drug · intermediate · complete
Tominersen
Tominersen is an antisense oligonucleotide given into the spinal fluid to lower huntingtin, the protein behind Huntington disease; it lowered the protein but patients did not improve.
Also called: RG6042, GENERATION HD1 drug
This describes published medicine for readers who want to understand how it works and what happened. It is not medical advice and says nothing about what any person should take.
In brief
Tominersen tried to lower huntingtin, the protein behind Huntington disease. It was injected into the spinal fluid. It did lower the protein, yet the people dosed every 2 months did worse than placebo and dosing was stopped. The reason is not established. 12
Reported, not yet replicated
What happened in GENERATION HD1 more
A phase 3 trial first dosed 108 people monthly or every 2 months; after a change of schedule, 791 adults were randomised to 120 mg every 2 months, every 4 months or placebo (the sponsor summary also gives arm sizes of 260, 261 and 260, which add to 781; the gap is unexplained). In March 2021 the sponsor stopped dosing on the advice of an independent data monitoring committee.
At week 69 the every-2-months group had worse composite clinical scores and function than placebo; the every-4-months group was similar to placebo. Mutant huntingtin in the spinal fluid fell by 47 and 27 percent against +0.4 percent. This is a description of a trial result, not advice. 12
Reported, not yet replicated
What moved and what did not more
The target engagement marker moved (mutant huntingtin down), the neuron-injury marker NfL went the wrong way in the more frequent arm (+10 percent against +6 percent on placebo), and ventricles grew more (+19 percent against +9 percent). Serious side effects were 48 of 260, 25 of 261 and 34 of 260. Only the clinical endpoint counts as benefit, and it did not improve. 1
Reported, not yet replicated
Why it failed: a hypothesis, not a finding more
The sources do not establish the cause. The patient-community explainer lists inflammation from the high dose and a disease stage too advanced as candidates, labelled as hypotheses. A sponsor-affiliated modelling study suggests lower exposure may give a better benefit-risk balance. A review by Huntington researchers says only that the early termination invites reflection on lessons learned. An earlier 46-person study had shown dose-dependent lowering and no serious adverse events, so the failure was not predicted from it. 3456
Contested
What came later deep
A recent journal abstract reports that the sponsor discontinued the phase 2 GENERATION HD2 programme. No clinical figures were read from it. A peer-reviewed letter on HD1 exists, but only its record was read; the numbers on this page come from the sponsor-written plain-language summary and should be checked against the peer-reviewed paper before reuse. 781
Reported, not yet replicated
The lesson for reasoning about drugs deep
Two edges are easy to confuse: the drug lowers mutant huntingtin, and lowering mutant huntingtin slows Huntington disease. Only the first was measured here. This is the same shape as a good docking score that fails to predict assay activity. Which of the explanations above applies cannot be said from the sources opened. 14
Our inference
Modality
Targets
Pathways
Indications
Identifiers
Roche study code phase3
BN40423
Phase3 trial name
GENERATION HD1
Mechanism
Huntington disease is caused by an expanded CAG repeat in the huntingtin gene, which makes a toxic mutant huntingtin protein (mHTT). Tominersen was meant to lower the output of huntingtin RNA and so of mutant huntingtin, delivered into the spinal fluid. That it does this by recruiting RNase H to cut the RNA (a gapmer) is standard background and was not confirmed from a source opened in this run. The sponsor summary describes the aim as lowering HTT and mHTT output. A later sponsor drug is described as allele-selective, largely preserving the normal protein, which implies tominersen was not. 1
Reported, not yet replicated
Delivery route
Must survive
The same obstacles as other antisense drugs: nuclease attack and the blood-brain barrier, handled by intrathecal injection. Whether injection into the spinal fluid reaches deep brain regions well enough in Huntington disease is not established in the sources opened. 1
Unverified: from our own report, source not opened
Assays used
CSF mutant huntingtin
Measured as a marker that the drug engaged its target. In HD1 at week 69 it changed by -47 percent (every 2 months), -27 percent (every 4 months) and +0.4 percent (placebo). 1
Reported, not yet replicated
CSF neurofilament light (NfL)
A marker of damaged neurons. In HD1 at week 69: +10 percent (every 2 months), -3 percent (every 4 months), +6 percent (placebo). 1
Reported, not yet replicated
Brain MRI ventricular volume
+19 percent, +13 percent and +9 percent for the same three groups. The sponsor summary says the cause and meaning of the increase are not clear. 1
Reported, not yet replicated
Composite clinical score (cUHDRS) and Total Functional Capacity
The primary clinical measures: a composite of motor, cognitive and functional scores, and a function scale. 1
Reported, not yet replicated
Off targets
Side effects
Serious adverse events in GENERATION HD1
Serious side effects were reported in 48 of 260 people (about 19 percent) on tominersen every 2 months, 25 of 261 (about 10 percent) every 4 months and 34 of 260 (about 13 percent) on placebo. Roche reported no new safety signals when it stopped dosing in March 2021. 12
Reported, not yet replicated
Rises in neurofilament light and ventricle volume
NfL rose by 10 percent and ventricular volume by 19 percent in the every-2-months arm, compared with 6 and 9 percent on placebo. Their meaning is not clear according to the sponsor summary. 1
Reported, not yet replicated
Early phase 1/2a safety
In the earlier 46-person phase 1/2a study, no serious adverse events were reported in treated participants (abstract). 6
