How biotech is treating epilepsy beyond seizures

Epilepsy is one of the most common neurological disorders, with more than 50 million people suffering from it globally. Epilepsy might be common, but the disease is also complex and layered, and despite decades of drug development and nearly 30 anti-seizure medicines (ASMs) approved, about a third of epilepsy patients still experience seizures that cannot be sufficiently controlled.

For much of modern epilepsy treatment, seizure control has been the central objective, as it is the most visible consequence of a complex disease. Historically, they represented the aspect of the disease that researchers could most directly measure and influence.

The result of this focus was the development of successive generations of anti-seizure medicines designed to put neuronal excitability under control and stabilize electrical activity in the brain. Most of these treatments work by modulating neuronal excitability through ion channels, neurotransmitter receptors, and synaptic activity. For many patients, those therapies remain highly effective and continue to form the backbone of epilepsy care. 

However, a study led by researchers at the University of Melbourne found that patients who achieve seizure freedom with medication usually do so early in treatment, often with the first or second anti-seizure regimen. After repeated treatment failures, the chances of achieving sustained seizure control tend to decline. This doesn’t mean that later options never work, but that the probability of achieving seizure freedom tends to fall after early treatment failures.

This also means the field’s historical focus on seizure control should not be treated as a mistake. “Anti-seizure medicines have made a real difference. For many patients, they work well and remain the foundation of treatment. But they mostly act by controlling neuronal activity. They don’t address what’s actually driving the disease,” noted Alistair Henry, executive vice president and chief scientific officer at UCB. Henry said we now understand that seizures are really the visible outcome of something deeper. “They reflect underlying changes in biology, whether that’s genetic, synaptic, or at the level of neural circuits.”

It’s no longer just about suppressing seizures. It’s about understanding what’s driving them and asking whether we can intervene earlier. This is particularly important for patients experiencing drug-resistant epilepsies, which, according to The International League Against Epilepsy (ILAE), represent one-third of cases overall.

From epilepsy to epilepsies

The difficulty with epilepsy is that the same clinical endpoint, recurrent seizures, can emerge from very different biological problems. A broad anti-seizure medicine can reduce neuronal excitability in many forms of epilepsy, but it does not address the reason that excitability emerged in the first place for a specific patient. 

Henry noted that when those drivers vary from patient to patient, whether it’s genetics, structural changes, or how neural networks are functioning, there’s a natural limit to how far a broad approach can go.

“In the short term, newer approaches to epilepsy treatment are likely to be more targeted. It’s becoming clear that different patients are driven by different mechanisms. So, it makes sense that the next generation of treatments will be developed for more defined groups. If you can match the treatment to the underlying biology, you have a much better chance of making a meaningful difference, especially for patients who don’t respond to current options,” explained Henry.

Genetic therapies may be most relevant where the cause is molecularly defined. Cell-based or circuit-level interventions may make more sense in focal epilepsies with identifiable seizure networks. More selective pharmacological approaches may still have an important role across broader patient groups.

Ryan Arnold, SVP, Head of Global Medical Affairs at Stoke Therapeutics confirmed that we are in the midst of a shift in the approach to epilepsy research with the advancement of an increasing number of investigational genetically targeted treatments for developmental epileptic encephalopathies (DEEs), which are the company’s focus. 

Arnold also noted that this means diagnostics will play an important part in the future of epilepsy treatment. “Alongside this shift, we are seeing an increase in genetic testing, which will be important to help the right medicine get to the right patient.”

Repairing the circuit

For focal epilepsies, moving beyond broad seizure suppression may mean intervening more directly in the neural circuits generating seizures. That is particularly relevant in mesial temporal lobe epilepsy (MTLE), a common form of focal epilepsy associated with seizure-generating networks in temporal lobe structures. It is often drug-resistant, and the therapeutic logic here is more about altering or repairing a dysfunctional local circuit.

MTLE involves structures such as the hippocampus and amygdala, and in some patients, the seizure-generating region can be identified precisely enough for surgery to become an option. Surgical resection can lead to seizure freedom in a substantial proportion of drug-resistant MTLE patients, although the procedure remains invasive and irreversible.

Newer therapeutic strategies are also focusing on local seizure networks. Some approaches aim to restore inhibitory signaling within dysfunctional circuits. Circuit-level approaches start from the idea that, in at least some epilepsies, the problem may lie within identifiable dysfunctional networks.

