All patients undergoing resection of damaged brain tissue due to DRE were recruited from Wuhan Children's Hospital. The decision to perform lesionectomy was made according to the hospital's standard clinical protocols, which involves a comprehensive assessment of seizure characteristics (at least one seizure per month with a significant impact on quality of life, meeting the criteria for DRE), disease progression, electroencephalogram (EEG) findings, and structural magnetic resonance imaging (MRI) results, in order to accurately localize the epileptogenic or suspected epileptogenic focus and network. Patient guardians’ consent and consensus from the epilepsy surgery conferences were also considered. This study included patients who underwent surgical resection for DRE between September 2021 and January 2023. Inclusion criteria were seizure onset before 14 years of age and guardian consent for comprehensive genetic testing; Patients with incomplete pathological specimens or other factors that could interfere with the assessment of brain lesions were excluded. Based on these criteria, a cohort of 21 pediatric DRE patients was established for subsequent analyses.

The study was approved by the Ethics Committee of Wuhan Children's Hospital to ensure compliance with ethical standards. Targeted WES was performed to comprehensively investigate the underlying etiologies of these patients, all of whom had previously undergone clinical and neuroimaging assessments.

Prior to participation, written informed consent was obtained from all patients or their legal guardians. The consent forms clearly explained the purpose of the study and confirmed their understanding and agreement regarding data analysis and management. The study strictly adhered to ethical and regulatory guidelines to safeguard the rights and welfare of all participants.

Brain tissue was collected during the surgical procedure, specifically from the resected pathological regions. Upon collection, the samples were promptly fixed in formalin and subsequently embedded in paraffin. The paraffin-embedded tissues were sectioned into approximately 5 µm-thick slices, which were then stained and examined under a microscope to assess tissue architecture and cellular morphology.

In this study, peripheral blood was collected from 21 pediatric patients with DRE using EDTA anticoagulant tubes, with 2 mL obtained per patient. Genomic DNA was extracted from the blood samples using the DNeasy Blood & Tissue Kit (Cat. No. 69504, QIAGEN, Germany), and WES libraries were prepared using the QIAseq Human Exome Kit (Cat. No. 333939, QIAGEN, Germany) following the manufacturer's instructions. High-throughput sequencing was performed on the Illumina HiSeq 2000 platform at a sequencing depth of 100–200×.

Resected lesional brain tissues were immediately immersed in RNAlater, incubated overnight at 4  °C, and then stored at −80  °C. Genomic DNA was subsequently extracted using the AllPrep DNA/RNA/miRNA Universal Kit (Cat. No. 80224, QIAGEN, Germany), and WES libraries were constructed with the QIAseq Human Exome Kit according to the manufacturer's protocols. High-throughput sequencing was performed on the Illumina HiSeq 2000 platform with a sequencing depth of 200–300×. The variant allele frequency (VAF) threshold was set at 1%–35% to identify somatic mosaic variants, while variants with VAF >35% were classified as heterozygous somatic mutations. To minimize false positives, analyses were performed at high sequencing depth (≥200–300×) with stringent quality control, including filtering of low-quality bases and removal of duplicate reads.

The resulting sequencing data were aligned to the hg19 reference genome for bioinformatic analyses, enabling the identification of single-nucleotide variants (SNVs), insertions, and deletions (indels), following previously described methods (17). Subsequently, the clinical significance of the detected variants was evaluated in accordance with the American College of Medical Genetics and Genomics (ACMG) guidelines (18) and the somatic variant interpretation guidelines jointly developed by AMP (Association for Molecular Pathology), ASCO (American Society of Clinical Oncology), and CAP (College of American Pathologists) (19).

To further verify the somatic mosaic variants identified in the lesional brain tissue, droplet digital PCR (ddPCR) was performed. An appropriate amount of frozen lesional brain tissue was used for genomic DNA extraction with the Blood/Cell/Tissue Genomic DNA Extraction Kit (centrifuge column type, YDP304, TIANGEN, China), following the manufacturer's instructions. The concentration of the extracted DNA was measured using the dsDNA HS Assay Kit (12640ES60, YEASEN, China), and DNA purity was assessed with a NanoDrop spectrophotometer. DNA samples with a concentration >5 ng/μL and an OD_260/280 ratio between 1.6 and 2.1 were considered acceptable for subsequent analysis.

Droplet digital PCR (ddPCR) was performed using the QX200 AutoDG Droplet Digital PCR System (Bio-Rad, Hercules, CA, USA) following the manufacturer's instructions. Each 20 µL ddPCR reaction contained 10 µL ddPCR Supermix for Probes (without dUTP), 1 µL of each 10 µM primer, 1 µL of 10 µM probe, 2 µL DNA template, and 5 µL nuclease-free water. Samples were loaded into a 96-well plate, centrifuged briefly, and processed with the ddPCR Automated Droplet Generator. After droplet formation, plates were sealed and subjected to PCR amplification using the following cycling conditions: initial denaturation at 95 °C for 10 min; 40 cycles of 94 °C for 30 s (denaturation) and 55 °C for 60 s (annealing); final extension at 98 °C for 10 min; followed by storage at 4 °C. The 96-well plate was then read using the QX200™ Droplet Reader, and data were analyzed using the accompanying software to generate quantitative results. The primer and probe sequences used for ddPCR are listed in Supplementary Table 1.

