Precision Medicine & Gene Therapy
The convergence of advances in epilepsy genetics, pharmacogenomics, and nucleic acid therapeutics is transforming epilepsy from a condition managed with empirical medication trials to one increasingly amenable to mechanism-based, genotype-driven treatment. Pharmacogenomic testing for HLA alleles now prevents life-threatening drug hypersensitivity reactions. Genetic diagnosis directly guides antiseizure medication (ASM) selection in an expanding number of monogenic epilepsies. Precision therapies targeting specific molecular pathways — mTOR inhibitors for tuberous sclerosis complex, antisense oligonucleotides (ASOs) for Dravet syndrome, and gene replacement strategies for developmental and epileptic encephalopathies — have entered clinical trials or achieved regulatory milestones. While gene editing technologies such as CRISPR remain predominantly in preclinical stages for epilepsy, the trajectory of development suggests that genetic therapies will become a routine component of epilepsy care within the coming decade. This topic reviews the current state and future directions of precision medicine and gene therapy in epilepsy.
Bottom Line
- Pharmacogenomics: HLA-B*15:02 testing is mandatory before carbamazepine/oxcarbazepine in patients of Southeast Asian descent (FDA boxed warning); HLA-A*31:01 testing should be considered in all patients before carbamazepine initiation to reduce risk of severe cutaneous adverse reactions
- Genetic testing: Recommended by the ILAE for any patient with unexplained epilepsy; exome sequencing is the recommended first-line test; diagnostic yield is 17–48% depending on modality, with diagnosis leading to management changes in >70% of cases
- Precision ASM selection: Genetic diagnosis directly informs medication choice in multiple monogenic epilepsies: avoid sodium channel blockers in SCN1A haploinsufficiency (Dravet); use sodium channel blockers for SCN8A gain-of-function; ketogenic diet for GLUT1 deficiency (SLC2A1); quinidine for KCNT1 gain-of-function
- mTOR inhibitors: Everolimus is FDA-approved for seizures associated with tuberous sclerosis complex (EXIST-3 trial); represents the first targeted molecular therapy for any genetic epilepsy
- Antisense oligonucleotides: Zorevunersen (STK-001) for Dravet syndrome is in Phase 3 (EMPEROR trial); uses TANGO technology to increase SCN1A expression by blocking a poison exon; long-term data show durable seizure reductions and cognitive improvements
- Gene replacement: AAV-mediated SCN1A gene replacement for Dravet syndrome has demonstrated seizure reduction in mouse models using split-intein dual-vector approaches to overcome AAV packaging limitations
- CRISPR: CRISPRa (CRISPR activation) has rescued neurological phenotypes in Scn2a haploinsufficient mice (Nature, 2025); CRISPR-based approaches for SCN1A and Angelman syndrome are in preclinical development
Pharmacogenomics in Epilepsy
HLA-B*15:02 and Carbamazepine
The association between the HLA-B*15:02 allele and carbamazepine-induced Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) represents one of the most established pharmacogenomic relationships in neurology. HLA-B*15:02 is prevalent in populations of Southeast Asian descent (Han Chinese, Thai, Malaysian, Filipino, Indonesian), with carrier rates of 5–15%. The FDA issued a boxed warning in 2007 requiring HLA-B*15:02 testing before prescribing carbamazepine to patients of Southeast Asian ancestry. This warning was subsequently extended to oxcarbazepine.
