Evidence status for regulators

What stage is the clinical evidence at?

Two interventional records are on ClinicalTrials.gov, both early phase, both with safety as a primary outcome. Dates and figures below are as the records read on 31 October 2026, and both carry sponsor-stated estimates that have moved before.

PrProfile (NCT06153966) is a Phase 1/2a randomised, double-blind, parallel study of intrathecal ION717 run by Ionis Pharmaceuticals. It started on 4 January 2024, reports actual enrolment of 85, and is listed as active and not recruiting, last updated 15 September 2026. The single primary outcome is the incidence of treatment-emergent adverse events from baseline to week 33. Primary completion is estimated for April 2027 and study completion for August 2030. Eligibility, as registered, requires a confirmed diagnosis of probable or definite prion disease, early-stage disease at screening, age 18 or over, and a caregiver willing to take part for the duration.

The ALN-6457 trial (NCT07832370), sponsored by Regeneron Pharmaceuticals, is a Phase 1/2 open-label dose escalation with a cohort expansion part and extended follow-up. It was first submitted on 16 September 2026, is listed as not yet recruiting, and gives an estimated start of 23 October 2026 with primary and study completion on 27 February 2031. Estimated enrolment is 60. Three primary outcomes are registered: occurrence of treatment-emergent adverse events and their severity through week 156, and percentage change from baseline in plasma neurofilament light chain through week 130. Registered eligibility is sporadic CJD or genetic prion disease as described in the protocol, with a body mass index under 35.

A sponsor’s registered dates and enrolment are not an audited record. Figures quoted in press material and in secondary summaries drift from the registry within months, so a decision should rest on the record itself, read on a stated date.

What has a controlled comparison shown?

The PRION-1 study remains the largest completed therapeutic trial in human prion disease and the clearest lesson about design. Patients recruited through the UK national referral system were offered quinacrine at 300 mg daily, no quinacrine, or randomisation between immediate and deferred treatment. Of 107 patients enrolled, 45 had sporadic disease, 42 inherited, 18 variant and 2 iatrogenic. Only two accepted randomisation. Choice of treatment tracked disease severity, with the least and most severely affected more likely to decline the drug. Adjusted mortality showed no difference between the groups, and the apparent advantage in those who chose the drug was attributed to that confounding. Four of 40 patients who took quinacrine had a transient response on neurological rating scales.

A patient-preference design in a disease this rapid produces an uninterpretable survival comparison. Regulators and funders assessing new proposals can reasonably ask how severity at entry will be measured and balanced.

How solid is the preclinical base?

Several of the most quoted preclinical results in PrP-lowering and gene therapy are visible only in funder-hosted grantee reports, conference slide decks and preprints. These state percentage reductions in prion protein and survival extensions in mouse models without peer review, methods sections or reported variance. They are not fraudulent and they are not evidence of the kind a dossier can rest on. A reasonable test before relying on such a figure is whether the same result appears in a peer-reviewed paper with the animal numbers, the strain and the dosing stated.

What is unknown?

• Whether a change in plasma or cerebrospinal fluid biomarkers will be accepted as evidence of benefit in this disease.
• How many eligible patients can be identified early enough to enrol, given diagnosis usually comes late. Surveillance capacity bears directly on this; see surveillance systems and the surveillance centres.
• How patient-level data from small international cohorts can be shared lawfully; see data protection.

Return to policy or the regulations atlas.