Clinical and genetic analysis of 23 Chinese children with epilepsy associated with KCNQ2 gene mutations

Abstract

Objective

To summarize the clinical features and genetic mutation characteristics of Chinese children with KCNQ2‐related epilepsy.

Methods

A cohort of children with genetically caused epilepsy was evaluated at Linyi People's Hospital from January 2017 to December 2023. After next‐generation sequencing and pathogenicity analysis, we summarized the medical records and genetic testing data of the children who had KCNQ2 gene mutations.

Results

We identified 23 KCNQ2 gene mutations. 73.9% (n = 17) of the mutation sites were located in S5–S6 segments and the C‐terminal region. In addition to the common phenotypes, 2 new phenotypes were identified: infantile convulsion with paroxysmal choreoathetosis (ICCA) and febrile seizure plus (FS+). Of all the cases with abnormal video‐electro‐encephalography, three cases with self‐limited familial infantile epilepsy (SeLNE) exhibited a small number of multifocal discharges. Of the patients who have taken a particular antiepileptic drug, the statistics on the number of patients who have responded to the drug are as follows: oxcarbazepine (8/9, 88.9%), levetiracetam (5/7, 71.4%), phenobarbital (9/16, 56.3%), and topiramate (2/5, 40.0%). However, the efficacy of phenobarbital varied widely in treating SeLNE and KCNQ2‐DEE. At the final follow‐up, 1 case with SeLNE had a transient developmental regression and 7 cases with KCNQ2‐DEE had mild to severe developmental backwardness.

Significance

Although clinically rare, we report 10 new KCNQ2 mutations and two new phenotypes: ICCA and FS+. This further expands genetic and phenotypic spectrum of KCNQ2‐related epilepsy. The gene mutation sites are mostly located in S5–S6 segments and the C‐terminal region, and the former is usually associated with KCNQ2‐DEE. Sodium channel blockers (including oxcarbazepine and topiramate) and levetiracetam should be prioritized over phenobarbital for KCNQ2‐DEE. Some cases with KCNQ2‐related epilepsy may have transient developmental regression during periods of frequent seizures. Early treatment and early seizure control may be beneficial for willing outcomes in children with KCNQ2‐DEE.

Plain Language Summary

This article reports 23 cases of children with KCNQ2‐related epilepsy, including 10 new mutation sites and 2 new phenotypes. It further expands the genetic and phenotypic spectrum of KCNQ2‐related epilepsy. In addition, the article summarizes the gene mutation characteristics and clinical manifestations of children with KCNQ2‐related epilepsy, with the expectation of providing a certain theoretical basis for the diagnosis and treatment of such patients.

Key points.

• In addition to the common phenotypes, KCNQ2 genetic mutations are also associated with infantile convulsions with paroxysmal choreoathetosis and febrile seizures plus.

• Some typical seizure characteristics and Video‐electro‐encephalography features help the clinic to recognize children with KCNQ2‐related epilepsy at an early stage.

• Phenotype was associated with the location of variants within the protein.

• There were differences in treatment options and prognosis between children with KCNQ2‐related epilepsy.

1. INTRODUCTION

The KCNQ2 gene encodes the Kv7.2 subunit, which consists of six transmembrane domains (S1–S6) and intracellular N‐ and C‐termini, with S1–S4 considered to be the voltage‐gated sensor and S5, S6, and the loop between them forming the ion pore. KCNQ2 mutations can result in impaired or abnormal function of Kv7.2, leading to epilepsy. The predicted incidence for KCNQ2 disorders is 2.93–3.59:10 000. ¹ Among them, the most common phenotypes are self‐limited familial neonatal epilepsy (SeLNE) and KCNQ2 encephalopathy (KCNQ2‐DEE). In recent years, some rare phenotypes have been discovered, including self‐limited familial infantile epilepsy (SeLIE), infantile epileptic spasm syndrome (IESS), and myokymia. ² , ³

KCNQ2‐related epilepsy often begins in the first few days of life. ⁴ The most common initial seizure types are tonic seizures with or without autonomic symptoms. In other cases, the children have similar initial seizures which begin with tonic seizures, followed by clonic seizures, and are usually accompanied by respiratory arrest, head tilt, or gaze. ⁵ The most frequently used and effective antiepileptic drugs are sodium channel blockers and phenobarbital. ⁶ Other effective antiepileptic drugs include levetiracetam, topiramate, and sodium valproate. ⁷ Ezogabine was shown to be effective in KCNQ2‐related refractory epilepsy, ⁸ but was withdrawn from the market because of its side effects.

