Pipeline epilepsy therapies
38 development programs · Updated 27 September 2026
Pipeline indications are investigational targets, not approved treatment recommendations. Review each program’s development status.
Ownership and licensing history, preclinical findings, and human trial results for selected oral drugs, acute treatments and IV formulations, and RNA, gene and cell therapies.
Seizure reductions are relative to baseline unless stated otherwise.
Broader categories include their listed subtypes. See each program’s stage for its current status.
Showing 38 programs
No programs match this indication. Choose another indication or clear the filter.
Oral maintenance
| Drug / mechanism / stage | Proposed indication(s) | Ownership & development history | Preclinical evidence | Human results & current status | References |
|---|---|---|---|---|---|
| AUT00206Kv3 programEarly clinicalPositive allosteric modulator of Kv3.1/3.2 potassium channels; supports rapid firing of parvalbumin-positive inhibitory interneurons. |
| Autifony Therapeutics (Stevenage, UK). | Kv3 modulation aims to improve inhibitory circuit function in disorders with cortical hyperexcitability. | AUT00206 target-engagement studies involved healthy volunteers and people with schizophrenia. | |
| AzetukalnerXEN1101NDA submitted · focal-onset seizuresOpens neuronal Kv7.2–Kv7.5 (KCNQ2–5) potassium channels, with preference for Kv7.2/7.3; augments outward potassium current and limits repetitive firing. |
| Xenon Pharmaceuticals (Burnaby, BC, Canada); acquired from 1st Order Pharmaceuticals (Research Triangle Park, NC, USA) in 2017. | Protected against maximal-electroshock seizures in mice; approximately 15-fold lower brain exposure than ezogabine in the reported comparison. | Published randomized P2b X-TOLE: eight weeks of treatment without titration; 323 participants in the modified intention-to-treat analysis. Median focal-seizure reductions were 52.8% at 25 mg (n=112) and 18.2% with placebo (n=114; p<0.001). The ≥50% responder rates were 54.5% and 14.9%, respectively (odds ratio 7.30, 95% CI 3.77–14.10). Adverse events caused discontinuation in 18/114 treated at 25 mg and 4/114 receiving placebo. Sponsor-reported randomized P3 X-TOLE2: 380 randomized; 374 in the safety/efficacy population over 12 weeks. Median focal-seizure reductions were 53.2% at 25 mg (n=124), 34.5% at 15 mg (n=125), and 10.4% with placebo (n=125); both active comparisons p<0.0001. Adverse-event discontinuations were 14.5%, 4.8%, and 3.2%, respectively. The current September presentation reports an NDA submission in Q3 2026 for focal-onset seizures; P3 X-ACKT continues in primary generalized tonic-clonic seizures. | |
| BexicaserinLP352Phase 3Highly selective serotonin 5-HT2C receptor agonist, with superagonist activity in functional assays and little 5-HT2A/5-HT2B activity; intended to strengthen inhibitory network signaling. |
| Originated at Arena Pharmaceuticals (San Diego, CA, USA; former headquarters); transferred to Longboard Pharmaceuticals (La Jolla, CA, USA; acquired), acquired by Lundbeck (Valby, Denmark) in 2024. | Increased pentylenetetrazol seizure threshold and reduced hippocampal epileptiform discharges in a kainate model. | Published randomized P1b/2a PACIFIC: 75 days (15-day titration; 60-day maintenance). The efficacy analysis included 35 bexicaserin and nine placebo participants who entered maintenance with seizure data. Median countable motor-seizure reductions over treatment were 59.8% and 17.4%, respectively (nominal p=0.0538). A post-hoc comparison of mean changes was significant (nominal p=0.0206); the trial had no formal statistical testing hierarchy or multiplicity adjustment. The safety population comprised 43 active and nine placebo participants. Adverse events led to bexicaserin discontinuation in 9/43: seven during titration and two during maintenance. P3 DEEp development followed. |
|
| BMB-101Phase 2 results · open labelSelective serotonin 5-HT2C receptor agonist biased toward Gq-protein signaling, with minimal β-arrestin recruitment; intended to enhance inhibitory neuronal activity. |
| Bright Minds Biosciences (Chicago, IL, USA / Vancouver, BC, Canada); internally developed serotonergic program. | Sponsor reports antiseizure activity in absence and convulsive seizure models. | Sponsor-reported open-label P2 BREAKTHROUGH enrolled 24 participants (15 in the absence-seizure arm; nine in the DEE arm). The evaluable EEG analysis (n=11, including one DEE participant) showed a 73.1% median reduction in absence seizures lasting ≥3 seconds on 24-hour recordings (p=0.012, within-patient comparison). The evaluable DEE diary analysis (n=6) showed a 63.3% median reduction in major motor seizures. The registered endpoint windows were six weeks for absence seizures and ten weeks for DEE motor seizures, within treatment lasting up to three months. Three participants in each enrolled arm discontinued. The January 2026 presentation reports these separate analysis populations; enrollment totals are not efficacy denominators. | |