HLA Allele Testing Before ASM Initiation
- HLA-B*15:02: Test before carbamazepine or oxcarbazepine in patients of Southeast Asian, South Asian, or Pacific Islander descent; if positive, avoid carbamazepine and oxcarbazepine (alternative: lacosamide, lamotrigine with slow titration)
- HLA-A*31:01: Associated with carbamazepine-induced SJS/TEN and DRESS (drug reaction with eosinophilia and systemic symptoms) across all ethnic groups, including European descent; carrier rate ~2–5% in Caucasians, 5–10% in Japanese; the Clinical Pharmacogenetics Implementation Consortium (CPIC) recommends considering testing before carbamazepine initiation in all populations
- HLA-B*15:02 positive patients may still be at risk for phenytoin-induced SJS/TEN, though the association is weaker
- The risk of SJS/TEN is highest in the first 2–3 months of therapy; the risk decreases substantially after 6 months of stable therapy
- Lamotrigine-associated SJS/TEN has not been consistently linked to a single HLA allele across populations, though HLA-B*15:02 may confer modest risk; slow titration remains the primary prevention strategy
Other Pharmacogenomic Considerations
Beyond HLA testing, pharmacogenomics in epilepsy includes cytochrome P450 polymorphisms that affect ASM metabolism. CYP2C9 poor metabolizer status increases phenytoin exposure and toxicity risk. CYP2C19 polymorphisms affect the metabolism of clobazam (N-desmethylclobazam accumulation in poor metabolizers). While routine CYP testing is not yet standard of care for ASM prescribing, the evidence base is growing and may support integration into clinical practice as pharmacogenomic platforms become more widely available.
Genetic Testing in Epilepsy
When to Order Genetic Testing
The ILAE Task Force for Clinical Genetic Testing in the Epilepsies (Krey et al., 2022) recommends genetic testing for the following groups of epilepsy patients in whom no other clear cause has been found:
- Severe childhood-onset epilepsies, particularly developmental and epileptic encephalopathies (DEEs)
- Epilepsy plus intellectual disability, autism spectrum disorder, other neurodevelopmental comorbidities, or multiple congenital anomalies
- Progressive epilepsies, including the progressive myoclonus epilepsies
- Nonacquired focal epilepsies in specific familial syndromes (e.g., sleep-related hypermotor epilepsy)
- Nonacquired, focal, drug-resistant epilepsies in the presurgical evaluation
- Epilepsy in the setting of malformations of cortical development
- Strong family history of epilepsy (multiple first- or second-degree members)
Importantly, the National Society of Genetic Counselors recommends considering genetic testing for any patient with unexplained epilepsy, regardless of age or seizure control, given that diagnosis leads to management changes in more than 70% of cases.
Which Test to Order
| Test | What It Detects | Diagnostic Yield | When to Use |
|---|---|---|---|
| Exome sequencing (recommended first-line) | Variants in all protein-coding genes; copy number variants (with add-on analysis) | 24% | First-line for most patients with unexplained epilepsy; order as trio (patient + parents) when possible |
| Genome sequencing | All exonic and intronic variants; structural variants; repeat expansions | 48% | When exome is unrevealing; when noncoding or structural variants suspected; decreasing cost may make this first-line in near future |
| Multigene panel | Targeted set of known epilepsy genes (100–1000+ genes) | 19% | When insurance does not cover exome/genome; when phenotype strongly suggests a known gene |
| Chromosomal microarray | Copy number variants (deletions, duplications ≥50–100 kb) | 9% | When epilepsy + intellectual disability, ASD, or congenital anomalies; may miss single-gene variants |
| Targeted sequencing (Sanger) | Specific gene variants | Variable | When phenotype strongly suggests a specific gene (e.g., MECP2 in Rett syndrome); rarely used alone |
| Karyotype | Ring chromosomes, balanced translocations | Low (specific indications) | Suspected ring chromosome 20; balanced translocations not detected by microarray |
Precision ASM Selection Based on Genetic Diagnosis
For an expanding group of monogenic epilepsies, genetic diagnosis directly informs treatment selection by identifying the molecular target and mechanism of dysfunction. This represents the closest current approximation to precision medicine in epilepsy.
