In a letter sent to global Huntington’s disease patient community leaders on July 9, 2026, Swiss pharmaceutical giant Roche announced the simultaneous discontinuation of its two leading Huntington’s disease drug candidates \u2014 tominersen and RG6496 \u2014 ending more than a decade of development and leaving roughly 30,000 Americans, and hundreds of thousands of people worldwide, with an inherited and uniformly fatal neurological condition that has no approved disease-modifying treatment.<\/p>\n
The announcement delivered a second consecutive blow to a patient community that watched the same company’s earlier Phase 3 tominersen program collapse in 2021. This time the science offered something worse than a straightforward failure: the drug appeared to work biologically while patients did not improve clinically, a gap that scientists say raises fundamental questions about trial design \u2014 and may ultimately say as much about how the field studies Huntington’s disease as it does about the drugs themselves.<\/p>\n
Roche Scrapped Two Programs at Once \u2014 Here Is Why Each Failed<\/p>\n
The larger of the two terminations concerns tominersen, a drug with an unusually prominent history<\/a> in Huntington’s research. Tominersen is an antisense oligonucleotide (ASO) \u2014 a short, chemically modified strand of synthetic DNA that binds to the huntingtin gene’s messenger RNA transcript and recruits a cellular enzyme called RNase H, which cleaves and destroys the RNA before it can be translated into protein. That mechanism is how the drug lowers huntingtin levels throughout the brain. Delivery requires a lumbar puncture into the cerebrospinal fluid, which then circulates the drug across the central nervous system. Tominersen was the first drug to demonstrate that reducing huntingtin in the human brain was achievable at all \u2014 a historic proof of concept established in its first-in-human clinical study<\/a> more than a decade ago.<\/p>\n Roche’s latest trial of tominersen, called GENERATION HD2, was a Phase 2 placebo-controlled study in people with early Huntington’s disease. After a dose design change in April 2025 that focused the trial on the higher 100 mg dose, the study was designed to assess whether tominersen could influence huntingtin and neurofilament light (NfL) \u2014 a protein released into the bloodstream and spinal fluid when neurons are damaged \u2014 as well as whether it could slow disease progression on two validated clinical rating scales: the composite Unified Huntington’s Disease Rating Scale (cUHDRS) and Total Functional Capacity (TFC). The GENERATION HD2 trial design and endpoints<\/a> were prespecified for a 16-month primary assessment.<\/p>\n The top-line results presented a maddening near-miss. Participants receiving tominersen showed significant reductions in huntingtin in their cerebrospinal fluid, alongside falling NfL levels in both spinal fluid and blood plasma \u2014 the biomarker profile of a drug doing exactly what it was designed to do. There were no new safety concerns. And yet, at the 16-month mark, participants on tominersen showed no meaningful difference from those on placebo on the clinical scales that measure how people actually function. The hoped-for improvements in cUHDRS and TFC did not materialize.<\/p>\n Roche concluded there was no path forward and announced a complete termination \u2014 meaning no open-label extension, no compassionate use, and no further development of tominersen for Huntington’s disease.<\/p>\n The second program, RG6496, failed for an entirely different reason. Where tominersen is designed to lower both the mutant and wild-type copies of the huntingtin protein \u2014 a “total lowering” approach \u2014 RG6496 was engineered to target only the mutant copy by exploiting a specific single nucleotide polymorphism, a single DNA letter change, that some patients carry in their expanded gene. The goal was selective lowering: reducing the toxic protein while leaving the normal, protective form intact. The POINT-HD Phase 1 study had given a single dose to just three participants, none of whom experienced adverse effects.<\/p>\n But Roche had been running concurrent animal studies of long-term, repeated RG6496 dosing in parallel with the human study \u2014 a standard industry practice intended to continue generating safety information while getting treatments to patients faster. Those animal studies identified findings the company concluded would prevent the drug from being safely administered with the long-term, chronic dosing that a Huntington’s treatment would require. Roche has not yet disclosed the specific nature of those findings, and participants already enrolled will continue to receive follow-up monitoring.