| CarisbamateYKP509 / RWJ-333369Phase 3Inhibits voltage-gated sodium currents and repetitive firing and reduces excitatory neurotransmission in experimental systems. |
| Discovered by SK Biopharmaceuticals (Seongnam, South Korea); licensed to Johnson & Johnson (New Brunswick, NJ, USA)/Janssen Pharmaceutica (Beerse, Belgium) in 1999, subsequently returned to SK Biopharmaceuticals (Seongnam, South Korea) development. Now SK Biopharmaceuticals (Seongnam, South Korea)/SK Life Science (Paramus, NJ, USA). | Broad rodent anticonvulsant effects, including electrically induced seizures and kindling. | Earlier randomized focal-epilepsy studies produced mixed efficacy. LGS development included P1 pharmacokinetics. P3 DISCOVER in LGS is the pivotal program. | |
| ClemizoleEPX-100Phase 3Repurposed H1 antihistamine with serotonergic activity, including 5-HT2A/5-HT2B receptors. Zebrafish work implicates 5-HT2B signaling in seizure suppression. |
| Repurposed antihistamine identified in UCSF zebrafish screening. Epygenix Therapeutics (Paramus, NJ, USA; acquired) acquired by Harmony Biosciences (Plymouth Meeting, PA, USA) in 2024. | Suppressed behavioral seizures and electrographic abnormalities in scn1lab-mutant zebrafish, a Dravet syndrome model. | Sponsor and AES reports describe the same ARGUS open-label-extension cohort: 18 patients with Dravet syndrome treated for ≥6 months had approximately 50% median reduction in countable motor seizures per 28 days; approximately half achieved ≥50% reductions. Common reported adverse events included seizures, pyrexia, and upper respiratory infection. P3 programs include ARGUS (Dravet syndrome) and LIGHTHOUSE (LGS). |
|
| DarigabatCVL-865 / PF-06372865Phase 2 completed · primary endpoint not metGABAA receptor positive allosteric modulator with preference for α2/α3/α5-containing receptors over α1-containing receptors; enhances GABA-dependent inhibition. |
| Originated at Pfizer (New York, NY, USA); transferred to Cerevel Therapeutics (Cambridge, MA, USA; acquired) in 2018. AbbVie (North Chicago, IL, USA) acquired Cerevel Therapeutics (Cambridge, MA, USA; acquired) in 2024. | Suppressed epileptiform activity in a mouse temporal-lobe epilepsy model, comparable to diazepam under tested conditions. | Posted randomized P2 REALIZE results: 154 randomized, 143 in the modified intention-to-treat analysis; two-week titration and eight-week maintenance. The primary endpoint was change in weekly focal-seizure frequency expressed as response ratio 100×(T−B)/(T+B), where T is treatment and B is baseline frequency. Neither dose differed significantly from placebo: adjusted difference 1.42 at 7.5 mg twice daily (90% CI −9.04 to 11.87; p=0.823) and 0.11 at 25 mg twice daily (90% CI −10.38 to 10.61; p=0.986). Adverse events caused discontinuation in 2/51 at 7.5 mg, 5/52 at 25 mg, and 3/51 with placebo. A separate small photosensitivity study reported suppression of photoparoxysmal responses. | |
| ES-481LY3130481Phase 2a completedSelective antagonist of TARP-γ8-associated AMPA receptors, reducing excitatory transmission in forebrain circuits while relatively sparing receptor complexes without TARP-γ8. |
| ES Therapeutics (New York, NY, USA; reported company base). | Antiseizure activity reported in rodent focal and generalized seizure models, including hippocampal paroxysmal discharges and genetic absence seizures. | Published randomized crossover P2a: 22 participants; 17 completed double-blind treatment. Each four-week treatment period escalated weekly from 25 mg daily to 75 mg twice daily. Seizure reductions across dose weeks were 68–80% with ES-481 and 38–49% with placebo; the overall comparison was p=0.097. At week 4 (75 mg twice daily), reduction was 80% (90% CI 43–97%) versus 49% (90% CI 1–74%) during the corresponding placebo week (p=0.047). Five participants withdrew: three for adverse events during placebo, one for an adverse event during ES-481, and one by participant request during ES-481. | |
| GAO-3-02PreclinicalSynaptamide-related lipid derivative. Experimental antiseizure activity involves CB2-sensitive signaling; a CB2 antagonist reversed the effect in a rat model. |
| GAOMA Therapeutics (Bron (Lyon), France); lipid-derivative program. | Reduced seizures in pilocarpine and amygdala-kindling models. Findings support an endocannabinoid-related mechanism. | — | |
