uniQure Loses 37% in a Session After Its Huntington's Gene Therapy Shows 44% Slowing at Four Years on a Primary Endpoint That Missed Significance

by Ellis Vance · September 30, 2026

Small-Cap Markets Correspondent, The SmallCap Signal

uniQure N.V. (Nasdaq: QURE) closed at $24.51 on Tuesday, Sept. 29, down 37.33% from the prior session's $39.11 close, on volume of 29,813,992 shares against 1,853,457 the day before, according to the historical-data table on stockanalysis.com as read on Oct. 2. The catalyst was a pre-market release reporting that, at 48 months, the company's Huntington's disease gene therapy ifezuntirgene inilparvovec (AMT-130) showed a 44% slowing of disease progression on the composite Unified Huntington's Disease Rating Scale versus an external control, with a p-value of 0.144 that the company itself described as not statistically significant.

That figure matters because of what it is measured against. On Sept. 24, 2025, uniQure announced that the same high dose had produced a 75% slowing on cUHDRS at 36 months (p=0.003), the study's pre-specified primary endpoint and the basis for its biologics license application. The 48-month cUHDRS result is the pre-specified primary endpoint for the four-year analysis, per a footnote in the Sept. 29 release, and it did not clear the significance bar.

The release, furnished as Exhibit 99.2 to a Form 8-K dated Sept. 29, lays out the 48-month high-dose results for 12 patients. Treated patients had a mean change in cUHDRS from baseline of -0.90, against -1.61 for the propensity score-matched external control, a difference of 0.71. On Total Functional Capacity, the measure the company has made the primary endpoint of its planned confirmatory study, the slowing was 61% with a nominal p-value of 0.008: -0.37 in treated patients versus -0.94 in the control. Mean cerebrospinal fluid neurofilament light protein was 4% above baseline in the 11 patients with a reading, having been 8.2% below baseline in the 36-month readout a year earlier.

The company's explanation for the weaker 48-month result centres on the comparator. The external control was rebuilt from an updated ENROLL-HD natural history dataset with a September 2025 cutoff, matched to 1,337 patients for the high-dose comparison, up from 940 in the 2025 analysis. According to the release, missing data in that matched control reached 53% at 48 months, and 'the Company's analysis of the updated control showed that patients discontinuing follow-up were progressing materially faster than those remaining in the control group.' Those factors, the company said, 'likely understated disease progression in the control and the resulting treatment effect of ifezuntirgene inilparvovec at 48 months.'

To support that reading, uniQure ran a post-hoc sensitivity analysis against the prior, smaller ENROLL-HD control. On that basis the 48-month slowing was 53.5% on cUHDRS (nominal p=0.041) and 68.3% on TFC (nominal p=0.001). The release labels the analysis post hoc, and its risk factors include the possibility that additional patient data produce a different interpretation from the one drawn from the three- and four-year analyses.

The 36-month data, by contrast, got better. With three more high-dose patients reaching the timepoint since the 2025 cut, that group now numbers 15, and the company reported an 80% slowing on cUHDRS (nominal p=0.005) and 67% on TFC (nominal p=0.011), against 75% and 60% a year ago; CSF NfL was 6% below baseline (n=14). At 48 months the company also compared doses: mean cUHDRS change of -0.91 in high-dose patients versus -1.94 in the 12 low-dose patients, and TFC of -0.30 versus -0.70, which it called 'consistent with a dose-dependent treatment effect.' (The release gives the 48-month high-dose cUHDRS change as -0.90 in one bullet and -0.91 in another; it does not reconcile the two.)

Chief medical officer Walid Abi-Saab framed the update around the functional measure: 'At 48 months, Total Functional Capacity (TFC), the primary measure of our confirmatory study, demonstrated consistent slowing of functional decline, with the absolute treatment benefit maintained in Month 48,' he said in the release. Victor Sung of the University of Alabama at Birmingham, quoted in the same release, said the slower decline in the updated comparator 'appears to reflect some attrition in the longitudinal external control data.'

The regulatory position explains why the company is arguing about the comparator rather than the drug. On Sept. 2, uniQure announced it had submitted a BLA to the FDA seeking accelerated approval, plus a marketing authorisation application to the UK's MHRA, both supported by the three-year analysis, and said it had requested priority review, which if granted 'would shorten the FDA review cycle to six months following the FDA's 60-day BLA filing review period.' Tuesday's release states that at a June 2026 Type B meeting 'the FDA communicated that the 36-month data from 12 high-dose patients would be acceptable as the primary basis for the BLA submission under the accelerated approval pathway,' and that the four-year results 'were not part of the submission.'

Whether the FDA sees it that way is the open question. The company's July 29 results release noted that the agency 'seeks to align on the confirmatory study design prior to the BLA submission,' including a randomized standard-of-care control rather than a sham procedure, and that under draft guidance the confirmatory study 'should be feasible to conduct within a reasonable timeline and be well underway, and potentially fully enrolled, at the time of accelerated approval.' The Sept. 29 release does not say whether that study has begun, nor does it give a filing-acceptance date or a PDUFA target.

On safety, the release repeats that five high-dose participants (17%) had a treatment-related serious adverse event of central nervous system inflammation, all resolved. It also discloses that, since the September 2025 readout, one low-dose patient died by suicide approximately five years after treatment, assessed by the study investigator as unrelated to treatment; the company noted that suicide 'is among the leading causes of death in those with the disease.'

uniQure is a Lexington, Mass. and Amsterdam-based gene therapy developer with an approved hemophilia B therapy. For the quarter ended June 30, 2026, it reported revenue of $5.8 million and a net loss of $81.1 million, or $1.22 per share, according to its July 29 results release, which put cash, cash equivalents and current investment securities at $810.3 million as of June 30, after a $259 million follow-on offering, and said those resources should fund operations into 2030. The company has been hiring ahead of a launch: it attributed a $4.3 million increase in employee and contractor costs to staff 'recruited in the second half of 2025 to support the potential commercial launches of AMT-130.'

Tuesday's collapse landed on a stock that was already weakening. The stockanalysis.com table shows closes of $45.28 on Sept. 22, $42.76 on Sept. 23, $38.32 on Sept. 24 (a 10.38% one-day drop for which we found no company filing), $38.23 on Sept. 25 and $39.11 on Sept. 28. By our arithmetic, the $24.51 close is about 46% below the Sept. 22 level. No Sept. 30 price is used here; the session had not closed at the time of writing.

The 48-month analysis rests on 12 patients; the 36-month analysis on 15. Both are compared with a matched natural-history cohort rather than a concurrent randomized control, the design the FDA appears to want for confirmation. The company's case is that a comparator missing 53% of its data flatters the untreated population; the counter-reading, visible in Tuesday's volume, is that the primary-endpoint treatment difference narrowed from 1.14 points at three years (our arithmetic from the 2025 release's -0.38 and -1.52) to 0.71 at four, and the biomarker moved the wrong way.

What comes next is largely in the FDA's hands. The company has not announced acceptance of the filing, priority review status or an action date, and said it will present the four-year data 'at a future scientific meeting.' Not yet known: whether the agency asks for the 48-month analysis to be folded into the review, the start date and design of the confirmatory study, and whether the 36-month cohort holds its 80% as the last two high-dose patients reach that timepoint.