uniQure released four-year results for its Huntington’s disease gene therapy on Tuesday 29 September 2026. The disease progression slowing measured on the composite clinical scale narrowed at 48 months. It missed statistical significance. A second measure, Total Functional Capacity, held. Shares fell sharply. The application now sitting with the FDA, however, does not rest on this four-year cut.
What uniQure reported at 48 months
The company issued the additional four-year data through GlobeNewswire at 07:12 Eastern on 29 September 2026. The release is datelined Lexington, Massachusetts and Amsterdam. It covers ifezuntirgene inilparvovec, the gene therapy previously known as AMT-130.
Twelve high-dose patients reached the 48-month mark. On the composite Unified Huntington’s Disease Rating Scale, uniQure reported disease progression slowing of 44 per cent. The comparison ran against an updated external control. The p-value was 0.144, so it did not reach statistical significance. Mean change was minus 0.90 in treated patients and minus 1.61 in controls.
Total Functional Capacity gave a different answer. At the same timepoint uniQure reported disease progression slowing of 61 per cent, at a p-value of 0.008. Mean change was minus 0.37 against minus 0.94 in the control group.
A separate 36-month analysis covered 15 high-dose patients. That cut showed disease progression slowing of 80 per cent on the composite scale, at a p-value of 0.005. Total Functional Capacity showed disease progression slowing of 67 per cent, at a p-value of 0.011.
The release also set the two doses against each other at 48 months. On the composite scale, high-dose patients changed by minus 0.91 and low-dose patients by minus 1.94. On Total Functional Capacity the figures were minus 0.30 and minus 0.70. Each comparison used 12 patients per group.
Disease progression slowing is measured against a moving control
The external control comes from ENROLL-HD, a long-running natural history study. uniQure matches treated patients to those records using propensity scores. The company said missing data in the control rose with the duration of follow-up.
Its release put that figure at 53 per cent in the updated matched controls at 48 months. uniQure also said patients who stopped follow-up were progressing materially faster than those who stayed.
What the post-hoc analysis shows
uniQure ran a sensitivity analysis against the prior ENROLL-HD control. On that comparator the 48-month composite result was disease progression slowing of 53.5 per cent. The nominal p-value was 0.041. Total Functional Capacity gave disease progression slowing of 68.3 per cent, at a nominal p-value of 0.001. The company labelled the analysis post-hoc.
One point deserves care, and uniQure has not framed it this way. The 36-month comparison also moved between the two readouts. In September 2025 the company reported 75 per cent on the composite scale, at a p-value of 0.003. That analysis used 12 high-dose patients against 940 matched ENROLL-HD controls.
The September 2026 release gives 80 per cent at the same timepoint. It uses 15 patients against 1,337 matched controls. So the updated control did not depress every comparison. The treated population grew at the same time. The two changes cannot be separated from the published figures, and LSDN draws no conclusion about which mattered more.
There is a wider reading here for anyone filing on an external control. A matched natural history cohort is not a fixed object. It grows, it loses patients, and the patients it loses may not be typical. uniQure’s own explanation makes that point. Sponsors may find the comparator itself becomes a review question.
Why the application rests on the 36-month data
uniQure submitted the biologics license application on 2 September 2026. A marketing authorisation application went to the MHRA the same day. LSDN covered the 2 September filing with the FDA and the MHRA at the time.
The company requested priority review. It said a grant would shorten the review to six months after the FDA’s 60-day filing period. LSDN found no announcement of a filing acceptance or a priority review decision as at 30 September 2026.
The 29 September release restates what the agency agreed earlier in the year. At the June 2026 Type B meeting, the FDA said 36-month data from 12 patients would be acceptable. Those were high-dose patients. It would serve as the primary basis for the submission under the accelerated approval pathway.
That framing matters for how the four-year numbers should be read. On LSDN’s reading, the 48-month cut is additional evidence rather than the primary basis of the file. The dataset the FDA agreed to review has not changed since 2 September.
The FDA has moved on this question twice
In June 2025 uniQure said the FDA continued to support its earlier agreement on the composite scale. The agency said it may serve as an acceptable registrational, intermediate clinical endpoint for accelerated approval. It also agreed that ENROLL-HD could serve as the external control dataset.
That position did not hold. uniQure announced on 2 March 2026 that a Type A meeting had gone the other way. The meeting itself took place on 30 January 2026. The FDA said it could not agree that the Phase I/II data were sufficient. That comparison was against an external control. The agency called the evidence insufficient to support a marketing application.
It strongly recommended a prospective, randomised, double-blind, sham surgery-controlled study instead. Three and a half months later the position moved again. The June 2026 Type B meeting produced the agreement that underpins the current filing.
LSDN sets out that sequence because it bears on the four-year numbers. The new disease progression slowing figures arrive against that history. The endpoint that missed at 48 months is the composite scale. That is the same measure the agency had backed as the intermediate clinical endpoint by June 2025. It is also the measure whose external-control comparison the agency rejected in early 2026, then accepted in June.