The duration of seizures and the development of the nervous system related to KCNQ2 vary greatly. KCNQ2‐DEE usually presents with severe neurological developmental delay, and these patients and families require individualized treatment and genetic counseling. Therefore, early and correct identification of the potential phenotypic severity of the variants is critical. Gene mutation sites and EEG characteristics have been correlated with disease phenotype and prognosis. Patients with variants located in S4 segment or the ion pore may be associated with a poorer prognosis. ⁹ Video‐electro‐encephalography (VEEG) showing burst suppression patterns, hypsarrhythmia, multifocal epileptiform activity, or background discontinuity is usually suggestive of KCNQ2‐DEE. ¹⁰ Recently, the construction of prognostic prediction models for KCNQ2‐related epilepsy has been a research interest. Saez‐Matia et al. ¹¹ have proposed that combining genomic information with a novel variant frequency index (VFI) will develop a MLe‐KCNQ2 model that can predict the prognosis of affected children. Zeng et al. ¹² showed that extracting depth features from the EEG spectrums also has the potential to evaluate prognosis.

Limited by the number of cases, the genotypic spectrum and disease phenotypic spectrum of KCQN2‐related epilepsy remain to be expanded. In addition, some of the investigators' cases were derived from open databases, which makes it difficult to predict the severity of associated diseases in the absence of detailed clinical information. In this study, we applied next‐generation sequencing to detect KCNQ2 mutations, summarized the clinical features and antiepileptic drug responses, and characterized the prognosis of KCNQ2‐related epilepsy. Furthermore, we analyzed and explored the correlation between the clinical phenotypes and genotypes associated with this condition. We also aimed to discover new mutation sites and clinical phenotypes to expand the genetic and phenotypic profiles of the KCNQ2 gene.

2. METHODS

2.1. Subjects

For our retrospective study, we included children aged 0–14 years old who had epilepsy and attended Linyi People's Hospital from 2017 to 2023. Children with KCNQ2 gene mutations were selected, and their medical records were collected and followed up. Exclusion criteria included seizures caused by non‐genetic etiology, such as acquired brain injury or metabolic disease, and single‐gene disorders defined by clinical phenotype, such as tuberous sclerosis. Seizure types and epilepsy syndromes were categorized according to the International League Against Epilepsy (ILAE) classification and definition of epilepsy syndromes with onset in neonates and infants in 2022. ¹³ The definition of treatment effect as follows: significant: greater than 75% reduction in seizures; Effective: greater than 50% reduction in seizures; Ineffective: less than 50% reduction in seizures or worsening of seizures. Neurological development of the children was assessed using the Gesell development scale (0–3 years), the Wechsler preschool and primary scale of intelligence (4–6 years), or the Wechsler intelligence scale (6 years and older). The study protocol was approved by the Ethics Committee of Linyi People's Hospital (No. 13003). All guardians signed the informed consent form.

2.2. Next‐generation sequencing (NGS) and DNA sequence analysis

Using EDTA anticoagulation tubes, 2 mL of peripheral blood was collected from each of the children and major family members. Genomic DNA was extracted using the QlAamp Whole Blood DNA Kit (Qiagen, Germany), and the samples were sent to gene companies for whole‐exome sequencing. Sequencing results were aligned with the human reference genome sequence (GRCh38, hg38) from the UCSC database using BWA software (http://bio‐bwa.sourceforge.net). Single nucleotide polymorphisms (SNPs) and insertion–deletion variants (Indel) were analyzed using GATK software (https://software.broadinstitute.org/gatk/). Candidate variants were searched in the National Center for Biotechnology Information Clinical Mutation Database (ClinVar, https://www.ncbi.nlm.nih.gov/clinvar/). The frequencies of candidate variants were searched in the Exome Integration Database (ExAC, https://gnomad.broadinstitute.org/) and the Genome Aggregation Database (gnomAD, https://gnomad.broadinstitute.org/). Pathogenicity was predicted with the help of online software such as SIFT, PolyPhen, and Mutation Taster. Protein 3D structures were predicted with Swiss‐Model software. The pathogenicity of the variants was assessed according to the American College of Medical Genetics (ACMG) guidelines. ¹⁴ Sanger sequencing was performed to validate the suspected pathogenic variant loci.

3. RESULTS

3.1. Clinical features

We identified KCNQ2‐related mutations in 23 patients: 18 males and 5 females. The clinical phenotypes include 10 SeLNE, 2 SeLIE, 1 infantile convulsions with paroxysmal choreoathetosis (ICCA), 1 febrile seizures plus (FS+), 5 KCNQ2‐DEE, 2 atypical KCNQ2‐DEE. The clinical information of the patients is summarized in Table 1. Of the 23 pre‐diagnosed cases, most (19/23, 82.6%) had the onset of disease within the first week of life, and the remainder (4/23, 17.4%) of the children had the onset of disease in infancy or early childhood. The initial seizure type was mostly tonic seizures of focal origin (19/23, 82.6%). 10 cases (10/23, 43.5%) were characterized by sequential seizures, of which 8 progressed from tonic to clonic and 2 from tonic to spastic. Most of the cases were associated with autonomic seizures (19/23, 82.6%), and some cases were associated with automatisms (2/23, 8.7%). The VEEG of children with SeLNE were either normal or indicated a small number of multifocal discharges predominantly in the frontal‐occipito‐temporal region. Those with multifocal discharges converted to focal discharges at 2–3 months of age. In one of these cases, epileptic discharges disappeared by the age of 3.

TABLE 1.

The clinical manifestations with KCNQ2‐ related epilepsy.

No.	Sex	Mutation site	Reported/novel	Phenotype	Onset age of epilepsy (m.d)	Initial seizure types	Interictal VEEG	VEEG of episodes
1	F	c.1545G>C (p.E515D)	Y	SeLNE	0.1	Focal tonic, autonomic	Normal	Not detected
2	M	c.1545G>C (p.E515D)	Y	SeLNE	0.3	Focal tonic evolve into focal clonic, autonomic	Normal	Not detected
3	M	c.913_915delTTC (p.F305del)	Y	KCNQ2‐DEE	0.5	Bilateral asymmetric tonic, automatisms, autonomic	Atypical burst inhibition	Right prefrontal cortex and frontal lobe onset
4	M	c.2288_2306del (p.M763Rfs*161)	N	KCNQ2‐DEE	0.1	Bilateral asymmetric tonic evolve into bilateral symmetrical clonic, autonomic	Multifocal discharges	Not detected
5	M	c.776delA (p.D259Afs*14)	N	SeLNE	0.4	Bilateral asymmetric tonic evolve into bilateral symmetrical clonic, autonomic	Normal	Not detected
6	F	c.749T>C (p.V250A)	N	KCNQ2‐DEE	0.1	Bilateral asymmetric tonic, automatisms, autonomic	Multifocal discharges with partial burst suppression	Uncertain
7	F	c.365C>T (p.S122L)	Y	KCNQ2‐DEE	0.1	Bilateral asymmetric tonic evolve into bilateral symmetrical clonic	Intermittent hypsarrhythmia	Migrating focal seizure, one side of the parietal lobe onset
8	M	c.1149‐2A>G	N	SeLNE	0.4	Focal tonic evolve into focal clonic Bilateral asymmetric clonic	Few multifocal discharges predominantly in the posterior head	Left parietal lobe onset
9	M	c.917C>T (p.A306V)	Y	KCNQ2‐DEE	0.2	Focal tonic evolve into bilateral symmetrical tonic, autonomic	Multifocal discharges with burst inhibition during sleep periods	Left onset
10	M	c.1988A>G (p.E663G)	N	SeLIE	11	Focal clonic Focal‐bilateral tonic clonic	Normal	Not detected
11	M	c.794C>T (p.A265V)	Y	KCNQ2‐DEE	0.1	Focal tonic, autonomic Bilateral symmetrical epileptic spasms	Burst inhibition	Not detected
12	M	c.1639C>T (p.R547W)	Y	SeLNE	0.2	Focal tonic, autonomic	Normal	Not detected
13	M	c.2228delC (p.P743Rfs*187)	N	SeLNE	0.2	Focal tonic, autonomic	Normal	Not detected
14	F	c.386T>C (p.L129P)	N	SeLNE	0.5	Focal tonic evolve into bilateral symmetrical clonic, autonomic	Normal	Not detected
15	M	Exon11–12 (Absence of protein synthesis)	N	SeLNE	0.5	Focal tonic, autonomic	Normal	Not detected
16	F	c.439delG (p.A147Profs*24)	Y	SeLNE	0.2	Focal tonic, autonomic	Multifocal discharges, significant on right side	Not detected
17	M	c.821C>T (p.T274M)	Y	KCNQ2‐DEE	0.2	Bilateral asymmetric tonic evolve into bilateral symmetrical epileptic spasms, automatisms, autonomic	Burst inhibition	Left central, parietal and occipital region onset
18	F	c.1966G>A (p.E656K)	Y	SeLIE	10	Focal‐bilateral tonic clonic, autonomic	Normal	Not detected
19	M	c.1684T>C (p.Y562H)	N	SeLNE	0.1	Focal tonic, autonomic	Few multifocal discharges predominantly in the frontal and temporal region	Left posterior head onset
20	M	c.1073C>T (p.S358F)	Y	Atypical KCNQ2‐DEE	0.3	Focal tonic evolve into focal clonic, autonomic	Multifocal discharges, intermittent bursts inhibition	One‐sided central, parietal (1 left, 2 right) or right anterior middle temporal region onset
21	M	c.1073C>T (p.S358F)	Y	Atypical KCNQ2‐DEE	0.3	Focal tonic evolve into focal clonic, autonomic	Multifocal discharges, intermittent bursts inhibition	One‐sided central, parietal (4 left, 2 right) or left middle posterior temporal region onset
22	M	c.1545G>C (p.E515D)	Y	FS+	24	Generalized tonic–clonic, autonomic	Multifocal and widespread discharges	Not detected
23	M	c.631A>G (p.M211V)	N	ICCA	11/145	Focal clonic, autonomic	Normal	No abnormal discharges during clinical episodes

