Abnormal metabolic covariance patterns associated with multiple system atrophy and progressive supranuclear palsy

Area of Science:

• Neuroimaging
• Neurology
• Metabolic Brain Imaging

Background:

• Differentiating neurodegenerative parkinsonisms like MSA and PSP is challenging due to similar early symptoms.
• Metabolic brain imaging offers potential for improved diagnostic accuracy.

Purpose of the Study:

• To identify characteristic metabolic patterns of MSA and PSP using network analysis of 2-[18F]FDG PET scans.
• To develop a tool for estimating metabolic brain heterogeneity in patients.

Main Methods:

• Applied scaled subprofile model/principal component analysis to 2-[18F]FDG PET scans from 20 MSA patients, 20 PSP patients, and 20 healthy controls.
• Validated patterns in a larger cohort (56 MSA, 45 PSP, 116 PD, 61 HC).
• Utilized innovative heat-map analysis for graphical display of regional pattern sub-scores in individuals.

Main Results:

• Identified MSARP (hypometabolism in cerebellum, putamen) and PSPRP (hypometabolism in medial prefrontal cortex, caudate nucleus, frontal cortex, mesencephalon).
• Patterns effectively discriminated between MSA/PSP patients and controls, and between different parkinsonian cohorts (p < 0.001).
• Patterns demonstrated high sensitivity and specificity (AUC for MSARP/PSPRP: 0.96/0.99). Heat-map analysis revealed individual heterogeneity.

Conclusions:

• Replication and validation confirm MSARP and PSPRP as robust metabolic biomarkers for clinical and research use.
• Heat-map analysis provides insights into regional pattern contributions and individual patient heterogeneity in MSA and PSP.