Novo Nordisk’s EVOKE and EVOKE+ phase 3 trials delivered a clear result. Oral semaglutide did not slow cognitive or functional decline in early symptomatic Alzheimer’s disease as measured by the CDR-SB over 104 weeks.

That conclusion is unambiguous.

What matters just as much is what the trials did produce.

The studies enrolled more than 3,800 participants with confirmed amyloid pathology and generated a large, well-controlled dataset across clinical outcomes, biomarkers, and safety measures. While oral semaglutide showed favorable movement in several Alzheimer’s-related biomarkers, those changes did not translate into measurable clinical benefit. The safety profile remained consistent with prior experience, and the extension phase was discontinued once the primary endpoint was missed.

This result does not diminish the value of the trial. It clarifies it.

The EVOKE findings reinforce a critical lesson in Alzheimer’s research: biomarker improvement does not automatically mean clinical improvement. That gap matters. Understanding where and why that disconnect occurs helps refine trial design, endpoint selection, and expectations for future therapies.

They also help reset expectations for GLP-1–based treatments in neurodegenerative disease. Semaglutide has demonstrated clear benefit in diabetes, obesity, cardiovascular disease, kidney disease, and metabolic liver disease. Alzheimer’s is different. Strong biological rationale alone is not enough.

Some observers have noted that the oral formulation delivers lower systemic exposure than injectable versions and may not provide sufficient central nervous system effect. Others have focused on what the results imply about Alzheimer’s biology itself rather than the drug. Both interpretations are informed by the data produced here.

Why the Data Still Matter

Richard DiMarchi offered one of the more measured responses to the results. While others immediately questioned dose selection, trial design, or strategy, he focused on what the study actually contributed.

His point was simple and important: entering an unproven therapeutic area carries real scientific risk, and that risk is only reduced by running large, well-designed trials. When those trials fail to meet their endpoints, they still provide something essential. They tell the field what does not work and help define where the next attempts should focus.

That perspective matters.

The EVOKE trials generated a substantial body of data that will inform future Alzheimer’s research. They help clarify the limits of biomarker-driven assumptions, refine expectations around metabolic therapies in neurodegeneration, and guide how future studies are designed.

A trial does not need to succeed to be useful.

In this case, the value lies in what was learned, not in what was hoped for.

Sources

Novo Nordisk
Fierce Pharma
LinkedIn - Doustdar (Video)
LinkedIn - DiMarchi