Abbreviations: F, female; m.d, months and days; M, male; N, The mutation site was novel; Y, The mutation site was reported.

Children with KCNQ2‐DEE exhibited burst suppression patterns (7/9, 77.8%), hypsarrhythmia (1/9, 11.1%), or multifocal discharges (1/9, 11.1%) on VEEG. The latest age at which burst suppression was detected at 42 days old. Burst suppression patterns were replaced by hypsarrhythmia or multifocal discharges at 2–3 months of age. Most of the children were treated with 1 to 3 antiepileptic drugs, with the exception being one case that was drug‐refractory (Table 2). Of the patients who have taken a particular antiepileptic drug, the statistics on the number of patients who have responded to the drug are as follows: oxcarbazepine (8/9, 88.9%), levetiracetam (5/7, 71.4%), phenobarbital (9/16, 56.3%), and topiramate (2/5, 40.0%) (Tables 2 and 3). Among them, the efficacy of phenobarbital varied greatly in patients with SeLNE (8/10, 80.0%) and KCNQ2‐DEE (1/6, 16.7%). At the final follow‐up, seizures were controlled in most patients (22/23, 95.7%). The age of seizure control ranged from 9 days to 24 months after birth, with 7 cases subsequently discontinuing medication, and the other 11 cases continuing on one medication. Of the cases with SeLNE, 1 case had transient mild developmental regression, and 1 case had hypertonia, which was significantly relieved by regular rehabilitation from 1 to 5 months of age. Of the cases with KCNQ2‐DEE, 2 cases had mild motor impairment and 7 cases had moderate to very severe cognitive impairment. In addition, 2 cases were accompanied by quadriplegia (Table 3).

TABLE 2.

Treatment options and follow‐up for KCNQ2‐ related epilepsy.