A total of 38 patients diagnosed with DRE who had undergone resective epilepsy surgery were initially included in this study. Ten patients were excluded because their parents did not provide consent for genetic testing. Preliminary assessment of the quality of resected brain tissue led to the exclusion of two additional cases that failed post-sequencing quality control. The remaining 26 patients underwent centralized evaluation of preoperative magnetic resonance imaging (MRI) and neuropathological findings. Among them, five cases were excluded according to the inclusion criteria—two due to MRI evidence of hypoxic–ischemic injury and three because the age at seizure onset exceeded 14 years.

The final cohort therefore consisted of 21 pediatric patients with DRE. The patients’ ages ranged from 0.2 to 10.7 years, with a mean age of 5.2 years. Eleven were male and ten were female. MRI revealed no detectable lesions in nine cases, while multiple abnormalities were identified in eleven cases. All patients underwent surgical resection of epileptogenic brain tissue, followed by a one-year postoperative follow-up. During this period, twelve patients (12/21, 57.1%) remained seizure-free, and nine patients (9/21, 42.9%) showed a marked reduction in seizure frequency, indicating a significant improvement after surgery (Table 1).

To investigate the genetic etiology of the 21 pediatric patients with DRE, all patients first underwent clinical evaluations, MRI, electroencephalography (EEG), and histopathological examination (Figure 1 and Supplementary Table 2). Based on these assessments, one patient was clinically diagnosed with tuberous sclerosis complex (TSC), yielding a diagnostic rate of 4.8% (1/21). Histopathological examination of resected brain tissue identified focal cortical dysplasia type IIb (FCD IIb) in patient 21, increasing the diagnostic rate to 9.5% (2/21) (Figure 1 and Supplementary Table 2). The remaining 19 patients did not receive a definitive genetic diagnosis (Supplementary Table 2). Subsequently, WES was performed on peripheral blood samples from all patients, revealing pathogenic or likely pathogenic variants in patients 2, 7, 14, and 20, involving DEPDC5, and TSC2, with a diagnostic yield of 19.0% (4/21). WES of the resected brain tissue identified pathogenic or likely pathogenic variants in patients 2, 7, 8, 14, 16, and 20, involving DEPDC5, BRAF, and TSC2, resulting in a markedly higher diagnostic yield of 28.6% (6/21) (Figure 1 and Table 2). The increased diagnostic rate in brain tissue WES was primarily attributable to the detection of somatic mosaic mutations in patients 8 and 16, which were not detected in peripheral blood (Table 2).

Patient 2 began experiencing unprovoked seizures at the age of 5, characterized by sudden upward eye deviation, generalized hypotonia, unresponsiveness, and limb flaccidity, with each episode lasting approximately 10 s and resolving spontaneously. Subsequently, seizure frequency gradually increased, becoming refractory to antiepileptic drugs, with up to seven episodes per day. MRI revealed no obvious abnormalities, while EEG demonstrated frequent multifocal spikes, polyspikes, spike-and-slow-wave complexes, and fast-wave rhythms, predominantly in the left anterior frontal region, with increased discharges during sleep and frequent partial seizures during wakefulness. Histopathological examination of the resected brain tissue showed clear gray-white matter demarcation, preserved cortical lamination, mildly disorganized neuronal arrangement, some neuronal atrophy or degeneration, and notable “neuronophagic” changes; gliosis was prominent without significant cellular atypia. Following surgical resection of the lesion, the patient experienced a marked reduction in seizure frequency (Supplementary Table 2). WES of both peripheral blood and the lesion tissue identified a DEPDC5 deletion mutation (NM_001242896.3:c.793_797del:p.Glu265MetfsTer9), which was classified as likely pathogenic (LP) according to the ACMG guidelines (Table 2).

Patient 7 began experiencing unexplained seizures at the age of 2, characterized by vacant staring, limb stiffness without obvious tremor, and slowed responsiveness. The child's mental status returned to normal after each episode, and no associated symptoms such as fever, vomiting, or diarrhea were reported throughout the disease course. MRI showed no obvious abnormalities, while EEG revealed abundant spikes, sharp waves, polyspikes, spike rhythms, fast-wave rhythms, spike-and-slow-wave complexes, and irregular slow waves predominantly in the right anterior frontal region. Multiple partial seizures and frequent electrical discharges were recorded during both awake and sleep states. Histopathological examination of the resected brain tissue demonstrated preserved cortical lamination, gliosis, neuronal degeneration, and notable neuronophagic changes. Following surgical resection of the lesion, the patient remained seizure-free (Supplementary Table 2). WES of both peripheral blood and the lesion tissue identified a DEPDC5 splice-site variant (NM_001242896.3:c.1287 + 1G > C), which was classified as LP according to the ACMG guidelines (Table 2).

