uniQure has submitted its gene therapy AMT-130 to the FDA for accelerated approval in Huntington’s disease — a fatal inherited neurodegenerative disorder with no treatment that slows its course. The filing was made on September 2, 2026.

The disease

Huntington’s is caused by a single faulty gene, and it is inherited in a pattern that makes it uniquely cruel: a child of an affected parent has a 50% chance of carrying it, and a genetic test can tell someone decades in advance that they will develop a disease with no treatment.

The mutation produces an abnormal huntingtin protein that accumulates and kills neurons, particularly in the striatum, a deep brain structure central to movement control. Patients progressively lose motor control, cognition and independence over years.

Because the genetic cause is unambiguous and identified in a single gene, Huntington’s has attracted repeated therapeutic attempts — and defeated them. The most prominent, a huntingtin-lowering antisense drug from a major developer, was halted in Phase 3 in 2021 when an interim analysis found it was not helping and might be causing harm.

What AMT-130 is

A one-time gene therapy delivered directly into the brain in an hours-long neurosurgical procedure. Rather than adding a working gene, it delivers genetic material designed to reduce production of the harmful huntingtin protein in the cells that need it most.

The surgical delivery is a direct response to what has defeated other approaches. Getting a therapeutic into deep brain structures at meaningful concentration is extraordinarily difficult; injecting into cerebrospinal fluid distributes unevenly and reaches deep tissue poorly. Placing the therapy into the striatum under imaging guidance solves the distribution problem — at the cost of requiring brain surgery.

The data

In three-year data, AMT-130 appeared to significantly slow signs of disease progression — a potentially historic result in a disease where nothing has worked. uniQure is seeking a priority review that could produce a decision in roughly eight months, with confirmatory four-year data expected soon.

The regulatory fight that nearly stopped it

The path here was extraordinary, and it is a substantial part of the story.

The three-year results compared treated patients against an external natural-history database — records of how Huntington’s patients decline without treatment. Under FDA leadership in place from late 2025 to mid-2026, the agency rejected that comparison and demanded a new sham-surgery-controlled trial.

That demand carried a genuine ethical problem. A sham control for this therapy means performing a fake brain surgery — anaesthesia, incision, skull burr holes — on patients with a fatal degenerative illness, delivering nothing. The scientific rationale is real, because surgical procedures produce powerful placebo effects and unblinded assessment is unreliable. But the cost imposed on the control group is severe in a way a sugar pill is not.

The dispute became public. The FDA’s then-commissioner and a senior official subsequently resigned, and by mid-2026 interim leadership reversed course and accepted the original three-year dataset for filing.

Why external controls are contentious

The underlying question is legitimate and will follow this application. Comparing treated patients to historical records is vulnerable to bias: trial participants are often healthier, more closely monitored, and more motivated than the general patient population, which can make a therapy look effective when it is not.

Against that, Huntington’s has a well-characterised and relatively predictable progression, and large prospective natural-history datasets exist — which is precisely why an external comparison is more defensible here than in most diseases.

Reasonable regulators could disagree. What was unusual was not the disagreement but the reversal, and the leadership turmoil surrounding it.

What accelerated approval would mean

Accelerated approval permits authorisation on a measure reasonably likely to predict benefit, with confirmation required afterwards. For a therapy involving irreversible brain surgery, that carries more weight than usual: a patient who receives it cannot un-receive it if confirmatory data disappoints. The four-year dataset therefore matters considerably.

What comes next

This is a filing, not an approval. The FDA’s decision is still ahead, and priority review would put it roughly eight months out.

What a one-time surgical therapy means in practice

Even with approval, delivery would be a serious constraint on how many patients could actually receive AMT-130.

The therapy requires hours-long stereotactic neurosurgery with imaging guidance, which restricts administration to specialised centres with the equipment and surgical expertise to perform it reliably. That is a very different distribution problem from a drug shipped to pharmacies, and it means access would initially depend heavily on geography and referral pathways.

The one-time nature cuts both ways. There is no adherence problem and no recurring administration, which for a disease that progressively impairs cognition and independence is a genuine advantage. But it also removes the ability to stop, adjust the dose, or reverse course if something goes wrong — and it front-loads the entire risk of the treatment into a single irreversible procedure, which is exactly why the confirmatory four-year data carries so much weight.

For families, the significance is hard to overstate — not only a first treatment, but a shift in what a positive genetic test means. The episode also illustrates something less comfortable: whether a therapy for an untreatable disease reaches patients depended, in this case, on which officials happened to be running the agency. Regulatory news, not medical advice.