| LorcaserinEPX-200 · investigational liquid formulationEPX-200: IND-enabling development Reviewed 27 September 2026. Harmony lists EPX-200 in its 2026 pipeline. The separate Eisai Phase 3 lorcaserin trial has terminated. No US antiseizure indication. Selective serotonin 5-HT2C receptor agonist; serotonergic modulation is proposed to suppress seizures. |
| Epygenix Therapeutics (Paramus, NJ, USA; acquired) developed EPX-200; Harmony Biosciences (Plymouth Meeting, PA, USA) acquired Epygenix Therapeutics (Paramus, NJ, USA; acquired) in 2024. Earlier lorcaserin studies, including Eisai’s LORIS study, are distinct from the current EPX-200 development program. | Lorcaserin suppressed behavioral and electrographic seizure activity in scn1lab-mutant zebrafish modeling Dravet syndrome. This is a preclinical rationale, not evidence that the new liquid formulation has established clinical efficacy. | Historical evidence: small uncontrolled epilepsy reports preceded Eisai’s randomized, placebo-controlled LORIS trial (NCT04572243). LORIS enrolled 22 participants aged at least two years with Dravet syndrome and ended in August 2024; the sponsor states termination was unrelated to safety or efficacy. Registry results posted October 2025 report a median 78.88% reduction in convulsive seizure frequency with lorcaserin versus an 18.78% increase with placebo over 14 weeks (10 analyzed per arm). One of 11 lorcaserin-treated participants had a serious adverse event versus none of 10 placebo recipients; adverse events led to discontinuation in two versus one, respectively. These small, prematurely terminated trial data do not establish an approved treatment or efficacy of EPX-200. Current program: Harmony’s 2026 pipeline retains EPX-200 for developmental and epileptic encephalopathies; the sponsor describes IND-enabling work. An available lorcaserin expanded-access record (NCT04457687, last updated January 2025) is access for eligible patients, not proof of a recruiting efficacy trial. Safety context: FDA requested withdrawal of the former weight-loss products Belviq/Belviq XR in February 2020 after an increased occurrence of cancer in a large safety trial. The short epilepsy studies cannot resolve that long-term risk. Lorcaserin remains investigational for epilepsy. | |
| LRP-661Cannabidiol sulfateLate preclinicalSulfated cannabidiol derivative. |
| London Research & Pharmaceuticals (London, ON, Canada); proprietary sulfate platform. | Sponsor reports protection across acute and chronic animal seizure models. | — | |
| MSCA-7136Preclinical · Phase 1 plannedBrain-penetrant MEK inhibitor that reduces downstream ERK signaling and neuronal hyperexcitability in experimental epilepsy. |
| Mosaica Medicines (Boston, MA, USA; reported company base). | Suppressed hippocampal paroxysmal discharges in a mouse temporal-lobe epilepsy model during repeated dosing. | First-in-human studies planned, according to the 2026 development report. | |
| OpakalimBHV-7000 / formerly KB-3061 (BPN-25203)Phase 3 · enrollment hold reportedSelective Kv7.2/7.3 potassium-channel activator that enhances the neuronal M-current and reduces excitability. |
| Knopp Biosciences (Pittsburgh, PA, USA) platform acquired by Biohaven (New Haven, CT, USA) in 2022. SK Biopharmaceuticals (Seongnam, South Korea) announced a worldwide licensing agreement in August 2026. | Rodent electroshock antiseizure activity with separation from motor impairment reported by the sponsor. | P1: EEG evidence of CNS activity. RISE-3 enrollment completed. September 2026 reporting described an FDA hold on new RISE-2 enrollment; existing dosing continued and RISE-3 was unaffected. | |
| OV329Phase 2Potent, irreversible, mechanism-based inactivator of GABA aminotransferase (GABA-AT). Reduces GABA degradation and increases inhibitory GABA availability. |
| Ovid Therapeutics (New York, NY, USA); originated from Richard Silverman’s Northwestern University work and licensed to Ovid Therapeutics (New York, NY, USA). | Potent GABA-AT inhibition and animal seizure suppression. | P1 healthy volunteers: increased cortical inhibition. Randomized P2 focal-seizure and photosensitivity studies initiated in 2026. | |
| PaxalisibSVG103Phase 1b/2a · recruitingBrain-penetrant dual PI3K/mTOR small-molecule inhibitor; targets excessive growth-pathway signaling in TSC and related mTORopathies. |
| Sovargen (Daejeon, South Korea); SVG103 is the epilepsy development program for paxalisib. | The rationale is suppression of overactive PI3K/mTOR signaling in epileptogenic brain lesions. | Epilepsy-specific registry NCT07287202 describes an open-label P1b/2a SVG103 study in adults with focal cortical dysplasia type II, tuberous sclerosis syndrome, or hemimegalencephaly. Target enrollment is 12. Primary outcomes assess safety and treatment-emergent adverse events over up to 12 weeks; secondary outcomes include diary-recorded seizure-frequency change. The registry lists a 14 July 2026 start and recruiting status in its 10 August update. The separate preclinical pharmacokinetic paper concerns oncology models. | |
| PTI5803Extended-release probenecidPhase 2a authorized · FranceSustained-release probenecid targeting pannexin-1 (PANX1) channels and ATP release implicated in ictogenesis. Probenecid also inhibits organic-anion and urate transporters. |
| Panntherapi (Nîmes, France); formulation developed for drug-resistant focal cortical dysplasia-associated epilepsy. | PANX1 blockade suppressed seizure-like activity in resected human epileptic tissue and experimental models. | ANSM authorized a P2a study on 8 September 2026 after healthy-volunteer P1 work. Planned dose escalation in adults and adolescents with FCD prioritizes safety, with preliminary seizure outcomes. | |
| RadiprodilPhase 3Negative allosteric modulator of GluN2B-containing NMDA receptors; reduces glutamate-driven excitation, including excessive signaling from relevant GRIN gain-of-function variants. |
| Previously developed by UCB (Brussels, Belgium); now GRIN Therapeutics (New York, NY, USA)/Neurvati Neurosciences (New York, NY, USA). Angelini Pharma (Rome, Italy) licensed rights outside the US, Canada and Mexico in 2025. | Suppressed seizures in models including Grin2a gain-of-function mice; experimental context, including sex, affected response. | Sponsor-reported open-label P1b/2a Honeycomb: seven participants with GRIN gain-of-function variants and countable motor seizures had an 86% median seizure-frequency reduction; 71% achieved >50% reduction. During eight-week maintenance, six of seven were seizure-free on ≥80% of days. P3 BEELINE began dosing in January 2026. Separate studies address TSC and focal cortical dysplasia. |
|
| RAP-219JNJ-64300912Phase 3Negative allosteric modulator of AMPA receptors associated with the auxiliary protein TARP-γ8. Targets forebrain-enriched receptor complexes rather than blocking all AMPA receptors equally. |
| Originated at Janssen Pharmaceutica (Beerse, Belgium); licensed to Rapport Therapeutics (Boston, MA, USA). | Strong mouse corneal-kindling and other seizure-model activity; reported separation from motor effects. | Open-label P2a in adults with implanted responsive neurostimulation devices: 30 received treatment over eight weeks. Median electrographic-long-episode reduction was 71% in 27 evaluable participants; median diary-recorded clinical-seizure reduction was 77.8% in 25 evaluable participants. These are separate endpoints with different analysis populations. Three participants discontinued because of adverse events. The linked AES poster reports a post-hoc baseline-severity analysis of this study. P3 FOCUS-1/2 were initiated in 2026. | |
| RelutriginePRAX-562NDA under reviewPreferential inhibitor of persistent sodium current, including current generated by gain-of-function Nav1.6 variants; also produces use-dependent block. Greater inhibition of persistent than resting peak current in vitro. |
| Praxis Precision Medicines (Boston, MA, USA); internally developed, initially for SCN2A/SCN8A developmental and epileptic encephalopathies (DEEs). | Dose-dependent seizure suppression in genetic epilepsy models, including SCN2A/SCN8A models. | Randomized, blinded P2/3 EMBOLD in SCN2A-/SCN8A-DEE: participants received either 16 weeks of relutrigine or 12 weeks of relutrigine plus a blinded four-week placebo period. Sponsor-reported placebo-adjusted reductions in monthly motor-seizure frequency were 46% in cohort 1 (efficacy n=15; p=0.0354) and 53% in cohort 2 (efficacy n=51; p<0.0002). The cohort-2 poster reports one adverse-event discontinuation during active treatment; placebo-period discontinuations included one withdrawal of consent and one death. NDA review is ongoing, with a sponsor-reported target action date of 27 December 2026. Recruitment to the broader-DEE EMERALD study was complete in the September presentation. |
|
| SimufilamTSC study on clinical holdInvestigational filamin A modulator. The epilepsy rationale targets abnormal filamin A-related neuronal signaling in TSC/FCDII. |
| Filana Therapeutics (Austin, TX, USA) (formerly Cassava Sciences (Austin, TX, USA)); repurposed for TSC-related epilepsy. | In Tsc1 conditional-knockout mice, higher doses attenuated worsening seizure frequency and ictal duration. | The proposed TSC proof-of-concept study remains on an FDA clinical hold in the July 29, 2026 company update. Additional preclinical data and protocol changes were being prepared. | |
| SN-2000PreclinicalSelective allosteric inhibitor of phosphodiesterase-4B (PDE4B), modifying intracellular cAMP signaling. |
