Skip to content

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.

·

Primary Outcome
Change in CDR-SB score from baseline to week 104
Key Finding
Oral semaglutide 14 mg did not slow clinical progression in early Alzheimer's disease compared with placebo in either the evoke (estimated CDR-SB difference -0.08, 95% CI -0.35 to 0.20, p=0.57) or evoke+ (0.10, 95% CI -0.17 to 0.38, p=0.46) trials.

AI-generated research brief — verify at source

Oral Semaglutide Fails to Slow Clinical Progression in Early Alzheimer’s Disease

In two large phase 3 placebo-controlled trials, oral semaglutide 14 mg produced no meaningful reduction in cognitive and functional decline in early Alzheimer’s disease — the estimated difference in CDR-SB score change was negligible and statistically indistinguishable from zero in both evoke (−0.08, 95% CI −0.35 to 0.20, p=0.57) and evoke+ (0.10, 95% CI −0.17 to 0.38, p=0.46). These definitively neutral results close a major translational hypothesis and challenge the assumption that GLP-1 receptor agonist benefits observed in metabolic and epidemiological contexts extend to Alzheimer’s disease pathology.

What Was Studied

The evoke and evoke+ trials investigated whether oral semaglutide — a GLP-1 receptor agonist already established in type 2 diabetes and obesity — could slow clinical progression in individuals with early symptomatic Alzheimer’s disease, as measured by change in the Clinical Dementia Rating-Sum of Boxes (CDR-SB) score over two years. The question was motivated by convergent signals from animal models, clinical observations, and real-world data in people with metabolic disease suggesting that GLP-1 receptor agonists may reduce dementia risk.

How It Was Studied

Both evoke and evoke+ were multicentre, randomised, double-blind, placebo-controlled phase 3 trials conducted across 566 sites in 40 countries, enrolling a combined 3,808 participants. Eligible individuals were aged 55–85 years with amyloid-confirmed Alzheimer’s disease presenting as either mild cognitive impairment or mild dementia; evoke+ additionally enrolled participants with significant cerebral small vessel pathology to probe a potentially distinct disease subgroup. Participants were randomised 1:1 to receive oral semaglutide up to 14 mg once daily (flexible dose) or matched placebo for up to 156 weeks, with the primary endpoint assessed at week 104. Mean age at baseline was 72.2 years and mean baseline CDR-SB score was 3.7 (SD 1.6), confirming an early-stage symptomatic population.

What Was Observed

  • No effect on the primary outcome in either trial: CDR-SB scores rose by approximately 2.3 points in both the semaglutide and placebo arms of evoke, and by approximately 2.2 and 2.1 points respectively in evoke+. The estimated treatment differences were trivially small and the confidence intervals comfortably crossed zero in both cases, confirming no benefit (evoke: −0.08, 95% CI −0.35 to 0.20; evoke+: 0.10, 95% CI −0.17 to 0.38).
  • Higher rate of adverse events with semaglutide: Treatment-emergent adverse events were reported in 91.2% of participants receiving semaglutide compared with 84.8% of those on placebo — a pattern consistent with the known gastrointestinal tolerability profile of GLP-1 receptor agonists seen in other indications, rather than an unexpected safety signal.
  • Treatment-related fatalities were rare and not concentrated in the semaglutide group: Five deaths were considered treatment-related by investigators — one in the semaglutide group and four in the placebo group — providing no evidence of excess mortality attributable to the active drug.
  • The small vessel pathology subgroup was severely underpowered: Only 54 participants (2.8%) in evoke+ carried significant small vessel pathology, making it impossible to draw conclusions about whether this subgroup might respond differently to treatment.

Why This Matters

These trials represent the largest and most rigorous test to date of GLP-1 receptor agonist therapy in Alzheimer’s disease, and their negative outcome has immediate implications for the field. The results directly challenge the translational logic that linked metabolic and vascular benefits of semaglutide to potential neuroprotection in Alzheimer’s pathology. Researchers and funders will need to reconsider whether the epidemiological signal reflects confounding, a disease-stage-dependent window of opportunity not captured here, or biological mechanisms that do not operate in established amyloid-driven neurodegeneration.

How to Read This Result

This is a high-quality, adequately powered phase 3 trial with a robust design, and the neutral findings across both trials are consistent and credible; however, the very small proportion of participants with small vessel pathology limits any conclusions about that subgroup, and early discontinuation means the full 156-week follow-up window was not completed for all participants.

Limitations

Both trials were discontinued early following the negative clinical outcome, meaning data beyond the primary timepoint may be incomplete or absent for a portion of participants. The evoke+ design intended to examine a small vessel disease subgroup, but only 2.8% of enrolled participants carried that pathology, rendering this analysis effectively underpowered and inconclusive. The published abstract does not detail dropout rates or the specific reasons for individual participant discontinuations, which limits a full assessment of potential attrition bias.

Quality: High High-impact journal Research Article
Source
Lancet· PMID: 41865758
View full study
Disclaimer: Content on MEDITELI is AI-generated for informational purposes only. It does not constitute medical advice. Always consult a qualified healthcare professional before making health-related decisions. Original research should be reviewed in full before clinical application.