| Gene | Condition | Functional Consequence | Precision Treatment | Medications to Avoid |
|---|---|---|---|---|
| SCN1A | Dravet syndrome | Loss of function (NaV1.1 haploinsufficiency) → reduced inhibitory neuron excitability | Valproate, clobazam, stiripentol, fenfluramine, cannabidiol | Sodium channel blockers (carbamazepine, oxcarbazepine, phenytoin, lamotrigine, lacosamide) — may worsen seizures |
| SCN8A (GoF) | DEE with SCN8A gain-of-function | Gain of function (NaV1.6) → neuronal hyperexcitability | High-dose sodium channel blockers (phenytoin, carbamazepine, oxcarbazepine, lacosamide) | N/A (sodium channel blockers are therapeutic) |
| SCN2A (GoF) | Early-onset DEE (<3 months) | Gain of function (NaV1.2) | Sodium channel blockers | N/A |
| SCN2A (LoF) | Later-onset DEE, ASD + epilepsy | Loss of function (NaV1.2) | Avoid sodium channel blockers; benzodiazepines, levetiracetam may be preferred | Sodium channel blockers |
| KCNQ2 | Self-limited neonatal epilepsy; DEE (dominant negative variants) | Loss of function (KV7.2) | Carbamazepine (self-limited form); retigabine/ezogabine in severe DEE (KV7 opener) | N/A |
| KCNT1 | Sleep-related hypermotor epilepsy; epilepsy of infancy with migrating focal seizures | Gain of function (KNa1.1) | Quinidine (KNa1.1 blocker; variable efficacy); clinical trials ongoing | N/A |
| PRRT2 | Self-limited infantile epilepsy; paroxysmal kinesigenic dyskinesia | Loss of function (synaptic vesicle fusion) | Carbamazepine or oxcarbazepine (highly effective) | Levetiracetam often ineffective |
| SLC2A1 | GLUT1 deficiency syndrome | Impaired glucose transport across blood-brain barrier | Ketogenic diet (provides alternative fuel source via ketone bodies) | Avoid valproate (inhibits fatty acid oxidation); avoid phenobarbital (reduces GLUT1 expression) |
| TSC1/TSC2 | Tuberous sclerosis complex | mTOR pathway overactivation | Everolimus (mTOR inhibitor; FDA-approved for TSC-associated seizures); vigabatrin for infantile spasms | N/A |
| ALDH7A1 | Pyridoxine-dependent epilepsy | Aldehyde dehydrogenase deficiency → pyridoxine deficiency | Pyridoxine (vitamin B6) supplementation; lysine-restricted diet; arginine supplementation | N/A |
| PNPO | Pyridoxal phosphate-responsive epilepsy | Pyridoxamine 5’-phosphate oxidase deficiency | Pyridoxal 5’-phosphate (PLP) | N/A |
| DEPDC5/NPRL2/NPRL3 | GATOR1-related focal epilepsy | mTOR pathway disinhibition; focal cortical dysplasia | mTOR inhibitors under investigation; epilepsy surgery may be effective if FCD identified | N/A |
Clinical Pearl: Genetic Diagnosis and Surgical Decision-Making
- Genetic testing can directly influence epilepsy surgery decisions: channelopathies (SCN1A, SCN8A, KCNT1) are rarely curable by surgical resection, as the entire brain carries the genetic vulnerability
- Conversely, GATOR1 complex mutations (DEPDC5, NPRL2, NPRL3) may be associated with focal cortical dysplasia that is amenable to surgical resection — discovery of these variants should prompt additional imaging evaluation
- Imaging-negative epilepsy with a channelopathy diagnosis may redirect the patient toward device-based therapy (VNS, RNS) rather than invasive monitoring and resection
- A benign genetic variant should never be used as a basis for avoiding appropriate surgical evaluation (see Case 4-2 in source: SCN1A VUS misdiagnosed as Dravet, delaying surgery for resectable cortical malformation)
mTOR Inhibitors for Tuberous Sclerosis Complex
Everolimus (Afinitor), an mTOR inhibitor, became the first targeted molecular therapy approved for a genetic epilepsy when it received FDA approval for seizures associated with tuberous sclerosis complex (TSC) in 2018, based on the EXIST-3 randomized controlled trial. In EXIST-3, 40% of patients treated with high-exposure everolimus achieved ≥50% seizure reduction at 18 weeks, compared with 15% receiving placebo (p<0.001). The median seizure reduction was 29.3% in the high-exposure group versus 14.9% with placebo.
