Semaglutide shows limited improvements in Alzheimer's disease at later time points in re-analyzed EVOKE trials

Current treatments for Alzheimer's disease (AD) offer limited efficacy, primarily managing symptoms rather than halting progression. Emerging evidence suggests a link between metabolic dysfunction and AD, positioning glucagon-like peptide-1 receptor (GLP-1R) agonists as potential therapeutic candidates. The Phase 3 EVOKE and EVOKE+ trials initially assessed oral semaglutide for AD, but were declared a failure based on an intermediate readout at week 104 of CDR-SB scores. This re-analysis addresses the gap by examining previously unanalyzed later time points.

This re-analysis focused on previously unanalyzed data from week 130 and 156 of the EVOKE and EVOKE+ Phase 3 trials, which investigated oral semaglutide 14 mg in subjects with early Alzheimer's disease. The original trials were randomized, double-blind, placebo-controlled studies. Using published means, converted SEMs to SDs, and patient numbers per group, the author re-analyzed the results. Statistical significance was determined using the Welch T-test (two-tailed), which accounts for potentially unequal standard deviations between groups.

Re-analysis of the EVOKE trial data at later time points revealed several instances of separation between the semaglutide and placebo groups. Specifically, the ADCS-ADL-MCI test in the Evoke trial at week 130 showed a statistically significant difference, with a p-value of 0.0039, indicating better outcomes for the semaglutide group. Other cognitive assessments, such as the ADAS-cog-13 results, demonstrated trends towards improvement by semaglutide at week 156, although specific p-values were not provided for these trends in the abstract. Furthermore, analyses of cerebrospinal fluid (CSF) biomarkers for Alzheimer's disease also showed significant differences in some markers, suggesting a biological impact. The abstract did not provide specific numerical values or p-values for the CSF biomarker changes. While the initial week 104 readout was negative, these later time points suggest a delayed or subtle effect.

The ADCS-ADL-MCI test in the Evoke trial at week 130 showed a significant difference with p = 0.0039, favoring semaglutide.