Imagine a copying error in a single gene causing brain cells to manufacture a protein that gradually piles up where it should not. The buildup damages the nervous system, and doctors can treat some of the consequences—but cannot switch off the source.
That has been the reality for people with Alexander disease, an extraordinarily rare neurological disorder that can affect movement, speech, swallowing and development.
Now, for the first time, the United States has approved a medicine designed to turn down production of the damaging protein itself.
On 3 September 2026, the US Food and Drug Administration approved Zanvastro, also known by its scientific name zilganersen, for children and adults with Alexander disease. The agency describes it as the first approved treatment for the condition and the first to directly target its underlying protein buildup. FDA approval announcement
This is a landmark for a tiny patient community—but it is not a cure, and it does not rewrite a patient's genes. Zanvastro is an RNA medicine that acts more like a temporary molecular dimmer switch. It must be injected into the fluid around the spinal cord every three months, and its evidence comes from one necessarily small clinical programme.
What is Alexander disease?
Alexander disease is caused by mutations in a gene called GFAP, short for glial fibrillary acidic protein.
GFAP is normally produced in astrocytes, star-shaped cells that help maintain the brain and spinal cord. In Alexander disease, altered GFAP can accumulate inside these cells. That buildup disrupts astrocytes and harms the wider nervous system.
The condition is estimated to affect fewer than one person in a million. It can appear in infancy, childhood or adulthood, and its course varies widely. Possible effects include seizures, loss of previously acquired abilities, difficulty walking, weakness, problems with speech or swallowing and increased pressure inside the skull. The FDA's overview describes the disease, its cause and its major symptoms.
Until now, treatment focused on symptoms and supportive care. That might include seizure medicines, physical therapy, feeding support or other measures tailored to the person. None could directly reduce the production of GFAP.
How Zanvastro targets RNA to lower GFAP protein
Zanvastro belongs to a class of medicines called antisense oligonucleotides, often shortened to ASOs.
The name sounds intimidating, but the principle is fairly simple.
DNA stores the cell's instructions. To make a protein, a cell first creates a temporary working copy of the relevant instruction called messenger RNA, or mRNA. A ribosome then reads that message and builds the protein.
An antisense medicine is a short, purpose-built strand of genetic material designed to recognise one particular RNA message. Zilganersen binds to GFAP pre-mRNA—the early version of the message used to make GFAP—and causes it to be broken down. With fewer usable messages, the cell makes less GFAP protein. It lowers overall GFAP production rather than selectively removing only the abnormal protein. Prescribing information
It is important to distinguish this from gene editing. Zanvastro does not cut, replace or permanently alter a patient's DNA. Its effect depends on the medicine remaining active in the nervous system, which is one reason treatment is repeated.
Because large RNA-targeting medicines do not readily cross from the bloodstream into the brain, Zanvastro is delivered by intrathecal injection—a procedure in which a trained professional injects the drug into the fluid surrounding the spinal cord. The approved dose is 50 milligrams every three months. Prescribing information
That is far less convenient than taking a tablet. But for medicines targeting the central nervous system, direct delivery can place the therapy much closer to the cells it needs to reach.
What did the Zanvastro trial show?
The FDA reviewed a randomized, controlled clinical study registered as NCT04849741, along with an open-label substudy involving very young children. The main study enrolled patients aged two and older, while the additional substudy included four children younger than two. ClinicalTrials.gov record
For participants aged five and older who had difficulty walking, the central question was practical: after 61 weeks, could they cover 10 metres faster than patients in the control group?
The walking-speed analysis compared 17 participants receiving the approved 50-milligram dose with 13 controls. At week 61, the adjusted average change in walking speed was −2.1% in the treated group versus −35.4% in controls: a 33.3-percentage-point difference. The result was statistically significant, although the estimate was imprecise because the groups were small. Ionis approval presentation
That figure needs careful interpretation. It does not mean patients became one-third faster. The main story was relative preservation of walking ability while the control group declined—not universal recovery or disappearance of the disease. The comparison covered approximately 14 months.
Among children aged two to four, researchers used a broader assessment of gross motor abilities. The treated group improved on average, while the control group declined, according to the FDA.
