Why uniQure’s AMT-130 For Huntington’s Is Not "Too Good To Be True"

This is the author’s opinion only, not financial advice, and is intended for entertainment purposes only. While the author aims for accuracy, no guarantee is made regarding the completeness or correctness of the data, assumptions, or statistical analysis provided. The author holds beneficial long positions in uniQure N.V. (NASDAQ: QURE) and ClearPoint Neuro, Inc. (NASDAQ: CLPT). The author receives no compensation for writing this article and has no business relationship with any of the companies mentioned.

On September 24, 2025, Dutch biotech company uniQure (NASDAQ: QURE) announced that its Phase I/II study of AMT-130, an investigational gene therapy for Huntington’s disease (HD), met its primary endpoint by demonstrating a statistically significant 75% slowing of disease progression at 36 months compared to an external control. At the time, I could not hold back my excitement and called AMT-130 "one giant leap for medicine," as these results would render AMT-130 the first disease-modifying therapy for HD, a devastating and fatal neurodegenerative disease.

However, the excitement was short-lived. On November 3, uniQure reported that "the FDA currently no longer agrees that data from the Phase I/II studies of AMT-130 in comparison to an external control [...] may be adequate to provide the primary evidence in support of a BLA submission." This position was confirmed in final meeting minutes disclosed on December 4, in which uniQure noted that "the FDA conveyed that data submitted from the Phase I/II studies of AMT-130 are currently unlikely to provide the primary evidence to support a BLA submission." Consequently, the timeline for a Biologics License Application (BLA) submission is now uncertain, highlighting the high regulatory bar for methodologies relying on propensity score matching against external controls.

At the same time, the FDA appears to be facing significant internal challenges. The Wall Street Journal, for example, warned on December 4, in an article explicitly mentioning uniQure titled "Turmoil at the FDA Threatens Biotech Recovery," that "chaos inside the agency [...] is undermining confidence and threatening investment." I previously discussed the FDA's reversal and its potential consequences in an article published here on November 11.

Despite the FDA’s agreement to uniQure’s study design in various prior meetings, this reversal appears driven by a new, dogmatic preference for conventional trial designs over pragmatic evidence generation. While randomized controlled trials (RCTs) are the gold standard, they are neither universally applicable nor always ethical. In special cases involving rare diseases like HD that lack alternative therapies, external controls have long proven to be a valid benchmark. For AMT-130, a traditional placebo control would necessitate a sham neurosurgical procedure, which is an invasive intervention without therapeutic intent and therefore hard to justify from an ethical standpoint.

In fact, the FDA's exact concerns on the AMT-130 study are unknown, as they have not been disclosed by uniQure, to my knowledge. Therefore, the assertion that the FDA is simply demanding an RCT is currently just speculation.

Consequently, I decided to take a step back and reviewed the study protocol and all information released by the company, where the following development caught my eye. Prior to the disclosure of the FDA's reversal, the study protocol was updated on October 21 and uniQure reported in third quarter report on November 10 that it "initiated and fully recruited a fourth cohort evaluating high-dose AMT-130 in six patients with lower striatal volumes compared to patients in previous cohorts" in October. This new cohort specifically targets patients with more advanced atrophy, defined as those with putamen volumes <2.5 cm³ or caudate volumes <2.0 cm³ per side, who were excluded from the initial cohorts. In contrast, the inclusion criteria for the first three cohorts required volumes of ≥2.5 cm³ per side for the putamen and ≥2.0 cm³ per side for the caudate, resulting in total minimum volumes of 5.0 cm³ and 4.0 cm³, respectively.

This might be significant because the impressive results of the Phase I/II study were derived using propensity score matching against the Enroll-HD dataset, which lacks systematic MRI data for all participants. Consequently, one could argue that the volumetric inclusion criteria for the first three cohorts introduced a selection bias: By enrolling only patients with less atrophic putamen and caudate volumes, and thus greater cerebral reserves, the study may have selected individuals with inherently slower disease progression compared to a control group lacking comparable volumetric stratification. Ultimately, the question arises whether volumetric bias could have artificially inflated the reported treatment effect.