LSDN draws no conclusion about how the agency will read the new data. Neither uniQure’s release nor the FDA had addressed that point in writing as at 30 September 2026.
The endpoint that held is the one the confirmatory study uses
Walid Abi-Saab, Chief Medical Officer, uniQure, addressed this directly in the release. He described Total Functional Capacity as “the primary measure of our confirmatory study”. He said the absolute treatment benefit was maintained at Month 48.
That detail matters for anyone modelling the accelerated approval route. Accelerated approval requires a confirmatory trial. The measure that missed at four years is not the measure that trial will be judged on.
That cuts both ways. The composite scale is the measure the FDA had backed for accelerated approval by June 2025. So the 48-month miss falls on the endpoint carrying the approval itself. The confirmatory measure is the one that held.
The design of that trial is still being settled. uniQure said in June 2026 that the FDA was considering a concurrent control on standard-of-care therapy. The alternative under discussion was a sham procedure.
Victor Sung, Professor of Neurology, University of Alabama at Birmingham, also commented in the release. He said Huntington’s disease “does not slow on its own”. He pointed to “the apparent stability of the functional capacity benefit through Month 48”.
How the programme is structured
The data come from two Phase I/II protocols. NCT05243017 is the randomised United States study. NCT04120493 is the European open-label study.
uniQure said 51 patients have been dosed across the programme. The analysis published on 29 September covers 29 patients from the first two cohorts. Seventeen received the high dose and 12 received the low dose.
The United States study enrolled 26 patients. Ten received the high dose, six received the low dose and 10 formed a sham control group. The company said four control patients later crossed over to active treatment.
A third cohort of 12 patients tested an optimised immunosuppression regimen. A fourth cohort of six United States patients received the high dose. Those patients had lower striatal volumes.
Phase, enrolment and dose detail here come from uniQure’s own statements. The therapy carries Breakthrough Therapy, Regenerative Medicine Advanced Therapy and Fast Track designations from the FDA.
Safety through four years
uniQure said the therapy continues to be generally well tolerated at both doses. It described the safety profile as manageable. The most common adverse events in the treatment groups related to the administration procedure, and all have resolved.
Five high-dose participants had a treatment-related serious adverse event. Each involved central nervous system inflammation. The company said all fully resolved, and that it had disclosed them previously.
The release also disclosed one death. A low-dose patient died by suicide about five years after treatment. uniQure recorded the event as occurring since the September 2025 data readout. The study investigator assessed it as unrelated to treatment, the company said.
Cerebrospinal fluid neurofilament light moved in both directions. At 36 months it sat 6 per cent below baseline in 14 patients. At 48 months it sat 4 per cent above baseline in 11 patients.
The UK arm and the route through the MHRA
Britain contributed patients and surgeons to the programme. University College London says its Huntington’s Disease Centre ran the UK trial site at UCL and UCLH. Professor Ed Wild was principal investigator at that site.
Professor Sarah Tabrizi was lead scientific advisor on the trial. She is based at the UCL Huntington’s Disease Research Centre and the UK Dementia Research Institute at UCL. In September 2025 she called the three-year results “the most convincing evidence in the field to date”.
UCL says the neurosurgeries for the UK arm took place at the University Hospital of Wales in Cardiff. Health and Care Research Wales funded that surgical work.
The regulatory route here is separate from the American one. uniQure announced a successful MHRA pre-submission meeting on 30 April 2026. The marketing authorisation application followed on 2 September 2026.
No NICE topic listed as at 30 September
LSDN searched the NICE guidance work programme on 30 September 2026. No topic naming Huntington’s disease, uniQure, AMT-130 or ifezuntirgene appeared on the awaiting-development list. That is a dated search finding, not a statement of NICE’s intentions. A licence in Britain would not by itself create NHS funding in England.
What the market did
The share reaction was severe. BioPharma Dive reported that uniQure shares fell about 40 per cent at the United States market open. That was on Tuesday 29 September. It put the level just below 24 US dollars. It estimated about 1 billion dollars of market value erased.
Published accounts of the size of the move varied widely. Figures ranged from about 37 per cent to about 65 per cent, depending on the window measured.
Analyst reaction was mixed rather than uniformly negative. BioPharma Dive reported that analysts at Leerink Partners expected investors to question durability. The same report quoted them calling slowing of any kind on Total Functional Capacity unheard of in this disease. Stifel analyst Paul Matteis called the four-year results “still impressive generally”. He described the three-year results as robust.
What to watch next
Four questions now sit ahead of this programme. Whether the FDA accepts the filing and grants priority review is the first. Whether the confirmatory study starts, and under which control arm, is the second.
The third is whether the MHRA validates the marketing authorisation application, and on what timetable. The fourth is whether NICE opens an appraisal once a British licence exists.
The disease progression slowing debate will not be settled by this readout. It will be settled by a confirmatory trial measured on Total Functional Capacity. LSDN tracks pending American decisions in its FDA drug approval decisions tracker.