<\/p>\n “These are independent, data-driven events, which have coincided by chance,” Mai-Lise Nguyen, patient partnership leader at Roche, wrote in Roche’s July 9 community letter<\/a>.<\/p>\n What the Biomarker Paradox Actually Means for the Huntingtin-Lowering Hypothesis<\/p>\n Scientists are urging the HD community to resist drawing the broadest possible conclusion from the GENERATION HD2 result. The failure of one drug in one dosing regimen over 16 months does not settle the question of whether lowering huntingtin can slow the disease \u2014 and the HDBuzz scientific team, whose articles are peer-edited by independent HD researchers, made that case explicitly in HDBuzz’s July 9 scientific analysis<\/a>.<\/p>\n The more significant finding, researchers say, may be the biomarker-clinical dissociation itself \u2014 the gap between what the drug did in measurable biology and what patients experienced in the exam room. That gap has a name in neurodegenerative disease research, and it recurs across Alzheimer’s, Parkinson’s, and ALS drug development: the phenomenon in which a pharmacodynamic marker responds as expected while functional outcomes do not, at least within the trial’s observation window. In Huntington’s disease, which progresses over decades and whose clinical symptoms can change imperceptibly from one year to the next, a 16-month trial may simply be an instrument too coarse to detect a meaningful signal \u2014 even if one is present.<\/p>\n That interpretation does not rescue tominersen. But it does change what the result means for every other huntingtin-lowering program that comes after it.<\/p>\n 16 Months May Not Be Long Enough to Test a Disease That Takes Decades<\/p>\n This is the largest structural implication of the GENERATION HD2 failure, and the one most often understated in clinical-trial focused coverage: the 16-month window may be insufficient to measure meaningful change in a disease whose clinical progression is measured in years and whose root cause begins silently, in many patients, a decade before symptoms appear.<\/p>\n NfL \u2014 the neuronal damage marker that tominersen successfully reduced \u2014 fell in both tominersen-treated patients’ spinal fluid and blood. If NfL reduction corresponds to reduced neuronal death, as most researchers believe, then the drug may have been slowing damage that would not manifest as measurable functional decline within a 16-month observation. The alternative interpretation \u2014 that NfL reduction does not adequately predict clinical benefit \u2014 would itself be a major scientific finding, because NfL has become one of the field’s most widely used biomarkers<\/a> across multiple neurodegenerative diseases.<\/p>\n Several additional questions remain unresolved: Was the drug reaching the striatum \u2014 the region of the brain most severely affected in early HD \u2014 in sufficient concentration? Intrathecal delivery distributes the drug broadly through the cerebrospinal fluid, but the striatum is less accessible via this route than cortical regions, and the adequacy of striatal huntingtin lowering was never confirmed with regional imaging. Was treatment beginning too late, at a stage when neurodegeneration has already progressed beyond the point where lowering huntingtin can reverse functional loss? Could a different delivery method achieve deeper penetration of the critical circuits?<\/p>\n Some participants were treated for considerably longer than 16 months. Their full outcomes have not yet been analyzed, and Roche has committed to presenting complete datasets at future scientific meetings.<\/p>\n How Tominersen’s Mechanism Differed From the Gene Therapy Now Closest to Approval<\/p>\n The most clinically advanced program in Huntington’s disease research right now is AMT-130 from uniQure, and it is worth understanding precisely why it works differently \u2014 because the distinction may explain the divergent outcomes. The full mechanism is detailed on uniQure’s AMT-130 program page<\/a>.