No.	Phenotype	Seizure free age (m.d)	Present age (y.m)	The last follow‐up
				Previous AEDS	Current AEDS	Development
1	SeLNE	2	6.6	PB	N	Normal
2	SeLNE	0.15	8	PB	N	Normal
3	KCNQ2‐DEE	9	4.10	PB/LEV/TPM	TPM	Very severe cognitive impairment
4	KCNQ2‐DEE	3	6.6	PB/LEV	LEV	Moderate cognitive impairment
5	SeLNE	1	4.2	PB	N	Normal
6	KCNQ2‐DEE	6	6.1	PB/LEV/OXC	OXC	Severe cognitive impairment
7	KCNQ2‐DEE	2	7.2	LEV	N	Very severe cognitive impairment and quadriplegia
8	SeLNE	0.9	4.10	PB	N	Normal
9	KCNQ2‐DEE	1	5.4	PB/TPM/OXC	OXC	Severe cognitive impairment
10	SeLIE	17	6.3	LEV	LEV	Normal
11	KCNQ2‐DEE	6	7.5	PB/OXC/TPM	TPM	Normal
12	SeLNE	1	4	PB	PB	Normal
13	SeLNE	0.5	4.4	PB	PB	Normal
14	SeLNE	1	4.5	PB/LEV	LEV	Normal
15	SeLNE	2	5.1	PB/OXC	N	Normal
16	SeLNE	4	3.11	PB/LEV/OXC	LEV/OXC	Transient mild cognitive/mild motor impairment
17	KCNQ2‐DEE	N	3.7	VPA/PB/OXC/CNZP/VGT/Perampanel	OXC/ZNCP/Perampanel	Very severe cognitive impairment and quadriplegia
18	SeLIE	20	5.3	VPA	VPA	Normal
19	SeLNE	1	0.3	PB/OXC	PB/OXC	High limb tension, improved with rehabilitation therapy
20	Atypical KCNQ2‐DEE	0.19	0.6	TPM	TPM	Mild motor impairment
21	Atypical KCNQ2‐DEE	2	0.6	TPM/OXC	TPM/OXC	Mild motor impairment
22	FS+	96	9	N	N	Normal
23	ICCA	24/152	15	OXC	OXC	Normal

Abbreviations: LEV, levetiracetam; m.d, months and days; N, no; OXC, oxcarbazepine; PB, phenobarbital; TPM, topiramate; VPA, valproic acid; y.m, years and months; ZNCP, zonisamide.

TABLE 3.

Effectiveness statistics of different antiepileptic drugs for KCNQ2‐ related epilepsy.

Groups	Significantly effective	Effective	Ineffective	Generally effective
PB	6 (37.5%)	3 (18.8%)	7 (43.8%)	9 (56.3%)
LEV	5 (71.4%)	0	2 (28.6%)	5 (71.4%)
OXC	7 (77.8%)	1 (11.1%)	1 (11.1%)	8 (88.9%)
TPM	2 (40.0%)	0	3 (60.0%)	2 (40.0%)

3.2. Genetic analyses

A total of 23 KCNQ2 mutations were identified by whole‐exome sequencing and Sanger validation (Table 1). These included 13 known mutations and 10 new mutations that are yet to be reported internationally. Among these, there were 16 missense mutations, 4 frameshift mutations, 1 splicing mutation, and 1 large segment deletion. The distribution of KCNQ2 mutation sites was primarily concentrated in S5–S6 segments and the C‐terminal region (Figure 1). The mutations in S5–S6 segments mainly manifested as KCNQ2‐DEE, while the mutations in the C‐terminal region could manifest various epileptic phenotypes. KCNQ2 mutations were present in 11 families. In examples 5 and 10 the mutation spans 3 generations. In the rest of the families the mutation was second generation. Two families showed incomplete epistasis (Figure S1). Figure S2 illustrated the impact of the 16 missense mutations on the corresponding protein's secondary structure. Figure S3 showed the results of the patient's sanger sequencing.

FIGURE 1.

Kv7.2 Structural Diagram. Markers indicate the location of the variant. Red markers: DEE; Green markers: SeLNE; Blue markers: ICCA; Yellow markers: FS+; Circle markers: Inframe mutations; Triangles markers: Missense mutations; Stars markers: Frameshift mutations.

