Buntanetap Phase 3

The research record

This is the project's file for this landmark, as written and checked. Sources are listed at the end; anything not yet confirmed is marked "Not yet verified".

Who

Annovis Bio, a small US company (Malvern, Pennsylvania). Results were announced by press release and posted on the registry in March 2025; I found no peer-reviewed paper of the efficacy result.

What happened

Buntanetap (earlier called Posiphen, ANVS401) is described by the company as a once-daily pill that blocks production of several neurotoxic proteins (amyloid precursor protein, tau, alpha-synuclein, TDP-43) by acting on the RNA that is translated into them. Its Phase 3 in early Parkinson's (NCT05357989; randomised, quadruple-masked; started August 2022, completed December 2023) enrolled 523 people (176 placebo, 174 on 10 mg, 173 on 20 mg daily) for six months. Annovis first named the combined MDS-UPDRS parts II and III as the primary measure, then moved to part II (daily-living symptoms) alone after FDA feedback, with part III (motor exam) secondary.

The registry's results give the whole-population primary result: the change from baseline in part II (lower is better) was −0.10 (standard error 0.31) on placebo, +0.57 (0.31) on 10 mg and −0.04 (0.32) on 20 mg. That is no benefit; the registry posts no P values. Part III fell by 2.90 on placebo, 1.04 on 10 mg and 2.66 on 20 mg, so the drug groups improved no more than placebo. In July 2024 the company reported that 20 mg improved scores in people diagnosed more than three years earlier and in a group with balance and walking problems, and better cognition (MMSE) in all treated groups. These were subgroup and exploratory analyses; the company's own release says people diagnosed less than three years earlier had too little part II deficit to measure change.

In January 2026 Annovis began a separate 36-month, single-arm, open-label safety study (NCT07284784, Phase 2/3, 30 mg a day, about 500 planned, recruiting; primary completion estimated September 2029). It is not an efficacy trial.

Why it moved us toward a cure

If the mechanism holds, it offers a pill that lowers alpha-synuclein by reducing its production, compared with the injected approaches in landmark 29. The six-month trial is too short to say anything about slowing, and the main result does not support even a symptom effect.

Outcome

Dead end for this trial. The primary measure showed no overall effect, and subgroup findings generate hypotheses, not evidence. As with LUMA (landmark 145), the trial is marked a dead end even though the sponsor has not dropped the drug; no Parkinson's efficacy trial is running. As of the 14 August 2026 update the open-label extension had enrolled 266 of about 500, and the main effort is an 850-person Alzheimer's Phase 3 with topline expected in the first quarter of 2027.

Sources

• ClinicalTrials.gov NCT05357989 (API, 19 and 22 Oct 2026, Phase 3, 523 enrolled; results first posted 3 Mar 2025) · NCT07284784
• Annovis release, 2 Jul 2024
• Clinical Trials Arena · NeurologyLive · Fierce Biotech, 2 Jul 2024 (blocked to automated reading; cited by title)
• Annovis Q2 2026 update, 14 Aug 2026
• Not yet verified: P values and subgroup sizes (the earlier "158" figure was removed as unsourced); independent evidence for the RNA mechanism (company-described); any peer-reviewed efficacy paper. NeurologyLive gives 450 people and 3 years, which conflicts with the registry's 523 and six months; I followed the registry.