AR1001 Phase 3 results infographic from the POLARIS-AD Alzheimer's trial

AR1001 Phase 3 results from AriBio’s POLARIS-AD trial, presented at CTAD 2026 in Boston on 16 November, show the once-daily oral Alzheimer’s drug slowed clinical decline by 31% over 52 weeks in 604 patients who were not taking other dementia medicines, but failed to separate from placebo across the full 1,535-person study. The split result leaves the South Korean developer with a statistically positive subgroup, a missed overall population, and a conversation with the FDA still to come.

Key takeaways on AR1001 Phase 3 results

  • Monotherapy group (n=604): CDR-SB difference of -0.44 points versus placebo at week 52, p=0.0266, equal to a 31% slowing of decline.
  • Overall population (n=1,535): -0.11 points, p=0.4015, not statistically significant.
  • The drug is a brain-penetrant PDE5 inhibitor taken as a 30 mg tablet once a day, not an antibody infusion.
  • No change in amyloid (CSF A-beta 42/40), but a nominal drop in the tau marker p-tau217 in the monotherapy group.
  • Adverse events were similar overall, with more diarrhea, dizziness and falls on drug, and 5.2% versus 1.7% stopping for side effects.

The trial at a glance

POLARIS-AD (NCT05531526) began in December 2022 and enrolled 1,535 people with early Alzheimer’s disease at about 230 sites in 13 countries, including the US, EU states, South Korea, China, Canada and the UK. Participants took AR1001 30 mg or placebo once daily for 52 weeks. The protocol named two primary populations in advance: everyone enrolled, and a monotherapy group defined as people not on an acetylcholinesterase inhibitor (such as donepezil) or memantine at baseline. Completion was high at 87.9%, and 95.2% of completers moved into a 52-week open-label extension expected to finish in the third quarter of 2027, according to AriBio’s announcement.

Efficacy and biomarker numbers

Measure at week 52Monotherapy (n=604)Overall (n=1,535)
CDR-SB vs placebo (primary)-0.44, p=0.0266-0.11, p=0.4015
ADAS-Cog 13 vs placebo-1.16, p=0.0798Not reported
MMSE treatment difference0.7, p=0.0232 (nominal)Not reported
CSF p-tau217 vs placebo-5.34 pg/mL, p=0.0402-3.52 pg/mL, p=0.1784
CSF cGMP (target engagement)Increased, p=0.0058Increased, p<0.0001
CSF A-beta 42/40 ratioNo differenceNo difference

The rise in cerebrospinal fluid cGMP confirms the drug reached its target in the brain. The p-tau217 signal is interesting because it suggests some effect on tau pathology without touching amyloid, but it was nominal and seen only in the subgroup.

Why the subgroup win is not the whole story

Because the monotherapy group was prespecified as a primary population, AriBio can say the trial met a primary efficacy objective. Still, regulators usually want consistent benefit across the population a drug will actually treat. The monotherapy group was about 39% of enrolment, by our calculation, and the remaining patients on standard dementia drugs saw little or no added effect, which dragged the overall result to a 0.11-point difference. One plausible explanation, which AriBio has not confirmed, is that cholinesterase inhibitors or memantine mask or overlap with AR1001’s effect. Another is chance. Secondary results pointed the same way but ADAS-Cog missed significance even in the subgroup, so independent replication would strengthen the case considerably.

How AR1001 works

AR1001 is based on mirodenafil, a phosphodiesterase-5 (PDE5) inhibitor from the same drug family as treatments for erectile dysfunction, redesigned for Alzheimer’s use because it crosses the blood-brain barrier. Blocking PDE5 raises cGMP, a signalling molecule linked to synaptic function, neuronal survival and blood flow. That is a very different approach from antibodies that clear amyloid plaques. Interest in the brain’s vascular and transport systems is rising; we recently covered a blood-brain barrier study showing early failures in protein import in Alzheimer’s.

Safety profile

52-week safetyAR1001Placebo
Any treatment-emergent adverse event69.0%66.4%
Serious adverse events10.0%9.8%
Fatal adverse events4 (0.5%)2 (0.3%)
Diarrhea7.7%4.4%
Falls7.3%5.9%
Dizziness6.1%3.1%
Stopped due to adverse events5.2%1.7%

The excess of dizziness and falls deserves attention in an older population, since PDE5 inhibitors can lower blood pressure. AriBio did not report ARIA-type brain swelling, which is expected given the drug does not target amyloid and is a key practical difference from antibody therapy.

How it stacks up against approved antibodies

Cross-trial comparisons are rough, but they help frame the result. Eisai’s lecanemab showed a 0.45-point CDR-SB difference over 18 months in its pivotal trial, a 27% slowing, in the full study population. AR1001’s 0.44-point figure came at 12 months but only in a subgroup, and the overall number was a quarter of that. The appeal of a pill without infusion visits or MRI monitoring is obvious, especially in countries such as India where infusion centres and PET imaging are scarce, but the evidence bar has not yet been cleared.

Market context

According to Acumen Research and Consulting, the global dementia drugs market was valued at USD 15.5 billion in 2021 and was forecast to reach USD 32.3 billion by 2030, an 8.7% compound annual growth rate, with Asia-Pacific the fastest-growing region. The report was last updated in 2022, before anti-amyloid antibodies were widely launched, so these figures are dated and indicative. Acumen also cites WHO estimates of around 50 million people living with dementia and nearly 10 million new cases a year, the scale that makes an affordable oral option commercially important.

What happens next

AriBio says it will discuss the complete data with the FDA and present further analyses at upcoming meetings. AR1001 is not approved anywhere for Alzheimer’s disease. Watch for full publication, the open-label extension in 2027, and whether regulators accept a monotherapy-only label or ask for another trial. Policy interest in earlier intervention is also growing, as seen in the FDA’s Alzheimer’s prevention initiative.

FAQ

Did AR1001 pass its Phase 3 trial?

Partly. It met its primary endpoint in the prespecified monotherapy population (31% slower decline) but not in the overall trial population.

Is AR1001 an erectile dysfunction drug?

It belongs to the same PDE5 inhibitor class and is based on mirodenafil, but it is being developed and dosed specifically for Alzheimer’s disease. Patients should not substitute other PDE5 drugs.

When could AR1001 be approved?

No timeline has been given. AriBio first needs to discuss the data with the FDA, and the open-label extension runs to the third quarter of 2027.

Follow more clinical trial news on MedGadget.

Sources: AriBio press release, CTAD 2026 presentation (16 November 2026). Market data: Acumen Research and Consulting, Dementia Drugs Market report. Not medical or investment advice. Last updated 17 November 2026.

By Simone Lamb

Simone Lamb is the editor of Medgadget.in, covering healthcare technology, medical devices, and the latest developments in digital health.

One thought on “AR1001 Phase 3 Results: Alzheimer’s Pill Slows Decline 31% in Subgroup but Misses Overall”
  1. […] Diranersen Alzheimer’s Results: Biogen Tau Drug Slows Decline 38% at Two Years Nelkarmi FDA Approval: GSK’s Neladalkib Offers a New Option After ALK Lung Cancer Resistance Apple Health AI Study: Who Can Join and What the OpenEvidence Chatbot Does NMN Supplements India: Maharashtra Seizes Rs 1.38 Crore Stock AR1001 Phase 3 Results: Alzheimer’s Pill Slows Decline 31% in Subgroup but Misses Overall […]

Leave a Reply

Your email address will not be published. Required fields are marked *