Ionis says other measures of function, communication and disease severity also generally favoured the medicine. Those secondary findings add context, but the strongest conclusion rests on the prespecified trial endpoints—not every individual measure or every patient's experience. Ionis trial and approval summary
The trial was unusually small by the standards of common diseases. The FDA describes 49 participants aged two and older in the controlled study, plus four children under two in an open-label substudy. Open-label means everyone knew the children were receiving the drug. The primary walking-speed comparison involved only a subset of the controlled study. That is a reminder of the central challenge in ultra-rare disease research: trials involving hundreds or thousands of people can be extraordinarily difficult to recruit. FDA approval announcement
For children under two, the FDA relied on limited direct safety data from four patients, evidence from older patients and pharmacokinetic modelling—computer-assisted estimates of how the medicine behaves in the body—to support the approved dosing. The evidence in this youngest age group is especially limited; there was no concurrent control group.
What the approval does—and does not—mean
The approval establishes that the FDA found Zanvastro's benefits to outweigh its risks for its approved use. It also transforms Alexander disease from a condition with no approved disease-targeting treatment into one with a medicine aimed at its molecular cause.
But several boundaries matter.
First, the drug reduces production of GFAP; it does not repair the GFAP mutation. It is not a one-time genetic cure.
Second, reducing a toxic protein may slow damage or improve certain abilities, but it cannot be assumed to reverse nervous-system injury that has already occurred. The pivotal comparison lasted 61 weeks, so longer follow-up will be needed to show how well benefits persist over years.
Third, patients may respond differently. Alexander disease varies by mutation, age of onset, affected brain regions and disease stage. An average trial result cannot predict precisely what will happen to an individual.
Finally, approval in the United States is not worldwide approval. Ionis says submissions in Europe and Japan are planned for 2027 through its partner Recordati. Those are future regulatory plans, not guarantees that other agencies will approve the medicine. Ionis announcement
The risks of treating the nervous system directly
The FDA lists vomiting, back pain, cough, headache and post-lumbar-puncture syndrome among the most common adverse reactions.
The prescribing information also carries an important warning about aseptic meningitis—inflammation of the membranes around the brain and spinal cord without the usual bacterial infection. Ionis reports that one patient experienced serious, recurring episodes that required treatment interruption and steroid premedication.
Some discomfort and headache can also come from the lumbar-puncture procedure used to deliver the drug. Repeated intrathecal treatment every three months creates a practical burden for patients, caregivers and specialist centres, especially when families live far from a hospital equipped to provide it.
Approval therefore does not make the decision automatic or risk-free. Treatment belongs in a specialist medical discussion based on the approved prescribing information and a patient's circumstances.
Why this matters beyond one rare disease
Alexander disease affects very few people, but the approval illustrates a larger change in medicine.
Traditional small-molecule drugs often work by fitting into pockets on proteins, like keys entering locks. Some disease-causing proteins have no convenient pocket and are extremely difficult to block directly. RNA medicines offer another route: intervene one step earlier, at the message used to manufacture the protein.
That does not make RNA medicines easy. Scientists still have to design a sequence that is selective, deliver it to the correct tissue, establish a useful dose and monitor immune or inflammatory effects. Direct spinal delivery also limits convenience.
But once a platform for designing and manufacturing antisense medicines exists, researchers can adapt the underlying strategy to different RNA targets. Each drug still requires its own rigorous testing, yet the approach turns genetic information itself into a starting point for drug design.
For families affected by Alexander disease, that broader platform story is secondary. What matters immediately is narrower and more concrete: a disorder that had no FDA-approved treatment now has one that addresses the biological process driving it.
The accurate headline is not that scientists have cured a fatal brain disease. They have built an RNA medicine that lowers a toxic protein, shown meaningful functional effects in a very small controlled trial, and persuaded regulators that the benefits justify the known risks.
For an ultra-rare disease, that is already a major shift.
Sources and further reading
- FDA: approval of the first drug to treat Alexander disease
- ClinicalTrials.gov: NCT04849741
- Ionis: US approval, trial results and safety information
- FDA: 2026 novel drug approvals
- Reuters: FDA approval and patient-population context
This article is for general information and does not provide medical advice.
Leave a Reply