One of the clearest examples, now being brought into a European epilepsy franchise through UCB’s planned acquisition, is Neurona Therapeutics’s NRTX-1001, recently renamed rezanecel. The Belgian company said it would pay $650 million upfront to buy the American company in a deal worth up to $1.15 billion. UCB framed the deal as an expansion of its epilepsy portfolio toward regenerative therapies after it had focused mostly on ASMs for the last 30 years.

NRTX-1001 is an investigational allogeneic cell therapy composed of human stem cell-derived GABAergic interneurons. The therapy has to be implanted into the seizure focus itself, where the cells are intended to provide inhibitory GABA signaling and rebalance hyperactive neural activity. 

“Neurona’s approach is interesting because it is grounded in restoring function at the level of neural circuits. The therapy is designed to introduce inhibitory neurons that aim to rebalance overactive networks, rather than simply suppressing activity pharmacologically. Most current pharmacological treatments aim to reduce seizures by modulating neuronal activity. Neurona Therapeutics approach is focused on restoring inhibitory control within the neural circuit itself,” explained Henry.

The acquisition also comes as Neurona moves further into clinical development. At the 2026 American Academy of Neurology meeting, the company presented updated phase 1/2 data from patients with drug-resistant unilateral and bilateral MTLE. According to the company, rezanecel showed encouraging seizure reduction signals with no serious adverse events attributed to the therapy or the procedure itself, although the studies remain early-stage, open-label, and involve relatively small patient numbers, only 15. The company also announced it planned to start the enrolment for a phase 3 study in the second half of 2026.

The therapy’s specificity is also part of its limitation. A localised cell therapy delivered directly into the brain is fundamentally different from prescribing another anti-seizure medicine, and it is unlikely to become a universal solution for epilepsy. “It is important to remain grounded,” said Henry. “This is an investigational approach and still needs to be validated through clinical trials. But it represents a clear example of how the field is beginning to explore new ways of thinking about epilepsy treatment.”

In the short term, Henry sees the therapy sitting alongside existing approaches, particularly for patients with few remaining treatment options beyond invasive procedures. “If clinical data continues to be encouraging, this type of intervention could become an important option for defined patient populations. Over time, it may also help reshape how we think about treating certain forms of epilepsy, but that will depend on long-term outcomes and careful patient selection.”

NRTX-1001 isn’t the only intervention with the idea of treating epilepsy through neural circuits rather than seizure suppression alone. Neurostimulation approaches such as responsive neurostimulation and deep brain stimulation target seizure-generating networks directly, while newer pharmacological programmes are increasingly trying to modulate inhibitory signaling more selectively. 

Companies such as Bright Minds Biosciences and Axsome Therapeutics, both based in the U.S., are developing therapies designed to influence specific inhibitory pathways or receptor subtypes involved in network stability. Neurona is unusual and pretty much the only one of its kind because it pushes that logic further toward restoration rather than modulation, aiming to rebuild inhibitory function within the circuit itself.

Circuit repair is only one direction the field is now exploring. In epilepsies driven by defined genetic defects, the therapeutic logic looks different.

Precision epilepsy: treating the mutation

In precision epilepsy, the premise is different than circuit-level repair. In some patients, the epilepsy can be traced to a defined molecular defect. This is particularly the case with rare DEES, where seizures are often only one part of a broader neurodevelopmental disorder.

“Dravet syndrome is one of these DEEs,” said Arnold. “Most cases are the result of a haploinsufficiency: insufficient protein production due to a mutation in one allele of the SCN1A gene.”

The disease points toward a relatively defined molecular problem, which is why companies such as Stoke Therapeutics and Encoded Therapeutics have pursued therapies aimed at restoring or regulating SCN1A expression. 

Stoke Therapeutics’ zorevunersen is one of the most advanced examples. It is an antisense oligonucleotide designed to increase functional NaV1.1 protein production from the unaffected copy of SCN1A, to reduce seizures and improve neurodevelopment. The company expects to complete enrollment of its phase 3 study in the second quarter of 2026, with a readout planned for mid-2027.

Arnold said he particularly enthousiastic about the drug’s phase 1/2 results. “Zorevunersen showed substantial and durable reductions in seizure frequency as well as ongoing improvements in cognition, behavior and adaptive functioning that children with Dravet syndrome do not typically experience. Importantly, the effects were seen in patients who were already taking standard-of-care anti-seizure medicines.”

