March 2026 Research Update

This month added two human studies testing newer add-on treatments for Dravet syndrome, and the picture was mixed but informative. A large phase 3 trial of soticlestat showed a numerical drop in convulsive seizures compared with placebo, but it did not meet its main statistical goal. An earlier-stage open-label program of zorevunersen reported large seizure reductions over longer follow-up, alongside important safety events, reminding readers that promising signals and unanswered questions can exist at the same time. [1] [2]

For soticlestat, the key result was that the main planned comparison did not reach conventional statistical significance, even though the seizure reduction favored soticlestat (for example, a median change of -22.16% with soticlestat vs -8.64% with placebo over the full treatment period). The study also reported several improvements on secondary measures such as caregiver and clinician impressions and ratings of seizure intensity and duration, which helps explain why some families may still find the results meaningful even when the primary analysis is not clearly positive. Side effects reported as related to the study drug included sleepiness, decreased appetite, insomnia, and changes in how seizures looked. [1]

For zorevunersen, participants in extension follow-up (up to 36 months) were reported to have substantial reductions in convulsive seizures along with improvements in overall clinical status, quality of life, and adaptive behavior, especially in the group described after receiving 70 mg and then lower extension dosing. Because this program did not include a control group, the study cannot fully separate drug effects from other influences over time, but it provides a longer view of what treatment and follow-up can look like. Safety findings were a major part of the report and included common procedure-related issues (post-lumbar puncture syndrome), elevated spinal fluid protein, and several serious events including reported SUDEP deaths and one death due to malnutrition. [2]

These studies may matter most as signposts for where Dravet treatment research is heading, rather than as settled answers. The soticlestat trial adds clarity that benefits were not strong enough to clearly pass the main test in this study, even though several secondary measures moved in a helpful direction. The zorevunersen program adds early evidence that a genetic-targeting approach might be linked with large seizure reductions in some children over time, but it also highlights that careful safety monitoring is central to this line of research and that uncertainties remain without a control group. [1] [2]

Studies using health records this month emphasized how changeable Dravet and other SCN1A-related disorders can look in day-to-day life, and how documentation choices can shape what gets captured. One analysis of 100 people reconstructed long histories and found frequent shifts in seizure patterns, high rates of developmental differences, and clear changes in medicine choices after genetic diagnosis. A separate study found that even when genetic epilepsy diagnoses are confirmed, the most specific billing codes are not used consistently, including for Dravet syndrome. [3] [4]

In the longitudinal SCN1A analysis, many participants had repeated ups and downs in seizure frequency over time, and some seizure types (such as myoclonic seizures) were very frequent for certain individuals. The study also reported that most participants had neurodevelopmental differences, including language delays and motor coordination problems, showing that seizures and development often travel together in these disorders. EEG findings were described as changing with age, with more abnormal findings appearing after early infancy. [3]

The same SCN1A dataset reported that medication patterns shifted after a genetic diagnosis, with increased use of cannabidiol, clobazam, and fenfluramine, and reduced use of sodium channel blockers. This does not prove any one medication change caused better outcomes, but it helps document how treatment choices often change once Dravet or another SCN1A-related diagnosis is recognized. For families, it can also explain why treatment histories often include many medication trials and changes over time. [3]

In the coding study, generic epilepsy codes were recorded more than twice as often as the Dravet-specific code among people with SCN1A variants, even in a setting where genetic information was available. When the Dravet-specific code was used, other epilepsy codes were used less, suggesting the specific code may sometimes replace broader ones rather than being added consistently. This matters because health systems often rely on these codes for research, tracking care patterns, and identifying groups of patients for quality improvement. [4]

These studies mainly add understanding of what living with Dravet and related SCN1A conditions looks like over years, and how health records may or may not reflect that experience. The SCN1A history study supports a plain-language message that seizures can fluctuate a lot, developmental challenges are common, and treatment plans often change after genetic answers are found. The coding study clarifies that even with genetic confirmation, Dravet may be under-labeled in medical records, which can affect how easily patients are found for research studies or how well systems measure Dravet care. [3] [4]

This month also included papers that do not test new treatments but can shape how families think about day-to-day risks and diagnosis labels. Two expert-led documents summarized approaches to fever, overheating, vaccination, and complex medication combinations involving stiripentol, reflecting shared practice rather than new trial results. A separate case report described a child whose condition looked like Dravet syndrome but was linked to a different gene, highlighting that a Dravet-like picture can sometimes have a non-SCN1A genetic cause. [5] [6] [7]

The fever and vaccination review stated that vaccinations may trigger an initial seizure in some children with Dravet syndrome but do not worsen long-term outlook, and it described aggressive fever and overheating management strategies. The authors noted that many of these approaches are based mainly on expert consensus rather than controlled trials, which is important for understanding the strength of the evidence. This kind of paper is meant to reduce uncertainty around common family concerns, even when high-quality trial data are limited. [5]

The stiripentol polytherapy consensus reported that most proposed statements reached agreement across an expert group and an international clinician panel, including statements emphasizing family involvement and practical steps around starting, adjusting, and managing side effects in multi-drug plans. Because it reports agreement rather than patient outcomes, it cannot tell how well any specific combination works, but it maps areas where clinicians largely align on practical management. For families, this can explain why care plans that include stiripentol may be discussed in a structured way that considers interactions and tolerability. [6]

The case report described a single child with a Dravet syndrome-like pattern who had a de novo CSNK2B mutation, which the authors presented as the first reported CSNK2B-related neurodevelopmental disorder presenting with a Dravet phenotype. The child had difficult-to-control seizures and developmental concerns, and genetic testing was central to identifying the cause. A single case cannot show how common this is, but it reinforces why genetic testing can matter when the clinical picture resembles Dravet syndrome. [7]

This set of papers may matter now mainly as context for conversations families already have about fever, vaccines, and complex medication plans, while making clear that much of the guidance comes from consensus rather than trials. The vaccination and fever review provides reassurance about prognosis in relation to vaccination while emphasizing careful handling of fever and overheating. The stiripentol consensus helps clarify that there is broad agreement on many practical aspects of multi-drug care, but it does not provide new proof of benefit. The single CSNK2B case mainly adds to diagnostic understanding by showing that a Dravet-like presentation can sometimes have a different genetic explanation. [5] [6] [7]

One laboratory mouse study added to the idea that Dravet syndrome biology involves more than neurons alone. In a Dravet mouse model, astrocytes (support cells in the brain) showed stage-dependent changes over time, including signs of increased activation and altered communication between astrocytes. These findings point to possible non-neuronal contributors to brain circuit changes in Dravet, but they do not directly predict what will happen in people. [8]

Across early to later disease stages in Scn1a+/- mice, the researchers observed higher levels of GFAP (a commonly used marker of astrocyte activation) and a temporary increase in astrocyte branching during a worsening stage that did not persist long-term. They also reported an expanded astrocyte network with changes in connexin proteins and impaired hemichannel function, without obvious large-scale tissue damage. The study also described changes in synaptic behavior (posttetanic potentiation), linking astrocyte changes to how brain signaling may shift in this model. [8]

This is early, animal-based research that mainly matters for future science rather than current care. It helps clarify that in a Dravet mouse model, brain support cells change over time and may affect how brain networks behave, which could open up new research directions beyond neuron-only explanations. What remains uncertain is whether the same astrocyte changes happen in people with Dravet syndrome and whether targeting these pathways would improve symptoms. [8]