| Stream Neuroscience (Calgary, AB, Canada). | Antiseizure activity in rodent focal and Dravet syndrome models and zebrafish seizure models. | — | |
| Sodium selenatePhase 2Enhances protein phosphatase 2A (PP2A) activity, promoting dephosphorylation of pathological tau; investigated as a disease-modifying approach to temporal-lobe epilepsy. |
| Monash University (Melbourne, VIC, Australia)-led academic program. | Disease-modifying signals in rat temporal-lobe epilepsy models informed the randomized human study. | The SeLECT protocol describes a randomized placebo-controlled study in drug-resistant TLE, with follow-up after treatment withdrawal. | |
| SPN-817Phase 2bExtended-release synthetic huperzine A, a reversible acetylcholinesterase inhibitor. Increases acetylcholine. |
| Supernus Pharmaceuticals (Rockville, MD, USA); obtained through acquisition of Biscayne Neurotherapeutics (Miami, FL, USA; acquired) in 2018. | Reduced hippocampal paroxysmal discharges in a mouse mesial temporal-lobe epilepsy model. | Sponsor-reported open-label P2a interim analysis (1 May 2024 cutoff): 41 enrolled; 19 had completed maintenance, including 16 with focal seizures. Across maintenance doses of 1–4 mg twice daily, median focal-seizure reduction was 58%; the selected 3–4 mg twice-daily subgroup had a 75% reduction. The treatment phase lasted 20 weeks, followed by an optional extension. Adverse events led to discontinuation in 22% during titration and 2.4% during maintenance, using the enrolled population. These are completer/subgroup efficacy results. P2b RENAISSANCE-2 is underway. | |
| VormatriginePRAX-628Phase 2/3 · primary endpoint missedInhibits Nav1.6 persistent sodium current and activity-dependent peak current. Preferential action in hyperexcitable neurons reflects channel-state dependence. |
| Praxis Precision Medicines (Boston, MA, USA); internally developed through the Cerebrum platform. | Activity in rodent seizure models, including maximal electroshock; designed to preferentially inhibit overactive neurons. | Earlier photosensitivity and open-label studies showed activity. Randomized P2/3 POWER-1 did not meet its primary endpoint of change in monthly focal-seizure frequency over 12 weeks (June 2026 sponsor report). A ≥50% seizure-reduction responder secondary endpoint was positive. The August update announced plans to restart POWER-2 and initiate POWER-3 in Q4 2026. |
Acute treatment & IV formulations
| Drug / mechanism / stage | Proposed indication(s) | Ownership & development history | Preclinical evidence | Human results & current status | References |
|---|---|---|---|---|---|
| AMO-01Ras–ERK pathway inhibitor · IV infusionPhase 2 completed · sponsor-listed development Reviewed 27 September 2026. AMO Pharma continues to list development for Phelan-McDermid syndrome; a new recruiting trial was not identified. Inhibits Ras–ERK signaling, a pathway implicated in SHANK3-related synaptic dysfunction and seizures. The proposed clinical mechanism remains investigational. |
| AMO Pharma (Leeds, UK; US office in New Jersey). The pilot study was conducted at the Icahn School of Medicine at Mount Sinai and Texas Children’s Hospital. | In a Shank3b-knockout mouse model, a single administration reduced seizure activity and behavioral abnormalities, with effects reported through five days. These model findings do not establish clinical disease modification. | Completed open-label Phase 2 pilot (NCT03493607): six participants with genetically confirmed Phelan-McDermid syndrome and epilepsy received a single six-hour IV infusion; eligibility was 12–45 years. The trial ended in March 2020; its publication appeared online in December 2024 (2025 journal issue). Five participants had fewer seizures after treatment; one had no baseline seizures. Seizure counts improved at weeks 2 and 4 (nominal p=0.043 at each visit). No serious adverse events were reported; events were mild or moderate, including appetite and repetitive-behavior changes. The uncontrolled, very small sample, short follow-up and uncorrected multiple comparisons prevent firm efficacy or long-term safety conclusions. A September 2025 sponsor update still identifies AMO-01 as under investigation for Phelan-McDermid syndrome. This supports sponsor-listed development, not a claim of ongoing enrollment or an announced next-phase trial. | |
| Captisol-enabled™ Topiramate InjectionIV topiramate · CURx / LigandNDA stage · sponsor-listedTopiramate solubilized with sulfobutylether-β-cyclodextrin (Captisol) for IV administration. Use-dependent sodium-channel inhibition, enhancement of GABAA-mediated inhibition, AMPA/kainate antagonism and carbonic anhydrase inhibition. |
