A retrospective chart review was conducted on pediatric patients with NDDs who had undergone WES and/or WGS at the Pediatric Neurology Clinic of Busan Paik Hospital between March 2018 and November 2024. Patients were included if they met clinical diagnostic criteria for one or more NDDs including GDD, ASD, ID, and epilepsy accompanied by developmental delay or ID, and remained without a definitive etiology after standard clinical and diagnostic evaluations.

GDD was defined as a delay in 2 or more developmental domains (gross motor, fine motor, social and personal, language, and cognition) in young children, as assessed using validated developmental tools, including the Korean Developmental Screening Test for Infants and Children and the Bayley Scales of Infant and Toddler Development, Third Edition [11,12]. ID was defined as an IQ score of less than 70 in children aged 5 years or older, with concurrent impairments in cognitive functioning and adaptive behavior, as measured by the Korean Wechsler Intelligence Scale [12,13]. ASD was diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition criteria [13]. Only patients with epilepsy in whom seizures were accompanied by developmental delay or ID were included; patients with isolated epilepsy without neurodevelopmental impairment were excluded.

All patients underwent comprehensive clinical evaluations, including neurologic examination, standardized developmental assessments, and neuroimaging. WES or WGS was performed in patients with unexplained NDDs after inconclusive results from prior diagnostic investigations. The decision to proceed with genomic sequencing was made by pediatric neurologists based on phenotype severity, clinical complexity, and the absence of an established molecular diagnosis.

All participants and their families received formal pretest genetic counseling, which included a detailed explanation of the potential benefits, limitations, and risks associated with genomic testing. Informed consent for genetic analysis was obtained in accordance with institutional guidelines. This study was approved by the Institutional Review Board of Busan Paik Hospital (approval No. BPIRB 2024-11-011).

Clinical data, including demographic characteristics, phenotypic features, laboratory results, EEG findings, and brain magnetic resonance imaging (MRI) results, were systematically collected from electronic medical records. Family history was defined as the presence of a definite history of NDDs in first- or second-degree relatives. Phenotypes of interest included GDD, epilepsy, facial dysmorphism, and other neurological or syndromic features. Abnormal growth and head circumference were defined as measurements below or above 2 standard deviations from the mean for age- and sex-matched populations. Multiple anomalies are defined as the presence of 2 or more major congenital malformations affecting different organ systems.

Prior diagnostic evaluation included genetic and metabolic investigations, such as conventional karyotyping, CMA, FMR1 gene testing for fragile X syndrome, methylation-specific polymerase chain reaction (PCR) for Prader-Willi and Angelman syndromes, MECP2 gene analysis, targeted multigene panels, and relevant biochemical assays.

WES and WGS were implemented as second- or third-tier diagnostic tests following inconclusive results from earlier investigations. Some patients underwent WES as a first-line test based on clinical urgency, phenotype severity, or test accessibility. In cases where WES did not yield a definitive molecular diagnosis, additional genetic tests were performed. CMA was selectively performed when clinically indicated, especially in cases where copy number variant (CNV) analysis was not sufficiently captured by initial sequencing and CMA had not been conducted before WES. Additional targeted assays, such as trinucleotide repeat expansion testing, were performed when indicated by specific clinical features identified after rephenotyping.

Genomic DNA was extracted from peripheral blood leukocytes for WGS and from buccal swabs for WES for each proband. WES was performed at 3billion, Inc. (Korea) using the IDT xGen Exome Research Panel v2 (Integrated DNA Technologies, USA) for exome capture. Sequencing was performed using the Illumina NovaSeq 6000 platform (Illumina, USA). WGS was performed by Macrogen, Inc. (Korea). DNA libraries were prepared using the TruSeq DNA PCR-free sample preparation kit (Illumina), and sequencing was performed using the Illumina NovaSeq 6000 system. WGS sequencing data analysis was performed at 3billion, Inc. to identify a comprehensive range of variants, including single-nucleotide variants, small insertions/deletions, CNVs, structural variants, repeat expansions, mobile element insertions, and regions of homozygosity [14,15].

Variants were annotated, filtered, and prioritized using EVIDENCE, 3billion’s proprietary system with a daily-updated database, variant classification module, and symptom similarity scoring module [14]. The variants were classified according to the guidelines of the American College of Medical Genetics and Genomics (ACMG) and Association for Molecular Pathology (AMP) into 5 categories: pathogenic, likely pathogenic, variant of uncertain significance (VUS), likely benign, or benign [16-18]. The candidate variants were manually reviewed by clinical geneticists and selected for reporting based on their correlation with the patient’s phenotype.

It should be noted that earlier WES analyses performed during 2018–2020 were based on capture-based exome sequencing platforms and analytical pipelines that were primarily optimized for single-nucleotide variants and small insertions/deletions. As such, the sensitivity for detecting CNVs, repeat expansions, and certain structural variants was limited, and these variant classes were not systematically evaluated at that time.