<\/p>\n AMT-130 is an adeno-associated virus serotype 5 (AAV5) vector carrying a microRNA designed to silence the huntingtin gene. It is delivered not by lumbar puncture but by direct stereotactic neurosurgical infusion into the caudate nucleus and putamen \u2014 the striatal structures most devastated by HD. That one-time surgical delivery reaches the affected brain regions directly, rather than relying on CSF circulation to distribute across the brain. The silencing mechanism is also distinct: where tominersen uses an RNase H enzyme to cleave the huntingtin mRNA once the ASO is bound, AMT-130’s microRNA engages the cell’s RNA-induced silencing complex (RISC), which degrades huntingtin transcripts through a different molecular pathway.<\/p>\n The clinical outcomes have been correspondingly different. At the three-year mark in Phase 1\/2 trials, the high-dose AMT-130 cohort showed a 75% slowing of disease progression compared to propensity score-matched external controls on the same cUHDRS scale used to measure tominersen’s failure \u2014 results confirmed in uniQure’s September 2025 data announcement<\/a>. The FDA confirmed in uniQure’s June 2026 FDA announcement<\/a> that this three-year data package is acceptable as the primary basis for a Biologics License Application seeking accelerated approval. uniQure plans to submit that application in the third quarter of 2026.<\/p>\n That comparison is not a controlled head-to-head. AMT-130 used external historical controls rather than a concurrent placebo group, and that methodology carries its own limitations. But the directional contrast \u2014 between a drug given once directly to the affected brain region over three years showing benefit, and a drug delivered quarterly to the spinal fluid over 16 months showing no clinical benefit \u2014 suggests that delivery route, treatment duration, and target-region access are not peripheral variables. They may be the central ones.<\/p>\n The HD Pipeline Has Not Gone Quiet<\/p>\n Roche’s exit from two programs is significant, but it does not represent the field’s last word on disease modification in Huntington’s. Multiple programs at different stages of development continue, using different mechanisms and delivery strategies.<\/p>\n Skyhawk Therapeutics is expanding its Phase 2\/3 FALCON-HD trial of SKY-0515 to US sites in July 2026. SKY-0515 is an oral small-molecule mRNA splicing modulator: rather than blocking or degrading the huntingtin transcript, it alters how the transcript is processed at the pre-mRNA level, reducing the production of toxic huntingtin protein without requiring intrathecal delivery. The FALCON-HD worldwide trial<\/a> had completed enrollment of its Australia-New Zealand arm and is now expanding globally.<\/p>\n Novartis initiated the Phase 3 INVEST-HD study of votoplam in March 2026, with the first participant dosed in late April 2026. Votoplam is an oral small molecule that targets splicing of the huntingtin pre-mRNA through a different mechanism than SKY-0515, also eliminating the need for lumbar punctures. The Phase 3 INVEST-HD trial<\/a> is enrolling up to 770 participants worldwide.<\/p>\n Roche itself continues its gene therapy program for HD. RG6662, formerly developed by Spark Therapeutics, remains active and unaffected by the ASO program terminations.<\/p>\n More than 1,500 HD families contributed to Roche’s programs since the first tominersen studies began over a decade ago. The biomarker validations, the CSF measurement protocols, the NfL assays, the refinements to the cUHDRS scale as a clinical endpoint \u2014 these will outlast the drugs that generated them and will shape how the next generation of trials is designed. Every patient who participated in a tominersen study helped establish whether the huntingtin protein can be measured and lowered in the human brain. The answer, definitively, is yes. The harder question \u2014 whether doing so can slow or halt the disease \u2014 remains open, and the field now has a more precise map of what a successful attempt will need to look like.<\/p>\n Frequently Asked QuestionsWhat is Huntington’s disease and why does it have no approved cure?<\/p>\n Huntington’s disease is an inherited neurological disorder caused by a mutation in the HTT gene \u2014 a CAG trinucleotide repeat that, when expanded beyond a threshold, produces a toxic form of the huntingtin protein that steadily destroys neurons, particularly in the striatum, over years and decades. Inheritance is autosomal dominant, meaning a child of an affected parent has a 50% chance of inheriting the mutation. Some drugs can manage specific symptoms \u2014 involuntary movements, psychiatric disturbances \u2014 but none has been approved to slow or halt the underlying neurodegeneration. That gap is what the clinical trial programs described in this article were attempting to close. The full genetic basis of the disease is explained on MedlinePlus’s Huntington’s disease genetics page<\/a>.