Genetic Convergence of Autism Spectrum Disorder and Epilepsy: From Molecular Mechanisms to Precision Therapeutics.

Genetic Convergence of Autism Spectrum Disorder and Epilepsy: From Molecular Mechanisms to Precision Therapeutics

Autism spectrum disorder (ASD) and epilepsy frequently co-occur, with the clinical overlap particularly evident in individuals with intellectual disability, developmental impairment, and genetically defined neurodevelopmental disorders. Advances in genomic diagnostics have revealed substantial genetic and molecular convergence between ASD and epilepsy. This convergence encompasses both recurrent copy-number variants (CNVs), including 15q11–q13, 16p11.2, and 22q11.2 rearrangements, and sequence-level alterations in genes implicated in both epilepsy and ASD-related neurodevelopmental phenotypes, including SCN1A, SCN2A, GRIN2A, SYNGAP1, STXBP1, TSC1/TSC2, MECP2, and UBE3A. These diverse genetic alterations affect gene dosage, ion-channel function, synaptic organization and neurotransmission, intracellular signaling, and gene regulation, ultimately converging on disruption of circuit homeostasis, excitation–inhibition regulation, synaptic plasticity, and network dynamics. Importantly, the relationship between epilepsy and ASD is neither uniform nor necessarily causal: epileptic activity may aggravate neurodevelopmental dysfunction in some disorders, whereas in others, seizures and autistic features represent partially independent consequences of a shared genetic defect. This heterogeneity has major therapeutic implications. Emerging precision approaches increasingly incorporate variant-specific functional effects and include ion-channel modulation, pathway-directed treatments, metabolic correction, and gene- or RNA-targeted therapies. However, successful seizure control does not necessarily translate into normalization of neurodevelopmental outcomes, emphasizing the importance of developmental timing and of distinguishing seizure-directed benefit from modification of the underlying neurodevelopmental disorder. This review integrates clinical, genetic, molecular, cellular, and network-level evidence linking ASD and epilepsy and examines how these convergent mechanisms may guide precision therapeutics and future early-intervention strategies.

1. Introduction

Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental condition defined according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM-5-TR), by persistent deficits in social communication and social interaction across multiple contexts, together with restricted, repetitive patterns of behavior, interests, or activities. A formal diagnosis requires the presence of all three social-communication criteria and at least two of the four restricted/repetitive behavior criteria; symptoms must be present in the early developmental period, cause clinically significant impairment, and not be better explained by intellectual disability or global developmental delay [1]. ASD may co-occur with intellectual disability; however, social-communication deficits must exceed those expected for the individual’s general developmental level.

Throughout this review, the term ASD is used when referring to formally diagnosed autism spectrum disorder or to evidence specifically derived from ASD cohorts. The term ‘autistic features’ refers to reported social-communication and/or restricted or repetitive behavioral characteristics resembling components of the ASD phenotype when the available evidence does not establish a formal ASD diagnosis. In genetically defined neurodevelopmental disorders, particularly severe developmental and epileptic encephalopathies (DEEs), such features may occur within a broader neurodevelopmental phenotype and should not be interpreted as synonymous with formally diagnosed ASD. Accordingly, the terms ‘autistic features,’ ‘ASD-related neurodevelopmental dysfunction,’ and ‘broader neurodevelopmental phenotype’ are used throughout the review according to the level and nature of the available phenotypic evidence.

Beyond its defining behavioral features, ASD is frequently accompanied by neurological and neurodevelopmental comorbidities that contribute substantially to clinical heterogeneity and long-term outcome. Among these, epilepsy is particularly important because of its association with intellectual disability (ID), developmental impairment, and increased clinical complexity [2,3].

The relationship between ASD and epilepsy extends beyond simple clinical co-occurrence. Advances in genomic medicine have revealed substantial overlap in their genetic architecture, with pathogenic variation affecting ion channels, synaptic proteins, neurotransmitter receptors, intracellular signaling pathways, and regulators of gene expression and chromatin organization. Many of these molecular defects can simultaneously influence neuronal excitability and the developmental organization of neural circuits, providing a biological framework through which epilepsy and ASD-related neurodevelopmental dysfunction may emerge from a shared molecular substrate [4,5].

Importantly, the relationship between the two phenotypes is neither uniform nor necessarily unidirectional. In some genetic disorders, epileptic activity may itself interfere with cognition and development, whereas in others, seizures and autistic features may represent partially independent consequences of an underlying molecular defect. Developmental timing, variant-specific functional effects, cell-type vulnerability, and compensatory network mechanisms further modify this relationship. Understanding this heterogeneity is increasingly important as genetic diagnosis begins to inform mechanism-based and molecularly targeted therapies [4,6].

This review examines the genetic convergence between ASD and epilepsy across molecular, cellular, and network levels. This review focuses on representative genetic disorders that illustrate major mechanistic pathways, evaluates how these mechanisms converge on altered circuit homeostasis and network function, and discusses their implications for precision therapeutics. Particular attention is given to the variable relationship between seizure control and neurodevelopmental outcomes and to the extent to which emerging precision therapies may modify the underlying disease process rather than seizures alone.

1.1. Epidemiological Scale and Phenotypic Modifiers of the Comorbidity

Epilepsy occurs substantially more frequently among autistic individuals than in the general population, although prevalence estimates vary widely according to age, intellectual functioning, ascertainment strategy, and clinical setting. An updated systematic review and meta-analysis reported a pooled epilepsy prevalence of approximately 10% among autistic individuals, with estimates of approximately 7% in cohort studies, 9% in population-based cross-sectional studies, and 19% in clinically ascertained cross-sectional samples [3]. Epilepsy prevalence also increases with age and is higher among individuals with ID and in females, illustrating the importance of phenotypic composition when comparing ASD cohorts [2,3].

Clinical cohorts enriched for neurological comorbidity may consequently show substantially higher rates. In a previous pediatric hospital-based cohort from our group, epilepsy was identified in 44.4% of children with ASD; however, the cohort was characterized by a high prevalence of ID and should therefore not be interpreted as representative of population-level prevalence [7]. Such observations illustrate the marked clinical heterogeneity of the ASD–epilepsy association and the importance of distinguishing population-based estimates from those derived from neurologically selected cohorts. Higher prevalence estimates in clinically ascertained samples likely reflect enrichment for individuals with greater neurodevelopmental and neurological complexity, whereas population-based studies include a broader and less clinically selected spectrum of autistic individuals [3].

Intellectual disability represents one of the most consistently demonstrated modifiers of epilepsy risk in ASD. Earlier meta-analytic evidence showed epilepsy in approximately 21.5% of autistic individuals with ID compared with 8% in those without ID, while more recent analyses continue to identify intellectual functioning as a major source of between-study heterogeneity [2,3]. Female sex and increasing age are also associated with higher epilepsy prevalence, suggesting that the classical male predominance observed in ASD cannot be directly extrapolated to the subgroup with epilepsy [2,3].

Together, these findings argue against a single fixed estimate of epilepsy prevalence in ASD. Instead, the co-occurrence should be regarded as a phenotype strongly modified by intellectual function, age, sex, ascertainment, and underlying etiology.

1.2. Clinical Overlap and Distinctive Manifestations of ASD and Epilepsy

Autism spectrum disorder and epilepsy are clinically distinct disorders despite their frequent co-occurrence. As defined above, ASD is characterized by persistent deficits in social communication and social interaction together with restricted, repetitive patterns of behavior, interests, or activities [1]. Epilepsy, by contrast, is diagnosed according to established clinical criteria reflecting an enduring predisposition to generate epileptic seizures [8]. Their coexistence should therefore not be interpreted as a single clinical phenotype.

The clinical intersection between ASD and epilepsy becomes particularly relevant in individuals with broader neurodevelopmental impairment. Intellectual disability is one of the most consistently demonstrated modifiers of epilepsy risk in ASD and contributes substantially to the heterogeneity of the combined phenotype [2,3]. Developmental regression may also occur in a subset of individuals with ASD, most commonly involving loss of previously acquired language and/or social skills [9]. However, developmental abnormalities and regression are heterogeneous phenomena and should not be assumed to reflect the same underlying mechanism across different neurodevelopmental disorders.

This distinction is particularly important in genetically defined neurodevelopmental disorders and developmental and epileptic encephalopathies (DEEs). In many genetic DEEs, the underlying molecular defect may itself contribute directly to developmental impairment, in addition to the potential adverse contribution of epileptic activity [10,11]. Consequently, developmental delay or regression in these disorders should not automatically be attributed to seizures or epileptiform activity alone. Similarly, autistic features or ASD-related behaviors occurring within complex genetic syndromes may show syndrome-specific characteristics and may differ from nonsyndromic ASD; severe intellectual or global developmental impairment may further complicate formal ASD assessment [12].

Core ASD diagnostic criteria concern social-communication and social-interaction deficits and restricted/repetitive patterns of behavior rather than neurological motor manifestations [1]. Motor abnormalities, hypotonia, movement disorders, and other neurological manifestations may instead occur within the broader phenotype of genetically defined neurodevelopmental and epileptic disorders [10]. When such manifestations accompany autistic features and epilepsy, they should therefore be interpreted within the broader neurological and genetic phenotype rather than as core manifestations of ASD [10,12].

Overall, the clinical overlap between ASD and epilepsy is best understood as a heterogeneous neurodevelopmental context in which distinct core phenotypes may coexist with shared developmental modifiers and manifestations of an underlying genetic disorder. The principal shared and distinctive clinical manifestations relevant to this spectrum are summarized in Table 1.

2. The Genetic Architecture and Diagnostic Yield in the ASD–Epilepsy Spectrum

Advances in chromosomal microarray analysis (CMA), next-generation sequencing (NGS), whole-exome sequencing (WES), and increasingly whole-genome sequencing (WGS) have substantially increased the ability to identify molecular etiologies in neurodevelopmental disorders. However, reported diagnostic yields vary considerably and should not be interpreted as fixed estimates applicable across all ASD or epilepsy populations. Differences in cohort composition and phenotypic severity, particularly the presence of ID, developmental impairment, syndromic features, and early-onset epilepsy, together with differences in sequencing strategy and variant interpretation, contribute substantially to this variability [13,14,15].

Early studies already demonstrated the influence of phenotypic complexity. In the study by Tammimies et al., combined CMA and WES produced an overall molecular diagnostic yield of 15.8% in children with ASD, with substantially higher yield among children with more complex morphological phenotypes [13]. This observation is consistent with subsequent evidence indicating that genetically enriched or clinically more complex neurodevelopmental cohorts generally have higher diagnostic yields than less selected ASD populations.

This pattern was confirmed in a systematic review and meta-analysis of 103 studies comprising 32,331 individuals, in which clinical sequencing yielded a molecular diagnosis in 17.1% of ASD cohorts, compared with 24.0% of epilepsy cohorts and 28.2% of ID cohorts. Within epilepsy populations, yield was higher in individuals with ID (27.9%) and reached 36.8% in early-onset seizures [14]. Thus, differences between diagnostic-yield estimates partly reflect differences in the underlying clinical populations rather than differences in test performance alone.

More recent evidence also demonstrates the impact of sequencing strategy. A 2025 systematic review and meta-analysis of 416 studies encompassing 124,937 individuals found substantially higher diagnostic yields for NGS-based approaches than for CMA. WES yielded molecular diagnoses in approximately 35.3% of NDD cohorts and 34.2% of epilepsy cohorts, while NDD accompanied by epilepsy had a pooled diagnostic yield of 35.6%. Dysmorphic features, syndromic presentations, early-onset epilepsy, and epileptic encephalopathy were associated with above-average yields [15]. The higher estimates in this broader meta-analysis should therefore not be interpreted as directly contradicting earlier ASD-specific estimates; rather, they reflect differences in cohort composition, phenotype severity, and the diagnostic technologies evaluated.

Taken together, these studies suggest that diagnostic yield across the ASD–epilepsy spectrum is best considered as a context-dependent range rather than a universal value. Lower yields are generally observed in less selected ASD cohorts and with approaches such as CMA, whereas substantially higher yields may be achieved with sequencing-based testing in individuals with complex neurodevelopmental phenotypes, ID, syndromic features, or early-onset epilepsy [13,14,15]. Direct comparison between studies nevertheless remains limited by substantial methodological and clinical heterogeneity, including differences in patient ascertainment and phenotypic composition, sequencing strategy and technical platform, genes included in targeted panels, analytical approaches such as trio-based versus proband-only sequencing and variant filtering, and the criteria used for variant classification and interpretation [14,15].

2.1. Epilepsy as a Marker of Increased Genetic Complexity in ASD

Within ASD populations, epilepsy should therefore be regarded not as proof of a monogenic disorder but as an important clinical feature that may increase suspicion of an identifiable genetic etiology, particularly when accompanied by ID, developmental regression, dysmorphism, abnormal head growth, early seizure onset, or other neurological abnormalities [14,16].

This association is supported by recent ASD-specific data. In a cohort of 523 children with ASD, an explanatory genetic diagnosis was obtained in 15.1% overall, with WES outperforming CMA (23.0% vs. 8.3%). Epilepsy was significantly more frequent among children with genetically defined ASD than among those without an identified genetic diagnosis (35.4% vs. 16.4%), and seizure onset occurred at a younger age in the genetically defined group. These findings support epilepsy as a clinically relevant marker of genetic enrichment while also demonstrating that it is neither necessary nor sufficient for establishing a molecular diagnosis [16].

2.2. Structural and Sequence-Level Genetic Variation

The shared genetic architecture of ASD and epilepsy encompasses both structural genomic variation and sequence-level pathogenic variants. CMA remains particularly effective for detecting pathogenic copy-number variants (CNVs), whereas WES and WGS substantially expand detection of single-nucleotide variants and small insertions or deletions. These approaches are therefore complementary rather than conceptually interchangeable [13].

Recurrent CNVs associated with neurodevelopmental phenotypes—including abnormalities involving 15q11–q13, 16p11.2, and 22q11.2—illustrate the complexity of genotype–phenotype relationships. Their manifestations are variably penetrant and may include ASD, ID, language impairment, epilepsy, and other neurological or systemic abnormalities. Importantly, neither epilepsy nor ASD is obligatory across all carriers, and phenotype severity may differ according to genomic architecture, parental origin, dosage, and additional genetic modifiers. This variability argues against describing individual CNVs as uniformly producing a fixed ASD–epilepsy phenotype [17,18,19].

Recurrent copy-number variants (CNVs) represent an important component of the genetic architecture shared by ASD, epilepsy, and related neurodevelopmental disorders. Several recurrent genomic regions, including 1q21.1, 15q11–q13, 15q13.3, 16p11.2, 16p13.11, 22q11.2, and 22q13.3, have been associated with variable combinations of ASD or autistic features, epilepsy, intellectual/developmental impairment, and other neurodevelopmental manifestations [17,18,19,20,21,22,23]. Importantly, the phenotypic consequences of reciprocal deletions and duplications are not necessarily equivalent. This is particularly evident for 16p11.2 and 22q11.2, where reciprocal dosage changes are associated with differences in penetrance and in the relative expression of neurodevelopmental and neurological manifestations [18,19,24,25,26]. In addition, recurrent CNVs frequently show incomplete penetrance and variable expressivity, indicating that genomic dosage alone does not fully determine the individual phenotype [18,20,26]. Common recurrent CNVs relevant to the ASD–epilepsy spectrum and their principal clinical associations are summarized in Table 2.

