In this study, children with SeLFE who were hospitalized in the Department of Paediatrics, the Second Affiliated Hospital of Xi'an Jiaotong University, from October 2011 to October 2021 were included for retrospective analysis. The inclusion criteria were as follows: (1) age of onset 0–24 months; (2) normal intelligence and motor development before and after the onset of the disease; (3) focal or focal to bilateral seizures; (4) focal epileptic discharge but no typical epileptic discharge during seizures; (5) no evidence of a lesion causing epilepsy; and (6) genetic testing results. Children diagnosed with any of the following were excluded from the study: (1) pediatric malignancies, (2) hypoglycaemia, (3) intracranial infections, (4) traumatic brain injuries, (5) hypocalcemia, and (6) symptomatic epilepsy. This study was approved by the Second Affiliated Hospital of Xi'an Jiaotong University (2023034). Informed consent was signed and provided by the patients' parents. This study was conducted in accordance with the guidelines of the Declaration of Helsinki (version 2013).

DNA was extracted using a Blood Genome Column Medium Extraction Kit (Kangweishiji, China) Kit. Liquid hybridization was performed using an IDT xGen® Exome Research Panel v2.0 capture probe with gDNA library sequences. The whole- exome library covered all coding and some non-coding regions of 19,396 genes in the human genome, with a capture interval size of 51 Mb. High-throughput sequencing was performed with the UWI DNBSEQ-T7 sequencing platform, and the coverage of target sequence sequencing was not less than 99%. The aligned sequences were then compared with the GRCh37/hg19 reference genome using the Burrows-Wheeler Aligner (BWA) software. Subsequently, single-nucleotide polymorphisms (SNPs) and insertions/deletions (indels) were analysed using the Genome Analysis Toolkit (GATK) software. Finally, the high-quality variants were annotated in major databases (e.g., dbSNP, 1,000 Genomes, ExAC, ESP and other frequency databases, as well as OMIM, HGMD, ClinVar, etc.) using variant annotation software developed independently. Provean, SIFT, PolyPhen2-HVAR, Polyphen2-HDIV, M-Cap, Revel, Mutation Taster, and MaxEntScan shear site prediction software were utilized to analyze the harmful effects of the variants and to screen out the variants that might have a detrimental effect on the protein structure.

As this study was concerned with the long-term outcome of SeLFE with neonatal/infantile onset, all children underwent developmental assessment at the last follow-up visit. Wechsler Preschool and Primary Scale of Intelligence (WPPSI) was used for children aged 4–6 years. The Wechsler Intelligence Scale for children (WISC) was used for patients aged 7–14 years. The Chinese version of the two instruments was revised by Gong Yaoxian of Hunan Medical University and divided into two parts: verbal and performance scales. The raw scores of the verbal and performance scales were recorded and then the verbal scale IQ (VIQ), performance scale IQ (PIQ) and full-scale IQ (FSIQ) were converted according to the age.

The criteria for selecting the first antiseizure medication were as follows: (1) diagnosed with epilepsy; (2) selection according to syndrome and seizure type (since the seizures of the children in this study were essentially partial seizures, the medication selected were OXC, LEV, PB, and topiramate (TPM); and (3) individual variations on the principle of medication selection. These variations included avoidance of antiseizure medication containing benzene rings, such as PB and carbamazepine, for allergic children; avoidance of LEV for unstable and easily agitated children; and avoidance of medication that are complicated to take and have side effects that require monitoring for parents with poorer cognition.

The initial medications and dosages used were as follows: used included LEV, 25∼60 mg/kg/day; PB, 3∼5 mg/kg/day; sodium valproate (VPA; oral solution, Sanofi Pharmaceutical Co., Ltd., 40 mg/ml), 20∼30 mg/kg/day; OXC,15∼30 mg/kg and TPM, 3∼5 mg/kg/day. If there was seizures recurrence after administration of the first medication, another medication was added or the initial medication was replaced until no more seizures occurred. Combined medication regimens included LEV + TPM, LEV + VPA, PB + OXC, PB + VPA, VPA + OXC, PB + TPM, LEV + VPA and OXC + TPM.

All the children were regularly followed up in the outpatient department (once every 4–6 months if their condition was stable and once every 2–4 weeks if their condition was unstable). The following follow-up data were included in this analysis: medication, missed medication, excessive medication, medication withdrawal without permission, serum medication concentration, and adverse medication reactions. Data on seizure status during the treatment period included in the analysis were the number of seizures, duration of each seizure, comorbidities and developmental assessment.

We collected basic information at birth, disease information, treatment information and follow-up information from the children's electronic medical records.

SPSS 24.0 statistical software was used for data analysis. If the continuous variables conformed to a normal distribution, they are expressed as the mean ± standard deviation. If the variables conform to non-normal distribution, they are expressed as the median (interquartile range). The categorical variables are expressed as frequencies and percentages. p < 0.05 was considered to indicate statistical significance.

