Predicting Conversion from MCI to AD Combining Multi-Modality Data and Based on Molecular Subtype.

Area of Science:

• Neuroscience
• Genetics
• Medical Imaging

Background:

• Alzheimer's disease (AD) is a leading cause of neurodegeneration in the elderly.
• Early identification of mild cognitive impairment (MCI) patients at high risk of progressing to AD is crucial for intervention.
• Alzheimer's disease is recognized as a heterogeneous condition, suggesting subtype-specific approaches may enhance prediction.

Purpose of the Study:

• To develop and validate a subtyping-based prediction strategy for forecasting the conversion of MCI to AD within three years.
• To investigate the utility of structural magnetic resonance imaging (sMRI) and multi-omics data in predicting MCI to AD progression.
• To assess the performance of a novel prediction model incorporating MCI patient subtypes.

Main Methods:

• Utilized structural magnetic resonance imaging (sMRI) and multi-omics data (genotype, gene expression) from MCI patients in the ADNI-1 and ADNI-GO/2 datasets.
• Employed a variational Bayes approximation model with multiple kernel learning for prediction.
• Conducted internal fivefold cross-validation and external validation using independent datasets.

Main Results:

• The subtyping-based model achieved an AUC of 0.83 in internal validation, outperforming the non-subtyping model (AUC 0.78).
• External validation demonstrated an AUC of 0.78 for the subtyping model.
• The study identified significant sMRI, single nucleotide polymorphism (SNP), and mRNA expression data from peripheral blood as valuable, noninvasive predictors.

Conclusions:

• Subtyping MCI patients, integrated with omics data, significantly enhances the accuracy of predicting conversion to Alzheimer's disease.
• Peripheral blood-derived data (sMRI, SNP, mRNA) offers a noninvasive and cost-effective approach for predicting MCI to AD progression.
• This strategy holds promise for early detection and personalized intervention in AD prevention.