3.3. Typical cases

Case 16 was diagnosed with SeLNE, a frameshift mutation of maternal origin. She has been having seizures since 2 days after birth, presenting as a focal tonic with autonomic seizures. The VEEG showed multifocal discharges. Seizures reoccurred after 2 months of phenobarbital control. Then, levetiracetam adequately controlled them, resulting in a further 1 year without any seizures. However, seizures reappeared later due to self‐medication reduction. Seizures were subsequently again controlled by levetiracetam in combination with oxcarbazepine. However, seizures reappeared later due to self‐medication reduction and were controlled by levetiracetam in combination with oxcarbazepine. The child was evaluated at 5 months old by Gesell, which showed borderline verbal ability, mild lag in gross and fine motor skills, and the ability to respond to both objects and people. Fortunately, the child's development was normal at the last follow‐up.

Case 17, diagnosed with EIDEE that evolved into IESS, began experiencing seizures on the second day after birth. These manifested as bilateral asymmetric tonic seizures, sometimes evolving into bilateral symmetric epileptic spasms with automatisms and autonomic seizures. The VEEG exhibited burst inhibition patterns. The child had been given phenobarbital, various sodium channel blockers, valproic acid, and other medications. However, the seizures remained uncontrolled. After 5 months of age, his seizures manifested as isolated or cluster of epileptic spasms. He underwent ketogenic therapy for 3 months starting at 15 months of age. Seizures were reduced but not controlled. The therapy was discontinued at 18 months of age due to poor parental compliance. ACTH was attempted at 20 months of age; seizures were not reduced. At the last follow‐up at 3 years and 7 months of age, frequent seizures were still present. He was quadriplegic. He can't sit or speak. This was also the only case of an uncontrollable seizure in our group.

Case 22 is a KCNQ2‐related FS+. The father had a history of febrile convulsions before the age of 1 year, which were untreated and resolved spontaneously after the age of 2 years. The father had normal development. The child began to have seizures at the age of 2 years and had four febrile seizures before the age of 6 years. These manifested as generalized tonic clonic seizures. The VEEG showed multiple and widespread discharges. No antiepileptic drugs were given at that time. At the age of 7 years, the child presented again with similar symptoms and was diagnosed with FS+, but the parents refused to take antiepileptic drugs. At the last follow‐up, the child had been free of seizures for 1 year and had normal growth and development.

Case 23 is a KCNQ2‐related ICCA, a missense mutation of maternal origin. His mother was asymptomatic. The child began experiencing focal clonic seizures in clusters at the age of 11 months, which resolved spontaneously by the age of 2 years. Starting at 12 years and 8 months old, the child experienced limb immobility when rising from a sitting position or while walking, which manifested itself as flexion contracture of the left toes and flexion of the left upper limb, lasting for about 10 s and occurring 7–8 times a day. His VEEG showed no abnormal discharges during the period of clinical seizures. The child was diagnosed with ICCA; His seizures decreased with oxcarbazepine. Occasional seizures occurred during fever or emotional stress, and growth and development were normal.

4. DISCUSSION

Most cases of KCNQ2‐related epilepsy begin with seizures in the neonatal period. ¹⁵ , ¹⁶ In our study, we also identified three patients with late infantile onset of the disease and one with early childhood onset, suggesting that mutations in the KCNQ2 gene are not rare in infantile epilepsy. The children typically present with focal or bilateral asymmetric tonic seizures as the initial seizure. Some of these children may progress to focal or bilateral symmetric clonic seizures. And some of those with KCNQ2‐DEE may progress to bilateral symmetric epileptic spasms. Seizures are most often accompanied by autonomic symptoms, and some may be accompanied by automatisms. Identifying a typical sequential epileptic pattern with distinct tonic seizures helps clinicians recognize KCNQ2‐related epilepsy at an early stage. ¹⁷ Children with different epileptic phenotypes can show multifocal discharges on VEEG during the neonatal period. However, those with SELNE tends to show a small number of multifocal discharges predominantly in the frontal‐occipito‐temporal region. Their seizures can be controlled in early infancy, leading to a significant reduction in epileptiform discharges. Xu et al. ⁹ suggested that sleep cycles on VEEG are generally not easily distinguishable for patients with KCNQ2‐DEE. This idea was also verified in our study. One child with SLENE showed transient developmental regression, suggesting that some children with KCNQ2‐related epilepsy with clustered seizure periods can have developmental regression and may have a good outcome after epilepsy control. Cases of atypical KCNQ2‐DEE have been reported in the past. ¹⁸ We also found 2 children with KCNQ2‐DEE who had only mild motor backwardness. It indicates that not all cases of KCNQ2‐DEE result in severe developmental delays and that early intervention with cessation of seizures, may improve long‐term outcomes.