Following screening on the basis of inclusion and exclusion criteria, 26 children were included in the final analysis (Table 1). Among these children, there were 13 males and 13 females. The mean follow-up was 54.0(49.0, 58.5) months. The mean age at seizure onset was 9.0 (4.8, 17.3) months. Twelve patients had a clearly relevant family history, including 9 patients with direct relatives and 3 patients with collateral relatives with SeLFE. Analysis of brain magnetic resonance imaging (MRI) results revealed no structural abnormalities. In the children's EEG examination results, abnormal EEG waves of different degrees in different regions were noted, mainly in the bilateral frontal and central areas (Table 2). Analysis of the developmental evaluation results revealed that 24 children developed normally, 1 patient had mild language and motor development delay, and 1 patient had mild motor development delay. Among the 26 children, 4 had fever triggered seizure.

Among the 26 patients, 3 (11.5%) had focal motor seizures, and 23 (88.5%) had focal to bilateral seizures. Seventeen patients had isolated seizures, and 9 patients had no isolated seizures. All 26 children had at least one seizure cluster. The age at the time of the first seizure cluster ranged from 1 day to 36 months. Among them, the first seizure cluster occurred within 1 month of birth (1 day, 30 days) in 2 children (7.7%), within 6 months of birth in 9 children (34.6%), within 12 months of birth in 15 children (57.7%), and after the age of 12 months in 11 children (42.3%). At the time of the first seizure cluster, the mean number of consecutive seizures was 3.0 ± 1.10. Sixteen children (61.5%) had a second seizure cluster episode, with a mean of 3.2 ± 2.1 consecutive seizures, and 9 children (34.6%) had a third seizure cluster episode.

All 26 children had undergone genetic testing. Pathogenic genes were absent for 16 children but present for 10 children. Among them, 3 patients were PRRT2 positive, 2 patients were SCN2A positive, 3 patients were KCNQ2 positive, 1 patient was ADA2 positive, and 1 patient was GRIN2A positive. Details of the genetic test results are shown in Table 3.

Outcomes Among the 26 children, seizures ceased without treatment for one child (no pathogenic gene). Among the other 25 children, the first-line medication was VPA for 2 children (8.0%), TPM for 2 children (8.0%), OXC for 3 children (12.0%), LEV for 11 children (44.0%), and PB for 7 children (28%). The seizures of 13 children (52.0%) ceased after the first-line medication treatment, whereas 12 children (48.0%) still experienced seizures. Among the 13 children whose seizures stopped, LEV was used for 5, VPA was used for 1, TPM was used for 2, OXC was used for 2, and PB was used for 3. Among the 10 children with pathogenic genes, seizures stopped for 4 children, including 2 who were PRRT2 positive, 1 who was SCN2A positive and 1 who was KCNQ2 positive, while the other 6 patients still experienced seizures. Of the 11 children who still had seizures, 4 (36.4%) received secondary monotherapy with OXC, and 7 received combination therapy, 1 (9.1%) with OXC, 2 (18.2%) with VPA, 1 (9.1%) with TPM and 1 (9.1%) with TPM and OXC.

Of the 26 patients, one patient had untreated seizures that ceased spontaneously, and 13 (50.0%) patients exhibited complete seizure control after the first monotherapy. In total, 12 (46.2%) patients did not achieve seizure control with their first monotherapy. Four patients achieved seizure control with secondary OXC monotherapy, and 4 patients achieved seizure control with the addition of OXC to their current therapy. An additional 3.8% of patients were administered OXC as add-on medication to a two-medication combination to achieve final seizure control. Ultimately, only one patient continued to experience sporadic seizures. These results suggest that seizures were controlled in all patients who preferred oxcarbazepine monotherapy and that the majority of patients who did not prefer oxcarbazepine had their seizures controlled with the addition of oxcarbazepine.

Seizures ceased in all the children with pathogenic genes. Only one medication was used for three patients who were PRRT2 positive, including 2 patients for whom OXC was used and 1 patient for whom PB was used. Among the 2 SCN2A-positive patients, only OXC was used for 1 patient, while LEV + OXC was used for the other patient. Among the 3 KCNQ2-positive patients, only LEV was used for 1 patient, PB + OXC was used for 1 patient, and PB + TPM + OXC was used for 1 patient. LEV-VPA was used for the ADA2-positive patient, while VPA + OXC was used for the GRIN2A-positive patient. At the final follow-up, only one child (female, 6 years old, no pathogenic gene detected) still experienced seizures (1–2 times per month), and the current treatment plan was OXC (20 mg/kg) + TPM(5 mg/kg)+ lamotrigine (LTG,5 mg/kg)+ketogenic diet. Although seizures stopped for all 10 children with pathogenic genes, these children continued to take medication. Among the 25 children whose seizures stopped, 15 were still taking medication orally, including 9 who were taking 1 medication, 5 who were taking 2 medication and 1 who was taking 3 medication. Details are shown in Table 4.