Patient 8 began to experience intermittent convulsions of unknown cause at the age of 5. The seizures were characterized by upward eye deviation and leftward head tilt with backward extension, without limb tremors. During the episodes, the patient was unresponsive, with each seizure lasting from several seconds to more than 10 s, and without accompanying symptoms such as fever or vomiting. Initially, the seizures occurred once every 7–8 days, but the frequency increased to once every 3–4 days as drug treatment proved ineffective. MRI examination revealed a localized softening lesion in the left occipital lobe. EEG recordings showed frequent spikes and polyspike–slow waves in the left occipital and posterior temporal regions, either sporadic or in clusters, with increased abnormal discharges during sleep and a few generalized spike and polyspike-slow waves. One partial seizure was captured during monitoring. Histopathological examination of the resected lesion revealed a mixture of atypical glial cells and neurons. The glial cells were densely arranged and enlarged in size, with microcystic changes and small foci of calcification, suggesting a possible tumor-like lesion in the left occipital lobe. Based on these features, the lesion was classified as a diffuse astrocytoma (Figures 2B,C and Supplementary Table 2). After surgical resection of the lesion, the patient's seizures were markedly improved, and no further seizures were observed during postoperative follow-up. WES of both peripheral blood and lesional brain tissue identified a missense variant in the NPRL2 gene (NM_006545.5:c.1037C > T:p.Thr346Ile); however, according to ACMG guidelines, this variant was classified as a variant of uncertain significance (VUS). In addition, a somatic mosaic BRAF variant (NM_001354609.2:c.1799T > A:p.Val600Glu) was detected exclusively in the lesional brain tissue, with a VAF of 20.2% (Table 2 and Figure 2A). Based on ACMG and somatic variant interpretation guidelines, this variant was classified as pathogenic (P).

Patient 14 began experiencing epileptic seizures at two months of age with an unknown etiology. The seizures manifested as clusters of nodding, bending forward, and head-shaking movements, often accompanied by crying. Antiepileptic drug treatment showed no significant improvement. MRI examination revealed widened sulci on the surface of both cerebellar hemispheres. EEG recordings showed frequent multifocal and generalized discharges, predominantly in the left frontotemporal region, with 10 partial seizures and 2 electrographic seizures captured during monitoring. Histopathological analysis of the resected brain tissue showed poorly demarcated gray and white matter structures, localized glial proliferation, neuronal degeneration, neuronophagia, and balloon cells. Focal proliferation of glial and neuronal cells formed nodular lesions. Following surgical resection of the lesion, the frequency of seizures was markedly reduced (Supplementary Table 2). WES of both peripheral blood and lesional brain tissue identified a missense variant in the TSC2 gene (NM_000548.5:c.3598C > T:p.Arg1200Trp), which was classified as LP according to ACMG guidelines (Table 2).

Patient 16 had experienced seizures since the age of 3 years, with no apparent precipitating factors. The seizures were characterized by brief eyelid twitching, nodding, and rapid limb jerks, often accompanied by a backward tilt of the body. These episodes occurred daily, without associated fever, vomiting, or diarrhea. Antiepileptic medications failed to achieve adequate seizure control. MRI revealed a cystic cortical lesion in the right temporal lobe. EEG recordings demonstrated frequent multifocal and generalized discharges, including spikes, polyspikes, sharp waves, spike-slow waves, and slow waves. Histopathological examination of the resected lesion confirmed a ganglioglioma in the right temporal lobe (Figures 2E,F and Supplementary Table 2). Following surgical resection, the frequency of seizures markedly decreased. WES of peripheral blood showed a negative result. In contrast, WES of the lesional brain tissue detected a somatic mosaic variant in BRAF (NM_001354609.2:c.1799T > A:p.Val600Glu) with a VAF of 13.5% (Table 2 and Figure 2D). According to the ACMG and the somatic variant interpretation guidelines, this variant was classified as pathogenic (P).

Patient 20 began experiencing seizures at the age of 2 years and 8 months, with no identifiable etiology. The seizures were characterized by eye blinking triggered by fear, limb stiffness or tremor, occasionally accompanied by head and eye deviation, and mild limb twitching. Each episode lasted from several tens of seconds to a few minutes and occurred frequently. MRI revealed tuberous sclerosis. EEG showed frequent generalized and multifocal discharges, most prominent in the bilateral occipital regions, which increased during sleep. Histopathological examination of the resected brain tissue demonstrated gliosis, neuronal degeneration, and the presence of neuronophagia and balloon cells (Supplementary Table 2). Following surgical resection of the lesion, the patient remained seizure-free. WES of both peripheral blood and lesional brain tissue identified a nonsense variant in TSC2 (NM_000548.5:c.3532C > T:p.Gln1178Ter) (Table 2). According to the ACMG guidelines, this variant was classified as pathogenic.