KCNQ2‐related epilepsy has a wide spectrum of clinical phenotypes. In addition to the common phenotypes SeLNE and KCNQ2‐DEE, less common phenotypes have also been described, including SeLNIE, SeLIE, neonatal encephalopathy with non‐epileptic myoclonus, infantile or childhood‐onset DEE, isolated intellectual disability without epilepsy, and myokymia. 1 Our study also identified 2 new phenotypes: ICCA and FS+. ICCA is a rare neurological disorder that was first proposed in 1997 by Szepetowski et al. ¹⁹ PPRT2 is the most common pathogenic gene for ICCA. ²⁰ ZHAO et al. ²¹ reported that oxcarbazepine was effective in treating PRRT2‐related diseases and patients were completely free of seizures after 10 months of follow‐up. In our group, children with KCNQ2 mutation‐related ICCA responded well to oxcarbazepine, although it did not completely control the seizures. A possible explanation for this is that oxcarbazepine has variable efficacy in ICCA associated with different pathogenic genes but is generally effective. Reported pathogenic genes for GEFS+ include sodium channel‐related genes (SCN1A, SCN2A, SCN1B), chloride channel‐related genes (GABRG2, GABRD), and others. ²² GEFS+ manifestations range from febrile seizures (FS) and FS+, which generally have a good prognosis, to myoclonic‐astatic epilepsy and severe myoclonic epilepsy of infancy, which are associated with a poor prognosis. ²³ Our patient diagnosed with FS+ achieved seizure control without the use of antiepileptic drugs, leading us to speculate that the KCNQ2 mutation may predispose to a milder phenotype of FS+. Our findings expanded the genetic spectrum of ICCA and the phenotypic spectrum of KCNQ2.

Statistically, missense variants located in the intracellular structural domains between S2 and S3 segments are prone to cause SeLNE. The hotspots of gene mutations in DEE are in the vicinity of S4 segments, ion pore, and the C‐terminal region of helix A and helix B. ²⁴ Most of the mutation sites in this group are located within S5–S6 segments and the C‐terminal region. In our case, SeLNE was mostly distributed in the C‐terminal region, which is inconsistent with literature reports. Although this may be attributable to the small number of SeLNE cases, it is still important to consider the presence of KCNQ2 gene mutation hotspots in Chinese children with SeLNE. DEE mutations were predominantly located in the S5–S6 segments, potentially due to the specialized function of the ion pore. The distribution of epilepsy phenotypes in the C‐terminal region did not show significant differences, indicating that further studies are required to elucidate the correlation between the functional domains of the C‐terminal and clinical phenotypes. It was previously believed that self‐limited epilepsy predominantly involved missense and truncating mutations. In contrast, DEE was mostly characterized by de novo missense mutations or very small in‐frame deletions. ²⁵ In our group, we found one case of KCNQ2‐DEE caused by a KCNQ2 frameshift mutation, indicating that a subset of children with KCNQ2 frameshift mutations may have a poor prognosis. We found KCNQ2 mutations in a pair of identical twins, both of whom had mild developmental delay, while the father, carrying the same pathogenic gene, had a normal developmental. Patients with the same mutation, even within the same family, may have diverse clinical phenotypes, suggesting a complex connection between genotype and phenotype. This emphasizes the need for clinics to create individualized treatment plans based on each child's symptoms and ancillary test results.