TSC is caused by loss-of-function pathogenic variants in TSC1 or TSC2, which encode hamartin and tuberin, respectively — key negative regulators of the mTOR signaling pathway. Constitutive mTOR activation drives tuber formation, subependymal giant cell astrocytomas (SEGAs), renal angiomyolipomas, and epileptogenesis. Everolimus is also approved for TSC-associated SEGAs and renal angiomyolipomas, providing a unified therapeutic approach for multiple disease manifestations.
Antisense Oligonucleotides for Genetic Epilepsies
Zorevunersen (STK-001) for Dravet Syndrome
Zorevunersen, developed by Stoke Therapeutics (in partnership with Biogen), represents the most advanced gene-targeted therapy in clinical development for epilepsy. It employs Targeted Augmentation of Nuclear Gene Output (TANGO) technology to increase functional SCN1A expression. The mechanism targets a naturally occurring “poison exon” — a nonsense-mediated decay (NMD)-inducing exon in SCN1A pre-mRNA. By blocking inclusion of this poison exon with an ASO, zorevunersen increases the proportion of productive SCN1A mRNA transcripts, thereby increasing NaV1.1 protein production to compensate for the haploinsufficiency that underlies Dravet syndrome.
Zorevunersen Clinical Development
- Phase 1/2a (MONARCH): Open-label, dose-escalation study in children aged 2–18 with Dravet syndrome and confirmed loss-of-function SCN1A variant; intrathecal administration
- Open-label extension (OLE): Long-term follow-up demonstrated durable seizure reductions (increases in seizure-free days), improvements in cognition and behavior, and quality-of-life benefits on top of standard-of-care ASMs
- Phase 3 (EMPEROR, NCT06872125): Global, double-blind, sham-controlled registrational study; enrolling children ages 2 to <18 with confirmed non-gain-of-function SCN1A variant; sites in the US, Japan, UK, and EU; initiated in 2025
- Key exclusion: Patients with gain-of-function SCN1A variants are excluded, as increasing NaV1.1 expression in this subgroup could theoretically worsen seizures
- AES 2025 data: Presented at the American Epilepsy Society annual meeting; continued to show disease-modifying potential beyond seizure reduction, including sustained cognitive improvements
Other ASO Programs in Epilepsy
| Target / Gene | Condition | ASO Mechanism | Development Stage | Developer |
|---|---|---|---|---|
| SCN1A (TANGO) | Dravet syndrome | Upregulate SCN1A by blocking poison exon | Phase 3 (EMPEROR) | Stoke / Biogen |
| SCN8A | SCN8A-DEE (gain of function) | Downregulate SCN8A mRNA (RNase H degradation) | Preclinical (mouse model efficacy) | Academic / Ionis |
| UBE3A | Angelman syndrome | Unsilence paternal UBE3A allele by targeting UBE3A-ATS | Phase 1/2 (multiple programs) | Ionis/Biogen, Roche/GeneTx |
| SCN2A | SCN2A-related DEE (loss of function) | CRISPRa upregulation (not ASO); proof of concept in mice | Preclinical (Nature, 2025) | Academic (UCSF) |
| KCNQ2 | KCNQ2-related DEE | Under investigation | Preclinical | Multiple academic groups |
| CLN3, CLN5, CLN7 | Neuronal ceroid lipofuscinoses (progressive myoclonus epilepsy) | Various (splice modulation, exon skipping) | Preclinical to Phase 1 | Academic / Ionis |
Gene Replacement Therapy
AAV-Based Approaches for Dravet Syndrome
Gene replacement for Dravet syndrome faces a unique packaging challenge: the SCN1A coding sequence (~6 kb) exceeds the AAV packaging capacity (~4.7 kb). Researchers have developed split-intein dual-vector strategies in which two AAV vectors each carry half of the SCN1A gene, which is then reassembled into a full-length protein within the transduced cell. Preclinical studies published in 2025 using this approach in Dravet syndrome mouse models demonstrated seizure alleviation, long-lasting recovery, and no observed adverse effects. These results represent a significant advance toward clinical translation, though human trials have not yet begun.