Collectively, these CNVs illustrate several features of the genetic convergence between ASD and epilepsy. First, the same genomic imbalance may be associated with a broad spectrum of neurodevelopmental outcomes rather than a single disorder. Second, reciprocal dosage changes within the same locus may produce partially distinct phenotypic profiles, as exemplified by 16p11.2 and 22q11.2 deletions and duplications [18,19,24,25,26]. Third, many recurrent CNVs encompass multiple dosage-sensitive genes, making attribution of the complete phenotype to a single gene inappropriate. Nevertheless, individual genes within these regions, including UBE3A within 15q11–q13, CHRNA7 and OTUD7A within 15q13.3, and SHANK3 within 22q13.3, provide important mechanistic links to pathways involved in synaptic function, neuronal excitability, and neurodevelopment [17,21,23].

Sequence-level pathogenic variants likewise span multiple functional classes. Large-scale genomic studies of neurodevelopmental disorders demonstrate enrichment of damaging variants in genes involved in neuronal excitability, synaptic function, transcriptional regulation, and chromatin remodeling. Within the ASD–epilepsy spectrum, representative examples include voltage-gated ion-channel genes (SCN1A, SCN2A, SCN8A, KCNQ2, and KCNT1), synaptic genes (SHANK3, SYNGAP1, and STXBP1), neurotransmitter-receptor genes (GRIN2A, GRIN2B, and GABRB3), and regulators of intracellular signaling and gene expression (TSC1/TSC2, PTEN, MECP2, UBE3A, FMR1, and CHD8) [4,32,33].

The clinical implication is therefore not that epilepsy in ASD automatically indicates a “druggable genetic syndrome,” but that co-occurring epilepsy—particularly when early-onset or accompanied by ID, developmental regression, dysmorphism, or other neurological features—strengthens the rationale for comprehensive genomic evaluation. Molecular diagnosis can then provide the basis for mechanism-informed interpretation and, in selected disorders, precision therapeutic strategies [14,15,16].

3. Functional Classification and Molecular Mechanisms of Shared Genes

Despite their considerable functional diversity, genes associated with ASD and epilepsy converge on a limited set of neurodevelopmental processes. These include regulation of intrinsic neuronal excitability and excitation–inhibition balance, synaptic formation and transmission, activity-dependent signaling and synaptic plasticity, and the maturation and homeostatic regulation of neural circuits [5,34,35]. Importantly, ASD-related circuit dysfunction should not be interpreted simply as a uniform shift toward excessive excitation. Genetic perturbations may alter excitatory or inhibitory neurons, synaptic strength, intrinsic excitability, or compensatory homeostatic mechanisms in a cell type-, circuit-, and developmental stage-dependent manner [34,35]. Neuronal activity itself is a critical regulator of synapse development, stabilization, elimination, and activity-dependent gene expression; consequently, genetic disturbances of neuronal excitability or synaptic signaling during sensitive developmental periods may alter circuit maturation and plasticity, providing a mechanistic route to ASD-related neurodevelopmental dysfunction [5,36]. The same molecular and cellular processes regulate network stability and synchronization and, when sufficiently disrupted, may also promote epileptogenesis. Thus, ASD and epilepsy may represent parallel and interacting consequences of disturbed circuit development and homeostasis rather than a simple unidirectional relationship in which one disorder necessarily causes the other. Representative genes illustrating these convergent molecular and cellular mechanisms are summarized in Table 3.

3.1. Voltage-Gated Ion Channelopathies

Voltage-gated sodium and potassium channels are major determinants of intrinsic neuronal excitability, action-potential generation, and repetitive firing. Pathogenic variants in genes encoding these channels provide a particularly direct link between altered neuronal excitability and the overlapping epilepsy–neurodevelopmental spectrum. Importantly, channel dysfunction may influence not only seizure susceptibility but also activity-dependent circuit maturation and plasticity, thereby providing a mechanistic route to ASD-related neurodevelopmental dysfunction. The resulting phenotype cannot be predicted simply from whether channel activity is increased or decreased, but depends on the direction and magnitude of the functional alteration, the affected neuronal population and subcellular compartment, and the developmental stage at which channel dysfunction alters circuit function [39,41].

SCN1A encodes the voltage-gated sodium-channel α-subunit NaV1.1, which is particularly important for the excitability of GABAergic inhibitory interneurons. In Dravet syndrome, most pathogenic SCN1A variants result in loss of function. Experimental models have demonstrated impaired sodium currents and reduced excitability of inhibitory interneurons, providing a mechanistic basis for network disinhibition and seizure susceptibility [37]. Thus, loss of sodium-channel function at the cellular level may paradoxically produce network hyperexcitability when the predominantly affected neuronal population is inhibitory. Importantly, the same interneuron dysfunction may contribute directly to the neurodevelopmental phenotype rather than acting exclusively through recurrent seizures. In Scn1a-haploinsufficient mice, impaired GABAergic neurotransmission is associated with deficits in social behavior and cognition, while selective reduction in NaV1.1 in forebrain interneurons is sufficient to reproduce these abnormalities. Their pharmacological rescue by enhancement of GABAergic transmission further supports disrupted inhibitory circuit function as a mechanistic link between SCN1A dysfunction and ASD-related behavioral phenotypes [38].

SCN2A, encoding NaV1.2, illustrates a particularly clear relationship between variant function, developmental timing, and divergent neurological phenotypes. Gain-of-function variants are most strongly associated with neonatal or early-infantile epileptic phenotypes, in which enhanced NaV1.2 activity increases the excitability of developing excitatory neurons, whereas many loss-of-function variants are associated with ASD, intellectual disability, and later-onset neurodevelopmental phenotypes, sometimes without epilepsy [39,74]. Importantly, the role of NaV1.2 extends beyond action-potential generation. In cortical pyramidal neurons, NaV1.2 contributes to dendritic excitability and action-potential backpropagation, thereby influencing synaptic strength and activity-dependent plasticity. Experimental reduction in NaV1.2 function impairs dendritic backpropagation and synaptic plasticity, providing a mechanistic link between SCN2A haploinsufficiency and ASD-related neurodevelopmental dysfunction [40]. Thus, SCN2A-related epilepsy and ASD may arise from distinct but developmentally related consequences of altered NaV1.2 function, with the phenotype shaped by the direction and magnitude of channel dysfunction, neuronal compartment, and developmental stage.

SCN8A encodes NaV1.6, a voltage-gated sodium channel highly enriched at the axon initial segment and nodes of Ranvier and critically involved in action-potential initiation, repetitive firing, and propagation. Many pathogenic missense variants associated with early developmental and epileptic encephalopathy produce gain-of-function effects, including enhanced persistent or resurgent sodium currents, thereby promoting neuronal hyperexcitability and seizure susceptibility [41,75]. However, the functional and phenotypic spectrum of SCN8A-related disorders extends beyond gain-of-function epileptic encephalopathy, and loss-of-function or functionally complex variants may be associated with intellectual disability, autistic features, and other neurodevelopmental phenotypes with variable seizure involvement [42,76]. Experimental evidence further supports convergence at the level of neuronal and circuit excitability: an Scn8a mouse model carrying the R1620L variant exhibits altered neuronal excitability and spontaneous seizures together with impaired social behavior and learning, linking NaV1.6 dysfunction to both epileptic and ASD-related phenotypes [77]. Thus, SCN8A-related epilepsy and neurodevelopmental dysfunction may represent partially overlapping consequences of disturbed NaV1.6-dependent neuronal and network function rather than a simple sequence in which seizures alone produce the developmental phenotype.

Potassium-channel disorders further illustrate how both reduced and enhanced channel activity can disrupt developing neural circuits.

KCNQ2 encodes the KV7.2 subunit, which contributes to the neuronal M-current that stabilizes membrane potential and limits repetitive firing. Loss-of-function variants reduce M-current and are associated with a spectrum ranging from self-limited neonatal epilepsy to severe developmental and epileptic encephalopathy, with dominant-negative effects particularly associated with severe phenotypes [43]. Importantly, KCNQ2 dysfunction may affect neurodevelopment through mechanisms that extend beyond seizure generation. Human iPSC-derived neuronal models demonstrate that reduced M-current can alter membrane potential, action-potential generation, homeostatic conductance, and early network oscillations, resulting in delayed neuronal maturation and disrupted network development [44,45]. Conversely, specific gain-of-function variants have been associated with developmental disability, prominent language impairment, and autistic features, sometimes in the absence of neonatal seizures [46]. Thus, both insufficient and excessive KV7.2 activity may disturb the developmental regulation of neuronal excitability and circuit maturation, providing a mechanistic link between KCNQ2 dysfunction, epilepsy, and ASD-related neurodevelopmental phenotypes.

KCNT1 encodes the sodium-activated potassium channel KNa1.1 (Slack), which contributes to the regulation of firing adaptation and neuronal excitability during repetitive activity. Pathogenic variants associated with severe early-onset epilepsies, particularly epilepsy of infancy with migrating focal seizures, commonly produce gain-of-function effects [47]. Although increased potassium conductance might be expected to suppress neuronal firing, KCNT1 gain of function can paradoxically promote network hyperexcitability by altering action-potential repolarization, firing patterns, and neuronal subtype-specific excitability. The neurodevelopmental consequences of KCNT1 dysfunction may also extend beyond the effects of recurrent seizures. Slack channels participate in signaling pathways relevant to neuronal plasticity and interact with the fragile X messenger ribonucleoprotein FMRP, suggesting a potential link between KCNT1-dependent excitability, activity-dependent signaling, and cognitive development [48]. Experimental manipulation of Kcnt1 further produces abnormalities in learning-related behavior and seizure susceptibility, supporting a broader role for Slack channels in both network excitability and neurodevelopment [49]. However, a direct KCNT1-specific mechanism for ASD remains insufficiently established, and the neurodevelopmental phenotype is therefore better interpreted as reflecting disruption of excitability and neuronal plasticity rather than a defined ASD-specific pathway.

Collectively, sodium- and potassium-channel disorders demonstrate that the shared pathophysiological substrate of epilepsy and neurodevelopmental dysfunction cannot be reduced to a simple increase or decrease in neuronal excitation. Variant-specific alterations in ion-channel function may disrupt intrinsic excitability, inhibitory control, dendritic integration, activity-dependent plasticity, and the maturation and temporal organization of developing neural circuits. Depending on the affected channel, neuronal population, subcellular compartment, and developmental stage, these disturbances may promote epileptic hypersynchrony, ASD-related neurodevelopmental dysfunction, or both. Ion channelopathies therefore provide a particularly clear example of how a primary molecular defect can generate partially overlapping epileptic and neurodevelopmental phenotypes through disruption of circuit homeostasis.

3.2. Synaptic Scaffolding, Presynaptic Release, and Activity-Dependent Synaptic Regulation

A second major group of genes shared between ASD and epilepsy affects synaptic organization, maturation, plasticity, and neurotransmitter release. Unlike ion channelopathies, in which altered intrinsic membrane excitability provides the primary molecular disturbance, synaptic disorders can disrupt the formation, strength, timing, and experience-dependent refinement of neuronal connections. These abnormalities may interfere directly with developmental circuit assembly and cognitive or social function, while concomitant disturbances in excitatory–inhibitory regulation and network stability may increase seizure susceptibility. Synaptic dysfunction therefore provides another mechanistic route through which a single genetic defect can generate partially overlapping ASD-related neurodevelopmental and epileptic phenotypes [5].

SHANK3 encodes a multidomain scaffolding protein of the postsynaptic density at excitatory glutamatergic synapses. SHANK3 haploinsufficiency, classically associated with Phelan–McDermid syndrome, disrupts postsynaptic organization and alters glutamatergic synaptic transmission, dendritic-spine structure, and activity-dependent plasticity. In Shank3-deficient mice, reduced AMPA receptor-mediated transmission and impaired long-term potentiation are accompanied by abnormalities in social interaction and communication, directly linking postsynaptic dysfunction to ASD-related behavioral phenotypes [31,55,78]. Studies of patient-derived neurons similarly demonstrate impaired excitatory synaptic transmission and altered glutamate-receptor function, supporting a primary synaptic mechanism for the neurodevelopmental phenotype. Epilepsy occurs only in a subset of individuals with SHANK3-related disorders [6], indicating that profound disruption of excitatory synaptic development is sufficient to produce neurodevelopmental dysfunction but does not invariably cross the threshold for epileptic network instability. Thus, in SHANK3 alterations, ASD-related dysfunction and epilepsy are better regarded as variably expressed consequences of abnormal synaptic and circuit organization rather than as obligatorily coupled phenotypes.

SYNGAP1 encodes SynGAP, a postsynaptic Ras/Rap GTPase-activating protein highly enriched at excitatory glutamatergic synapses, where it regulates signaling downstream of NMDA receptors and controls synaptic maturation and plasticity. SYNGAP1 haploinsufficiency is associated with intellectual disability, epilepsy, and frequent ASD-related behavioral features. Experimental studies have shown that reduced SynGAP expression accelerates dendritic-spine maturation and the functional development of excitatory synapses during critical postnatal periods, resulting in premature stabilization of neural circuits and impaired experience-dependent synaptic plasticity [56,57]. These developmental abnormalities provide a direct mechanistic link to persistent cognitive and behavioral dysfunction. At the same time, altered excitatory synaptic function and disruption of excitation–inhibition regulation increase network excitability and seizure susceptibility [56,57,58,79]. Thus, SYNGAP1 haploinsufficiency provides a particularly clear example of mechanistic convergence between ASD-related neurodevelopmental dysfunction and epilepsy: both phenotypes arise from disruption of the timing and plasticity of excitatory circuit development, rather than one necessarily being secondary to the other.

In contrast to these predominantly postsynaptic mechanisms, STXBP1 encodes syntaxin-binding protein 1 (Munc18-1), an essential component of the presynaptic machinery controlling synaptic vesicle fusion and neurotransmitter release. STXBP1 haploinsufficiency disrupts synaptic transmission and causes a broad neurodevelopmental disorder characterized by developmental delay or intellectual disability, epilepsy, movement disorders, and variable behavioral abnormalities. Because Munc18-1 is required at both excitatory and inhibitory synapses, the consequences of STXBP1-related dysfunctions cannot be attributed to a selective defect of a single neurotransmitter system. Nevertheless, experimental models demonstrate particular vulnerability of inhibitory circuitry: Stxbp1 haploinsufficiency reduces cortical inhibitory neurotransmission through distinct abnormalities affecting parvalbumin- and somatostatin-expressing interneurons, resulting in cortical hyperexcitability and increased seizure susceptibility [59]. The same models exhibit cognitive and behavioral abnormalities, including impaired learning and repetitive behaviors, supporting a broader effect of presynaptic dysfunction on neurodevelopmental circuit function. Thus, STXBP1 alterations provide a mechanistic link between impaired neurotransmitter release, disturbed excitation–inhibition regulation, and the coexistence of epileptic and neurodevelopmental phenotypes, although linking these alterations to ASD-related features remain incompletely understood.

