a rapid regression in language and limited progress in the psychomotor development. They may also be accompanied by the related complications such as autism, hand stereotypes, and respiratory pattern disorders (63). While more than 95% of patients carry de novo mutation(s) in the methyl-CpG-binding protein 2 (MECP2) gene (classical RTT), a small fraction of the patients (atypical RTT) may carry genetic mutations in other genes, such as the cyclin-dependent kinase-like 5 (CDKL5) and FOXG1 (64, 65).

The role of SCN1A dysfunction in RTT has also been highlighted by a few cases (46). Henriksen and colleagues (46) reported two patients with RTT caused by mutations in SCN1A. The first case is a 19-year-old female who developed febrile seizures at 5 months of age and subsequently developed afebrile focal seizures and intractable generalized seizures, including myotonic, tonic, and tonic–clonic. She also had several episodes with convulsive status epilepticus. She manifested normal hand functions and started to use a few words until she was 15 months old, but after that, her development slowed down. She stopped using her hands, her gait became broad and ataxic, and her speech disappeared. Between 1 and 2 years of age, she developed autism. At the age of 19, she still had dysmotility of hands and ataxia and suffered from breath holding and teeth grinding. Her height was only 141 cm. Her clinical signs and symptoms were consistent with classic RTT. Genetic testing showed that she was negative for MECP2, CDKL5, and FOXG1 genes, which are common to RTT, but SCN1A mutations were found. The second case occurred in a 32-year-old woman. She had her first febrile bilateral tonic–clonic seizure when she was 7 months old. The seizures worsened between the ages of one and two. Like the first patient, she grew normally until 12 to 15 months of age, but later acquired developmental disabilities and began to lose acquired skills. Her hand functions gradually declined, her speech disappeared, and she no longer seemed interested in her surroundings. She also suffered from bruxism and hand-washing stereotypes. At age of 32, she could walk for a few meters with support but still had ataxic and apraxic hand movements. She could not speak and had slight scoliosis. Epilepsy was always present. She also met the classic diagnostic criteria for RTT. Whole-exome sequencing unveiled the variant in SCN1A (Figure 2, Table 1).

Developmental and epileptic encephalopathies (DEEs), also known as early onset epileptic encephalopathies, early infantile epileptic encephalopathies (EIEEs), or early infantile-onset developmental and epileptic encephalopathies (EIODEEs) (45, 51, 66), comprise a kind of refractory epileptic encephalopathy that is mainly characterized by early-onset in neonates or infants, refractory epileptic seizures, and severe abnormal electroencephalogram discharge, psychomotor retardation, or regression. DEEs include early myoclonic encephalopathy (EME), Otahara syndrome, EIMFS, West syndrome, and DS (57). Nonsyndromic epileptic encephalopathy (NEE) can be referred to as clinically diagnosed epileptic encephalopathy without the inclusion of a specific syndrome or epileptic disorder (51).

In 2014 and 2016, Japanese scholars Ohashi and Kobayashi et al. described a distinct SCN1A phenotype called early infantile SCN1A encephalopathy, in which the patient had an apparent movement disorder (49, 51). Sadleir et al. also reported eight cases of SCN1A mutation with hyperkinetic movement disorder in 2017 (53) (Figure 2, Table 1). This may become a new type of epileptic encephalopathy shortly. Similarly, SCN1A mutations are rarely found in other cases of NEE (48, 52, 54).

Hemiplegic migraine is the most common neurological disorder that often presents with aura, which is associated with sensory and motor disturbances (67). Familial (FHM) and sporadic (SHM) hemiplegic migraines are severe subtypes of migraine associated with transient hemiparesis (68). FHM, a rare autosomal dominant genetic disorder, is a subtype of migraine with aura (MA) (69). The common classification and pathogenic genes are CACNAIA (FHM1), ATP1A2 (FHM2), and SCN1A (FHM3) (70). Familial hemiplegic migraine type 3 (FHM3) is seldom caused by mutations in SCN1A (71). Martin et al. first identified the SCN1A mutation in 2005 in the three familial migraine families (5). Subsequently, numerous SCN1A mutations have been found in FMH3 and, currently, about 60 patients carry SCN1A mutations (67, 72–85). In addition to FHM, SCN1A mutations are also found in a very small number of sporadic hemiplegic migraine patients (68, 86, 87) (Figure 3, Table 2). Therefore, it is confirmed that SCN1A is one of the pathogenic genes for hemiplegic migraine.

