Scientific Publications

A key issue to Alzheimer's disease clinical trial failures is poor participant selection. Participants have heterogeneous cognitive trajectories and many do not decline during trials, which reduces a study's power to detect treatment effects. Trials need enrichment strategies to enroll individuals who will decline. We developed machine learning models to predict cognitive trajectories in participants with early Alzheimer's disease (n=1342) and presymptomatic individuals (n=756) over 24 and 48 months respectively. Baseline magnetic resonance imaging, cognitive tests, demographics, and APOE genotype were used to classify decliners, measured by an increase in CDR-Sum of Boxes, and non-decliners with up to 79% area under the curve (cross-validated and out-of-sample). Using these prognostic models to recruit enriched cohorts of decliners can reduce required sample sizes by as much as 51%, while maintaining the same detection power, and thus may improve trial quality, derisk endpoint failures, and accelerate therapeutic development in Alzheimer's disease.

We present our new algorithmic solution to enrich presymptomatic cohorts in Alzheimer's disease.

A significant proportion of trial participants will not exhibit cognitive decline during a trial. This can negatively impact a trial's ability to detect differences between placebo and treatment arms, leading to failure to meet its endpoints. We present a new algorithmic enrichment solution for early AD clinical trials.

We present a new multimodal subtype approach to address challenges related of Alzheimer’s disease trials to enroll presymptomatic individuals who are likely to decline cognitively.

We propose to use highly specific neuroimaging and genetic signatures that are indicative of the risk of cognitive decline in individuals with mild Alzheimer’s disease and mild cognitive impairment (MCI).

Subjects enrolled in Alzheimer's disease clinical trials need to have a significant amount of β-Amyloid deposits in their brain since most drugs under investigation are currently targeting that protein. The in-vivo gold standard biomarkers are amyloid PET imaging and cerebrospinal fluid (CSF) from a lumbar puncture. Those methods have significant drawbacks, namely being very expensive, invasive and not readily available. Moreover, the prevalence of β-Amyloid+ subjects decreases as we investigate earlier phases of the disease, stressing the importance of an enrichment strategy to increase prevalence. We propose to use cheaper and more broadly available modalities in order to make a pre-screening of the subjects most likely to be β-Amyloid+ prior to the most expensive and invasive modalities.

We propose to use highly specific signatures (based on neuroimaging and cognitive tests) that are indicative of the risk of cognitive decline in the MCI population. Our first goal was to subdivide a group of subjects with MCI showing significant β-Amyloid deposit into two cohorts of high-and low-risk of decline and obtain a less heterogeneous cohort with more drastic cognitive changes. The second objective was to compare the high-risk and low-risk groups with the initial cohort which was based only on β-Amyloid positive criteria. Lastly, we wanted to evaluate if the predicted risk was confirmed using common cognitive endpoints when the subjects are followed longitudinally.