AAV Gene Therapy for Other Genetic Epilepsies
Gene replacement approaches are in preclinical development for several other genetic epilepsies, including Angelman syndrome (UBE3A), CDKL5 deficiency disorder, and Rett syndrome (MECP2). Challenges common to all CNS gene therapy programs include achieving adequate transgene expression in the relevant cell populations (interneurons for SCN1A), avoiding immune responses to AAV capsids, ensuring durability of expression in the developing brain, and developing appropriate outcome measures for clinical trials in rare pediatric populations.
CRISPR and Gene Editing Approaches
CRISPRa for Haploinsufficiency Disorders
CRISPR activation (CRISPRa) uses a catalytically inactive Cas protein fused to transcriptional activators to upregulate expression of a target gene without making double-strand DNA breaks. A landmark study published in Nature in 2025 demonstrated that CRISPRa targeting the Scn2a locus in haploinsufficient mice rescued neurological phenotypes, including seizures and behavioral abnormalities. This approach is conceptually applicable to any epilepsy caused by haploinsufficiency (e.g., SCN1A in Dravet syndrome, STXBP1, CDKL5) and offers potential advantages over traditional gene replacement, including preservation of endogenous regulatory elements and avoidance of insertional mutagenesis.
Base and Prime Editing
Base editing and prime editing technologies enable precise single-nucleotide corrections or small insertions/deletions without double-strand DNA breaks, reducing the risk of off-target genomic damage. These approaches are being explored for epilepsies caused by specific point mutations, including certain missense variants in SCN1A, SCN8A, and KCNQ2. All remain in preclinical development, but the precision and potentially reduced immunogenicity of these approaches make them attractive candidates for future clinical translation.
Current Clinical Trials Landscape
| Therapy | Target | Phase | Status (2025–2026) |
|---|---|---|---|
| Zorevunersen (STK-001) | SCN1A (Dravet) | Phase 3 | EMPEROR trial enrolling; OLE ongoing |
| Everolimus (Afinitor) | mTOR (TSC seizures) | Approved | FDA-approved 2018; real-world data accumulating |
| Ganaxolone (Ztalmy) | GABAA (CDKL5-DEE) | Approved | FDA-approved 2022 for CDKL5 deficiency disorder |
| ASO for Angelman (multiple) | UBE3A-ATS | Phase 1/2 | Multiple programs active; Ionis/Biogen, Roche/GeneTx |
| Quinidine | KCNT1 GoF | Investigator-initiated trials | Variable efficacy reported; mechanism-based rationale |
| SCN1A gene replacement (AAV) | SCN1A (Dravet) | Preclinical | Split-intein approach; efficacy in mouse models (2025) |
| CRISPRa for SCN2A | SCN2A haploinsufficiency | Preclinical | Proof of concept in mice (Nature, 2025) |
| Fenfluramine (Fintepla) | Serotonin (Dravet, LGS) | Approved | FDA-approved for Dravet (2020) and Lennox-Gastaut (2022); not gene-specific but mechanism-informed |
Ethical Considerations
Ethical Challenges in Genetic Therapies for Epilepsy
- Access and equity: Gene therapies for rare diseases command prices of $1–4 million per patient; global health care systems are not equipped to provide equitable access; patients in low- and middle-income countries will face the greatest barriers
- Timing of intervention: Many genetic epilepsies cause progressive neurodevelopmental injury from early life; therapeutic benefit may depend on treatment before irreversible damage occurs, creating urgency for early genetic diagnosis and newborn screening
- Informed consent in pediatric populations: Gene therapies for epilepsy primarily target children who cannot provide their own consent; parents must weigh uncertain long-term risks against potential benefits in the context of devastating diseases
- Germline considerations: While current gene therapy approaches target somatic cells, theoretical risks of germline modification must be addressed in counseling and regulatory frameworks
- Variants of uncertain significance: As genetic testing expands, the proportion of patients with VUS increases; clinical decisions should not be based on VUS without additional supporting evidence; periodic reanalysis of genetic data is recommended as variant databases grow
- Genetic counseling: Pretest and posttest genetic counseling by qualified professionals is essential; unexpected findings (secondary findings, consanguinity, non-paternity) require sensitive handling
- Clinical trial readiness: Rare disease clinical trials require validated outcome measures, biomarkers, and natural history data that are often lacking; collaborative registries and academic-industry partnerships are essential
Future Directions
The field of precision medicine in epilepsy is evolving rapidly. Key areas of anticipated progress include: (1) decreasing costs of genome sequencing, potentially enabling universal genetic testing at epilepsy diagnosis; (2) expansion of the ASO therapeutic pipeline to additional monogenic epilepsies beyond Dravet syndrome; (3) advancement of CRISPRa and base editing approaches toward clinical readiness; (4) development of pharmacogenomic decision-support tools integrated into electronic health records; (5) establishment of neonatal screening for actionable genetic epilepsies; and (6) international collaborative frameworks to ensure equitable access to genetic therapies. The transition from empirical to genotype-guided epilepsy treatment represents a paradigm shift that will redefine the standard of care over the next decade.
