Predicting cognitive function in Alzheimer's clinical trials via amyloid β-protein biomarkers

This study investigates the association between amyloid-β (Aβ) biomarkers and clinical cognitive outcomes and quantitatively elucidates their relationship, providing robust evidence supporting the amyloid hypothesis and advancing the development of novel anti-amyloid therapeutics, such as aducanumab, lecanemab and donanemab.

Placebo-controlled randomized clinical trials reporting Aβ-related biomarkers and cognitive function clinical outcomes were retrieved from PubMed, EMBASE and Cochrane Library. Pearson correlation analysis was first used to screen indices, and then a model-based meta-analysis (MBMA) using non-linear mixed-effect modelling was established to predict cognitive function based on biomarkers while examining relevant factors affecting the relationship.

Primary outcomes included changes in the Clinical Dementia Rating Scale Sum of Boxes (CDR-SB) and the Alzheimer's Disease Assessment Scale-Cognitive Subscale (ADAS-COG-11). The analysis included 57 articles representing 93 417 subjects, with a modelling subset of 18 246 patients providing paired biomarker-endpoint data for Alzheimer's disease or mild cognitive impairment. Results showed significant correlations between the standard uptake value ratio (SUVR) of β-amyloid plaques and CDR-SB, centiloid and CDR-SB, and SUVR and ADAS-COG-11. Three prediction models of cognitive function scales based on imaging index of β-amyloid plaques were established and found to be significantly impacted by factors such as baseline CDR-SB, treatment duration and the proportion of patients receiving basic treatment.

This study clarified the correlation and established predictive models for CDR-SB and ADAS-COG-11 based on amyloid imaging. This research identifies potential biomarkers and model-derived benchmarks for future dose selection and decision-making in Alzheimer's drug development, potentially accelerating the development of new treatments.