Phenobarbital is the current first‐line drug, controlling 43% to 80% of electrical seizures in children with EEG abnormalities. ²⁶ Levetiracetam is also a recently recommended drug. ²⁷ Sodium channel blockers are considered an effective treatment for KCNQ2‐related DEE. ²⁸ The common effective therapeutic agents in our study included oxcarbazepine, levetiracetam, phenobarbital, and topiramate. However, the effectiveness rate of phenobarbital differed significantly between SeLNE and KCNQ2‐DEE, suggesting that sodium channel blockers and levetiracetam may be preferred over phenobarbital for KCNQ2‐DEE. EIDEE is the most common epileptic syndrome in KCNQ2 encephalopathy and rarely evolves into IESS. ⁵ Valproic acid has been suggested to be more effective in children transitioning to IESS. ²⁹ In our study, one child with EIDEE evolving into IESS presented with drug‐refractory epilepsy. However, the ketogenic diet led to a reduction in seizures. It is suggested that the prognosis for such children may be worse than for those with EIDEE. Ketogenics may have some effect, and the effectiveness of valproic acid remains to be evaluated. There are various treatment options for KCNQ2‐related epilepsy. Corticosteroids, folinic acid, and pyridoxine are also effective in patients with refractory epilepsy. ³⁰ , ³¹ Ezogabine, a potassium channel opener, was withdrawn from the market due to the side effects such as skin bluing and retinal pigmentation. ³² 5‐HT6 receptors, calmodulin, and so on may be the new candidate targets for the treatment of KCNQ2‐related epilepsy. ³³ However, these are still in the basic experimental stage. We hope that these treatments will be applied clinically for the benefit of such patients as soon as possible.

We investigated the association between the type and distribution of KCNQ2 mutations and clinical phenotypes. This may provide clues for clinical diagnosis and treatment. However, this study has some limitations. First, the cases came from a single research center and the sample size was small, potentially causing a selection bias. Second, the retrospective study may have omitted information, possibly leading to recall bias. Third, the relationship between KCNQ2‐related epilepsy genotypes and clinical phenotypes is complex, and the amount of KCNQ2 protein expression, the number of potassium channels in axon initiation segments, and the M‐current are closely associated with the severity of the clinical phenotypes. ³⁴ , ³⁵ In the future, basic experiments will be needed to explore the potential pathogenic mechanisms.

5. CONCLUSIONS

In this retrospective study, we report 10 new KCNQ2 mutation sites and two new phenotypes: ICCA and FS+. This further expands genetic and phenotypic spectrum of KCNQ2‐related epilepsy. By analyzing the characteristics of the genetic mutations and clinical phenotypes of the affected children, we found the following features. (1) Such patients tend to present with sequential seizures starting with tonic seizures. (2) Although different epileptic phenotypes can present with multifocal discharges, SeLNE children have their own characteristics. (3) The gene mutation sites are mostly located in S5–S6 segments and the C‐terminal region, and the former is usually associated with KCNQ2‐DEE. (4) Sodium channel blockers and levetiracetam should be prioritized over phenobarbital for KCNQ2‐DEE. (5) Some cases with self‐limited epilepsy can present with transient developmental regression during periods of frequent seizures. However, some cases with KCNQ2‐DEE may have only mild developmental lag. Our study can help clinicians diagnose and treat such patients.

AUTHOR CONTRIBUTIONS

Xixi Yu wrote the first draft of the manuscript. Na Xu and Liping Zhu performed the acquisition of data. Xin Zhang performed the data analyses. Fengyuan Che, Li Yang, Shiyan Qiu, and Yufen Li contributed to the conception and design of the study. All authors helped to revise the manuscript regarding crucial intellectual content. All authors approved the final version for publication.

FUNDING INFORMATION

This research was funded by the following organizations. (1) Science and Technology Development Fund Project of XuZhou Medical University (XYFY202220). (2) Fund Project: Key R&D Plan Project in Linyi City (2023YX0005). (3) Science and Technology Development Program of Shandong Province (202306010670). (4) Science and Technology Development Fund Project of WeiFang Medical University (2023FYM038).

CONFLICT OF INTEREST STATEMENT

None of the authors has any conflict of interest to disclose. The authors confirm that they have read the journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

ETHICAL APPROVAL

Ethics approval for the study was granted by the Ethics Committee of Linyi People's Hospital.

CONSENT FOR PUBLICATION

Informed consent for clinical study and molecular genetic analysis was obtained from the patients' parents.

Supporting information

Appendix S1

Yu X, Che F, Zhang X, Yang L, Zhu L, Xu N, et al. Clinical and genetic analysis of 23 Chinese children with epilepsy associated with KCNQ2 gene mutations. Epilepsia Open. 2024;9:1658–1669. 10.1002/epi4.13028

DATA AVAILABILITY STATEMENT

The original sequencing data used to support the findings of this study are restricted by the Ethics Committee of Linyi People's Hospital in order to protect patient privacy. Data are available from the corresponding author Li Yang (docyangli@163.com) for researchers who meet the criteria for access to confidential data.