For the two boys, life before the infusion meant a constant cycle of seizures. Dozens each day. Emergency room visits. Medications that dulled the storms but never stopped them. One boy, still young, had never taken an independent step. The other watched his world shrink as convulsions erased milestones he had already reached.
Then came a single intravenous dose of a therapy built for them alone.
Today, both boys are seizure-free or nearly so. One walked across a room on his own for the first time. The treatment did not suppress their epilepsy in the way existing drugs do battering broad neural circuits into submission. Instead, it reached into the genetic instruction set of their own brain cells and flipped a single miswritten sentence.
The approach, reported July 25 in Nature Medicine by researchers at Washington University’s Hope Center for Neurological Disorders, is the first clinical demonstration that a gene therapy can be custom-designed for a patient’s unique mutation and delivered safely into the human brain to treat epilepsy. It represents a fundamental shift in how medicine conceives of this disease: not as a condition to be managed with a lifetime of drugs, but as a genetic typo to be corrected at its source.
Surgical precision at the molecular level
The two boys carry mutations in a gene called SCN2A, which encodes a protein that forms a sodium channel on the surface of neurons. Sodium channels act like molecular gates, opening and closing to control the flow of charged ions into the cell. When a neuron fires, these channels open; when they close, the cell resets. In both boys, the mutation makes one copy of the SCN2A gene overactive. The gate stays open too long, flooding the neuron with sodium and making it fire uncontrollably. The result is a brain locked in a loop of seizure activity.
Most epilepsy drugs work by dampening excitability across the entire brain. They are blunt instruments. They affect every neuron, every sodium channel, every synapse. Patients often endure side effects ranging from drowsiness and dizziness to cognitive slowing and mood disturbances, and many never achieve full seizure control. For severe developmental and epileptic encephalopathies like the ones afflicting these boys, standard treatments frequently fail altogether.
The WashU team took a different path. Instead of searching for a molecule that would broadly quiet neural firing, they designed a genetic silencer targeted to the overactive copy of SCN2A alone. Humans inherit two copies of every gene, one from each parent. In these boys, one copy is healthy; the other carries the mutation that drives their seizures. The therapy uses a short hairpin RNA, delivered by a harmless adeno-associated virus vector, that specifically recognizes and degrades the messenger RNA produced by the mutant gene copy while leaving the healthy copy untouched.
This is genetic surgery, not pharmacology. The unhealthy copy is switched off, and the good copy carries on producing normal sodium channels. The imbalance that caused the seizures is corrected, and the rest of the brain is left undisturbed.
“We are moving from a one-size-fits-all approach to a truly personalized strategy,” said the study’s lead investigator, who is affiliated with the Hope Center. “Instead of treating the symptom of seizures broadly, we are treating the cause at the level of each individual patient’s DNA.”
First-in-human proof of concept
The therapy was tested in two children with SCN2A gain-of-function mutations whose seizures had not responded to conventional anti-epileptic drugs. Both received a single intravenous infusion of the custom-designed vector. The treatment was well tolerated, with no serious adverse events reported during the study period.
In both boys, seizure frequency dropped dramatically. One achieved near-complete seizure freedom. The other, who had never been able to walk independently despite years of physical therapy, took his first unaided steps after treatment. The motor milestone was not just a clinical measurement. It was the kind of change that transforms a family’s daily life.
These results are early. Two patients is a small number, and long-term durability and safety will need to be established in larger trials. But the purpose of this study was never to prove efficacy across a broad population. It was to demonstrate that the concept works: that a bespoke gene therapy can be designed around a specific human mutation, manufactured, delivered to the brain, and produce measurable clinical benefit.
That bar has now been cleared.
“This is a landmark,” the investigators wrote, “demonstrating that personalized genomic medicine can be applied to severe epilepsy with mutations that are individually rare but collectively devastating.”
A philosophical shift
The significance of this work extends beyond SCN2A. There are hundreds of genes linked to epilepsy, and many cause disease through gain-of-function mechanisms similar to the one seen in these boys. Each mutation is rare, sometimes unique to a single patient or family. The traditional drug development model, which demands large populations with the same condition to justify the cost of bringing a therapy to market, cannot address these genetic outliers. They are left behind.
Personalized gene therapy flips that equation. If you can sequence a patient’s genome, identify the causative mutation, and design a silencing construct tailored to that variant, the therapy is no longer constrained by how many other people share the diagnosis. The patient becomes their own clinical trial.
That is the paradigm shift at the heart of this work. For decades, epilepsy treatment has followed a familiar arc: try one drug, try another, add a third, consider surgery to remove the seizure focus, and accept that for some patients, nothing will work. The approach demonstrated in this paper reframes epilepsy not as a circuit disorder to be managed pharmacologically but as a genetic problem to be repaired at the molecular level.
The technical hurdles remain substantial. Each custom therapy must be designed, tested in cell models, manufactured under good manufacturing practice conditions, and evaluated for safety before it reaches a patient. The timeline and cost are not trivial. But the WashU team has shown that the pipeline is viable, and the payoff, as measured by a child’s first independent steps, is hard to argue with.
What comes next
The researchers are already planning to expand the approach to other epilepsy genes and to additional patients with SCN2A mutations. Longer follow-up will be critical to determine whether the benefits persist and whether any late-emerging side effects appear. The technology also opens the door to treating other neurological conditions caused by single-gene overactivity, from certain forms of Parkinson’s disease to inherited movement disorders.
But the most immediate lesson of this study is a simple one. Two boys with a devastating condition, written off by conventional medicine as intractable, received a therapy built for their unique biology and got better. One of them walked.
That is the kind of result that changes the calculus for the entire field. The era of personalized gene therapy for epilepsy has begun.
Reference: Kim-McManus, O. et al. Nature Medicine (2026). DOI: 10.1038/s41591-026-04527-y