A 78-week, placebo-controlled phase 3 trial of the oral drug valiltramiprosate (ALZ-801), restricted to people with two copies of the APOE4 gene, did not meet its primary goal of slowing cognitive decline across the full study group, according to peer-reviewed results from the APOLLOE4 trial published by Abushakra and colleagues in the journal Drugs. Among 325 participants, the placebo-adjusted difference on the ADAS-Cog13 cognitive scale favored the drug by a nonsignificant margin (p=0.607). In a smaller, prespecified subgroup of 125 participants with mild cognitive impairment, the same measure showed a nominal, statistically unadjusted 52 percent slowing of decline (p=0.041), with hippocampal volume loss also slowed more in this group than in the trial as a whole.

The drug, sponsored by Alzheon and registered as NCT04770220 on ClinicalTrials.gov, works differently from the infused antibodies lecanemab and donanemab that dominate current anti-amyloid treatment. Rather than clearing amyloid plaques already deposited in the brain, valiltramiprosate is designed as a small molecule taken by mouth that blocks the formation of soluble amyloid oligomers, an earlier step some researchers consider central to nerve cell damage in Alzheimer's disease. The trial enrolled only APOE4/4 homozygotes, who make up roughly 15 percent of Alzheimer's patients and carry the clearest genetic signal of amyloid buildup, but who also face the highest rates of brain swelling or microbleeds, called ARIA, when treated with antibody drugs. Published appropriate-use guidance for lecanemab reported ARIA-E in 32.6 percent of APOE4 homozygotes in that drug's pivotal trial, compared with 5.4 percent of people with no APOE4 copies.

In the APOLLOE4 trial, participants treated with valiltramiprosate showed 18 percent less hippocampal atrophy than those on placebo across the full study group (nominal p=0.0174), and no increased rate of ARIA was observed despite a cohort in which roughly 30 percent of participants already had microhemorrhages or similar MRI findings at the start of the study. An earlier readout of these results was presented at the AD/PD 2025 conference in April 2025, ahead of the peer-reviewed publication.

The subgroup finding in mild cognitive impairment has not been independently confirmed and was not the trial's primary analysis; Alzheon has said it intends to pursue further study focused on that earlier-stage population.