Sanger sequencing was performed for variant confirmation when sufficient DNA sample was available. Parental and familial segregation analyses were performed whenever possible. For the purposes of analysis, patients were classified as “positive” if a causative variant was identified by WES or WGS, and as “negative” if no causative variant was identified by these sequencing methods (Table 1). Novel variants were defined as those absent from ClinVar and without prior disease-associated reports in the literature at the time of analysis.

All statistical analyses were performed using IBM SPSS Statistics ver. 29.0 (IBM Co., USA). Continuous variables were reported as means with standard deviations, whereas categorical variables were expressed as percentages. For comparison of continuous variables, Student t test was employed. Categorical variables were analyzed using Pearson chi-square test or Fisher exact test, with the latter applied when the expected cell frequency was <5. Two-tailed null hypotheses of no difference were rejected if the P values were <0.05.

A total of 70 patients were enrolled during the study period. After excluding individuals with infantile seizures, paroxysmal dyskinesia, or skeletal deformities in the absence of neurodevelopmental delay, 64 patients (30 males [46.9%], 34 females) were included in the final analysis (Fig. 1). The mean age at symptom onset (mean±standard deviation) was 1.2±2.0 years, and the mean age at WES or WGS testing was 4.6±5.0 years (Table 1).

The most prevalent clinical phenotype was GDD, observed in 57 of the 64 patients (89.1%). Epilepsy was present in 25 patients (39.1%) and ASD in one (1.6%) (Table 1). Hypotonia was noted in 32 patients (50.0%) and short stature in 24 (37.5%). Abnormal movements, including ataxia, tremor, and myoclonus, were reported in 5 patients (7.8%).

Brain MRI was performed in 62 patients, of whom 33 (53.2%) demonstrated structural abnormalities, including cerebellar dysplasia, reduced cerebral white matter volume, cerebellar vermis hypoplasia, dysgenesis or agenesis of the corpus callosum, diffuse cerebral atrophy, ventriculomegaly, Dandy-Walker variant, white matter injury, cortical dysplasia, and polymicrogyria. EEG was conducted in 43 patients, of whom 22 (51.2%) demonstrated abnormal findings, including focal epileptiform discharges, background slowing, and generalized slow-wave activity. An overview of the genetic investigations conducted in our cohort is presented in Table 1.

WES was performed in 56 patients and WGS in 12 (Fig. 1). Among those who underwent WES, a molecular diagnosis was achieved in 21 individuals (37.5%). Of the 12 patients who underwent WGS, 4 (33.3%) were diagnosed with a pathogenic or likely pathogenic variant. WES was performed as a first-tier test in 6 patients (patient numbers 4, 6, 11, 14, 17, and 24) of these positive cases. Of the 39 patients who remained negative, 10 had undergone WES as an initial investigation. Among the 4 individuals with negative WES results, subsequent WGS revealed a pathogenic variant in 1 patient, leading to a genetic diagnosis. Twelve patients, along with 2 affected siblings and 23 unaffected family members, underwent WGS as part of the National Project of Bio Big Data. Among these, 10 families were enrolled as proband-parent trios, one family as a proband-sibling-parent quartet, and one family as a proband-sibling-single-parent trio. All patients who underwent WGS had previously received genetic evaluations, including WES or other genetic tests such as conventional karyotyping, CMA, FMR1 gene testing for fragile X syndrome, MECP2 gene analysis, and targeted multigene panels. Periodic reanalysis of WES/WGS data was not performed in this study.

Overall, 25 of the 64 patients (39.1%) received a molecular diagnosis through WES or WGS. Of the variants identified through WES/WGS, 16 were classified as pathogenic, 10 as likely pathogenic, and one as a VUS (Table 2). Of the 25 genetically and clinically confirmed diagnoses, inheritance patterns included autosomal dominant in 13 cases, X-linked in 9 cases, and autosomal recessive in 3 cases. Sanger sequencing in 21 probands and their family members confirmed 14 de novo variants. Notably, 15 of the pathogenic or likely pathogenic variants were novel, 7 of which had been previously reported in our earlier study [19-21]. Comprehensive phenotypic data of patients with pathogenic or likely pathogenic variants are provided in Supplementary Table 1. Detailed information on population allele frequencies, ClinVar reporting status, and the ACMG/AMP criteria applied for variant classification is summarized in Supplementary Table 2.

Patient 19 harbored a novel pathogenic variant presented with distinctive clinical features. The patient was born at 40 weeks and 5 days of gestation with a birth weight of 3.06 kg. At 9 months of age, the patient developed focal impaired awareness seizures characterized by altered consciousness, cyanosis, and vomiting, with episodes lasting >1 hour, leading to a diagnosis of epilepsy. The patient exhibited GDD and progressive microcephaly, with head circumference declining to below the 3rd percentile, despite being within the normal range at birth. Dysmorphic facial features included synophrys, hypertelorism, narrow nasal bridge, bulbous nasal tip, and thin upper lip. Brain MRI revealed ventricular dilatation and mild cerebral atrophy (Fig. 2A and B).