Autism spectrum disorder is a complex psychiatric disorder characterized by impaired communication and social skills, and also restricted and repetitive behaviors (95). ASD can occur by itself or as a complication of epilepsy such as DS (6, 88, 96). DS caused by SCN1A gene mutation is associated with ASD (22, 96–98). Li et al. evaluated 37 patients with DS, nine of whom (24.3%) met autism criteria. They also found that people with autism had more severe intellectual disabilities than people without autism (97). Han et al. found an autism-like phenotype in SCN1A-mutated DS model mice (99). Interestingly, low-dose Clonazepam (a positive allosteric regulator of GABAAR) was used to mitigate this symptom, suggesting that GABAgic neurons may be directly related to ASD (99). Autism spectrum disorders can last from childhood to adulthood and even throughout life. Berkvens et al. conducted a follow-up on 13 patients with DS, among whom eight (61.5%) were classified as having ASD (96). Furthermore, ASD can occur in isolation from epilepsy. Weiss et al. found five missense mutations in patients with autism (6). Roak et al. also found one case of SCN1A missense mutation (p.Pro1894Leu) in 20 patients with ASD, and this mutation may be inherited from its parent (88). A recent study of 134 cases of autism identified 16 variants and 12 genes with evidence of pathogenicity, including three SCN1A mutations (91). In summary, SCN1A is closely related to ASD and has been considered as an ASD candidate gene (6, 7, 11, 89, 90) (Figure 3, Table 2).

Epilepsy-related deaths include seizures leading to asphyxia, injury, drowning, the occurrence of epileptic status, suicide, and SUDEP, which is a common cause of death in patients with epilepsy (100). SUDEP is a sudden, accidental death of a person with epilepsy, with or without witnesses, not from trauma or drowning, and with or without epileptic seizures; an epilepsy status must be ruled out and no structural or toxic cause of death is found at autopsy (101, 102). SUDEP generally occurs in 1.2 per 1,000 people with epilepsy per year (101). The sodium channels SCN1A, SCN1B, and SCN5A are considered as genes related to SUDEP (8, 103–106). DS, which is mainly caused by the SCN1A mutation gene, is the best model for studying the SCN1A gene (107). The mortality rate in patients with DS is about 20%, with SUDEP generally present in the deaths of children and adults with epileptic status (108). SUDEP occurs at a higher rate in DS than in other childhood epilepsies, accounting for up to about 50–60% of mortality (109, 110).

Sudden death associated with SCN1A mutations has also been reported in nonepileptic patients. In 2016, Halvorsen et al. (92) found one SCN1A mutant aged 20.8 months among nine children with sudden disease who died of the unknown causes. The child developed normally with a history of febrile convulsions but, interestingly, her siblings were diagnosed with DS. In 2018, Brownstein et al. (9) found an association between SCN1A mutation and sudden death in younger infants. The first case is a girl who died suddenly at the age of 2 months, with the cause of death recorded as sudden infant death syndrome (SIDS). Gene sequencing revealed an SCN1A mutation. Microscopic examination of the hippocampus revealed focal bilamination of the dentate gyrus. The other case occurred in a 7-week-old female with two SCN1A mutations (92) (Figure 3, Table 2). These results suggest that SCN1A mutations are not only closely related to SUDEP but also associated with nonepileptic-related sudden death.

The exact mechanism of SUDEP remains unclear. In systemically knockout heterozygous SCN1A+/- mice, severe arrhythmias were found to be characterized by prolonged PR interval, increased heart rate variability, and even atrioventricular block, suggesting that changes in the cardiac SCN1A may be related to SUDEP (111). In another study, paroxysmal chronic bradycardia and associated ventricular electrical dysfunction were found in heterozygous SCN1A^+/− mice; notably, atropine and N-methyl scopolamine were effective in preventing sudden death in mice (112). In addition, respiratory dysfunction was also found in mouse models of DS, which may also be one of the causes of SUDEP in SCN1A mutant mice (109).

Arthrogryposis multiplex congenita refers to an etiologically heterogeneous condition that is characterized by the congenital joint contractures in two or more body areas (113). AMC is generally thought to be the downstream result of a reduction in the fetal movements. AMC has an overall incidence of one in 3,000 to 5,000 (114).

Although over 320 genes have been implicated, exemplifying the genetic heterogeneity of the condition (115), AMC is poorly related to SCN1A, with only two reports documented (93, 94). The first report described SCN1A mutations in three infants with AMC from three different families (93). During the fetal period, they are characterized by abnormal development of different joints and a lack of fetal movements (in family 1, bilateral flexion of both hands, hyperextension of knees, and reduced swallowing; in family 2, arthrogryposis of the upper limbs and microretrognathism; in family 3, bilateral camptodactyly, hyperextension of knees, and hallux valgus of feet). It is noteworthy that one of the infants (family 1) developed refractory epilepsy 2 days after birth, while the other two patients both died due to early termination of pregnancy. This suggests that in addition to peripheral joint dysplasia, AMC patients may also have abnormalities of the central nervous system, such as epilepsy, which may be similar to DS. The other description was reported by Laquerriere et al., who sequenced 315 patients with AMC and found 51 gene mutations in 166 (52.7%), including the rare SCN1A (94) (Figure 3, Table 2).