CDKL5 encodes a brain-enriched serine/threonine kinase that regulates phosphorylation-dependent processes involved in neuronal maturation, dendritic morphogenesis, synapse formation, and synaptic function. Pathogenic loss-of-function variants cause CDKL5 deficiency disorder (CDD), a severe developmental and epileptic encephalopathy characterized by very early-onset, frequently drug-resistant seizures and profound neurodevelopmental impairment. Autistic features are reported in CDD, although formal ASD diagnosis can be difficult in individuals with severe global developmental impairment. Experimental data indicate that loss of CDKL5 function disrupts dendritic and synaptic development, supporting a primary neurodevelopmental mechanism that is not reducible to seizure burden alone [60,61]. CDD therefore extends the synaptic convergence model beyond structural scaffolds and vesicle-release proteins to activity-dependent kinase signaling.

Together, SHANK3, SYNGAP1, STXBP1, and CDKL5 illustrate how disturbances at different levels of synaptic function can converge on a common set of downstream processes linking epilepsy and neurodevelopmental dysfunction. Altered postsynaptic organization and receptor signaling, abnormal timing of excitatory-synapse maturation, defective presynaptic neurotransmitter release, and disrupted activity-dependent kinase signaling can all impair synaptic plasticity, excitation–inhibition regulation, and the developmental refinement of neural circuits [55,56,57,58,59,60,61]. These disturbances may simultaneously interfere with the circuit adaptations required for cognitive, behavioral, and social development and increase susceptibility to pathological network synchronization and seizures. Importantly, however, the strength and nature of this mechanistic convergence differ among genes. SYNGAP1 provides a particularly clear example in which disruption of excitatory-synapse maturation and critical-period plasticity is linked to both neurodevelopmental dysfunction and seizure susceptibility [56,57,58], whereas in SHANK3- and STXBP1-related disorders the evidence for a single shared mechanism specifically linking ASD and epilepsy is less direct [55,59]. CDD further illustrates how a primary disturbance of neuronal and synaptic maturation can coexist with severe early-onset epilepsy without implying that the neurodevelopmental phenotype is secondary to seizure activity [60,61]. Thus, synaptic dysfunction represents a convergent biological substrate for ASD-related or broader neurodevelopmental dysfunction and epilepsy, rather than a uniform causal pathway producing both phenotypes in every genetic disorder.

3.3. Glutamatergic and GABAergic Receptor Dysfunction

Excitatory glutamatergic and inhibitory GABAergic neurotransmission represent another major point of mechanistic convergence between ASD and epilepsy. Pathogenic variants affecting ligand-gated neurotransmitter receptors can alter receptor expression and trafficking, agonist sensitivity, channel gating, synaptic current kinetics, and activity-dependent plasticity. Because these receptors also regulate synapse maturation and circuit refinement, their dysfunction may directly affect neurodevelopment while simultaneously destabilizing network activity and increasing seizure susceptibility. Importantly, the resulting phenotype cannot be reduced to a simple model of excessive excitation or insufficient inhibition, because both gain- and loss-of-function effects may disrupt circuit function in a receptor-, cell type-, and developmental stage-dependent manner [50,80].

GRIN2A encodes the GluN2A subunit of N-methyl-D-aspartate receptors (NMDARs), which are central to excitatory neurotransmission, synaptic plasticity, learning, and activity-dependent circuit maturation. Pathogenic GRIN2A variants are particularly associated with epilepsy–aphasia spectrum disorders and related developmental and epileptic phenotypes, but functional effects are heterogeneous and include both gain- and loss-of-function abnormalities [51,52]. Because GluN2A-containing NMDARs contribute to the developmental maturation of excitatory synapses and experience-dependent plasticity, altered receptor function may impair language- and cognition-related circuit development while also modifying neuronal excitability and seizure susceptibility. Thus, GRIN2A-related epilepsy and neurodevelopmental dysfunction can reflect parallel consequences of abnormal NMDAR-dependent synaptic signaling rather than a purely sequential effect of seizures on development.

GRIN2B encodes the GluN2B NMDAR subunit, which is highly expressed during early brain development and plays an important role in neuronal differentiation, dendritic morphogenesis, synaptogenesis, and synaptic plasticity. GRIN2B-related disorders commonly present with developmental delay or intellectual disability, behavioral abnormalities including ASD, and variable epilepsy. Pathogenic variants show marked functional heterogeneity, including both gain- and loss-of-function effects that may alter receptor kinetics, agonist sensitivity, surface expression, or downstream signaling [50,53]. Experimental studies further indicate that disease-associated GluN2B variants can interfere directly with neuronal differentiation, dendritic development, synapse formation, and plasticity, providing a mechanistic basis for ASD-related neurodevelopmental dysfunction that is not dependent on seizure activity. The same alterations in NMDAR-mediated synaptic signaling can also disturb network excitability and stability, creating a shared mechanistic substrate for epilepsy and neurodevelopmental impairment [50,53].

On the inhibitory side, GABRB3 encodes the β3 subunit of the GABA\(_A\) receptor, a major mediator of fast inhibitory neurotransmission in the developing and mature brain. Pathogenic GABRB3 variants are associated with a broad phenotypic spectrum ranging from relatively mild epilepsies to severe developmental and epileptic encephalopathies, often accompanied by intellectual disability and behavioral or autistic features. Functional consequences are heterogeneous and include altered receptor expression, trafficking, GABA sensitivity, gating, and current amplitude [54]. Reduced GABAergic signaling can lower seizure threshold through impaired inhibitory control; however, pathogenic GABRB3 variants are functionally heterogeneous, and developmental and epileptic phenotypes cannot be explained by reduced inhibition alone [54,81]. Thus, disruption of β3-containing GABA\(_A\) receptors may affect both network inhibition and developmental circuit organization, providing a plausible mechanistic link between GABRB3 dysfunction, epilepsy, and ASD-related neurodevelopmental abnormalities [35,54].

Together, GRIN2A, GRIN2B, and GABRB3 illustrate how receptor dysfunction can couple altered synaptic transmission to abnormal circuit development. In glutamatergic disorders, impaired NMDAR-dependent signaling may disrupt synaptic maturation, plasticity, and cognitive circuit formation while also destabilizing excitatory network activity. In GABAergic disorders, altered inhibitory receptor function can affect both seizure threshold and the developmental organization of inhibitory control. These disorders therefore reinforce the concept that excitation–inhibition dysregulation is not a uniform directional shift, but a disturbance in the magnitude, timing, localization, and developmental integration of excitatory and inhibitory signaling that may produce epilepsy, ASD-related neurodevelopmental dysfunction, or both.

The relative contribution of these receptor-level mechanisms to epilepsy and ASD-related neurodevelopmental dysfunction nevertheless differs among genes. GRIN2A-related disorders are weighted predominantly toward epilepsy–aphasia and speech/language phenotypes, indicating that disruption of GluN2A-dependent signaling does not uniformly produce an ASD–epilepsy phenotype [51,52]. In contrast, GRIN2B dysfunction more consistently affects early neurodevelopment, with developmental delay or intellectual disability and behavioral abnormalities including ASD, whereas epilepsy is variably expressed [50,53]. GABRB3 further illustrates the functional complexity of receptor-mediated convergence: pathogenic variants can produce heterogeneous effects on receptor expression, trafficking, gating, and GABA sensitivity, demonstrating that developmental and epileptic phenotypes cannot be explained simply by a uniform reduction in inhibition [54]. These genotype-specific differences reinforce the concept that a shared molecular pathway may converge on common processes—synaptic plasticity, developmental excitation–inhibition regulation, and circuit maturation—while producing different relative burdens of epilepsy and ASD-related or broader neurodevelopmental dysfunction.

3.4. PI3K–AKT–mTOR Pathway Dysregulation

The PI3K–AKT–mTOR signaling pathway represents a major point of mechanistic convergence between ASD and epilepsy. It regulates protein synthesis, cellular growth, neuronal morphology and migration, synapse formation, and activity-dependent plasticity, thereby influencing both the structural organization and functional maturation of developing neural circuits. Dysregulated signaling can consequently produce abnormal neuronal growth, connectivity, and synaptic plasticity relevant to ASD-related neurodevelopmental dysfunction, while alterations in intrinsic excitability, synaptic transmission, and network organization may simultaneously promote epileptogenesis [62,63].

TSC1 and TSC2 encode hamartin and tuberin, respectively, which form a functional complex that negatively regulates mTORC1 signaling. Loss-of-function variants disrupt this inhibitory control, resulting in constitutive mTORC1 activation. Tuberous sclerosis complex (TSC) is consequently associated with a broad neurological phenotype that includes epilepsy, intellectual disability, and neuropsychiatric manifestations, with ASD occurring in a substantial subgroup of affected individuals. Importantly, mTOR hyperactivation can directly alter neuronal growth, morphology, synaptic development, and plasticity, providing a primary molecular mechanism for neurodevelopmental dysfunction that does not require epilepsy as an intermediary [62,63]. At the same time, abnormal neuronal morphology and connectivity, altered synaptic function, and changes in network excitability create a substrate for epileptogenesis, while cortical tubers and surrounding dysplastic tissue provide an additional structural contribution to seizure generation [64].

The relationship between epilepsy and neurodevelopmental dysfunction in TSC is therefore not simply causal or unidirectional. The primary TSC1/TSC2–mTOR defect can independently disrupt developmental and synaptic processes relevant to cognition and ASD, whereas early and severe epileptic activity may further modify the developing network and may contribute to less favorable developmental outcomes. Experimental and clinical observations consequently support an interactive model in which primary pathway dysregulation, structural brain abnormalities, and epileptic activity contribute to neurodevelopmental outcome to varying degrees across individuals and developmental stages [62,64].

PTEN acts upstream in the PI3K–AKT–mTOR pathway as a major negative regulator of intracellular signaling. Germline loss-of-function variants are associated with PTEN hamartoma tumor syndrome and a neurodevelopmental phenotype frequently characterized by macrocephaly, developmental abnormalities, and ASD, while epilepsy occurs in a subset of affected individuals. Loss of PTEN activity enhances PI3K–AKT–mTOR signaling and disrupts neurogenesis, neuronal growth and migration, dendritic and axonal morphology, synapse formation, connectivity, and synaptic plasticity, providing a direct mechanistic route from impaired PTEN function to ASD-related neurodevelopmental dysfunction [65,66]. Experimental Pten loss also produces neuronal hypertrophy, increased excitatory synaptic input, hyperexcitability, and seizures, demonstrating how the same upstream signaling defect can simultaneously generate structural and functional substrates for epileptogenesis [67].

TSC1/TSC2- and PTEN-related disorders therefore illustrate how dysregulation of a shared intracellular signaling pathway can simultaneously affect neuronal growth, migration, synaptic organization, plasticity, circuit connectivity, and network excitability. These effects provide partially overlapping but not necessarily identical mechanistic routes to ASD-related neurodevelopmental dysfunction and epilepsy. The relative contribution of primary developmental abnormalities, structural pathology, and epileptic activity differs among disorders and individuals, emphasizing that “mTORopathies” should not be regarded as a uniform biological or clinical entity. Rather, they exemplify how pathway-level convergence can generate divergent phenotypes according to the spatial distribution, magnitude, and developmental timing of signaling dysregulation.

3.5. Chromatin, Transcriptional, and Translational Regulation

A further level of convergence occurs upstream of ion channels and synapses, through genes that regulate chromatin organization, transcription, protein turnover, and activity-dependent translation. Unlike variants that directly alter membrane conductance or synaptic receptors, these disorders can modify broad developmental gene-expression programs and thereby affect multiple downstream processes simultaneously. Dysregulation of neuronal differentiation, synapse formation and maturation, activity-dependent plasticity, and excitation–inhibition regulation can consequently generate ASD-related neurodevelopmental dysfunction while also altering circuit stability and seizure susceptibility [5,36].

CHD8 encodes an ATP-dependent chromatin-remodeling protein and is one of the most consistently implicated high-confidence ASD-associated genes. Disruptive variants cause a recognizable neurodevelopmental syndrome frequently characterized by ASD, developmental delay or intellectual disability, macrocephaly, and gastrointestinal and sleep abnormalities [73]. CHD8 regulates broad transcriptional programs involved in neuronal differentiation, axon guidance, synapse development, connectivity, and plasticity, providing a direct mechanistic route from chromatin dysregulation to ASD-related neurodevelopmental dysfunction. Seizures occur in a minority of affected individuals and are not a defining feature of CHD8-related disorder. CHD8 is therefore most informative for the ASD–epilepsy interface at the level of pathway convergence: disruption of an upstream chromatin regulator can alter downstream neurodevelopmental and synaptic programs that overlap with those affected by established epilepsy-associated genes, without necessarily producing epilepsy in every disorder or individual [73].

MECP2 encodes methyl-CpG-binding protein 2, a chromatin-associated regulator that links neuronal activity to gene-expression programs required for neuronal maturation and circuit function. Pathogenic loss-of-function variants cause Rett syndrome, characterized by severe neurodevelopmental impairment, loss of acquired skills, communication deficits, stereotypic hand movements, motor abnormalities, and frequent epilepsy. MeCP2 dysfunction disrupts activity-dependent transcription, neuronal and synaptic maturation, connectivity, and experience-dependent plasticity and can alter excitation–inhibition regulation at the circuit level [68]. These effects provide a direct developmental mechanism for cognitive and behavioral dysfunction, while abnormalities of excitatory and inhibitory circuit function can simultaneously promote network instability and seizures. MECP2-related disease therefore illustrates how disruption of activity-dependent transcriptional regulation can generate both neurodevelopmental and epileptic phenotypes through partially overlapping effects on circuit maturation and homeostasis.

UBE3A encodes an E3 ubiquitin ligase with a parent-of-origin-dependent expression pattern in neurons, where the paternal allele is largely silenced. Loss of function of the maternally expressed neuronal UBE3A allele causes Angelman syndrome, characterized by severe developmental impairment, minimal or absent speech, characteristic behavioral features, motor dysfunction, and a high prevalence of epilepsy. UBE3A deficiency disrupts protein homeostasis, synaptic development and plasticity, and neuronal excitability, providing a mechanistic basis for both developmental dysfunction and network instability. Conversely, maternally derived duplications of chromosome 15q11–q13 are strongly associated with ASD, developmental impairment, and epilepsy. However, the duplicated interval contains multiple dosage-sensitive genes, including GABAA-receptor subunit genes; consequently, the Dup15q phenotype cannot be attributed exclusively to increased UBE3A dosage. Disorders involving altered UBE3A gene expression or copy number therefore illustrate how genomic imprinting and dosage imbalance can influence synaptic and circuit development while producing markedly different neurodevelopmental and epileptic phenotypes [69,70].

Finally, FMR1 encodes fragile X messenger ribonucleoprotein (FMRP), an RNA-binding protein that regulates the translation of numerous neuronal and synaptic mRNAs in response to activity. Loss of FMRP in fragile X syndrome produces dysregulated synaptic protein synthesis and alters synapse maturation and activity-dependent plasticity. These abnormalities provide a direct molecular link to intellectual disability and ASD-related behavioral phenotypes by disrupting the experience-dependent modification of developing neural circuits. Altered synaptic signaling and circuit excitability may also contribute to the increased susceptibility to seizures observed in a subset of affected individuals. FMRP-related translational dysregulation therefore illustrates how abnormal control of activity-dependent protein synthesis can simultaneously influence neurodevelopmental plasticity and network stability [71,72].

