evoke and evoke+ results published in The Lancet

The Lancet has published the complete results of the evoke and evoke+ trials, released the same day Novo Nordisk presented the full data at the AD/PD 2026 conference on March 19 ^[3]. The verdict is unambiguous: once-daily oral semaglutide up to 14 mg did not slow cognitive or functional decline in people with early Alzheimer's disease, and both trials have been discontinued because of the negative outcome ^[1].

What the trials found

The two phase 3 trials ran across 566 sites in 40 countries and screened 9,981 people, of whom 3,808 were randomly assigned — 1,855 in evoke (928 semaglutide, 927 placebo) and 1,953 in evoke+ (976 semaglutide, 977 placebo) ^[1]. Participants were 55 to 85 years old with amyloid-confirmed Alzheimer's disease and either mild cognitive impairment or mild dementia. Mean age was 72.2 years and the mean baseline Clinical Dementia Rating–Sum of Boxes (CDR-SB) score was 3.7 ^[1]. Enrollment ran from May 18, 2021 to September 8, 2023 ^[1].

The main measure was change in CDR-SB from baseline to week 104. In evoke, scores worsened by 2.3 points with semaglutide and 2.3 with placebo; in evoke+, by 2.2 and 2.1. The estimated differences were −0.08 (95% CI −0.35 to 0.20, p=0.57) and 0.10 (95% CI −0.17 to 0.38, p=0.46) ^[1]. Neither came close to statistical significance.

Secondary measures told the same story. There was no separation on the ADCS-ADL-MCI daily-function scale, the MMSE, or ADAS-Cog, and a pooled analysis found no delay in progression from mild cognitive impairment to mild Alzheimer's ^[3]. Exploratory subgroup analyses by sex, age, race, body mass index and type 2 diabetes status turned up no strong signal in any group ^[3]. At the earlier CTAD presentation, Novo Nordisk's Peter Johannsen said of the curves, "When you look at the curves, they are exactly on top of each other" ^[4].

One design detail matters: evoke+ was meant to enroll people with cerebrovascular disease, but only 54 participants (2.8%) had small vessel pathology, making the two study populations nearly identical ^[1][4].

Safety and biomarkers

Treatment-emergent adverse events occurred in 1,729 of 1,896 people on semaglutide (91.2%) versus 1,613 of 1,902 on placebo (84.8%) ^[1]. Five deaths were judged treatment-related by investigators — one in the semaglutide group and four on placebo ^[1]. The authors concluded safety was consistent with semaglutide's use in other conditions ^[1]. Common side effects were the familiar ones: roughly a quarter of participants on the drug reported nausea, 14% diarrhea and 12% vomiting ^[4]. Participants on semaglutide lost an average of 5.8% of body weight over two years, while the placebo group gained 0.6% ^[4].

In a small spinal-fluid substudy, seven biomarkers including p-tau181 and p-tau217 dropped by 10% or less at week 78 on semaglutide, and blood C-reactive protein fell about 30% ^[4]. None of those shifts translated into clinical benefit ^[3][4]. Blood markers GFAP and NfL actually rose slightly, a result presenter Jeffrey Cummings said he could not explain ^[4].

Why it matters for patients

For people taking semaglutide for type 2 diabetes or weight, this changes nothing about those approved uses — evoke tested an unapproved use in a different population ^[1]. But it closes the door, for now, on the hope that taking a GLP-1 drug will slow an existing Alzheimer's diagnosis. Observational studies linking GLP-1 use to lower dementia risk looked at people with diabetes or obesity and measured time to a new diagnosis, while evoke enrolled people with biomarker-confirmed Alzheimer's already underway — two different questions ^[3]. Patients or families weighing a GLP-1 medication specifically for memory benefit now have a large, negative, placebo-controlled answer.

What happens next

Novo Nordisk announced topline results on November 24, 2025, and said the planned one-year extension in both trials would be discontinued and the program terminated ^[3][4]. Panelists at AD/PD argued GLP-1 research in dementia should continue, possibly in people with both Alzheimer's and metabolic or vascular disease, or as prevention started earlier — but any such trials would require years and substantial new investment, and none have been announced ^[3][4].