The genetic spectrum features of 2261 Chinese children with epilepsy and intellectual disability

Abstract

Background: Epilepsy (EP) and intellectual disability (ID) are two highly correlated diseases that seriously impact neurodevelopment in children. Precision diagnosis of EP and ID remains challenging due to their clinical and genetic heterogeneity, necessitating a profound understanding of disease characteristics.

Methods: We provide a clinical and genetic landscape of 2261 Chinese patients performed chromosome microarray analysis (CMA) or next-generation sequencing to uncover causal copy number variants (CNVs) or single-nucleotide variants (SNVs). Patients were stratified into three groups: EP (374 cases), ID (863 cases), and EP + ID (1024 cases).

Results: We reported a 24.3% diagnostic yield from 496 causal CNVs and SNVs, including 182 novel variants, in which updated 33 previously reported VUS. Significant intergroup differences emerged: EP patients were predominantly caused by autosomal dominant SNVs, showing the highest rates of incomplete penetrance and family history. ID patients were more likely caused by CNVs and autosomal recessive SNVs, with the highest genetic heterogeneity. EP + ID patients displayed the earliest onset ages and highest diagnostic yields. We prioritized genes by diagnostic efficiency and revealed that X-linked SNVs disproportionately affected females, particularly in the EP + ID group, under current diagnostic paradigms. This real-world dataset informs genetic counseling, testing strategies, precision therapies, and long-term management for EP/ID.

Conclusions: The clinical and genetic profiles from this study provided a reliable baseline reference for diagnosing EP and ID.

Keywords: Clinical assessment; Epilepsy; Intellectual disability; Molecular diagnosis; Sex difference.

Fig. 1

Genetic feature of children with genetic EP and/or ID. a Variant classification summary. b Inheritance pattern of confirmed diagnosis across subgroups, AD inheritance dominated all groups, with X-linked contributions at 3% (isolated EP), 13% (isolated ID), and 14% (EP + ID). X-linked findings do not explain the male predominance. c Statistical comparison of inheritance patterns between groups, no AR-CNV was identified. d Gender difference, although male patients count 1.56 times more than female patients, no statistical difference in the diagnostic yield was observed between males and females (left). X-linked positivity in females was expected to be higher than in males (1.97-fold for X-linked-positive vs. autosomal-positive, middle, and 2.27-fold for X-linked-diagnostic yield, shown in Additional file 1: Table 7), mainly contributed by the X-linked gene in the EP + ID group (right). *P < 0.05, **P < 0.01

Fig. 2

Etiological architecture of EP and/or ID. a Case counts for CNV/SNV testing: positive (red) and negative (orange: isolated EP; green: isolated ID; blue: EP + ID). b CNV composition, 157 (likely) pathogenic CNVs included 89 syndromes and 68 non-syndromic CNVs. c Top 10 genes by case count per subgroup. d Etiological composition, top 5 causes per subgroup, *including SNV and deletion of UBE3A, causing Angelman syndrome. e Gene capture range-outcome calculation, number of confirmed cases divided by the CDS length of each gene, ranked highest to lowest (X-axis), and cumulatively summed to model diagnostic returns. Each point represents an individual gene, with the top 20 genes labeled. The curve demonstrates diminishing marginal returns as more genes are involved in the NGS test

Fig. 3

Clinical features of children with genetic EP and/or ID. a Onset age in 3 groups, EP + ID patients had a much earlier onset age than isolate EP and isolate ID patients (left), the onset age of the top 5 in 3 groups (middle to right), PWS: Prader-Willi syndrome, WBS: Williams-Beuren syndrome, UBE3A: including SNV and deletion of UBE3A causing Angelman syndrome. b Diagnostic yield difference in all patients grouped by CNV test (606 patients), SNV test (876 patients), and both tests (779 patients), this difference is mainly contributed by the EP + ID group, shown in Additional file 1: Table 11. c Medication profiles for all patients diagnosed with variants in the top 10 genes, the Y-axis indicates the medication count per gene-confirmed patient, abbreviations of antiepileptic drugs and therapies: VPA, valproic acid; TPM, topiramate; CLB, clobazam; LEV, levetiracetam; KD, ketogenic diet; CZP, clonazepam; LCM, lacosamide; PER, perampanel; NZP, nitrazepam; CBZ, carbamazepine; OXC, oxcarbazepine; LTG, lamotrigine; VGB, vigabatrin; ACTH, adrenocorticotropic hormone; PB, phenobarbital; IVMP, intravenous methylprednisolone; ZNS, zonisamide