Taken together, CHD8, MECP2, UBE3A, and FMR1 demonstrate that ASD–epilepsy convergence may originate upstream of membrane excitability and synaptic transmission. Disturbances in chromatin remodeling, transcriptional regulation, protein homeostasis, and activity-dependent translation can alter neuronal differentiation, synaptic maturation, receptor and channel expression, connectivity, and circuit plasticity. These upstream abnormalities may therefore converge on many of the same cellular and network processes that are affected more directly by ion-channel, synaptic, and receptor disorders, while the relative expression of ASD-related neurodevelopmental and epileptic phenotypes remains gene- and context-dependent.

Across these mechanistic categories, no single molecular abnormality explains the coexistence of ASD and epilepsy. Pathogenic variants may primarily alter intrinsic membrane excitability, synaptic architecture and neurotransmitter release, receptor-mediated signaling, intracellular growth pathways, or transcriptional and translational programs. Yet these apparently distinct molecular defects converge on a smaller set of systems-level consequences: impaired circuit homeostasis, disturbed excitation–inhibition regulation, altered synaptic plasticity, and abnormal maturation and temporal coordination of neuronal networks. The relative contribution of each process varies according to the affected gene, cell type, circuit, developmental stage, and functional consequence of the variant. This convergence provides the mechanistic bridge from genetic heterogeneity to the network-level pathophysiology considered in Section 4.

4. The Pathophysiological Core: Circuit Homeostasis, E/I Imbalance, and Network Dysregulation

The molecular mechanisms described above ultimately converge at the level of developing neural circuits. Although ion-channel dysfunction, altered synaptic transmission, receptor abnormalities, dysregulated intracellular signaling, and impaired gene-expression control originate at different molecular levels, each can interfere with the ability of neuronal networks to establish and maintain stable patterns of activity. The resulting phenotype is therefore better conceptualized as a disturbance of circuit homeostasis than as a uniform shift toward neuronal hyperexcitability. This distinction is particularly relevant to the ASD–epilepsy spectrum, in which the same genetic alteration may disturb developmental circuit homeostasis in ways that simultaneously influence seizure susceptibility, synaptic plasticity, sensory processing, cognition, and behavior [34,35].

4.1. Intrinsic Excitability and Failure of Network Homeostasis

Neural circuits must maintain activity within a functional dynamic range despite continuous changes in synaptic input, intrinsic membrane properties, and developmental connectivity. Homeostatic mechanisms normally compensate for such perturbations by adjusting ion-channel expression, synaptic strength, receptor composition, and neuronal firing thresholds. Genetic disruption of these processes may overwhelm or misdirect compensatory responses, producing persistent network instability. Importantly, compensatory adaptations may partially normalize average neuronal activity while introducing other abnormalities in circuit function, making it difficult to infer the network phenotype directly from the initial molecular defect [34].

Channelopathies provide a particularly clear illustration of this principle. As discussed in Section 3.1, both gain- and loss-of-function variants can destabilize neural networks, depending on the affected channel, neuronal population, subcellular compartment, and developmental stage. Importantly, the initial molecular disturbance may trigger compensatory changes in ion-channel expression, synaptic strength, and neuronal firing that partially restore average activity while altering the temporal or spatial organization of circuit function. The network phenotype therefore reflects not only the primary biophysical defect but also the capacity—and limitations—of developmental homeostatic compensation [34].

This framework also helps explain why epilepsy and neurodevelopmental impairment may coexist without being mechanistically identical. A genetic alteration that destabilizes circuit homeostasis can increase susceptibility to hypersynchronous epileptic activity while simultaneously interfering with activity-dependent processes required for normal synaptic refinement and cognitive development. Seizures may therefore represent one manifestation of network instability and may further modify developmental trajectories, rather than constituting the sole cause of neurodevelopmental dysfunction [11,34].

4.2. Interneuron Dysfunction and GABAergic Network Disinhibition

GABAergic interneurons are central regulators of cortical excitability, oscillatory activity, temporal precision, and experience-dependent plasticity. Despite representing a minority of cortical neurons, they comprise multiple functionally specialized populations that control distinct spatial and temporal aspects of pyramidal-cell activity. Their maturation also contributes critically to the regulation of developmental periods of enhanced cortical plasticity. Disruption of interneuron development or function can therefore have disproportionately large effects on circuit organization and information processing [35,82].

SCN1A provides a paradigmatic example of cell type-specific network dysfunction. Preferential impairment of NaV1.1-dependent excitability in inhibitory interneurons weakens inhibitory control and produces network disinhibition despite an underlying loss-of-function channel defect. At the circuit level, this illustrates why E/I dysregulation cannot be inferred simply from the direction of the molecular defect and why reduced excitability of one neuronal population may generate hyperexcitability of the network as a whole [34,37].

Interneuron dysfunction is not restricted to primary channelopathies. Molecular defects originating at other levels of synaptic organization can propagate through the network and secondarily alter inhibitory circuit function. STXBP1 haploinsufficiency, for example, can compromise inhibitory neurotransmission through abnormalities involving parvalbumin- and somatostatin-expressing interneurons [59]. Experimental SHANK3 deficiency has likewise been associated with reduced cortical interneuron activity accompanied by increased spontaneous and stimulus-evoked pyramidal-neuron firing. Selective disruption of Shank3 in cortical inhibitory interneurons can reproduce pyramidal-cell hyperactivity and sensory hyper-reactivity, illustrating how a molecular defect need not originate in a GABAergic-specific protein to produce functionally significant impairment of inhibitory circuit control [59,83].

The developmental dimension is equally important. GABAergic circuits undergo prolonged maturation and contribute to the regulation of critical periods of cortical plasticity. Alterations in interneuron maturation may therefore affect not only seizure susceptibility but also the timing of experience-dependent circuit refinement. Such disturbances provide a mechanistic link between inhibitory-circuit dysfunction, abnormal sensory processing, impaired developmental plasticity, and vulnerability to epilepsy [35,82,84].

Thus, GABAergic dysfunction in ASD–epilepsy disorders should be understood as a disturbance of when, where, and how inhibition is recruited within developing networks, rather than simply as a global reduction in inhibitory neurotransmission [34,35].

4.3. Synaptic Dysfunction, Network Coordination, and Impaired Plasticity

Normal neurodevelopment requires more than an appropriate average level of excitation and inhibition. Information processing depends on the precise temporal coordination of neuronal populations, appropriate synaptic strengthening and weakening, and the ability of circuits to reorganize in response to experience. Genetic abnormalities affecting synaptic proteins, neurotransmitter receptors, intracellular signaling, or transcriptional and translational regulation can disrupt these processes even when basal neuronal excitability is not markedly abnormal [34,35].

Across these molecular categories, disruption of synaptic architecture, neurotransmitter release, receptor signaling, intracellular pathways, or gene-expression programs can impair the temporal organization and adaptive modification of neuronal communication. The resulting network dysfunction may involve altered synaptic strength, mistimed recruitment of neuronal populations, impaired experience-dependent remodeling, or failure to maintain stable patterns of coordinated activity. Thus, molecularly distinct disorders can converge on common deficits in network integration and plasticity without requiring identical abnormalities of basal neuronal excitability.

These molecular disturbances can generate two apparently contrasting network consequences. On the one hand, excessive synchronization of neuronal populations can facilitate epileptiform activity and seizures. On the other, impaired physiological coordination and temporal precision may interfere with distributed information processing and developmental network integration. These phenomena are not mutually exclusive: the same network may become pathologically hypersynchronous during epileptic activity while exhibiting deficient or mistimed physiological coordination during normal cognitive and sensory processing. This distinction is particularly important in the ASD–epilepsy spectrum, because network dysfunction cannot be adequately described along a single axis of increased or decreased excitation [34,35].

For this reason, the term network desynchronization alone does not adequately describe the ASD–epilepsy phenotype. The broader concept of network dysregulation accommodates both pathological hypersynchrony associated with epilepsy and impaired physiological coordination relevant to neurodevelopmental dysfunction.

Synaptic plasticity provides a further point of convergence. Activity-dependent modification of synaptic strength is required for learning, memory, sensory adaptation, and developmental circuit refinement. Genetic disturbances that alter the timing or magnitude of plasticity may therefore produce lasting developmental effects even in the absence of continuous epileptic activity. Conversely, recurrent seizures and persistent interictal epileptiform activity may further disrupt plasticity and network maturation, particularly during sensitive developmental periods. Contemporary concepts of developmental and epileptic encephalopathy explicitly recognize these potentially separable but interacting contributions of the underlying etiology and epileptic activity [11].

Taken together, the ASD–epilepsy spectrum can be conceptualized as a disorder of developmental circuit homeostasis in which diverse molecular defects converge on abnormal regulation of neuronal excitability, inhibition, synaptic plasticity, and temporal network organization. The relative contribution of each mechanism varies according to genotype, variant function, brain region, cell type, and developmental stage. This heterogeneity helps explain why related molecular pathways may produce markedly different combinations of epilepsy, intellectual disability, language impairment, and autistic features.

At the network level, these mechanisms can be summarized as a convergent framework in which genetic alterations affecting intrinsic excitability, synaptic function, intracellular signaling, or developmental regulatory programs disrupt circuit homeostasis (Figure 1). The resulting abnormalities in E/I regulation, synaptic plasticity, and temporal coordination may diverge toward pathological hypersynchronization and epileptic activity, impaired physiological network coordination and ASD-related neurodevelopmental dysfunction, or both. Importantly, these outcomes may interact dynamically during development rather than representing a simple linear sequence from genetic defect to epilepsy and subsequently to autism.

5. From Gene to Bedside: Precision Therapeutics in Co-Occurring ASD and Epilepsy

The increasing identification of monogenic and pathway-defined disorders within the ASD–epilepsy spectrum is progressively shifting therapeutic strategies from predominantly syndrome-based seizure management toward mechanism-informed precision medicine. A molecular diagnosis may influence treatment at several levels: avoidance of potentially aggravating drugs, pharmacological modulation of a variant-specific functional defect, targeting of a dysregulated intracellular pathway, correction of a metabolic deficiency, or direct manipulation of gene expression or RNA processing. Precision therapy, however, should not be equated with gene-specific therapy. Rather, it encompasses a continuum of interventions that differ substantially in target specificity, level of mechanistic correction, and capacity to modify the underlying disease process.

This distinction is particularly important in developmental and epileptic encephalopathies (DEEs), in which seizure control, correction of the primary molecular abnormality, and modification of neurodevelopmental trajectory represent related but not necessarily equivalent therapeutic outcomes. Although several genetically or pathway-informed interventions have entered clinical practice and an increasing number of molecular therapies are under clinical development, the strength of evidence remains highly heterogeneous. Biological plausibility, target engagement, seizure reduction, improvement in developmental function, and true disease modification should therefore be regarded as distinct levels of therapeutic evidence.

Accordingly, the therapeutic approaches discussed below should be interpreted according to their level of clinical evidence. They range from established therapies supported by clinical trials and incorporated into current clinical practice, through mechanism-informed or off-label interventions supported primarily by observational or limited clinical evidence, to investigational therapies undergoing clinical evaluation and preclinical strategies supported predominantly by experimental data. Throughout this section, mechanistic rationale is therefore distinguished from demonstrated clinical efficacy, and early therapeutic signals are not interpreted as evidence of established disease-modifying benefit.

5.1. Sodium Channelopathies: Functional Variant–Guided Therapy

Voltage-gated sodium-channel disorders provide one of the clearest examples of why genetic diagnosis must be complemented by functional interpretation. Pathogenic variants in SCN1A, SCN2A, and SCN8A can produce markedly different—and sometimes opposing—effects on channel activity, with direct implications for treatment.

In Dravet syndrome, most pathogenic SCN1A variants result in loss of NaV1.1 function, with prominent impairment of inhibitory interneuron excitability. Maintenance treatment with sodium-channel-blocking antiseizure medications, particularly carbamazepine and lamotrigine, may aggravate seizures and is therefore generally avoided. Current syndrome-directed treatment includes valproate and clobazam, with stiripentol, fenfluramine, and pharmaceutical-grade cannabidiol representing additional established therapeutic options [85]. These interventions can substantially reduce seizure burden but do not directly correct the underlying SCN1A haploinsufficiency.

The therapeutic logic differs in SCN2A-related disorders. Variants associated with neonatal or early-infantile epilepsy frequently produce gain-of-function effects, and sodium-channel blockers may be particularly effective in this functional subgroup. Conversely, many loss-of-function variants are associated with later neurodevelopmental phenotypes, including intellectual disability and ASD, with or without epilepsy [39,74]. This genotype–function–phenotype relationship is clinically informative but not absolute, emphasizing that treatment decisions should integrate phenotype, developmental timing, and functional characterization when available.

A similar principle applies to SCN8A. Many variants associated with early-onset developmental and epileptic encephalopathy increase NaV1.6 activity, providing a mechanistic rationale for sodium-channel-blocking treatment in selected individuals. However, SCN8A-related disorders encompass gain-of-function, loss-of-function, and functionally complex variants associated with distinct neurological phenotypes. Recent international consensus recommendations therefore emphasize phenotype and functional variant interpretation rather than application of a uniform therapeutic strategy to all individuals with SCN8A-related disorders [41].

These examples illustrate a central principle of precision neurotherapeutics: identification of the affected gene defines the molecular context, but the direction, magnitude, cellular context, and developmental consequences of the functional disturbance determine how that target should be approached therapeutically. Strategies aimed at modifying sodium-channel gene expression itself are discussed in Section 5.7.

5.2. Potassium Channelopathies: Opportunities and Limitations of Targeted Modulation

Potassium channelopathies provide a complementary model of function-guided treatment. KCNQ2 encodes the KV7.2 subunit contributing to the neuronal M-current. Many variants associated with severe KCNQ2 developmental and epileptic encephalopathy reduce channel function, providing a mechanistic rationale for pharmacological enhancement of KV7 activity.

Ezogabine (retigabine), a KV7-channel opener, provided an early proof of concept. In an observational series including 11 treated individuals with KCNQ2 encephalopathy, improvement in seizures and/or development was reported in some patients, particularly among those treated early. However, the study was retrospective and uncontrolled and therefore could not establish therapeutic efficacy [43]. Moreover, the functional heterogeneity of KCNQ2 disorders—including gain-of-function as well as loss-of-function variants—means that channel enhancement cannot be assumed to be appropriate for every pathogenic variant.

KCNT1 illustrates an apparently opposite therapeutic problem. Pathogenic variants associated with severe early-onset epilepsies commonly produce gain of function of the sodium-activated potassium channel KNa1.1 (Slack), providing a rationale for channel inhibition. Quinidine therefore emerged as a mechanistically attractive candidate. Although individual reports suggested seizure improvement, a small randomized, blinded, placebo-controlled crossover trial failed to demonstrate significant seizure reduction and identified dose-limiting QT prolongation [86]. Thus, KCNT1 provides a particularly clear example in which strong molecular rationale did not translate straightforwardly into clinical efficacy.