| Developed at the University of Minnesota (Minneapolis, MN, USA). Ligand Pharmaceuticals (Jupiter, FL, USA) licensed the IV formulation rights and granted CURx Pharmaceuticals (La Jolla, CA, USA) a global development and commercialization license. | Parent topiramate showed broad antiseizure activity in rodent models and neuroprotective effects in experimental status epilepticus and hypoxia–ischemia. Captisol improves aqueous solubility for parenteral delivery. | Investigational replacement/bridging formulation for patients unable to take oral topiramate. Preliminary 2024 pharmacokinetic data from 25 healthy participants supported 1:1 IV-to-oral dose substitution. Bamgboye et al. (2026) modeled 246 concentrations from 20 adults with epilepsy or migraine after a single IV dose; enzyme-inducing co-medications increased clearance by 63%. The study evaluated pharmacokinetics and simulated loading/bridging regimens. | |
| Intranasal seletracetamPrevEp006Early proof of conceptHigh-affinity synaptic vesicle protein 2A (SV2A) ligand that modulates vesicular neurotransmitter release. Intranasal delivery is being investigated for rapid seizure prevention or intervention. |
| Developed as an oral tablet formulation by UCB (Brussels, Belgium) through Phase 2 clinical studies. PrevEp (Bethesda, MD, USA) developed a novel liquid intranasal spray formulation for on-demand administration. | Parent molecule showed potent activity in experimental seizure models. | Koepp et al. (2026): one patient with reading epilepsy despite levetiracetam 3,000 mg/day. Intranasal 30 mg delayed reading-induced seizure onset from 1:56 with placebo to 4:17; a second 30 mg dose prevented seizures during 25 minutes of reading. Serum drug was detectable within 2 minutes. Single-patient proof of concept. The AES abstract and subsequent publication describe the same patient. | |
| Intravenous topiramatePrevEp004Investigational IV formulationMeglumine-solubilized IV formulation of topiramate: use-dependent sodium-channel inhibition, enhancement of GABAA-mediated inhibition, AMPA/kainate antagonism and carbonic anhydrase inhibition. |
| PrevEp (Bethesda, MD, USA). | Builds on topiramate’s broad antiseizure pharmacology. | Being developed for neonatal seizures and status epilepticus. FDA orphan drug designation for treatment of neonatal seizures was granted to PrevEp on 28 February 2024. The 2025 Löscher–Soul review describes first-in-human administration of the meglumine formulation in one adult with epilepsy. | |
| Levetiracetam · neonatal developmentNEOLEV3 · IV dose-escalation programPhase 2 · recruiting Reviewed 27 September 2026. NEOLEV3 registry updated 15 June 2026. Neonatal use and the study doses are investigational; this is not a new drug approval. Binds synaptic vesicle protein SV2A, a target involved in neurotransmitter release. The neonatal program studies an existing antiseizure drug in an investigational population and dosing setting. |
| University of California, San Diego (San Diego, CA, USA) leads NEOLEV3 with US and New Zealand collaborators; funded by FDA’s Office of Orphan Products Development. This is an academic development program for an established drug. | SV2A binding was demonstrated in synaptic vesicles and expression systems; binding affinity of related compounds correlated with seizure protection in an audiogenic mouse model. These experiments do not establish neonatal dosing or long-term developmental safety. | Active study: NEOLEV3 (NCT05610085), an open-label Phase 2 dose-escalation study with a randomized phenobarbital control arm, plans 133 neonates. Eligibility includes corrected gestational age 35–44 weeks, postnatal age under 28 days and weight over 2,200 g; seizure-burden limits also apply. Its primary aim is to find the maximum safe, tolerated dose; seizure control, pharmacokinetics and developmental outcomes are also assessed. No results are posted. Estimated study completion is December 2027. Prior evidence: NEOLEV2, a blinded randomized Phase 2b trial, found 24-hour EEG-confirmed seizure freedom in 15/53 neonates given levetiracetam (28%) versus 24/30 given phenobarbital (80%; p<0.001). Adverse effects were more frequent with phenobarbital, but the difference was not statistically significant. These results do not support a claim that levetiracetam was superior. A 2024 analysis of NEOLEV1/2 found predictable pharmacokinetics at the studied doses and provided the rationale for higher-dose investigation. NEOLEV3’s control arm helps interpret safety and response; the study is not powered for a definitive comparison with phenobarbital. Study dosing is not a prescribing recommendation. | |
| Staccato alprazolamPhase 3 · recruitingAlprazolam enhances GABAA receptor responses at the benzodiazepine binding site. The inhaled thermal-aerosol formulation enables rapid pulmonary delivery. |
