The Generative Brain

Therapeutics landscape

No treatment for prion disease has human safety and efficacy data. Two PrP-lowering approaches, an antisense oligonucleotide and a divalent siRNA, are in early human trials, and gene editing and small molecules remain preclinical.

Where treatments in development try to step in
  1. The PRNP geneSwitch off or edit the gene that makes the protein.
    • Base editing
    • Epigenetic editing
    • Zinc finger repressors
  2. The RNA messageLower how much protein the cell makes.
    • ION717
    • PrP siRNA
  3. The prion proteinBind the protein or change how cells handle it.
    • PRN100 antibody
    • Small molecules
  4. The brain's responseSupport the cells reacting to the damage.
    • Early ideas

A simplified view: none of these is a proven treatment. ALN-6457 also targets PRNP, but the trial registry does not state its drug type, so it is not placed here.

Where approaches stand
  1. Early research
  2. First human studies
  3. Clinical trials
  4. Approved
Base editing
Epigenetic editing
Zinc finger repressors
Small molecules
PRN100 antibody
ION717 (PrProfile)
PrP siRNA (PRiSM)
ALN-6457

A simplified view. No treatment is proven yet. See the Atlas for each record and its date.

The common strategy in current programs is to lower the amount of cellular prion protein (PrP), because prion replication needs PrP as its substrate. Most of what follows concerns that strategy, with the limits of the evidence stated for each item. Trial eligibility is as stated by the named source on the date given and can change. Current registry entries are tracked on current trials and program entries on therapies in the atlas.

What human trials of PrP-lowering drugs are running?

Two trials were described by the Prion Alliance in its 16 July 2026 update as the PrP-lowering trials under way. Its author states that neither has human safety or efficacy data yet.

PRiSM (NCT07444580) tests a divalent siRNA. The Prion Alliance update says it was launched in April 2026, is recruiting, and at that date enrolled only symptomatic patients with diagnosed disease. It names NIH funding through NeuroNEXT and says the sponsor committed to sharing individual patient data. The CJD Foundation drug development page, which carries a statement dated 14 October 2026, describes a pre-symptomatic arm as now open. The two sources therefore differ by date on eligibility, and the sponsor’s registry entry is the authority.

PrProfile (NCT06153966) is a phase 1/2a trial of ION717, an antisense oligonucleotide against PRNP mRNA given intrathecally, run by Ionis. According to a March 2026 commentary on the cureffi.org blog, 56 symptomatic patients were enrolled in 2024 into two randomised cohorts, the last study visit of that group was in July 2025, and Ionis reopened recruitment on 17 March 2026 with a third, non-randomised regimen. Ionis has not disclosed what differs between regimens. The commentary’s view that the first two regimens lowered PrP less than hoped is the author’s interpretation and is not confirmed by Ionis. The Prion Alliance update reports that Ionis stated there would be no public results announcement until 2027.

The CJD Foundation page also lists a Regeneron study of ALN-6457 (NCT07832370), described there as a divalent siRNA that lowers PRNP expression, in adults with symptomatic prion disease. Phase and status are not given on that page. A drug aggregator that labels ALN-6457 a small molecule conflicts with that description and was not used.

What is preclinical?

In the Nature Medicine paper by An, Davis and colleagues (14 January 2025), a pair of AAVs delivered a base editor that installs the R37X change in PRNP in a humanised mouse model infected with prions. The Broad Institute reports an average of 37 percent of gene copies edited, about 50 percent lower PrP, and about 50 percent longer lifespan. It reports an improved system with 63 percent lower PrP at a sixfold lower AAV dose. Funders named are the NIH, Prion Alliance and HHMI. The authors state that this is preclinical and that dual AAVs are costly to produce. The approach was not tested in genetic or spontaneous disease models in this study.

The CJD Foundation page lists further preclinical work: an oral small molecule from Gate Bioscience (mouse survival testing), engineered zinc finger transcription factors from Sangamo (mouse studies), and a Nobelpharma compound with clinical studies stated to be expected “probably next year” in remarks made in July 2025. It lists PRN100, an anti-PrP monoclonal antibody given under a Specials Licence, as not currently enrolling. I could not open the primary UCL sources on PRN100, so no outcome data are given here.

What did earlier clinical trials show?

The cureffi.org review of completed trials covers flupirtine (28 patients), pentosan polysulfate (two small non-randomised series of 7 and 11), quinacrine (a French series of 32, the UK PRION-1 patient-preference study of 107, and a US randomised trial of 54) and doxycycline (a randomised French and Italian phase 2 trial of 121). The review describes the quinacrine and doxycycline randomised trials as negative. It attributes apparent survival gains in non-randomised designs to confounding, since survival varies with genotype, age and subtype.

The lessons it draws are stricter preclinical standards, including human PrP mouse models, earlier diagnosis with RT-QuIC, trials in asymptomatic PRNP carriers, and stratification by a disease rating scale. It also notes that placebo-controlled multinational trials proved feasible. For the genetic and diagnostic context see research overview and research resources.

What is uncertain?

No PrP-lowering drug has reported human efficacy data. The cureffi.org review notes that some earlier compounds worked in animals only early in disease, while trials enrolled symptomatic patients, so timing of treatment is a design question for every program. Biomarkers that could serve as trial endpoints are still being validated. Funding and company events listed in programs change often and are not evidence of efficacy.

Sources for this page