uniQure's Huntington's Gene Therapy Reaches the FDA on a Missed Endpoint

On November 2, the FDA accepted uniQure's application to approve AMT-130, a gene therapy for Huntington's disease. The agency granted Priority Review and set a target decision date of May 1, 2027. If approved, it would be the first treatment shown to slow the inherited, fatal brain disorder.

Five weeks earlier, the same therapy missed its primary endpoint at 48 months. The stock fell 41% in a day. A senior FDA official had described the treatment as a "failed product."

Both facts now sit in the same file. The question is what the agency is willing to count as evidence when a disease has no good options and the patient community has waited generations.

What the FDA actually accepted

The application does not rest on a randomized pivotal trial. It rests on a 36-month analysis of two Phase I/II studies, compared against a propensity-score-matched external control drawn from the Enroll-HD natural history registry.

That is the accelerated-approval bargain. The FDA can accept an early signal that is reasonably likely to predict benefit, then require confirmation after approval. For a disease that kills everyone it touches, the trade is deliberate.

uniQure has dosed 51 patients across the programme. The high-dose arm at 36 months numbered 12. The company reported a 60% slowing of progression on the Total Functional Capacity scale (p=0.033) against that external control. The therapy carries Breakthrough Therapy, RMAT and Fast Track designations, which speed review without promising approval.

What the FDA accepted is a file. What it decides by May is whether the file is enough.

Pole A: a 44% signal is worth a conditional yes

Huntington's has no approved therapy that delays onset or slows progression. About 75,000 people in the US, EU and UK live with it. The standard of care manages symptoms while the disease advances.

Against that baseline, a one-time treatment with a consistent direction of effect across three years is not a rounding error. The 36-month analysis was statistically significant on TFC. The FDA's accelerated pathway exists for serious disease, unmet need, early evidence and a confirmatory obligation.

"The FDA's acceptance of our BLA for review marks an important milestone for the Huntington's disease community, which has waited generations for a treatment that changes the course of this devastating disease."— Matt Kapusta, chief executive officer, uniQure

Patients and families have argued for years that waiting for a perfect trial is its own harm. An advisory committee, if it convenes, will hear that argument directly from the people the disease affects.

Pole B: an external control is not a randomized trial

The 48-month readout is the problem. On the composite Huntington's Disease Rating Scale, the high-dose group declined 44% more slowly than the Enroll-HD cohort. The result fell short of statistical significance. Investors treated that as a verdict and cut the stock by 41%.

The comparison is the weak point. Enroll-HD is a natural-history registry, not a randomized arm. Propensity-score matching can balance variables that someone thought to record. It cannot balance the ones nobody measured. And it cannot fix a high-dose group of 12 people at 36 months.

There is also the regulatory record. In 2022, dosing was paused after three patients were hospitalized. The pause lifted later that year. In 2025, the FDA told uniQure the Phase I/II data were not enough to file. It reversed that position in June 2026, after a change in CBER leadership, and accepted the application in November.

That sequence does not prove the bar was lowered. It proves the bar moved. A reviewer can reasonably ask whether the reversal reflects new evidence or a new mood.

Five high-dose patients developed central nervous system inflammation, which resolved with treatment. The surgical delivery into the striatum is itself a risk. The confirmatory trial will have to separate drug effect from procedure effect, and it will have to do so in patients who know whether they received the therapy.

The bridge: what May 1, 2027 decides

The PDUFA date is a target, not a guarantee. The FDA is tentatively planning an advisory committee, most likely the Cellular, Tissue, and Gene Therapies Advisory Committee. A negative vote would not bind the agency. It would shape the review and the public record around it.

The confirmatory trial is the real insurance policy. It is designed for about 200 patients, randomized 2:1 to the gene therapy or standard of care, with TFC at 36 months as the primary endpoint. Screening is supposed to begin in the US by the end of 2026, so the study is underway before any approval.

That timeline is the tell. If the trial slips, the approval rests on a smaller and older dataset. If it starts on schedule, the FDA gets its confirmation while the market gets a product.

Accelerated approval is reversible. The FDA can withdraw a product if the confirmatory data disappoint. That makes May less a finish line than a conditional start.

The number that does not go away

The Huntington's community has spent decades watching promising mechanisms fail in the clinic. As we wrote in October about Estuary's local gene therapy for polycystic kidney disease, the delivery method and the biology had to be judged together. That earlier filing faced the same question: how much certainty should a fatal disease demand before a regulator says yes?

If AMT-130 is approved, the label will carry a story about what the FDA counted. If it is rejected, the story will be about what it refused to count. Either way, the 48-month number stays in the file.

That is the bet on the table: a 44% signal, an external control, and a confirmatory trial that has not enrolled a patient yet. The FDA has agreed to look. May will show what looking means.