| Alexza Pharmaceuticals (Mountain View, CA, USA; historical base) delivery technology; epilepsy development by Engage Therapeutics (Summit, NJ, USA; acquired), acquired by UCB (Brussels, Belgium) in 2020. UCB (Brussels, Belgium) holds licensed development rights. | Established alprazolam anticonvulsant pharmacology. | Published randomized P2b trial under inpatient supervision: 116 participants received a single treatment for an observed seizure episode. The composite endpoint was seizure cessation within two minutes with no recurrence for two hours. Response occurred in 25/38 participants at 1 mg and 25/38 at 2 mg (65.8% each), compared with 17/40 receiving placebo (42.5%): an absolute difference of 23.3 percentage points for either dose (p=0.0392). No drug-related serious adverse events were reported. The Phase 3 registry remained recruiting in its 11 September 2026 update. |
RNA, gene & cell therapies
| Drug / mechanism / stage | Proposed indication(s) | Ownership & development history | Preclinical evidence | Human results & current status | References |
|---|---|---|---|---|---|
| AMT-260Phase 1/2aAAV9 gene therapy delivers two engineered microRNAs targeting GRIK2 mRNA, reducing GluK2-containing kainate receptors and pathological excitatory signaling in the epileptic hippocampus. |
| uniQure (Amsterdam, Netherlands / Lexington, MA, USA), via its 2021 acquisition of Corlieve Therapeutics (Paris, France; acquired). Originated from Bordeaux/CNRS and INSERM/Aix-Marseille academic work, with REGENXBIO (Rockville, MD, USA) involvement. | Suppressed seizures in chronic temporal-lobe epilepsy models; supporting activity in resected human epileptic tissue ex vivo. | Open-label P1/2a GenTLE, six low-dose participants: during months 4–6, three had 79–100% reductions in disabling-seizure frequency; the other three ranged from a 33% reduction to a 36% increase. These are individual outcomes, not a cohort median. No serious adverse events, dose-limiting events, or adverse-event-related discontinuations were reported at the 29 May 2026 cutoff. Higher-dose enrollment was ongoing in the June presentation. | |
| ElsunersenPRAX-222Phase 3 · EMBRAVE3 enrollingAntisense oligonucleotide that lowers SCN2A mRNA and Nav1.2 channel expression; intended for gain-of-function SCN2A epilepsy. |
| Praxis Precision Medicines (Boston, MA, USA), in collaboration with Ionis Pharmaceuticals (Carlsbad, CA, USA) and RogCon (Miami, FL, USA). | Reduced SCN2A expression, seizures and mortality in relevant genetic models, with associated phenotype improvements. | Sponsor-reported randomized, sham-controlled P1/2 EMBRAVE Part A: nine children aged 2–12 years, treated for 24 weeks. Sham-adjusted monthly seizure reduction was 77% (95% CI 33–92%; p=0.015). No discontinuations or drug-related serious adverse events were reported. Pivotal EMBRAVE3 is enrolling approximately 30 participants in a single-arm, baseline-controlled study. The primary treatment period is 24 weeks, followed by a 24-week extension; the primary efficacy endpoint is change in monthly motor-seizure frequency. This is a separate study from the nine-child randomized trial. | |
| ETX101Phase 1/2 · pivotal ENDEAVOR Part 2 ongoingAAV9-delivered engineered transcriptional regulator increases endogenous SCN1A expression selectively in GABAergic inhibitory neurons, restoring Nav1.1-dependent interneuron function. |
| Encoded Therapeutics (South San Francisco, CA, USA); internally developed gene-regulation platform. | Reduced spontaneous and hyperthermia-induced seizures and prolonged survival in Dravet syndrome mouse model. | Sponsor-reported open-label P1/2 POLARIS: the three-patient dose-level-3 cohort had approximately 76% median reduction in monthly countable-seizure frequency over weeks 5–52 after a single administration (10 April 2026 cutoff). Developmental measures were exploratory. Across dose levels, reported treatment-related transaminase elevations resolved; no treatment- or procedure-related serious adverse events were reported. Pivotal ENDEAVOR Part 2 is ongoing in children aged six months to four years. | |
| ION283GYS1 antisense therapyPhase 1/2 · Lafora diseaseRNase H1-recruiting antisense oligonucleotide lowers GYS1 mRNA and glycogen synthase 1, aiming to reduce formation of disease-driving Lafora bodies. |
| Elpida Therapeutics (Encino, CA, USA) licensed the program from Ionis Pharmaceuticals (Carlsbad, CA, USA); developed with UT Southwestern Medical Center (Dallas, TX, USA). | Gys1-targeting antisense treatment reduced glycogen aggregates and epileptiform discharges in Lafora mouse models; ION283 targets the human transcript. | Registered open-label intrathecal P1/2 study NCT06609889 plans to enroll ten participants with genetically confirmed Lafora disease. The primary endpoint is treatment-related adverse events over two years. Secondary outcomes over two years include Lafora Disease Performance Scale and pediatric disability scores, seizure frequency, EEG measures, and quality of life. This is a study protocol, not a report of treatment effects. | |