Reported, not yet replicated
Verdict
Outcome
failed
Why
The drug did lower its target and the patients did not benefit. In GENERATION HD1 the every-2-months arm had a worse composite clinical score and Total Functional Capacity than placebo at week 69, and the every-4-months arm was similar to placebo, while CSF mutant huntingtin fell by 47 and 27 percent. Dosing was stopped in March 2021 on the advice of an independent data monitoring committee. The lesson from the sources is that lowering the molecular target is not the same as treating the disease; a biomarker change is not a clinical result. 12
Reported, not yet replicated
Failure mode
Phase 3 GENERATION HD1: dosing stopped in March 2021 on the recommendation of an independent data monitoring committee after an overall benefit-risk assessment. At week 69 the more frequently dosed arm did worse than placebo on clinical measures and showed more neuron-injury and ventricle-enlargement signals. The cause is a hypothesis, not a finding. Candidate explanations in the sources (inflammation at the high dose, a disease stage too advanced to benefit, too much exposure, possibly lowering normal huntingtin too) were not proven; a post hoc look at younger people on the every-4-months dose looked slightly better, but the sponsor itself says this may be chance. A sponsor exposure-response model found the highest exposure quartile had a 54 percent CSF mutant huntingtin reduction with transient rises in neuroinjury and inflammation markers, the lowest quartile 24 percent with a favourable biomarker profile, which suggests lower exposure may give a better benefit-risk. This is a modelling suggestion, from Roche-affiliated authors, read as an abstract. Later, the phase 2 GENERATION HD2 programme was reported discontinued (abstract only). 123475
Contested
Teach this
In one sentence
Tominersen lowered the target protein in Huntington disease and the patients did not get better, which shows a biomarker win is not a clinical win. 1
Reported, not yet replicated
The five-minute version
Huntington disease is driven by a toxic mutant huntingtin protein. Tominersen, an antisense drug injected into the spinal fluid, was meant to lower huntingtin RNA. In the phase 3 trial the spinal-fluid level of mutant huntingtin fell by 47 and 27 percent in the two dose arms, but the more frequently dosed arm did worse than placebo on clinical scores and dosing was stopped in March 2021. The cause has not been established; inflammation at high exposure, too advanced a disease and lowering normal huntingtin are hypotheses. The contrast with nusinersen is that there the target change and the disease were tightly linked. 123
Reported, not yet replicated
What a listener will ask
Did it fail because it did not hit the target?
Do we know why it failed?
Is the lower-frequency arm a success?
The common wrong idea
If an antisense drug lowers the toxic protein, the disease must improve.
Lowering the protein was measured and worked here; whether that helps the patient is a separate result, and in this trial it did not.
The one citation to hand over: 1
Still open here
- Why the more frequent dose arm did worse than placebo is not established. Candidate explanations (too much exposure, lowering the normal huntingtin protein too, inflammation from the backbone class, treating too late) are hypotheses; none is proven in the sources opened.
- The exact chemistry of tominersen was not opened. That it is a gapmer that cuts huntingtin RNA by RNase H is standard background but is not confirmed from a source here.
- Whether tominersen also lowers the normal copy of huntingtin is not stated in the sources opened.
- The sponsor-written summary is the only source for the week-69 numbers; the peer-reviewed letter was not read beyond its record. No peer-reviewed phase 3 primary paper was found.
- The report that the phase 2 GENERATION HD2 programme ended rests on an abstract only; no clinical figures from it are given.
Linked from
Sources
- F. Hoffmann-La Roche Ltd (sponsor). GENERATION HD1 (BN40423) final results: plain-language summary, September 2023. Roche forpatients (2023). organisation-page, not peer reviewed; opened 2026-11-17 all records citing this
- BioSpace. Ionis and Roche halt Huntington's disease trials. BioSpace (2021). news, not peer reviewed; opened 2026-11-17 all records citing this
- HDBuzz. HDBuzz explainer on the GENERATION HD1 results (article 354). HDBuzz (patient-community science news) (2023). news, not peer reviewed; opened 2026-11-17 all records citing this
- Yamamoto, et al.. [Descriptive title, not as printed] Exposure-response analysis of tominersen across four studies. British Journal of Clinical Pharmacology (2025). paper; opened 2026-11-17 all records citing this
- Tabrizi, S. J., et al.. [Descriptive title, not as printed] Lancet Neurology review of huntingtin-lowering after early termination of the phase 3 trial. Lancet Neurology (2022). review; opened 2026-11-17 all records citing this
- Tabrizi, S. J., et al.. Targeting huntingtin expression in patients with Huntington disease. New England Journal of Medicine (2019). paper; opened 2026-11-17 all records citing this
- Farag, et al.. [Descriptive title, not as printed] Report on the discontinued GENERATION HD2 programme. Journal of Huntington's Disease (2026). paper; opened 2026-11-17 all records citing this
- McColgan, P., et al.. [Descriptive title, not as printed] GENERATION HD1 journal letter on tominersen. New England Journal of Medicine (2023). paper; opened 2026-11-17 all records citing this
Where it sits on the map
History of this record (1)
This record has been changed 1 time, on the dates below. The note kept against each change is written for the people who maintain the atlas, in their own shorthand, so it is not shown here; it is in the record itself.
- Year
- 2021
- Regulatory status
- failed-in-trials
- Status
- complete
- Last updated
- 2026-11-17
- Record
- content/drugs/tominersen.json