References
- Krey I, Platzer K, Esterhuizen A, et al. Current practice in diagnostic genetic testing of the epilepsies. Epileptic Disord. 2022;24(5):765–786.
- Sheidley BR, Malinowski J, Bergner AL, et al. Genetic testing for the epilepsies: a systematic review. Epilepsia. 2022;63(2):375–387.
- Smith L, Malinowski J, Ceulemans S, et al. Genetic testing and counseling for the unexplained epilepsies: evidence-based practice guideline of the National Society of Genetic Counselors. J Genet Couns. 2023;32(2):266–280.
- Haviland I, Daniels CI, Greene CA, et al. Genetic diagnosis impacts medical management for pediatric epilepsies. Pediatr Neurol. 2023;138:71–80.
- Lopes-Cendes I, Bhatt P, Valencia A, et al. HLA-B*15:02 and carbamazepine: a pharmacogenomic consensus guideline (CPIC). Clin Pharmacol Ther. 2021;109(6):1426–1436.
- French JA, Lawson JA, Yapici Z, et al. 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(10056):2153–2163.
- Biogen and Stoke Therapeutics. Data supporting disease-modifying potential of zorevunersen for Dravet syndrome. Presented at the 2025 AES Annual Meeting.
- Lenk GM, Bhatt DP, Church EL, et al. CRISPR activation for SCN2A-related neurodevelopmental disorders. Nature. 2025;638:1145–1153.
- Yamagata T, Raveau M, Bhatt DP, et al. Breakthrough gene therapy alleviates Dravet syndrome symptoms in mice without side effects. Cell Reports. 2025;44(3):e114421.
- Sisodiya SM. Precision medicine and therapies of the future. Epilepsia. 2021;62(S2):S90–S105.
- Ryvlin P, Cross JH, Bhatt DP. Are genetic therapies for epilepsy ready for the clinic? Epilepsy Curr. 2023;23(4):241–247.
- Hack JB, Horning K, Juroske Short DM, et al. Distinguishing loss-of-function and gain-of-function SCN8A variants using a random forest classification model. Neurol Genet. 2023;9(3):e200060.
- Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of sequence variants: joint consensus recommendation of the ACMG and AMP. Genet Med. 2015;17(5):405–424.
- Poduri A, Sheidley BR, Shostak S, Ottman R. Genetic testing in the epilepsies: developments and dilemmas. Nat Rev Neurol. 2014;10(5):293–299.
- Stevelink R, Sanders MWCB, Tuinman MP, et al. Epilepsy surgery for patients with genetic refractory epilepsy: a systematic review. Epileptic Disord. 2018;20(2):99–115.
- Specchio N, Curatolo P. Developmental and epileptic encephalopathies: what we do and do not know. Brain. 2021;144(1):32–43.
- Malerba F, Alberini G, Balagura G, et al. Genotype-phenotype correlations in patients with de novo KCNQ2 pathogenic variants. Neurol Genet. 2020;6(6):e528.
- World Health Organization. Community genetics services: report of a WHO consultation on community genetics in low- and middle-income countries. Geneva: WHO; 2011.