Predicting Progression from Normal to MCI and from MCI to AD Using Clinical Variables in the National Alzheimer's...

Area of Science:

• Neuroscience
• Gerontology
• Biostatistics

Background:

• Clinical trials increasingly target early-stage Alzheimer's disease (AD), including pre-manifest and mild cognitive impairment (MCI) stages.
• Accurate prediction of clinical progression (e.g., normal to MCI, MCI to dementia) is crucial for efficient clinical trial enrollment and avoiding unnecessary interventions.

Purpose of the Study:

• To develop a predictive model for clinical progression in Alzheimer's disease using only clinical variables.
• To create a user-friendly tool for estimating conversion probabilities to aid in pre-screening trial participants.

Main Methods:

• Utilized machine learning approaches to select a minimal set of clinical variables for predicting progression.
• Updated previous analyses using the National Alzheimer's Coordinating Center (NACC) Uniform Data Set Version 3 (V3).
• Developed a conversion probability calculator based on the predictive model.

Main Results:

• Successfully predicted clinical progressions using a limited set of clinical variables.
• The updated model incorporates the latest NACC dataset (V3) with enhanced analytical capabilities.
• A practical calculator for estimating conversion probabilities has been generated.

Conclusions:

• Clinical variables, when analyzed with machine learning, can effectively predict Alzheimer's disease progression.
• The developed calculator offers a valuable tool for pre-screening participants in AD clinical trials.
• This approach enhances the efficiency and accuracy of identifying individuals likely to progress, optimizing trial design and participant selection.