When you think of epilepsy, the first thing that comes to mind is still seizure. Arnold explained that this is the distinctive point with new therapies, including zorevunersen. “Zorevunersen aims to reduce seizures and also improve multiple other aspects of the disease including cognition, behavior and adaptive functioning – the things that people do to function in their everyday lives. Improvements in neurodevelopmental impacts of Dravet syndrome such as adaptive behavior have not been demonstrated in studies of anti-seizure medicines.”

Encoded Therapeutics is pursuing the same disease from a different angle. Its ETX101 is an AAV9-based, cell-selective gene regulation therapy, a one-time treatment to restore SCN1A expression in Dravet syndrome. Encoded dosed its first patient in part 2 of ETX101’s phase 1/2 and plans to share initial data by the end of 2027.

SCN1A-related epilepsies aren’t the only area where precision strategies seem to be the way to go. GRIN Therapeutics’ radiprodil, for instance, targets disorders linked to excessive NMDA receptor signaling. The therapy modulates NMDA receptor activity to reduce excitatory signaling more selectively. In 2025, Italy-based Angelini Pharma partnered with GRIN Therapeutics to develop and commercialise radiprodil outside North America.

More selective approaches are also making their way in other DEEs. “Dravet syndrome may be at the forefront of this shift in thinking and treatment in the epilepsy field, but there are other examples such as Syngap1 and SCN2A-DEE. Going forward, we can expect more genetically targeted treatments to be developed and for treatment to become more disease-specific, which hopefully will lead to improved outcomes for patients,” said Arnold.

Boston-based Praxis Precision Medicines, for instance, is developing relutrigine, a small molecule for SCN2A- and SCN8A-related epileptic encephalopathies. Last year, the company received FDA Breakthrough Therapy Designation after reporting strong seizure reduction data in phase 2. In March, the FDA accepted Praxis’ NDA for priority review, with a target decision in September 2026. If approved, the company says relutrigine could become the first approved therapy specifically for SCN2A/8A DEEs.

Praxis is also developing elsunersen, an antisense oligonucleotide for early-onset SCN2A developmental and epileptic encephalopathy. In April 2026, the company reported seizure reductions of 77% in a phase 1/2 study, alongside improvements in sleep, motor function, muscle tone and attention in treated patients.

Lundbeck’s acquisition of Longboard Pharmaceuticals in 2024 brought in bexicaserin, a selective 5-HT2C receptor agonist for DEEs, including Dravet syndrome and Lennox-Gastaut syndrome. The therapy targets serotonergic pathways involved in network regulation. In 2025, Lundbeck reported 12-month extension data of its phase 1b/2a study showing a median 59.3% reduction in motor seizures in patients with DEEs.

Many of the most visible precision epilepsy companies remain U.S.-based: Stoke, Encoded, GRIN, and Longboard all emerged from the U.S., even when their programmes involve European partners later down the road. Europe’s role is more visible through companies such as Angelini and Lundbeck, and through clinical and regulatory participation. 

That may partly reflect the nature of the field itself. Precision epilepsy overlaps with areas such as antisense oligonucleotides, gene regulation therapies, and stratified rare disease development, sectors where U.S. biotech ecosystems have ­historically been active and still ahead of Europe.

The end of broad epilepsy treatment?

For a long time, epilepsy treatment revolved around anti-seizure medicines that were built regardless of the underlying cause. Uncontrolled neuronal activity simply had to be suppressed. That logic made a meaningful difference for millions of patients and remains central to epilepsy treatment today.

What is changing is not the importance of seizure control, but the way seizures themselves are being understood. “Seizure control will remain essential, but it will not be the only goal,” said Henry. They are now being treated as the visible consequence of deeper biological dysfunctions that may differ from one patient to another.

Some epilepsies are becoming genetically tractable, others are being approached through dysfunctional circuits, developmental pathways, or highly specific receptor biology. That future also depends on diagnosis catching up with therapy. “Genetic testing to establish a diagnosis is critical for diseases like Dravet syndrome,” said Arnold. “For example, in order to be considered for entry into our zorevunersen clinical studies, patients must have a confirmed variant in SCN1A and a clinical diagnosis of Dravet syndrome.” Precision treatments require patients to be genetically or biologically characterised, and access to that kind of testing remains uneven.

Henry expects epilepsy care to become more personalised and more closely aligned to underlying biology in the future. “There will not be a single solution that can provide deep therapeutic responses in broad patient populations. Epilepsy is too complex for that.”

---

This article was originally published in EBM’s summer edition. Get it here.