Roche Just Walked Away From Huntington's. Now What?
Roche killed both of its Huntington's disease gene-silencing drugs after a decade of work, leaving patients without any approved disease-modifying treatment. The field isn't dead, but the biggest pharma player just left the building.
Imagine spending over a decade building a house only to discover the foundation can't hold it. That's roughly what happened to Roche's Huntington's disease program.
On July 9, the Swiss pharma giant announced it was permanently shutting down both of its Huntington's gene-silencing drugs: tominersen and RG6496. No further trials. No compassionate use. No open-label extensions. The lights are off and the doors are locked.
For the roughly 41,000 symptomatic Americans (and over 200,000 more who carry the gene), this is devastating. Huntington's disease has no approved treatment that slows or stops its progression. Every drug on the market just manages symptoms. Roche was supposed to change that.
The Drug That "Worked" But Didn't
Tominersen was an antisense oligonucleotide, or ASO: a short strand of synthetic genetic material designed to silence the huntingtin gene. Think of it like a mute button for the protein that slowly destroys the brains of Huntington's patients.
The concept was elegant. Early data from a small Phase 1/2a study showed the drug cut levels of the toxic protein by about 38% in spinal fluid. Roche licensed it from Ionis Pharmaceuticals in 2017 and charged into a massive Phase 3 trial called GENERATION HD1, enrolling roughly 791 patients across 18 countries.
Then things went sideways. In March 2021, an independent safety board pulled the plug on dosing. Patients who received the drug every two months were actually getting worse faster than those on placebo. Not just failing to improve; actively declining. The higher the exposure, the worse the outcome.
Roche didn't give up entirely. Post-hoc analyses (the biotech equivalent of reviewing game tape after a loss) hinted that younger patients with milder disease might still benefit. So the company designed a more targeted Phase 2 trial, GENERATION HD2, testing lower doses in earlier-stage patients.
The result, announced this month: no meaningful clinical benefit at 16 months. The drug still lowered the toxic protein. The safety profile was acceptable. But patients didn't get better. Roche called it what it was: "no path forward."
Get tomorrow's biotech intelligence before your competitors.
Join thousands of biotech professionals who start their day with our free, daily briefing.
The Backup Plan That Never Had a Chance
RG6496 was supposed to be the smarter sibling. Where tominersen silenced all huntingtin protein (both the mutant and the healthy version your brain actually needs), RG6496 was designed to target only the mutant copy. It was an allele-selective ASO, precision-guided to hit the bad stuff and leave the good stuff alone.
It barely got off the ground. The Phase 1 trial, called POINT-HD, had enrolled just three participants when Roche halted it. New animal studies showed the drug simply couldn't be given repeatedly over the long term. For a disease that lasts decades, a drug you can't take chronically is a non-starter.
Roche stressed that the two failures were "independent, data-driven events" that just happened to land at the same time. Cold comfort for patients and families who watched their best hope evaporate twice in a single announcement.
A Pattern That Should Worry Everyone
Roche's Huntington's collapse isn't an isolated incident. It fits a grim pattern across CNS gene-silencing programs in recent years.
In ALS, Biogen's BIIB078 targeting the C9orf72 gene showed target engagement by reducing CSF dipeptide repeat proteins, but key biomarkers like neurofilament light chain (NfL) actually worsened at higher doses, tissue-level pathology was not meaningfully reduced, and the drug missed on clinical outcomes. Wave Life Sciences' WVE-004 similarly failed to translate into clinical benefit. Even tofersen, the one ASO that made it to FDA approval for a rare form of ALS, technically failed its primary clinical endpoint in Phase 3; it squeaked through on biomarker evidence and desperate unmet need.
The recurring theme is uncomfortable: these drugs do what they're designed to do biologically, yet the patients don't get better. It's like acing every practice exam and still bombing the final. Something fundamental is getting lost in translation.
Experts point to several culprits. Getting ASOs deep into brain tissue through repeated spinal injections is imprecise and invasive. Higher doses improve delivery but also trigger inflammation, including serious events like myelitis and meningitis in some trials. And intervening after neurodegeneration has already taken hold may simply be too late, like trying to un-burn a building.
Who's Still in the Fight?
Roche's exit leaves Huntington's without its biggest pharma backer, but it doesn't leave the field empty. A handful of companies are pressing forward with different approaches.
uniQure has the most advanced alternative: AMT-130, a one-time gene therapy delivered directly to the brain via an engineered virus carrying a micro-RNA that silences huntingtin. Unlike ASOs, it doesn't require chronic spinal injections. In June 2026, the FDA signaled that uniQure's three-year Phase 1/2 data could serve as the primary basis for accelerated approval. The company has already stacked up Fast Track, Orphan Drug, Breakthrough Therapy, and RMAT designations.
Alnylam is running a Phase 1 trial of ALN-HTT02, an siRNA (a different flavor of gene silencing) that started dosing patients in December 2024. Wave Life Sciences continues developing WVE-003, an allele-specific ASO. Skyhawk Therapeutics is pushing SKY-0515, a small-molecule RNA-splicing modulator, toward Phase 2/3. And Vico Therapeutics began dosing an expanded cohort in its Phase 1/2a trial of VO659 in February 2026.
Earlier-stage programs from Atalanta Therapeutics, Incisive Genetics (CRISPR-based editing), and Passage Bio round out a pipeline that's scrappy but fragmented. None of these companies have Roche's resources.
The Billion-Dollar Question Nobody Can Answer Yet
The fundamental puzzle remains: if lowering the toxic protein doesn't help patients, what will?
Researchers and advocates are careful not to declare huntingtin-lowering dead as a strategy. One drug, one regimen, one 16-month window can't settle a mechanistic debate that's been raging for decades. Maybe the answer is treating patients before symptoms appear, which raises enormous ethical and regulatory questions. Maybe it's a different delivery method, like uniQure's one-shot gene therapy instead of chronic spinal injections. Maybe it's targeting different pieces of the genetic machinery entirely.
What's clear is that the bar has been raised. Regulators and investors will now demand proof that biomarker changes actually translate into clinical benefit before getting excited about the next gene-silencing candidate. The era of "we lowered the protein, trust us" is over.
For the Huntington's community, the wait continues. There are still no approved treatments to slow this disease. The average patient lives 15 to 20 years after symptoms begin, watching their motor control, cognition, and personality erode in sequence. They've been waiting a long time for a breakthrough.
Roche just told them it won't be coming from Basel.
Clinical & Regulatory5 min read
Wegovy Just Became a Liver Drug.
India just approved Wegovy for fatty liver disease, making it the second country to clear a GLP-1 drug for MASH. With nearly 40% of Indian adults affected by fatty liver and zero prior approved treatments, Novo Nordisk's latest label expansion could reshape both the Indian pharma market and the global MASH treatment landscape.