<\/p>\n Does the failure of tominersen mean that lowering huntingtin cannot work as a treatment strategy?<\/p>\n No \u2014 and most HD researchers are emphasizing that distinction. Tominersen did lower huntingtin measurably in the brain, and it reduced neurofilament light, a marker of neuronal damage. What it did not do, within 16 months of treatment, was change how patients functioned on composite clinical rating scales. That result may reflect the drug’s limitations, but it may equally reflect limitations in the trial: 16 months is a short window in a disease that progresses over decades, delivery to the most affected brain regions may have been incomplete, and treatment may have begun too late in the course of the disease. Other huntingtin-lowering programs \u2014 using different delivery mechanisms, including gene therapy and oral small molecules \u2014 remain in development. HDBuzz’s analysis of the huntingtin-lowering hypothesis<\/a> addresses this distinction directly.<\/p>\n Why is the 16-month window a problem for testing Huntington’s disease drugs?<\/p>\n Huntington’s disease progresses slowly and continuously over many years. The composite clinical scales used to measure its progression \u2014 cUHDRS and TFC \u2014 change gradually enough that detecting a meaningful difference between treated and placebo patients requires either a very large trial, a very long treatment period, or both. The GENERATION HD2 trial’s 16-month endpoint design<\/a> was the prespecified primary assessment \u2014 but even if tominersen was slowing neurodegeneration, that slowing might not be detectable as a change in cUHDRS or TFC within 16 months. Several participants in GENERATION HD2 were treated for considerably longer; their outcomes have not yet been analyzed, and those data could provide important additional context.<\/p>\n What Huntington’s disease therapies still have a realistic chance of reaching patients?<\/p>\n The most advanced is uniQure’s AMT-130, a one-time gene therapy delivered surgically to the striatum that showed 75% slowing of disease progression at three years in Phase 1\/2 trials. The FDA agreed in June 2026 that the existing clinical data are sufficient as the basis for a Biologics License Application; uniQure plans to file in uniQure’s Q3 2026 BLA filing announcement<\/a>. Two oral small-molecule programs are also in active Phase 2\/3 or Phase 3 testing: Skyhawk’s SKY-0515 (FALCON-HD trial, expanding to the US in July 2026) and Novartis’s votoplam\/PTC518 (INVEST-HD Phase 3, initiated March 2026). Roche’s own gene therapy candidate, RG6662, continues in its clinical program unchanged by the ASO terminations. More detail on the current state of the Huntington’s disease pipeline<\/a> is available from uniQure’s patient resources.<\/p>\n","protected":false},"excerpt":{"rendered":"In a letter sent to global Huntington’s disease patient community leaders on July 9, 2026, Swiss pharmaceutical giant…\n","protected":false},"author":2,"featured_media":101359,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","_share_on_mastodon":"0"},"categories":[124],"tags":[212,50077,30409,50078,206,134,50075,50076],"class_list":["post-101358","post","type-post","status-publish","format-standard","has-post-thumbnail","category-roche","tag-gene-therapy","tag-huntington-disease-drug-trial-2026","tag-huntingtons-disease","tag-neurofilament","tag-novartis","tag-roche","tag-tominersen-generation-hd2","tag-uniqure-amt-130"],"share_on_mastodon":{"url":"https:\/\/pubeurope.com\/@ch\/116897796983882560","error":""},"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/posts\/101358","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/comments?post=101358"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/posts\/101358\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/media\/101359"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/media?parent=101358"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/categories?post=101358"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/tags?post=101358"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}