Potassium channelopathies consequently demonstrate both the promise and limitations of precision pharmacology. Mechanistic plausibility is necessary but insufficient: variant function, pharmacokinetics and brain penetration, developmental timing, off-target effects, and clinical safety all determine whether a molecularly rational intervention translates into meaningful clinical benefit.

5.3. GABAergic Dysfunction and Receptor-Targeted Therapies

Disorders affecting GABAergic neurotransmission appear intuitively suited to enhancement of inhibitory signaling, but the mechanistic framework developed in Section 3 and Section 4 indicates substantially greater complexity. GABRB3-associated disorders, for example, encompass heterogeneous functional abnormalities involving receptor expression, trafficking, gating, GABA sensitivity, and current kinetics. A gene-level diagnosis therefore does not establish a uniform pharmacological direction.

Enhancement of GABA_A-receptor activity may be mechanistically appropriate in selected contexts, but it cannot be assumed to correct all disorders involving GABAergic signaling. Functional characterization and knowledge of the affected receptor population, neuronal context, and developmental stage are therefore important both for genotype–phenotype interpretation and for the rational design of receptor-targeted interventions.

Ganaxolone provides a clinically established example of a related but less gene-specific strategy. Ganaxolone is a neuroactive steroid positive allosteric modulator of GABA_A receptors. In the randomized, placebo-controlled phase 3 Marigold trial, ganaxolone significantly reduced major motor seizure frequency in patients with CDKL5 deficiency disorder [87]. This represents syndrome- and pathway-informed treatment rather than direct correction of the causal CDKL5 defect and illustrates an important distinction between precision treatment of downstream network dysfunction and molecular correction of the primary etiology. The underlying CDKL5-related developmental biology is discussed in Section 3.2 [60,61].

5.4. Targeting Intracellular Signaling: mTOR Pathway Modulation

The PI3K–AKT–mTOR pathway provides one of the strongest examples of successful pathway-directed treatment in genetic neurodevelopmental disease. Loss of TSC1 or TSC2 function results in excessive mTORC1 activity, creating a direct molecular rationale for mTOR inhibition [62,63].

Everolimus has demonstrated antiseizure efficacy in TSC-associated treatment-resistant focal-onset seizures. In the randomized, double-blind, placebo-controlled phase 3 EXIST-3 trial, adjunctive everolimus significantly reduced seizure frequency compared with placebo, providing high-level evidence that pharmacological targeting of a genetically dysregulated intracellular pathway can translate into clinically meaningful seizure reduction [88].

However, correction of pathway activity sufficient to improve seizure control should not automatically be equated with normalization of neurodevelopment. Studies of neuropsychiatric and cognitive outcomes with mTOR inhibition have yielded more limited and heterogeneous findings. Even in TSC—one of the clearest examples of pathway-directed precision treatment—target engagement, seizure reduction, and developmental benefit therefore remain distinct therapeutic outcomes [63,89].

TSC also illustrates the importance of therapeutic timing through a different mechanism. In the EPISTOP study, infants with TSC underwent serial EEG surveillance, and preventive vigabatrin initiated after detection of epileptiform EEG activity but before clinical seizures delayed seizure onset and reduced the risks of clinical seizures, drug-resistant epilepsy, and infantile spasms compared with conventional treatment [90]. EPISTOP does not demonstrate complete prevention of neurodevelopmental impairment, nor is vigabatrin a direct mTOR-correcting therapy. Rather, it provides evidence that intervening against emerging epileptic network dysfunction before recurrent clinical seizures may modify the epilepsy trajectory.

5.5. Glutamatergic and Synaptic Targets

NMDAR-related disorders provide a particularly instructive example of variant-specific pharmacological reasoning. Pathogenic variants in GRIN genes can produce gain- or loss-of-function effects, and treatment strategies should therefore ideally be informed by functional characterization rather than gene identity alone [50].

Memantine, an NMDAR antagonist, initially provided proof of concept in individual patients with functionally characterized gain-of-function GRIN variants. More recent retrospective evidence has extended this principle: in a 2026 cohort of 34 individuals with GRIN variants, clinical benefit from memantine was substantially more frequent among individuals with functionally classified gain-of-function variants than among those with loss-of-function, indeterminate, or functionally neutral variants, although responses remained heterogeneous [91,92]. These findings strengthen the principle of function-guided therapy while remaining insufficient to establish memantine as universally effective treatment for GRIN-related disorders.

Conversely, L-serine supplementation has been investigated as a strategy for GRIN loss-of-function disorders, based on enhancement of NMDAR co-agonist availability. Initial individualized observations in GRIN2B loss-of-function disease suggested improvement in motor, cognitive, and communication outcomes [93]. Subsequent open-label studies, including a phase 2A study of children with GRIN loss-of-function variants, have provided additional evidence of tolerability and signals of improvement in selected functional outcomes, while emphasizing the need for controlled validation [93,94].

Synaptic disorders have also stimulated attempts to modulate downstream signaling rather than the primary genetic defect. In Phelan–McDermid syndrome, recombinant human insulin-like growth factor-1 (IGF-1) was evaluated in a small pilot controlled trial, with preliminary improvement in selected social and repetitive-behavior measures [95]. Subsequent studies have explored growth hormone as an indirect strategy for increasing IGF-1 signaling. A 2022 proof-of-concept study by Sethuram et al. reported feasibility and exploratory clinical signals, but these approaches remain investigational and the small sample sizes preclude definitive conclusions [96].

These studies illustrate both the potential and limitations of synaptic precision therapy: molecularly rational interventions can generate clinically relevant signals, but small cohorts, functional heterogeneity, variable outcome measures, and limited controlled evidence frequently prevent firm conclusions about efficacy or disease modification.

5.6. Metabolic and Cofactor-Responsive Disorders

Metabolic disorders provide some of the clearest examples in which an etiological diagnosis can directly alter treatment. This section uses cerebral creatine deficiency syndromes as an illustrative model rather than as an exhaustive review of metabolic causes of epilepsy and neurodevelopmental dysfunction. Numerous other treatable inherited metabolic disorders can present with developmental impairment, epilepsy, and behavioral or autistic features, but their systematic discussion is beyond the scope of this genetically focused review [97]. Cerebral creatine deficiency syndromes comprise arginine:glycine amidinotransferase deficiency (GATM/AGAT), guanidinoacetate methyltransferase deficiency (GAMT), and creatine transporter deficiency (SLC6A8). These disorders may present with intellectual disability, severe speech impairment, autistic features, movement abnormalities, and epilepsy, yet their therapeutic responsiveness differs substantially according to the affected metabolic step [98].

In GAMT deficiency, creatine supplementation combined with strategies to reduce guanidinoacetate accumulation can substantially improve clinical outcome, particularly when treatment begins early. AGAT deficiency is likewise responsive to creatine replacement and represents an especially compelling example of the potential value of presymptomatic or very early etiological diagnosis [98,99].

In contrast, SLC6A8-related creatine transporter deficiency is considerably more difficult to treat because the primary defect involves cellular creatine transport, including transport into the brain. Creatine supplementation alone or in combination with precursor amino acids has produced variable and generally incomplete clinical responses, emphasizing that biochemical knowledge of a deficient pathway does not necessarily translate into effective substrate replacement when transport itself is impaired [100].

These disorders demonstrate particularly clearly why early etiological diagnosis matters. In selected metabolic diseases, treatment before substantial developmental dysfunction has accumulated may fundamentally alter outcome, whereas later biochemical correction may improve specific functions without fully reversing established neurodevelopmental abnormalities [98,99,100].

5.7. Gene- and Molecularly Targeted Therapies

The most direct form of precision therapy aims to modify expression of the disease-causing gene or its transcript. Antisense oligonucleotides (ASOs), RNA-based strategies, gene replacement, gene activation, and related molecular approaches are rapidly expanding the therapeutic landscape of genetic neurodevelopmental disorders. Unlike conventional antiseizure treatments, these strategies have the potential to intervene upstream of network dysfunction and therefore raise the possibility of modifying both epilepsy and associated developmental manifestations [89].

Dravet syndrome now provides an important example of an investigational molecularly targeted therapy. Zorevunersen is an investigational ASO designed to increase productive SCN1A expression from the unaffected allele and thereby increase NaV1.1 expression. In 2026, Laux et al. reported phase 1–2a open-label studies in children and adolescents with Dravet syndrome, with safety and pharmacokinetics as principal objectives and exploratory evidence of reductions in convulsive seizure frequency and improvement in measures of adaptive behavior. These early clinical findings support continued investigation of zorevunersen as a potential disease-modifying therapy but do not establish definitive efficacy, long-term clinical benefit, or an established disease-modifying treatment [101].

A parallel investigational molecular strategy has entered early-phase clinical evaluation in Angelman syndrome. In neurons, the paternal UBE3A allele is normally transcriptionally silenced by the UBE3A antisense transcript (UBE3A-ATS).

Rugonersen is an investigational ASO designed to bind and reduce UBE3A-ATS, thereby releasing this antisense-mediated silencing and increasing expression of the otherwise inactive paternal UBE3A allele. In the phase 1, open-label TANGELO trial, rugonersen was administered to children with Angelman syndrome and showed an acceptable safety and tolerability profile, dose-dependent partial normalization of the characteristic EEG abnormality, and exploratory signals of clinical improvement beyond expectations from natural-history data [102]. These findings support continued clinical development and demonstrate the feasibility of molecular reactivation of paternal UBE3A in humans; however, controlled evidence is still required to establish clinical efficacy, durability of benefit, and whether the intervention provides disease-modifying effects.

Gene- and RNA-targeted therapies should consequently be regarded as a heterogeneous and rapidly evolving therapeutic class, ranging from preclinical proof-of-concept strategies to clinically advanced investigational programs with early human efficacy signals, rather than as established disease-modifying therapies for genetic ASD–epilepsy disorders [89].

An additional challenge for gene- and RNA-targeted therapies is effective delivery to the central nervous system (CNS). Systemically administered molecular therapies may be limited by blood–brain barrier (BBB) penetration, peripheral exposure, and off-target effects, whereas direct CNS delivery can improve target exposure but introduces procedural and distributional constraints. These challenges are particularly relevant to nucleic acid-based therapies, including antisense oligonucleotides, for which systemic CNS delivery remains difficult because of both the BBB and the physicochemical properties of oligonucleotides [103]. Intrathecal or intracerebroventricular administration can bypass the BBB and facilitate CNS exposure, but therapeutic efficacy still depends on adequate biodistribution and uptake within disease-relevant brain regions and cell populations [103]. Similar considerations apply to CNS-directed gene therapy, in which route of administration, vector tropism, tissue and cell-type specificity, peripheral exposure, and distribution within the CNS influence both efficacy and safety [104]. Thus, successful molecular correction depends not only on target specificity but also on achieving sufficient and appropriately distributed CNS exposure, while the optimal route, dose, durability, and timing of intervention may differ according to the therapeutic platform and developmental stage.

5.8. The Clinical Correlation Paradox: Why Seizure Control Does Not Necessarily Normalize Neurodevelopment

The development of precision therapies exposes a fundamental clinical question in genetic ASD–epilepsy disorders: if seizures and neurodevelopmental impairment arise from a shared molecular defect, why does successful seizure control not consistently produce parallel improvement in cognition, language, or autistic features?

The concept of developmental and epileptic encephalopathy provides part of the answer. Neurodevelopmental impairment may reflect both direct developmental consequences of the underlying etiology and additional adverse effects of epileptic activity. Their relative contributions differ among disorders, individuals, and developmental stages. Seizure suppression may therefore have markedly different neurodevelopmental consequences depending on when, where, and to what extent epileptic activity participates in the disease process [11].

A second component of the paradox is developmental timing. Correction of a primary molecular defect after abnormal circuit maturation has occurred may improve neuronal function without fully reversing changes established during earlier sensitive periods. Dose, tissue distribution, cell-type specificity, degree of molecular correction, and especially the developmental stage at which treatment is delivered may therefore influence whether gene- or RNA-targeted therapy produces seizure reduction, broader neurodevelopmental benefit, or both [105,106].

This review proposes three complementary mechanistic paradigms to conceptualize this heterogeneity. These paradigms are intended as a conceptual framework rather than as established diagnostic categories.

5.8.1. Electrical Interference Paradigm

In the Electrical Interference Paradigm, epileptic seizures or persistent interictal epileptiform activity contribute directly to dysfunction of a developing network that retains substantial developmental potential. Abnormal electrical activity during sensitive periods may interfere with language acquisition, learning, memory, synaptic plasticity, and experience-dependent circuit refinement. Under these circumstances, early and effective suppression of pathological activity may permit meaningful developmental improvement [11,106].

The epilepsy–aphasia spectrum associated with GRIN2A and developmental and/or epileptic encephalopathy with spike-wave activation in sleep (DEE-SWAS) provide representative examples. In these disorders, the timing, distribution, and burden of epileptiform activity may contribute substantially to cognitive and language dysfunction, making rapid suppression of pathological electrical activity an important potential component of developmental rescue [10,51,52].

TSC is a particularly informative, but not pure, example of the Electrical Interference Paradigm. A systematic review and meta-analysis found that ASD in children with TSC was associated with seizures and, more specifically, with infantile spasms compared with other seizure types [107]. Longitudinal evidence published in 2026 further identified developmental trajectories in which infantile spasms, alone and in combination with greater cortical tuber burden, were associated with later autistic features in children with TSC [108]. These findings support the possibility that early severe epileptic activity may contribute to, or amplify, neurodevelopmental risk. However, they do not establish a direct causal relationship between infantile spasms and autism in TSC. In a prospective infant cohort, epilepsy was associated with increased ASD risk, but atypical developmental trajectories were already detectable before seizure onset in a subgroup, and early versus conventional vigabatrin treatment did not reduce ASD risk at 24 months [109]. EPISTOP nevertheless demonstrated that vigabatrin initiated after emergence of epileptiform EEG abnormalities but before clinical seizures reduced subsequent epilepsy-related outcomes, including infantile spasms [90]. Thus, early suppression of epileptic activity may potentially modify one component of neurodevelopmental risk, while primary mTOR-related developmental abnormalities remain independently relevant.

5.8.2. Kinetic Race Paradigm

In the Kinetic Race Paradigm, therapeutic outcome depends on the temporal competition between progressive pathological neurodevelopment and effective intervention. The primary molecular defect initiates abnormal cellular and circuit development, but there may be a developmental window during which correction can prevent or substantially reduce the accumulation of persistent dysfunction.

Treatable metabolic encephalopathies provide particularly clear examples. In GAMT and AGAT deficiencies, early metabolic correction can markedly improve outcome, whereas treatment initiated after prolonged developmental impairment may correct or improve the biochemical disturbance without fully reversing established cognitive, language, or behavioral abnormalities [98,99].

TSC may also partially overlap with this paradigm because epileptiform abnormalities and seizures frequently emerge during infancy, when neural circuits are undergoing rapid maturation. Earlier intervention may therefore have greater potential to preserve developmental function than later suppression of established epilepsy, although the extent to which seizure prevention or early seizure control independently modifies neurodevelopmental outcome remains uncertain. The critical variable is not simply whether an intervention is mechanistically appropriate, but when it is delivered relative to the evolving developmental pathology [64,90].