Initial genetic evaluations, including conventional karyotyping, CMA, MECP2 gene sequencing, and an epilepsy gene panel, failed to identify a causative variant. Subsequent WGS identified a novel heterozygous missense pathogenic variant in CDC42 (NM_001791.4:c.67T>C; p.Tyr23His), leading to a diagnosis of Takenouchi-Kosaki syndrome. This is the first reported case of Takenouchi-Kosaki syndrome in Korea. The variant was confirmed to be de novo, and appropriate genetic counseling was provided to the family. The patient was also registered in the Korean National System for Newly Identified Ultra-Rare Disorders.

A case of clinical interest is that of patient number 23, a female infant born at term. Prenatal ultrasonography had revealed corpus callosum hypoplasia and mega cisterna magna. Postnatal MRI confirmed corpus callosum hypoplasia and periventricular nodular heterotopia (Fig. 2C and D). Initial genetic investigations, including conventional karyotyping and CMA, performed during admission to the neonatal intensive care unit, were unremarkable. Because of the accompanying developmental delay, WES was performed, which identified a VUS in FLNA (NM_001456.4:c.7895G>T; p.Ser2640Ile). Although neither parent exhibited overt clinical features suggestive of an X-linked disorder, a segregation analysis was conducted under the presumption of a potential de novo variant. However, an identical variant was detected in the asymptomatic mother. Upon further clinical enquiry, the mother disclosed a history of mild developmental delay. A subsequent brain MRI revealed periventricular nodular heterotopia, consistent with the daughter’s neuroimaging findings. Based on phenotypic concordance and segregation evidence, the variant was reclassified as likely pathogenic, establishing a molecular diagnosis.

A comparison between patients who were genetically diagnosed using WES/WGS and those who remained negative revealed notable clinical differences (Table 1). Among the 64 patients included in this study, 25 (39.1%) received a definitive molecular diagnosis via WES or WGS.

The proportion of male patients was lower in the positive group than in the negative group (32.0% vs. 56.4%), although the difference was not significant (P=0.074). A positive family history of NDDs was more common in the positive group (20.0%) than in the negative group (2.6%, P=0.030), whereas a history of prematurity was more frequent in the negative group (33.3% vs. 8.0%, P=0.032). Among WES/WGS-positive cases, 3 patients had a family history of GDD/ID, and 2 had a family history of epilepsy accompanied by GDD/ID. Among WES/WGS-negative cases, 1 patient had a family history of epilepsy accompanied by GDD.

There were no significant differences between groups regarding age at symptom onset, timing of WES/WGS, or major clinical features such as developmental delay, epilepsy, microcephaly, hypotonia, short stature, facial dysmorphism and multiple anomalies. Neuroimaging and EEG abnormalities were comparably distributed between the groups, and no significant difference was observed in the use of WES or WGS.

Among the cohort of patients who remained negative following WES, 3 individuals were later diagnosed through CMA or targeted genetic testing.

The first case involved an 8.7-year-old girl with epilepsy, GDD, dysmorphic facial features, hypotonia, and intellectual disabilities. CMA performed 2 months after nondiagnostic WES revealed a 4.5-Mb deletion at 1q44 (chr1:244,737,375–249,224,684), consistent with 1q44 microdeletion syndrome. The second case involved an 11.8-year-old girl with epilepsy, GDD, dysmorphic features, and short stature. WES failed to identify pathogenic variants, but subsequent CMA detected a 10-Mb duplication at 15q11.2–q13.3 (chr15:22,770,421–32,915,723) involving 4 copies, leading to a diagnosis of 15q duplication syndrome. Importantly, both patients had undergone prior genetic evaluations, including conventional karyotyping, Prader-Willi/Angelman syndrome methylation-specific PCR, or targeted gene panel testing, all of which failed to yield a diagnosis prior to WES. The third patient was a 14.5-year-old boy with GDD, ID, dysmorphic features, spasticity, cerebral palsy, and macrocephaly. Brain MRI revealed bilateral lateral ventricular dilatation and suspected white matter volume loss. Despite negative PTEN gene analysis and WES results, further clinical evaluation revealed a high-arched palate, stiff tongue, and suspected handgrip myotonia prompting DMPK trinucleotide repeat analysis, which confirmed >150 CTG repeats, establishing a diagnosis of childhood-onset myotonic dystrophy type 1. In this case, myotonic dystrophy was not initially suspected, as the patient did not exhibit classic clinical features of the disorder at the time of the initial evaluation. The patient had a history of prematurity and periventricular leukomalacia, which likely contributed to central nervous system manifestations, including spasticity, and may have confounded the initial phenotypic interpretation. These cases underscore the importance of continued phenotypic re-evaluation and the utility of sequential genetic testing beyond WES in pediatric patients with complex neurodevelopmental phenotypes.