Screening and predicting progression from high-risk mild cognitive impairment to Alzheimer's disease

Area of Science:

• Neurology
• Neuroimaging
• Biostatistics

Background:

• Mild cognitive impairment (MCI) presents heterogeneous progression risks to Alzheimer's disease.
• Current follow-up methods involving MRI and PET scans are resource-intensive.
• Precise identification of high-risk MCI populations is crucial for efficient monitoring.

Purpose of the Study:

• To develop and evaluate a two-stage screening framework for predicting Alzheimer's disease progression in MCI patients.
• To assess the utility of additional MRI measurements in identifying high-risk MCI individuals.
• To optimize resource allocation for MCI patient monitoring.

Main Methods:

• A two-stage screening approach was implemented.
• Stage one utilized K-means clustering on clinical assessments for trajectory-template analysis.
• Stage two involved prognosis models incorporating neuroimaging data to filter high-risk individuals.

Main Results:

• The ADAS-13 score demonstrated high sensitivity in filtering high-risk MCI individuals.
• An optimal predictive model combined clinical assessment change rates and three neuroimaging measurements.
• This model showed significant Net Reclassification Improvement (NRI) and Integrated Discrimination Improvement (IDI), with the ADAS-13 longitudinal model achieving a concordance index of 0.830.

Conclusions:

• The ADAS-13 score is a sensitive biomarker for identifying MCI patients at higher risk of Alzheimer's disease progression.
• A combined approach using clinical data and specific MRI metrics enhances predictive accuracy.
• Recommended strategy: annual MRI for high-risk MCI patients and regular clinical assessments for low-risk individuals to conserve resources.