The same principle is increasingly relevant to molecular therapies. The early clinical experience with zorevunersen and rugonersen demonstrates that primary molecular abnormalities can now be manipulated in humans, but whether earlier correction will yield greater developmental rescue remains a central unanswered question [101,102].

5.8.3. Independent Dual-Track Paradigm

In the Independent Dual-Track Paradigm, epilepsy and ASD-related neurodevelopmental dysfunction are partially parallel consequences of the same underlying genetic alteration. The molecular defect affects multiple cellular or circuit processes, some of which promote seizures while others independently disrupt cognition, communication, behavior, or developmental plasticity.

This model may be particularly relevant to disorders involving synaptic and transcriptional regulators such as SHANK3, SYNGAP1, MECP2, UBE3A, and FMR1. As discussed in Section 3, these disorders can directly perturb synaptic maturation, activity-dependent plasticity, transcriptional or translational regulation, and circuit development. Seizures may therefore represent one consequence of the underlying molecular disorder rather than the necessary mediator of the developmental phenotype. Effective seizure control can consequently reduce one component of disease burden without correcting developmental abnormalities produced directly by the primary genetic defect [55,56,68,69,71].

Clinical observations from the previously reported pediatric ASD cohort provide additional support for this heterogeneous relationship between seizure control and neurodevelopmental outcome. Among children with ASD and epilepsy, 58% achieved seizure freedom and a further 27% experienced a >50% reduction in seizure frequency; however, improved seizure control was not accompanied by a corresponding improvement in autistic behavior [7]. These observations cannot establish mechanistic independence between epilepsy and ASD, but they illustrate that successful suppression of seizures does not necessarily translate into improvement in autistic behavior. This clinical dissociation is consistent with the possibility that, in at least a subset of individuals, epilepsy and ASD represent partially parallel manifestations of shared neurodevelopmental pathology.

5.8.4. Integrating the Three Paradigms

These three paradigms should not be regarded as mutually exclusive diagnostic categories. Rather, they represent complementary positions within a mechanistic continuum, and a single genetic disorder—or even an individual patient—may incorporate more than one mechanism at different developmental stages.

TSC provides a particularly informative example. Primary TSC1/TSC2–mTOR dysregulation directly alters brain development, cellular growth, and synaptic function, while structural abnormalities and emerging epileptic activity provide additional sources of network dysfunction. TSC may therefore combine features of the Independent Dual-Track, Electrical Interference, and Kinetic Race paradigms, with their relative contributions changing across development [62,64,90].

More broadly, the therapeutic implications of these paradigms differ. When epileptic activity is a major developmental disruptor, rapid seizure and EEG control may itself provide substantial developmental benefit. When the underlying molecular defect progressively alters circuit maturation, treatment may need to be delivered within an early developmental window. When epilepsy and neurodevelopmental dysfunction arise through partially parallel consequences of the same genetic defect, seizure suppression alone is unlikely to normalize cognition or autistic features, and upstream molecular or pathway-directed therapy may be required.

The central objective of precision therapeutics in the ASD–epilepsy spectrum should therefore extend beyond seizure reduction. A mechanistically informed strategy requires identification of which component of the disease process is being targeted, at what developmental stage, and with which expected outcome. Genetic diagnosis, functional variant interpretation, biomarkers of target engagement and network dysfunction, and developmentally sensitive clinical endpoints will increasingly need to be integrated if emerging molecular therapies are to achieve meaningful modification of both epileptic and neurodevelopmental trajectories.

The principal features and therapeutic implications of these three mechanistic paradigms are summarized in Table 4.

This distinction has major implications for precision therapeutics. Treatment success should not be defined solely by seizure freedom. For molecular and gene-targeted interventions, clinically meaningful outcomes should also include cognition, language, adaptive behavior, social communication, motor development, and quality of life, assessed over sufficiently long periods to detect changes in developmental trajectory [105,106]. Conversely, failure to normalize neurodevelopment despite excellent seizure control should not automatically be interpreted as failure of antiseizure treatment; it may instead indicate that the neurodevelopmental phenotype is driven substantially by the underlying molecular disorder [10,106].

Ultimately, the relationship between epilepsy and ASD is better conceptualized as a dynamic developmental interaction between shared genetic vulnerability, intrinsic network dysfunction, epileptic activity, therapeutic timing, and developmental plasticity, rather than as a simple unidirectional causal sequence. This framework reconciles shared molecular mechanisms with the variable clinical relationship between seizure burden and autistic or cognitive outcomes and provides a rationale for therapeutic strategies that address both epileptic activity and the underlying neurodevelopmental disorder.

6. Limitations of the Review

This review has several limitations inherent to its narrative design. Study identification and selection were relevance-based rather than systematic, and no formal risk-of-bias assessment or quantitative evidence synthesis was performed. Consequently, the selection and relative weighting of individual studies may have been influenced by author judgment, and the evidence discussed should not be interpreted as an exhaustive or systematically graded representation of the literature. In addition, the substantial heterogeneity of genetic disorders, phenotypic definitions, study designs, and therapeutic endpoints across the ASD–epilepsy spectrum limits direct comparison between studies.

7. Conclusions and Future Directions

The co-occurrence of ASD and epilepsy is increasingly understood as a complex neurodevelopmental spectrum arising from partially shared genetic and biological mechanisms. Advances in genomic medicine have revealed convergence across apparently distinct molecular disorders, including ion-channel dysfunction, abnormalities of synaptic organization and neurotransmission, altered receptor signaling, dysregulation of intracellular pathways, and disruption of transcriptional and translational control. At the systems level, these mechanisms converge on impaired circuit homeostasis, altered excitation–inhibition regulation, disrupted synaptic plasticity, and abnormal network dynamics. This convergence provides a mechanistic framework for understanding why epilepsy, ASD-related features, intellectual disability, and other neurodevelopmental manifestations frequently coexist, while remaining highly variable across individuals and genetic disorders [5,6].

Genetic diagnosis is therefore becoming an increasingly important component of clinical characterization in individuals with co-occurring epilepsy and ASD, particularly when accompanied by early seizure onset, intellectual disability, developmental regression, dysmorphism, or other neurological abnormalities. However, identification of a pathogenic variant should be regarded as the beginning rather than the endpoint of precision medicine. Interpretation increasingly requires consideration of variant-specific functional effects, cell-type and circuit vulnerability, developmental stage, and the extent to which the identified molecular mechanism is therapeutically accessible. Genomic information can therefore refine and complement clinical syndromic classification rather than simply replace it [6,13,14,15,16].

The therapeutic consequences of this molecular stratification are already evident. Function-guided treatment of ion channelopathies, pathway-directed therapy in mTOR-related disorders, correction of selected metabolic defects, and emerging gene- and RNA-targeted approaches demonstrate that molecular diagnosis can influence therapeutic decision-making. Nevertheless, the evidence supporting these strategies remains heterogeneous. For many proposed precision therapies, clinical experience is still based on small cohorts, observational studies, or early-phase trials, and mechanistic plausibility should not be equated with demonstrated clinical efficacy or disease modification [89].

A major challenge for the next generation of precision therapeutics will therefore be to determine whether correction of a molecular defect can alter neurodevelopmental trajectory rather than merely reduce seizures or modify an electrophysiological or molecular biomarker. This distinction is particularly important in DEEs, in which developmental impairment may reflect both the direct effects of the underlying etiology and the additional consequences of epileptic activity [10,106]. The therapeutic window may differ substantially among disorders: in some conditions, early suppression of epileptic activity may preserve developmental function, whereas in others, intervention may need to precede overt seizures or substantial circuit maturation to achieve meaningful disease modification.

The development of reliable biomarkers represents a parallel priority. EEG measures, quantitative neuroimaging, molecular biomarkers, digital phenotyping, and other physiological readouts may provide intermediate indicators of target engagement and network response. However, improvement in a biomarker should not automatically be interpreted as evidence of meaningful neurodevelopmental benefit. Biomarkers will be most valuable when validated against longitudinal cognitive, behavioral, language, motor, and adaptive outcomes and when their relationship to clinically meaningful change is established [105,106].

Clinical-trial design must evolve in parallel with these therapeutic advances. Conventional seizure-frequency endpoints remain essential but are insufficient for interventions intended to modify the underlying neurodevelopmental disorder. Trials should incorporate developmentally appropriate measures of cognition, communication, adaptive behavior, motor function, behavior, and quality of life, together with sufficiently long follow-up to detect changes in developmental trajectory. Robust natural-history data and controls, age-dependent endpoints, and strategies to address floor effects in individuals with profound impairment will be particularly important. Recent work on operational definitions of DEEs similarly emphasizes the need for precisely defined populations and targeted outcome measures as precision therapies enter clinical trials [106].

Finally, the rapid expansion of molecular therapeutics raises important practical challenges. Access to genomic diagnosis, functional testing, multidisciplinary expertise, and advanced therapies remains uneven, while increasingly individualized treatments may be associated with substantial financial and logistical demands. The future of precision medicine in ASD–epilepsy disorders will therefore depend not only on therapeutic innovation but also on international collaboration, standardized outcome assessment, long-term safety surveillance, and equitable access to diagnosis and treatment [89,105].

In conclusion, genetic convergence between ASD and epilepsy does not imply a single molecular mechanism or a uniform therapeutic strategy. Rather, diverse genetic alterations converge on common disturbances of developing neural circuits, while the resulting clinical phenotype is shaped by variant function, cell type, developmental timing, network plasticity, and epileptic activity. The central challenge for precision therapeutics is therefore no longer simply to identify which gene is affected, but to determine how, where, and when the molecular defect alters the developing brain—and when that process remains therapeutically modifiable. Addressing these questions will be essential if precision medicine is to progress from seizure-directed treatment toward genuine modification of neurodevelopmental outcome.

References

1.American Psychiatric AssociationDiagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM-5-TR) American Psychiatric Association PublishingWashington, DC, USA2022 10.1176/appi.books.9780890425787

2.Amiet C. Gourfinkel-An I. Bouzamondo A. Tordjman S. Baulac M. Lechat P. Mottron L. Cohen D. Epilepsy in autism is associated with intellectual disability and gender: Evidence from a meta-analysis Biol. Psychiatry 2008 64 577 582 10.1016/j.biopsych.2008.04.030 18565495

3.Liu X. Sun X. Sun C. Zou M. Chen Y. Huang J. Wu L. Chen W.X. Prevalence of epilepsy in autism spectrum disorders: A systematic review and meta-analysis Autism 2022 26 33 50 10.1177/13623613211045029 34510916

4.Lee B.H. Smith T. Paciorkowski A.R. Autism spectrum disorder and epilepsy: Disorders with a shared biology Epilepsy Behav. 2015 47 191 201 10.1016/j.yebeh.2015.03.017 PMC4475437 25900226

5.Bourgeron T. From the genetic architecture to synaptic plasticity in autism spectrum disorder Nat. Rev. Neurosci. 2015 16 551 563 10.1038/nrn3992 26289574

6.Specchio N. Di Micco V. Trivisano M. Ferretti A. Curatolo P. The epilepsy-autism spectrum disorder phenotype in the era of molecular genetics and precision therapy Epilepsia 2022 63 6 21 10.1111/epi.17115 34741464

7.Pacheva I. Ivanov I. Yordanova R. Gaberova K. Galabova F. Panova M. Petkova A. Timova E. Sotkova I. Epilepsy in children with autistic spectrum disorder Children 2019 6 1510.3390/children6020015 PMC6406948 30691036

8.Fisher R.S. Acevedo C. Arzimanoglou A. Bogacz A. Cross J.H. Elger C.E. Engel J. Jr.Forsgren L. French J.A. Glynn M. A practical clinical definition of epilepsy Epilepsia 2014 55 475 482 10.1111/epi.12550 24730690

9.Tan C. Frewer V. Cox G. Williams K. Ure A. Prevalence and age of onset of regression in children with autism spectrum disorder: A systematic review and meta-analytical update Autism Res. 2021 14 582 598 10.1002/aur.2463 33491292

10.Scheffer I.E. Zuberi S. Mefford H.C. Guerrini R. McTague A. Developmental and epileptic encephalopathies Nat. Rev. Dis. Primers 2024 10 61 10.1038/s41572-024-00546-6 39237642

11.Scheffer I.E. Berkovic S. Capovilla G. Connolly M.B. French J. Guilhoto L. Hirsch E. Jain S. Mathern G.W. Moshé S.L. ILAE classification of the epilepsies: Position paper of the ILAE Commission for Classification and Terminology Epilepsia 2017 58 512 521 10.1111/epi.13709 PMC5386840 28276062

12.Jenner L. Richards C. Howard R. Moss J. Heterogeneity of autism characteristics in genetic syndromes: Key considerations for assessment and support Curr. Dev. Disord. Rep. 2023 10 132 146 10.1007/s40474-023-00276-6 PMC10169182 37193200

13.Tammimies K. Marshall C.R. Walker S. Kaur G. Thiruvahindrapuram B. Lionel A.C. Yuen R.K.C. Uddin M. Roberts W. Weksberg R. Molecular diagnostic yield of chromosomal microarray analysis and whole-exome sequencing in children with autism spectrum disorder JAMA 2015 314 895 903 10.1001/jama.2015.10078 26325558

14.Stefanski A. Calle-López Y. Leu C. Pérez-Palma E. Pestana-Knight E. Lal D. Clinical sequencing yield in epilepsy, autism spectrum disorder, and intellectual disability: A systematic review and meta-analysis Epilepsia 2021 62 143 151 10.1111/epi.16755 PMC7839709 33200402

15.Chang Y.-M. Huang Y.-T. Lai P.-C. Genetic testing for diagnosing neurodevelopmental disorders and epilepsy: A systematic review and meta-analysis Syst. Rev. 2025 14 155 10.1186/s13643-025-02896-y PMC12302900 40722196

16.Lob K. Sawka D.M. Gaitanis J.N. Liu J.S. Nie D.A. Genetic diagnostic yield in autism spectrum disorder (ASD) and epilepsy phenotypes in children with genetically defined ASD J. Autism Dev. Disord. 2025 55 4324 4336 10.1007/s10803-024-06512-1 39136901

17.Bisba M. Malamaki C. Constantoulakis P. Vittas S. Chromosome 15q11-q13 duplication syndrome: A review of the literature and 14 new cases Genes 2024 15 130410.3390/genes15101304 PMC11507414 39457428

18.Rein B. Yan Z. 16p11.2 copy number variations and neurodevelopmental disorders Trends Neurosci. 2020 43 886 901 10.1016/j.tins.2020.09.001 PMC7606557 32993859

19.Zinkstok J.R. Boot E. Bassett A.S. Hiroi N. Butcher N.J. Vingerhoets C. Vorstman J.A.S. van Amelsvoort T.A.M.J. Neurobiological perspective of 22q11.2 deletion syndrome Lancet Psychiatry 2019 6 951 960 10.1016/S2215-0366(19)30076-8 PMC7008533 31395526

