---
type: source
title: "EVOKE/EVOKE+: Oral Semaglutide Fails Phase 3 in Alzheimer's — Biomarkers Improve But Clinical Benefit Absent"
author: "The Lancet / Alzheimer's Drug Discovery Foundation / NeurologyLive"
url: https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(26)00459-9/fulltext
date: 2026-04-15
domain: health
secondary_domains: []
format: research-article
status: unprocessed
priority: medium
tags: [GLP-1, semaglutide, Alzheimers, neurodegeneration, Phase-3, clinical-failure, biomarker-gap, CNS]
intake_tier: research-task
---

## Content

**Study:** "Efficacy and safety of oral semaglutide 14 mg (flexible dose) in early-stage symptomatic Alzheimer's disease (evoke and evoke+): two phase 3, randomised, placebo-controlled trials." Published in The Lancet (2026).

**Trials:** EVOKE and EVOKE+ — two parallel Phase 3, double-blind, placebo-controlled trials. n=3,808 total, ages 55-85 with mild cognitive impairment (MCI) or mild dementia due to Alzheimer's disease with confirmed amyloid positivity.

**Primary endpoints:** Not met in either trial. Neither trial demonstrated superiority of oral semaglutide 14mg to placebo in slowing cognitive or global decline from baseline to week 104.

**Key findings:**
- No delay in time to progression to dementia (MCI subgroup, pooled, up to week 156)
- Semaglutide improved AD-related biomarkers significantly: up to 10% reduction in CSF core AD biomarkers
- Biomarker changes also: significant reductions in CSF neuroinflammation markers
- BUT: 10% biomarker change was NOT large enough to produce measurable clinical benefit
- Novo Nordisk decision: discontinue 1-year extension periods for both trials

**Scientific interpretation (Alzheimer's Drug Discovery Foundation):**
- The biomarker improvement without clinical benefit suggests: either (a) the dose was insufficient, (b) the treatment window was too late, or (c) neuroinflammation/AD pathobiology is not the rate-limiting step in this population
- The finding provides data on which mechanisms to pursue in "next generation therapies targeting Alzheimer's pathobiology"
- Semaglutide may still be relevant in prevention (not treatment) of neurodegeneration — different trial design needed

**Context for GLP-1 scope expansion story:**
- Strong preclinical rationale (GLP-1 receptors in hippocampus, neuroprotective mechanisms)
- Multiple observational studies showing GLP-1 users have lower Alzheimer's incidence
- EVOKE failure: demonstrates observational signal ≠ causal benefit in clinical trial
- This is the ALZHEIMER'S entry in the GLP-1 CNS expansion story — distinguished from AUD (which has Phase 2 RCT success)

**Comparison to AUD evidence:**
- AUD: Phase 2 RCT (SEMALCO, 108 patients) — significant efficacy on primary endpoint
- Alzheimer's: Phase 3 RCT (EVOKE/EVOKE+, 3,808 patients) — no efficacy on primary endpoint
- Mechanism: AUD operates through VTA dopamine (reward), Alzheimer's through neurodegeneration/amyloid (distinct mechanism)
- Lesson: GLP-1 CNS effects are not uniform across conditions — mechanism specificity matters

## Agent Notes

**Why this matters:** This is a critical limiting factor on the GLP-1 scope expansion story. The same drug that shows promise for AUD (Phase 2) and appears protective for depression (large observational) fails in a Phase 3 Alzheimer's trial. This demonstrates: (1) GLP-1 CNS effects are mechanism-specific, not universal; (2) biomarker improvement does not guarantee clinical benefit; (3) Phase 3 failure is a reminder that Phase 2 AUD success needs replication.

**What surprised me:** The biomarker finding (10% CSF reduction in AD biomarkers) with NO clinical benefit is the most scientifically interesting result. It suggests either the 10% change was too small (threshold effect) or biomarkers are measuring a process that doesn't drive clinical decline. This is relevant far beyond GLP-1 — it's a statement about Alzheimer's biomarker utility more broadly.

**What I expected but didn't find:** Any signal of efficacy on cognitive outcomes, even secondary endpoints. The prior observational data was strong enough that many researchers expected at least a trend.

**KB connections:**
- [[AI compresses drug discovery timelines by 30-40 percent but has not yet improved the 90 percent clinical failure rate that determines industry economics]] — EVOKE is another example of Phase 3 failure despite promising preclinical/observational data
- [[GLP-1 receptor agonists are the largest therapeutic category launch in pharmaceutical history but their chronic use model makes the net cost impact inflationary through 2035]] — Alzheimer's failure creates a ceiling on GLP-1 scope expansion claims

**Extraction hints:**
1. This constrains any GLP-1 CNS expansion claim — should be cited as counter-evidence to broad CNS benefit claims
2. Possible claim: "Semaglutide Phase 3 Alzheimer's failure demonstrates GLP-1 CNS effects are mechanism-specific — VTA dopamine pathways (addiction, reward) respond while amyloid/neurodegeneration pathways (Alzheimer's) do not"
3. The biomarker-without-clinical-benefit finding connects to broader KB theme on surrogate endpoints

**Context:** Published in The Lancet (April 2026, approximate). Covered by NeurologyLive, Neurology Advisor, Alzheimer's Drug Discovery Foundation. Novo Nordisk stock impact. This was one of the most-watched Alzheimer's trials of the year.

## Curator Notes (structured handoff for extractor)
PRIMARY CONNECTION: [[GLP-1 receptor agonists are the largest therapeutic category launch in pharmaceutical history but their chronic use model makes the net cost impact inflationary through 2035]]
WHY ARCHIVED: Major limiting factor for the GLP-1 therapeutic scope expansion narrative. Success in addiction (VTA dopamine) + failure in Alzheimer's (amyloid/neurodegeneration) is the mechanistic boundary of GLP-1 CNS effects.
EXTRACTION HINT: Write as scope-qualification evidence for existing GLP-1 claim. Also potential standalone claim about mechanism-specific vs. universal CNS effects. Pair with SEMALCO archive.