To assess potential bias from volumetric inclusion criteria, the AMT-130 study population must be accurately characterized and benchmarked against established natural history data. Fortunately, van den Bogaard et al. published a detailed analysis in 2010 focusing on the atrophy of subcortical grey matter structures in HD, such as the putamen and caudate. This extensive study analyzed data from 366 participants in the Track-HD dataset, including 123 controls, 120 premanifest individuals, and 123 manifest patients. The findings revealed two crucial insights regarding the potential volumetric bias in uniQure's Phase I/II study:

1. Volume reduction of the putamen is a strong predictor of "motor disturbances in manifest HD."
2. The influence of caudate volume on predicting "motor disturbances in manifest HD" is "minimal," and regarding clinical symptoms, "the role of the caudate nucleus in the premanifest stage of the disease seems to diminish after manifestation."

This implies that in patients with early manifest HD, such as those in the Phase I/II AMT-130 study, only the inclusion threshold for putamen volume is likely to introduce a relevant bias, whereas the threshold for caudate volume is unlikely to have an effect.

To further evaluate this, let's compare the results reported by van den Bogaard et al. with the population of the AMT-130 study. Cohorts 1 and 2 of the AMT-130 study included only patients with a Total Functional Capacity (TFC) score of 9 to 13, while Cohort 3 required a TFC of 11 to 13. In van den Bogaard's analysis, patients with manifest HD were stratified into HD1 (TFC 11–13, n=77) and HD2 (TFC 7–10, n=46). Therefore, the AMT-130 population in the first three cohorts aligns directly with the HD1 group and overlaps with the milder portion (TFC 9–10) of the HD2 group. Crucially, the HD2 subgroup as a whole represents a more advanced disease state than the AMT-130 enrollment criteria permitted, as it includes patients with TFC scores of 7 and 8 who were ineligible for the AMT-130 study. The volumetric data from the analysis of the TRACK-HD dataset, presented in the table below with standard deviations (SD), provide the necessary context.

From van den Bogaard SJ, Dumas EM, Acharya TP, Johnson H, Langbehn DR, Scahill RI, Tabrizi SJ, van Buchem MA, van der Grond J, Roos RA (2010) Early atrophy of pallidum and accumbens nucleus in Huntington’s disease. Journal of Neurology 258(3):412–420

For the HD1 group, which serves as the most representative comparison for the majority of the initial AMT-130 population, the mean putamen volume was 7.55 cm³. The 5.0 cm³ threshold applied in the first three cohorts of the AMT-130 study lies approximately 2.58 standard deviations below this mean. Consequently, assuming a normal distribution, this threshold would exclude only an estimated 0.50% of the HD1 population based on putamen volume. Regarding the caudate volume in the HD1 group (mean 5.06 cm³), the 4.0 cm³ threshold falls 1.39 standard deviations below the mean, corresponding to an exclusion of approximately 8% of patients.

The criteria remain permissive even when compared to the more advanced HD2 group. Again, the HD2 group only partially overlaps with the AMT-130 study population. While cohorts 1 and 2 included patients with TFC scores of 9 and 10, cohort 3 was restricted to TFC scores of 11 to 13. Furthermore, the HD2 group includes patients with TFC scores of 7 and 8 who were ineligible for any cohort in the AMT-130 study. Despite the inclusion of these more advanced patients in the comparison data, the putamen volume threshold of the AMT-130 study falls 1.62 standard deviations below the HD2 mean (6.94 cm³), theoretically excluding only the lowest 5%. The caudate threshold lies 1.01 standard deviations below the HD2 mean (4.83 cm³), excluding approximately 16%.

Although the inclusion threshold for putamen volume appears sufficiently permissive, one could argue that the threshold for caudate volume was too restrictive. However, as van den Bogaard et al. reported, the volume reduction of the caudate in manifest HD appears to have minimal influence on predicting clinical symptoms. Consequently, excluding a small subset of patients with lower caudate volumes is unlikely to introduce a relevant bias. Thus, only the putamen volume remains a potential source of bias, yet this risk appears negligible as the threshold excludes only around 0.5% of patients with a TFC of 11 to 13.

One potential critique of this rationale is the use of total volumes (the sum of left and right) as a proxy for the inclusion criteria specific to each side. However, this assumption is statistically justifiable given the pathology of Huntington’s. HD is a systemic, genetically driven disorder characterized by bilateral and generally symmetric atrophy of subcortical grey matter (Tabrizi et al., 2009). While some neuroimaging studies note subtle lateralization, often exhibiting a slight left hemisphere dominance in atrophy (Rosas et al., 2001), the interhemispheric correlation for striatal volume remains very high.

Given that the putamen volume threshold excluded only a negligible fraction of the target population, it appears unlikely that the AMT-130 study selectively enrolled patients with larger "cerebral reserve," and thus potential slow progressors, in the first three cohorts. The criteria permitted the enrollment of the vast majority of patients with typical putamen volumes for their disease stage. Instead, the thresholds likely reflect pragmatic requirements for the neurosurgical administration of the therapy. AMT-130 utilizes an adeno-associated virus (AAV) vector to deliver a microRNA designed to silence the huntingtin gene. Intrastriatal delivery of AAV gene therapy requires adequate target volume to ensure safe and effective transduction. This technical constraint provides a clinical rationale for excluding patients with severe atrophy.

Furthermore, it was noted in the announcement of the AMT-130 study results that "various other supportive analyses of the results from the AMT-130 high dose treatment group, including those using a propensity score-weighted external control and comparisons to the TRACK-HD and PREDICT-HD datasets, were consistent with the primary analysis." This indicates that the findings were explicitly benchmarked against the TRACK-HD dataset, which contains the relevant volumetric data.

Additionally, the clinical effect of AMT-130 is corroborated by objective biomarker data. Neurofilament Light chain (NfL) serves as a validated biomarker of axonal damage. In contrast to the expected progressive rise observed in the natural history of HD, mean cerebrospinal fluid NfL levels in the high-dose group declined to 8.2% below baseline at 36 months. This finding provides independent biological evidence of reduced neurodegeneration.

In conclusion, skepticism regarding the statistical validity of the AMT-130 trial based on its initial volumetric inclusion criteria appears unsupported by the data. While the regulatory pathway remains uncertain, the convergence of statistically significant clinical endpoints and favorable changes in NfL biomarkers indicates that the observed treatment effect of AMT-130 is genuine. Data from the newly recruited Cohort 4 will be critical for assessing the therapy's efficacy in patients with more advanced atrophy.

On November 12, Dr. Vinay Prasad, the FDA's Director of the Center for Biologics Evaluation and Research, and Dr. Martin A. Makary, the Commissioner of Food and Drugs, published an article in The New England Journal of Medicine titled "FDA’s New Plausible Mechanism Pathway." In this piece, they wrote that "the FDA has heard [...] that current regulations are onerous and unnecessarily demanding, provide unclear patient protection, and stifle innovation." Additionally, they argued that "an appropriately designed study with a small sample size can support licensure of a product for which pharmacologic effect is aligned with biologic plausibility and congruent with observed clinical outcomes." They further stated that "the Food and Drug Administration (FDA) is committed to providing regulatory guidance and encouragement, outlining a path to market entry for products where a randomized trial is not feasible."

Despite this public commitment, their approach to uniQure's breakthrough therapy AMT-130 appears to diverge from this philosophy. Let's hope they reconsider their position, as time is running out for patients.

Follow me on X for frequent updates (@chaotropy).