20.Bernier R. Steinman K.J. Reilly B. Wallace A.S. Sherr E.H. Pojman N. Mefford H.C. Gerdts J. Earl R. Hanson E. Clinical phenotype of the recurrent 1q21.1 copy-number variant Genet. Med. 2016 18 341 349 10.1038/gim.2015.78 PMC7263044 26066539

21.Lowther C. Costain G. Stavropoulos D.J. Melvin R. Silversides C.K. Andrade D.M. So J. Faghfoury H. Lionel A.C. Marshall C.R. Delineating the 15q13.3 microdeletion phenotype: A case series and comprehensive review of the literature Genet. Med. 2015 17 149 157 10.1038/gim.2014.83 PMC4464824 25077648

22.de Kovel C.G.F. Trucks H. Helbig I. Mefford H.C. Baker C. Leu C. Kluck C. Muhle H. von Spiczak S. Ostertag P. Recurrent microdeletions at 15q11.2 and 16p13.11 predispose to idiopathic generalized epilepsies Brain 2010 133 23 32 10.1093/brain/awp262 PMC2801323 19843651

23.de Coo I.F.M. Jesse S. Le T.-L. Sala C. European Phelan-McDermid Syndrome ConsortiumBourgeron T. Consensus recommendations on epilepsy in Phelan-McDermid syndrome Eur. J. Med. Genet. 2023 66 104746 10.1016/j.ejmg.2023.104746 36967043

24.Steinman K.J. Spence S.J. Ramocki M.B. Proud M. Kessler S.K. Marco E.J. Snyder L.G. D’Angelo D. Chen Q. Chung W.K. 16p11.2 deletion and duplication: Characterizing neurologic phenotypes in a large clinically ascertained cohort Am. J. Med. Genet. Part A 2016 170 2943 2955 10.1002/ajmg.a.37820 27410714

25.Moufawad El Achkar C. Rosen A. Kessler S.K. Steinman K.J. Spence S.J. Ramocki M. Marco E.J. Snyder L.G. Spiro J.E. Chung W.K. Clinical characteristics of seizures and epilepsy in individuals with recurrent deletions and duplications in the 16p11.2 region Neurol. Genet. 2022 8 e200018 10.1212/NXG.0000000000200018 PMC9756306 36531974

26.Wenger T.L. Miller J.S. DePolo L.M. de Marchena A.B. Clements C.C. Emanuel B.S. Zackai E.H. McDonald-McGinn D.M. Schultz R.T. 22q11.2 duplication syndrome: Elevated rate of autism spectrum disorder and need for medical screening Mol. Autism 2016 7 27 10.1186/s13229-016-0090-z PMC4859984 27158440

27.Conant K.D. Finucane B. Cleary N. Martin A. Muss C. Delany M. Murphy E.K. Rabe O. Luchsinger K. Spence S.J. A survey of seizures and current treatments in 15q duplication syndrome Epilepsia 2014 55 396 402 10.1111/epi.12530 24502430

28.Tropeano M. Ahn J.W. Dobson R.J.B. Breen G. Rucker J. Dixit A. Pal D.K. McGuffin P. Farmer A. White P.S. Male-biased autosomal effect of 16p13.11 copy number variation in neurodevelopmental disorders PLoS ONE 2013 8 e6136510.1371/journal.pone.0061365 PMC3630198 23637818

29.Wither R.G. Borlot F. MacDonald A. Butcher N.J. Chow E.W.C. Bassett A.S. Andrade D.M. 22q11.2 deletion syndrome lowers seizure threshold in adult patients without epilepsy Epilepsia 2017 58 1095 1101 10.1111/epi.13748 28448680

30.Olsen L. Sparsø T. Weinsheimer S.M. Dos Santos M.B.Q. Mazin W. Rosengren A. Sanchez X.C. Hoeffding L.K. Schmock H. Baekvad-Hansen M. Prevalence of rearrangements in the 22q11.2 region and population-based risk of neuropsychiatric and developmental disorders in a Danish population: A case-cohort study Lancet Psychiatry 2018 5 573 580 10.1016/S2215-0366(18)30168-8 PMC6560180 29886042

31.Srivastava S. Sahin M. Buxbaum J.D. Berry-Kravis E. Soorya L.V. Thurm A. Bernstein J.A. Asante-Otoo A. Bennett W.E. Jr.Betancur C. Updated consensus guidelines on the management of Phelan-McDermid syndrome Am. J. Med. Genet. Part A 2023 191 2015 2044 10.1002/ajmg.a.63312 PMC10524678 37392087

32.De Rubeis S. He X. Goldberg A.P. Poultney C.S. Samocha K. Cicek A.E. Kou Y. Liu L. Fromer M. Walker S. Synaptic, transcriptional and chromatin genes disrupted in autism Nature 2014 515 209 215 10.1038/nature13772 PMC4402723 25363760

33.Satterstrom F.K. Kosmicki J.A. Wang J. Breen M.S. De Rubeis S. An J.-Y. Peng M. Collins R. Grove J. Klei L. Large-scale exome sequencing study implicates both developmental and functional changes in the neurobiology of autism Cell 2020 180 568 584.e23 10.1016/j.cell.2019.12.036 PMC7250485 31981491

34.Nelson S.B. Valakh V. Excitatory/inhibitory balance and circuit homeostasis in autism spectrum disorders Neuron 2015 87 684 698 10.1016/j.neuron.2015.07.033 PMC4567857 26291155

35.Sohal V.S. Rubenstein J.L.R. Excitation-inhibition balance as a framework for investigating mechanisms in neuropsychiatric disorders Mol. Psychiatry 2019 24 1248 1257 10.1038/s41380-019-0426-0 PMC6742424 31089192

36.Ebert D.H. Greenberg M.E. Activity-dependent neuronal signalling and autism spectrum disorder Nature 2013 493 327 337 10.1038/nature11860 PMC3576027 23325215

37.Yu F.H. Mantegazza M. Westenbroek R.E. Robbins C.A. Kalume F. Burton K.A. Spain W.J. McKnight G.S. Scheuer T. Catterall W.A. Reduced sodium current in GABAergic interneurons in a mouse model of severe myoclonic epilepsy in infancy Nat. Neurosci. 2006 9 1142 1149 10.1038/nn1754 16921370

38.Han S. Tai C. Westenbroek R.E. Yu F.H. Cheah C.S. Potter G.B. Rubenstein J.L. Scheuer T. de la Iglesia H.O. Catterall W.A. Autistic-like behaviour in Scn1a+/− mice and rescue by enhanced GABA-mediated neurotransmission Nature 2012 489 385 390 10.1038/nature11356 PMC3448848 22914087

39.Wolff M. Johannesen K.M. Hedrich U.B.S. Masnada S. Rubboli G. Gardella E. Lesca G. Ville D. Milh M. Villard L. Genetic and phenotypic heterogeneity suggest therapeutic implications in SCN2A-related disorders Brain 2017 140 1316 1336 10.1093/brain/awx054 28379373

40.Spratt P.W.E. Ben-Shalom R. Keeshen C.M. Burke K.J. Jr.Clarkson R.L. Sanders S.J. Bender K.J. The autism-associated gene Scn2a contributes to dendritic excitability and synaptic function in the prefrontal cortex Neuron 2019 103 673 685.e5 10.1016/j.neuron.2019.05.037 PMC7935470 31230762

41.Conecker G. Xia M.Y. Hecker J. Achkar C. Cukiert C. Devries S. Donner E. Fitzgerald M.P. Gardella E. Hammer M. Global modified Delphi consensus on diagnosis, phenotypes, and treatment of SCN8A-related epilepsy and/or neurodevelopmental disorders Epilepsia 2024 65 2322 2338 10.1111/epi.17992 38802994

42.Wagnon J.L. Barker B.S. Ottolini M. Park Y. Volkheimer A. Valdez P. Swinkels M.E.M. Patel M.K. Meisler M.H. Loss-of-function variants of SCN8A in intellectual disability without seizures Neurol. Genet. 2017 3 e170 10.1212/NXG.0000000000000170 PMC5499976 28702509

43.Millichap J.J. Park K.L. Tsuchida T. Ben-Zeev B. Carmant L. Flamini R. Joshi N. Levisohn P.M. Marsh E. Nangia S. KCNQ2 encephalopathy: Features, mutational hot spots, and ezogabine treatment of 11 patients Neurol. Genet. 2016 2 e96 10.1212/NXG.0000000000000096 PMC4995058 27602407

44.Simkin D. Marshall K.A. Vanoye C.G. Desai R.R. Bustos B.I. Piyevsky B.N. Ortega J.A. Forrest M. Robertson G.L. Penzes P. Dyshomeostatic modulation of Ca2+-activated K+ channels in a human neuronal model of KCNQ2 encephalopathy eLife 2021 10 e64434 10.7554/eLife.64434 PMC7864629 33544076

45.Rosa F. Theiss S. Krepp S. Loeffler H. Mendis D.C. Klingenstein S. Liebau S. Weckhuysen S. Alber M. Petrou S. KV7.2 channel dysfunction delays neuronal maturation and undermines early network development in a hiPSC model of KCNQ2-DEE Neurobiol. Dis. 2025 216 107120 10.1016/j.nbd.2025.107120 40998073

46.Miceli F. Millevert C. Soldovieri M.V. Mosca I. Ambrosino P. Carotenuto L. Schrader D. Lee H.K. Riviello J. Hong W. KCNQ2 R144 variants cause neurodevelopmental disability with language impairment and autistic features without neonatal seizures through a gain-of-function mechanism eBioMedicine 2022 81 104130 10.1016/j.ebiom.2022.104130 PMC9254340 35780567

47.Barcia G. Fleming M.R. Deligniere A. Gazula V.R. Brown M.R. Langouet M. Chen H. Kronengold J. Abhyankar A. Cilio R. De novo gain-of-function KCNT1 channel mutations cause malignant migrating partial seizures of infancy Nat. Genet. 2012 44 1255 1259 10.1038/ng.2441 PMC3687547 23086397

48.Kim G.E. Kaczmarek L.K. Emerging role of the KCNT1 Slack channel in intellectual disability Front. Cell. Neurosci. 2014 8 20910.3389/fncel.2014.00209 PMC4112808 25120433

49.Quraishi I.H. Mercier M.R. McClure H. Couture R.L. Schwartz M.L. Lukowski R. Ruth P. Kaczmarek L.K. Impaired motor skill learning and altered seizure susceptibility in mice with loss or gain of function of the Kcnt1 gene encoding Slack (KNa1.1) Na+-activated K+ channels Sci. Rep. 2020 10 3213 10.1038/s41598-020-60028-z PMC7035262 32081855

50.Korinek M. Candelas Serra M. Abdel Rahman F.E.S. Dobrovolski M. Kuchtiak V. Abramova V. Fili K. Tomovic E. Hrcka Krausova B. Krusek J. Disease-associated variants in GRIN1, GRIN2A and GRIN2B genes: Insights into NMDA receptor structure, function, and pathophysiology Physiol. Res. 2024 73 S413 S434 10.33549/physiolres.935346 PMC11412357 38836461

51.Carvill G.L. Regan B.M. Yendle S.C. O’Roak B.J. Lozovaya N. Bruneau N. Burnashev N. Khan A. Cook J. Geraghty E. GRIN2A mutations cause epilepsy-aphasia spectrum disorders Nat. Genet. 2013 45 1073 1076 10.1038/ng.2727 PMC3868952 23933818

52.Lemke J.R. Lal D. Reinthaler E.M. Steiner I. Nothnagel M. Alber M. Geider K. Laube B. Schwake M. Finsterwalder K. Mutations in GRIN2A cause idiopathic focal epilepsy with rolandic spikes Nat. Genet. 2013 45 1067 1072 10.1038/ng.2728 23933819

53.Platzer K. Yuan H. Schütz H. Winschel A. Chen W. Hu C. Kusumoto H. Heyne H.O. Helbig K.L. Tang S. GRIN2B encephalopathy: Novel findings on phenotype, variant clustering, functional consequences and treatment aspects J. Med. Genet. 2017 54 460 470 10.1136/jmedgenet-2016-104509 PMC5656050 28377535

54.Møller R.S. Wuttke T.V. Helbig I. Marini C. Johannesen K.M. Brilstra E.H. Vaher U. Borggraefe I. Talvik I. Talvik T. Mutations in GABRB3: From febrile seizures to epileptic encephalopathies Neurology 2017 88 483 492 10.1212/WNL.0000000000003565 PMC5278942 28053010

55.Bozdagi O. Sakurai T. Papapetrou D. Wang X. Dickstein D.L. Takahashi N. Kajiwara Y. Yang M. Katz A.M. Scattoni M.L. Haploinsufficiency of the autism-associated Shank3 gene leads to deficits in synaptic function, social interaction, and social communication Mol. Autism 2010 1 15 10.1186/2040-2392-1-15 PMC3019144 21167025

56.Clement J.P. Aceti M. Creson T.K. Ozkan E.D. Shi Y. Reish N.J. Almonte A.G. Miller B.H. Wiltgen B.J. Miller C.A. Pathogenic SYNGAP1 mutations impair cognitive development by disrupting maturation of dendritic spine synapses Cell 2012 151 709 723 10.1016/j.cell.2012.08.045 PMC3500766 23141534

57.Clement J.P. Ozkan E.D. Aceti M. Miller C.A. Rumbaugh G. SYNGAP1 links the maturation rate of excitatory synapses to the duration of critical-period synaptic plasticity J. Neurosci. 2013 33 10447 10452 10.1523/JNEUROSCI.0765-13.2013 PMC3685838 23785156

58.Aceti M. Creson T.K. Vaissiere T. Rojas C. Huang W.C. Wang Y.X. Petralia R.S. Page D.T. Miller C.A. Rumbaugh G. Syngap1 haploinsufficiency damages a postnatal critical period of pyramidal cell structural maturation linked to cortical circuit assembly Biol. Psychiatry 2015 77 805 815 10.1016/j.biopsych.2014.08.001 PMC4326604 25444158

59.Chen W. Cai Z.-L. Chao E.S. Chen H. Longley C.M. Hao S. Chao H.-T. Kim J.H. Messier J.E. Zoghbi H.Y. Stxbp1/Munc18-1 haploinsufficiency impairs inhibition and mediates key neurological features of STXBP1 encephalopathy eLife 2020 9 e48705 10.7554/eLife.48705 PMC7056272 32073399

60.Zhu Y.-C. Xiong Z.-Q. Molecular and synaptic bases of CDKL5 disorder Dev. Neurobiol. 2019 79 8 19 10.1002/dneu.22639 30246934

61.Leonard H. Downs J. Benke T.A. Swanson L. Olson H. Demarest S. CDKL5 deficiency disorder: Clinical features, diagnosis, and management Lancet Neurol. 2022 21 563 576 10.1016/S1474-4422(22)00035-7 PMC9788833 35483386

62.Ehninger D. de Vries P.J. Silva A.J. From mTOR to cognition: Molecular and cellular mechanisms of cognitive impairments in tuberous sclerosis J. Intellect. Disabil. Res. 2009 53 838 851 10.1111/j.1365-2788.2009.01208.x PMC2844770 19694899

63.Winden K.D. Ebrahimi-Fakhari D. Sahin M. Abnormal mTOR activation in autism Annu. Rev. Neurosci. 2018 41 1 23 10.1146/annurev-neuro-080317-061747 29490194