| RezanecelNRTX-1001Phase 1/2 · Phase 3 plannedAllogeneic, human embryonic stem cell-derived medial ganglionic eminence-type GABAergic interneurons. Hippocampal grafts are intended to integrate into local circuits and restore synaptic inhibition. |
| Developed by Neurona Therapeutics (South San Francisco, CA, USA); UCB (Brussels, Belgium) announced its acquisition in April 2026. | Transplanted interneurons integrated into host circuits and sustained seizure suppression in chronic temporal-lobe epilepsy models. | Sponsor-reported open-label P1/2 in unilateral MTLE with sclerosis: median disabling-seizure reductions during months 7–12 were 89% in the low-dose cohort (n=9) and 58% in the high-dose cohort (n=9). These were separate, nonrandomized dose cohorts. P3 EPIC was planned in the April update. | |
| ZorevunersenSTK-001Phase 3Splice-modifying antisense oligonucleotide that reduces nonproductive SCN1A splicing, increasing full-length transcript and Nav1.1 protein from the functional allele in SCN1A haploinsufficiency. |
| Stoke Therapeutics (Bedford, MA, USA) TANGO platform. Biogen (Cambridge, MA, USA) licensed rights outside the US, Canada and Mexico in 2025; Stoke Therapeutics (Bedford, MA, USA) retained those markets. | Increased functional NaV1.1 expression; reduced seizures and premature mortality in Dravet syndrome mice. | Published open-label P1/2a MONARCH/ADMIRAL enrolled 81 patients; 75 entered extensions by 30 May 2025. Primary analyses concerned safety and pharmacokinetics. In the subgroup initially given 70 mg, followed by extension doses ≤45 mg, median convulsive-seizure reductions ranged from 58.82% to 90.91% across one-month intervals during the first 20 months of extension. This range describes subgroup results across time, not an overall four-year effect. The publication reported post-lumbar-puncture syndrome in 25%, elevated CSF protein in 45% during extensions, and one adverse-event-related withdrawal. Three deaths were reported (two sudden unexpected deaths in epilepsy; one from malnutrition). A separate September 2026 sponsor update described follow-up through four years, with 58/75 extension entrants remaining at that cutoff. In this later report, elevated CSF protein was recorded as a treatment-emergent adverse event in 59%. P3 EMPEROR is ongoing. |
Redevelopment / discontinued programs
| Drug / mechanism / stage | Proposed indication(s) | Ownership & development history | Preclinical evidence | Human results & current status | References |
|---|---|---|---|---|---|
| RozanolixizumabAnti-FcRn antibodyAutoimmune encephalitis program discontinuedHumanized monoclonal antibody blocks the neonatal Fc receptor (FcRn), reducing IgG recycling and circulating pathogenic autoantibodies; studied in LGI1-antibody autoimmune encephalitis. |
| UCB (Brussels, Belgium). Approval for myasthenia gravis is a separate indication. | Immunological rationale is reduction of pathogenic IgG. | UCB reported in 2024 that P2a AIE001/LEGIONE did not demonstrate efficacy and the autoimmune encephalitis program was terminated; the decision was not attributed to safety. | |
| SoticlestatTAK-935 / OV935Negative Phase 3 · rights transferredSelective cholesterol 24-hydroxylase (CH24H/CYP46A1) inhibitor. Lowers brain 24S-hydroxycholesterol, an endogenous positive modulator of NMDA receptors, to reduce excitatory signaling. |
| Takeda (Tokyo, Japan) discovery; co-developed with Ovid Therapeutics (New York, NY, USA), then Takeda (Tokyo, Japan) worldwide control in 2021. After Takeda (Tokyo, Japan) discontinuation, global rights transferred in 2026 to a Perceptive Advisors (New York, NY, USA)-backed new company. Ovid Therapeutics (New York, NY, USA) retains equity and contingent economics. | Lowered brain 24S-hydroxycholesterol and reduced experimental seizure susceptibility. | P2 ELEKTRA supported advancement, particularly in Dravet syndrome. P3 SKYLINE (Dravet syndrome) missed its primary endpoint (p=0.06); P3 SKYWAY (LGS) also failed. |
Abbreviations: CI, confidence interval; CSF, cerebrospinal fluid; DEE, developmental and epileptic encephalopathy; EEG, electroencephalogram; FCD, focal cortical dysplasia; LGS, Lennox-Gastaut syndrome; MTLE, mesial temporal-lobe epilepsy; TSC, tuberous sclerosis syndrome (tuberous sclerosis complex); NDA, new drug application; P1/P2/P3, Phase 1/2/3. A dash in the human-evidence column indicates a preclinical program without clinical results summarized here.