Image

APOLLO I Study

graphical divider

Validating Blood-Based Prediction Of Alzheimer’s Dementia Progression

APOLLO I was AgenT’s first clinical validation study designed to demonstrate that blood-based multiomics can predict progression to Alzheimer’s dementia.

The study analyzed plasma samples from 603 individuals across 7 independent cohorts in the United States, France, Spain, and Australia. Participants were clinically followed to determine whether they progressed to Alzheimer’s dementia, developed another brain disorder, or, for cognitively unimpaired individuals, remained cognitively healthy during follow-up.

APOLLO I led to the development of two blood-based predictive algorithms: B-HEALED, designed to predict progression to Alzheimer’s dementia in individuals with cognitive impairment, and B-AHEAD, designed to predict progression in cognitively unimpaired individuals.

B-HEALED™: Predicting Progression in Cognitively Impaired Individuals

B-HEALED was developed and validated in 345 cognitively impaired individuals at baseline, including 193 individuals with mild cognitive impairment and 152 individuals with mild dementia. The algorithm was trained on 236 individuals and externally validated on 109 individuals.

Individuals were clinically followed over time to determine whether they progressed to Alzheimer’s dementia or to another brain disorder, making B-HEALED a predictive algorithm based on longitudinal outcomes rather than a cross-sectional biomarker test.

B-HEALED is AgenT’s most advanced predictive algorithm and has been published in the Journal of Prevention of Alzheimer’s Disease following APOLLO I clinical validation (Souchet et al., 2024).

B-AHEAD™: Predicting Progression in Cognitively Unimpaired Individuals

B-AHEAD was developed and validated in 258 cognitively unimpaired individuals at baseline. The algorithm was trained on 176 individuals and externally validated on 82 individuals.

Individuals were clinically followed over time to determine whether they progressed to Alzheimer’s dementia, developed another brain disorder, or remained cognitively healthy, making B-AHEAD a predictive algorithm based on longitudinal outcomes rather than a cross-sectional biomarker test.

B-AHEAD is designed to identify cognitively unimpaired individuals at risk of progressing to Alzheimer’s dementia before cognitive symptoms become clinically apparent. Results were presented at CTAD 2025 (B-AHEAD CTAD 2025 Poster).

Research Partners

icon
icon
icon
icon
icon
icon
icon
arrow-up icon