64.Curatolo P. Moavero R. de Vries P.J. Neurological and neuropsychiatric aspects of tuberous sclerosis complex Lancet Neurol. 2015 14 733 745 10.1016/S1474-4422(15)00069-1 26067126

65.Zhou J. Parada L.F. PTEN signaling in autism spectrum disorders Curr. Opin. Neurobiol. 2012 22 873 879 10.1016/j.conb.2012.05.004 22664040

66.Spina Nagy G. Kawamoto E.M. Bridi J.C. The role of PTEN signaling in synaptic function: Implications in autism spectrum disorder Neurosci. Lett. 2021 759 136015 10.1016/j.neulet.2021.136015 34111509

67.LaSarge C.L. Danzer S.C. Mechanisms regulating neuronal excitability and seizure development following mTOR pathway hyperactivation Front. Mol. Neurosci. 2014 7 1810.3389/fnmol.2014.00018 PMC3953715 24672426

68.Ip J.P.K. Mellios N. Sur M. Rett syndrome: Insights into genetic, molecular and circuit mechanisms Nat. Rev. Neurosci. 2018 19 368 382 10.1038/s41583-018-0006-3 PMC6402579 29740174

69.Margolis S.S. Sell G.L. Zbinden M.A. Bird L.M. Angelman syndrome Neurotherapeutics 2015 12 641 650 10.1007/s13311-015-0361-y PMC4489961 26040994

70.Urraca N. Cleary J. Brewer V. Pivnick E.K. McVicar K. Thibert R.L. Schanen N.C. Esmer C. Lamport D. Reiter L.T. The interstitial duplication 15q11.2-q13 syndrome includes autism, mild facial anomalies and a characteristic EEG signature Autism Res. 2013 6 268 279 10.1002/aur.1284 PMC3884762 23495136

71.Hagerman R.J. Berry-Kravis E. Hazlett H.C. Bailey D.B. Jr.Moine H. Kooy R.F. Tassone F. Gantois I. Sonenberg N. Mandel J.-L. Fragile X syndrome Nat. Rev. Dis. Primers 2017 3 17065 10.1038/nrdp.2017.65 28960184

72.Bagni C. Zukin R.S. A synaptic perspective of fragile X syndrome and autism spectrum disorders Neuron 2019 101 1070 1088 10.1016/j.neuron.2019.02.041 PMC9628679 30897358

73.Bernier R. Golzio C. Xiong B. Stessman H.A. Coe B.P. Penn O. Witherspoon K. Gerdts J. Baker C. Vulto-van Silfhout A.T. Disruptive CHD8 mutations define a subtype of autism early in development Cell 2014 158 263 276 10.1016/j.cell.2014.06.017 PMC4136921 24998929

74.Ben-Shalom R. Keeshen C.M. Berrios K.N. An J.Y. Sanders S.J. Bender K.J. Opposing effects on NaV1.2 function underlie differences between SCN2A variants observed in individuals with autism spectrum disorder or infantile seizures Biol. Psychiatry 2017 82 224 232 10.1016/j.biopsych.2017.01.009 PMC5796785 28256214

75.Meisler M.H. SCN8A encephalopathy: Mechanisms and models Epilepsia 2019 60 S86 S91 10.1111/epi.14703 PMC6953611 31904118

76.Larsen J. Carvill G.L. Gardella E. Kluger G. Schmiedel G. Barisic N. Depienne C. Brilstra E. Mang Y. Nielsen J.E.K. The phenotypic spectrum of SCN8A encephalopathy Neurology 2015 84 480 489 10.1212/WNL.0000000000001211 PMC4336074 25568300

77.Wong J.C. Grieco S.F. Dutt K. Chen L. Thelin J.T. Inglis G.A.S. Parvin S. Garraway S.M. Xu X. Goldin A.L. Autistic-like behavior, spontaneous seizures, and increased neuronal excitability in a SCN8A mouse model Neuropsychopharmacology 2021 46 2011 2020 10.1038/s41386-021-00985-9 PMC8429750 33658654

78.Shcheglovitov A. Shcheglovitova O. Yazawa M. Portmann T. Shu R. Sebastiano V. Krawisz A. Froehlich W. Bernstein J.A. Hallmayer J.F. SHANK3 and IGF1 restore synaptic deficits in neurons from 22q13 deletion syndrome patients Nature 2013 503 267 271 10.1038/nature12618 PMC5559273 24132240

79.Vlaskamp D.R.M. Shaw B.J. Burgess R. Mei D. Montomoli M. Xie H. Myers C.T. Bennett M.F. XiangWei W. Williams D. SYNGAP1 encephalopathy: A distinctive generalized developmental and epileptic encephalopathy Neurology 2019 92 e96 e107 10.1212/WNL.0000000000006729 PMC6340340 30541864

80.Paoletti P. Bellone C. Zhou Q. NMDA receptor subunit diversity: Impact on receptor properties, synaptic plasticity and disease Nat. Rev. Neurosci. 2013 14 383 400 10.1038/nrn3504 23686171

81.Absalom N.L. Liao V.W.Y. Johannesen K.M.H. Gardella E. Jacobs J. Lesca G. Gokce-Samar Z. Arzimanoglou A. Zeidler S. Striano P. Gain-of-function and loss-of-function GABRB3 variants lead to distinct clinical phenotypes in patients with developmental and epileptic encephalopathies Nat. Commun. 2022 13 1822 10.1038/s41467-022-29280-x PMC8983652 35383156

82.Hensch T.K. Critical period plasticity in local cortical circuits Nat. Rev. Neurosci. 2005 6 877 888 10.1038/nrn1787 16261181

83.Chen Q. Deister C.A. Gao X. Guo B. Lynn-Jones T. Chen N. Wells M.F. Liu R. Goard M.J. Dimidschstein J. Dysfunction of cortical GABAergic neurons leads to sensory hyper-reactivity in a Shank3 mouse model of ASD Nat. Neurosci. 2020 23 520 532 10.1038/s41593-020-0598-6 PMC7131894 32123378

84.Hensch T.K. Fagiolini M. Excitatory-inhibitory balance and critical period plasticity in developing visual cortex Prog. Brain Res. 2005 147 115 124 10.1016/S0079-6123(04)47009-5 15581701

85.Wirrell E.C. Hood V. Knupp K.G. Meskis M.A. Nabbout R. Scheffer I.E. Wilmshurst J. Sullivan J. International consensus on diagnosis and management of Dravet syndrome Epilepsia 2022 63 1761 1777 10.1111/epi.17274 PMC9543220 35490361

86.Mullen S.A. Carney P.W. Roten A. Ching M. Lightfoot P.A. Churilov L. Nair U. Li M. Berkovic S.F. Petrou S. Precision therapy for epilepsy due to KCNT1 mutations: A randomized trial of oral quinidine Neurology 2018 90 e67 e72 10.1212/WNL.0000000000004769 29196578

87.Pestana Knight E.M. Amin S. Bahi-Buisson N. Benke T.A. Cross J.H. Demarest S.T. Olson H.E. Specchio N. Fleming T.R. Aimetti A.A. Safety and efficacy of ganaxolone in patients with CDKL5 deficiency disorder: Results from the double-blind phase of a randomised, placebo-controlled, phase 3 trial Lancet Neurol. 2022 21 417 427 10.1016/S1474-4422(22)00077-1

88.French J.A. Lawson J.A. Yapici Z. Ikeda H. Polster T. Nabbout R. Curatolo P. de Vries P.J. Dlugos D.J. Berkowitz N. Adjunctive everolimus therapy for treatment-resistant focal-onset seizures associated with tuberous sclerosis (EXIST-3): A phase 3, randomised, double-blind, placebo-controlled study Lancet 2016 388 2153 2163 10.1016/S0140-6736(16)31419-2 27613521

89.Wang S. Perucca E. Berkovic S.F. Perucca P. Precision therapies for genetic epilepsies in 2025: Promises and pitfalls Epilepsia Open 2025 early view10.1002/epi4.70065 PMC13394099 40411479

90.Kotulska K. Kwiatkowski D.J. Curatolo P. Weschke B. Riney K. Jansen F. Feucht M. Krsek P. Nabbout R. Jansen A.C. Prevention of epilepsy in infants with tuberous sclerosis complex in the EPISTOP trial Ann. Neurol. 2021 89 304 314 10.1002/ana.25956 PMC7898885 33180985

91.Pierson T.M. Yuan H. Marsh E.D. Fuentes-Fajardo K. Adams D.R. Markello T. Golas G. Simeonov D.R. Holloman C. Tankovic A. GRIN2A mutation and early-onset epileptic encephalopathy: Personalized therapy with memantine Ann. Clin. Transl. Neurol. 2014 1 190 198 10.1002/acn3.39 PMC4019449 24839611

92.Karnstedt M. Perszyk R.E. Myers S.J. McDaniels E. Somorai M. Borggraefe I. Veenma D.C.M. Schoonjans A.-S. Striano P. Fantaneanu T.A. Memantine treatment in individuals with GRIN gain-of-function variants is associated with improvements in behavior, development, and seizure frequency Epilepsia 2026 67 1961 1974 10.1002/epi.70090 PMC13075606 41489401

93.Soto D. Olivella M. Grau C. Armstrong J. Alcon C. Gasull X. Santos-Gómez A. Locubiche S. Gómez de Salazar M. García-Díaz R. L-Serine dietary supplementation is associated with clinical improvement of loss-of-function GRIN2B-related pediatric encephalopathy Sci. Signal. 2019 12 eaaw0936 10.1126/scisignal.aaw0936 31213567

94.Juliá-Palacios N. Olivella M. Sigatullina Bondarenko M. Ibáñez-Micó S. Muñoz-Cabello B. Alonso-Luengo O. Soto-Insuga V. García-Navas D. Cuesta-Herraiz L. Andreo-Lillo P. L-serine treatment in patients with GRIN-related encephalopathy: A phase 2A, non-randomized study Brain 2024 147 1653 1666 10.1093/brain/awae041 38380699

95.Kolevzon A. Bush L. Wang A.T. Halpern D. Frank Y. Grodberg D. Rapaport R. Tavassoli T. Chaplin W. Soorya L. A pilot controlled trial of insulin-like growth factor-1 in children with Phelan-McDermid syndrome Mol. Autism 2014 5 54 10.1186/2040-2392-5-54 PMC4326443 25685306

96.Sethuram S. Levy T. Foss-Feig J. Halpern D. Sandin S. Siper P.M. Walker H. Buxbaum J.D. Rapaport R. Kolevzon A. A proof-of-concept study of growth hormone in children with Phelan-McDermid syndrome Mol. Autism 2022 13 6 10.1186/s13229-022-00485-7 PMC8800321 35093163

97.Hoytema van Konijnenburg E.M.M. Wortmann S.B. Koelewijn M.J. Tseng L.A. Houben R. Stöckler-Ipsiroglu S. Ferreira C.R. van Karnebeek C.D.M. Treatable inherited metabolic disorders causing intellectual disability: 2021 review and digital app Orphanet J. Rare Dis. 2021 16 170 10.1186/s13023-021-01727-2 PMC8042729 33845862

98.Stockler-Ipsiroglu S. van Karnebeek C.D.M. Cerebral creatine deficiencies: A group of treatable intellectual developmental disorders Semin. Neurol. 2014 34 350 356 10.1055/s-0034-1386772 25192512

99.Stöckler-Ipsiroglu S. van Karnebeek C. Longo N. Korenke G.C. Mercimek-Mahmutoglu S. Marquart I. Barshop B. Grolik C. Schlune A. Angle B. Guanidinoacetate methyltransferase (GAMT) deficiency: Outcomes in 48 individuals and recommendations for diagnosis, treatment and monitoring Mol. Genet. Metab. 2014 111 16 25 10.1016/j.ymgme.2013.10.018 24268530

100.Dunbar M. Jaggumantri S. Sargent M. Stockler-Ipsiroglu S. van Karnebeek C.D.M. Treatment of X-linked creatine transporter (SLC6A8) deficiency: Systematic review of the literature and three new cases Mol. Genet. Metab. 2014 112 259 274 10.1016/j.ymgme.2014.05.011 24953403

101.Laux L. Sullivan J. Perry M.S. Brunklaus A. Desurkar A. Schreiber J.M. Roberts C.M. Knupp K.G. Wheless J.W. Wirrell E.C. Zorevunersen in children and adolescents with Dravet syndrome N. Engl. J. Med. 2026 394 969 982 10.1056/NEJMoa2506295 41780062

102.Hipp J.F. Bacino C.A. Bird L.M. Bruenig-Traebert I. Chan D. de Wit M.C. Fontoura P. Hooper G. Jagasia R. Krishnan M.L. The UBE3A-ATS antisense oligonucleotide rugonersen in children with Angelman syndrome: A phase 1 trial Nat. Med. 2025 31 2936 2945 10.1038/s41591-025-03784-7 40646322

103.Goto A. Yamamoto S. Iwasaki S. Biodistribution and delivery of oligonucleotide therapeutics to the central nervous system: Advances, challenges, and future perspectives Biopharm. Drug Dispos. 2023 44 26 47 10.1002/bdd.2338 36336817

104.Ling Q. Herstine J.A. Bradbury A. Gray S.J. AAV-based in vivo gene therapy for neurological disorders Nat. Rev. Drug Discov. 2023 22 789 806 10.1038/s41573-023-00766-7 37658167

105.Specchio N. Trivisano M. Aronica E. Balestrini S. Arzimanoglou A. Colasante G. Cross J.H. Jozwiak S. Wilmshurst J.M. Vigevano F. The expanding field of genetic developmental and epileptic encephalopathies: Current understanding and future perspectives Lancet Child Adolesc. Health 2024 8 821 834 10.1016/S2352-4642(24)00196-2 39419567

106.Scheffer I.E. French J. Valente K.D. Auvin S. Cross J.H. Specchio N. Operational definition of developmental and epileptic encephalopathies to underpin the design of therapeutic trials Epilepsia 2025 66 1014 1023 10.1111/epi.18265 PMC11997937 40013914

107.Mitchell R.A. Barton S.M. Harvey A.S. Ure A.M. Williams K. Factors associated with autism spectrum disorder in children with tuberous sclerosis complex: A systematic review and meta-analysis Dev. Med. Child Neurol. 2021 63 791 801 10.1111/dmcn.14787 33432576

108.McEwen F.S. Runicles A.K. Liang H. Woodhouse E. Underwood L. Shephard E. Barker E.D. Sheerin F. Steenbruggen J.W. Tuberous Sclerosis 2000 Study GroupDevelopmental pathways to autism in tuberous sclerosis complex: Evidence from a longitudinal cohort Epilepsia 2026 67 e191 e197 10.1002/epi.70431 PMC13675775 42573453

109.Moavero R. Kotulska K. Lagae L. Benvenuto A. Emberti Gialloreti L. Weschke B. Riney K. Feucht M. Krsek P. Nabbout R. Is autism driven by epilepsy in infants with tuberous sclerosis complex? Ann. Clin. Transl. Neurol. 2020 7 1371 1381 10.1002/acn3.51128 PMC7448162 32705817