Longevity & AgingPress Release

Erectile dysfunction drug AR1001 slows early Alzheimer's decline in one subgroup but misses overall

In a 1,535-patient phase III trial, the PDE5 inhibitor AR1001 missed its overall endpoint but slowed decline 31% in patients on no other Alzheimer's drugs.

Wednesday, November 18, 2026 0 views
Published in MedPage Today
Article visualization: Erectile dysfunction drug AR1001 slows early Alzheimer's decline in one subgroup but misses overall

Summary

Mirodenafil (AR1001), an oral drug approved in South Korea for erectile dysfunction, was tested in early Alzheimer's disease in the phase III POLARIS-AD trial. It blocks phosphodiesterase-5, boosting cyclic GMP signaling in the brain, and in preclinical models it improved cognition and reduced amyloid and tau. In 1,535 people with amyloid-positive mild cognitive impairment or mild dementia, it did not significantly beat placebo overall on the CDR-SB scale at 52 weeks (-0.11 points, P=0.40). In a prespecified monotherapy subgroup of 604 people not taking standard Alzheimer's drugs, it slowed decline by 0.44 points, or 31% (P=0.027). Side effects included diarrhea, falls and dizziness, and more people stopped the drug than placebo (5.2% vs 1.7%). This is conference-reported data, not yet peer reviewed, and people should not self-treat with Viagra.

Detailed Summary

Alzheimer's disease is a leading threat to healthy aging, and few treatments meaningfully slow it. Phosphodiesterase-5 (PDE5) inhibitors, best known for treating erectile dysfunction, have drawn interest because they raise cyclic GMP signaling and relax blood vessels, which may improve brain blood flow. Observational studies of sildenafil users have given conflicting hints of lower Alzheimer's risk.

The phase III POLARIS-AD trial, presented at the CTAD conference, tested AR1001 (mirodenafil), a brain-penetrant PDE5 inhibitor approved in South Korea for erectile dysfunction. It randomized 1,535 amyloid-positive people with mild cognitive impairment or mild dementia, average age 73 and just over half women, to 30 mg daily or placebo for 52 weeks.

In the overall population, the primary endpoint was missed: the CDR-SB difference was -0.11 points versus placebo (P=0.4015). In the prespecified monotherapy subgroup of 604 people not taking a cholinesterase inhibitor or memantine, the difference was -0.44 points (P=0.0266), a 31% slowing of decline. Secondary cognitive and functional measures trended in favor of AR1001 in both groups, with a nominally significant Mini-Mental State Examination difference in the monotherapy group.

Safety was broadly acceptable. The most common events were diarrhea, falls and dizziness. Discontinuation due to adverse events was 5.2% with AR1001 versus 1.7% with placebo. Serious adverse events were similar (10.0% vs 9.8%). Four deaths occurred on AR1001 and three on placebo.

The mechanism is plausible and worth watching, but the evidence is mixed. The positive result comes from a subgroup after the overall primary endpoint failed, so it needs confirmation. The data were presented at a meeting, and full results are pending. Lifespan and biomarker effects were not reported here. Experts caution that people should not take Viagra to treat Alzheimer's.

Key Findings

  • Phase III POLARIS-AD (n=1,535) missed its primary endpoint overall: CDR-SB difference -0.11 vs placebo at 52 weeks, P=0.4015.
  • In the 604-person monotherapy subgroup, AR1001 slowed decline by 0.44 CDR-SB points (31%, P=0.0266).
  • Secondary measures trended favorably in both groups; MMSE was nominally significant only in the monotherapy group.
  • Adverse-event discontinuation was 5.2% on AR1001 vs 1.7% on placebo; serious events were similar, with four vs three deaths.
  • Mechanism: brain-penetrant PDE5 inhibition raises cGMP signaling; preclinical models showed less amyloid and tau and better cognition.

Methodology

This is a news report on a conference presentation of a randomized, placebo-controlled phase III trial, 52 weeks, in 1,535 people with early Alzheimer's. MedPage Today is a reputable outlet, but the data are not yet peer reviewed and the article text is truncated. The monotherapy result is a prespecified subgroup finding; the overall primary endpoint was a null.

Study Limitations

The benefit appeared in a subgroup after the overall primary endpoint failed, so it may reflect chance or subgroup effects and needs confirmation. Full peer-reviewed data, biomarker results and long-term outcomes are unavailable, and the article was cut off before describing the primary endpoint in full. Verify details in the eventual publication.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.

Enter your email to subscribe: