A data-driven model of brain volume changes in progressive supranuclear palsy.

Area of Science:

• Neuroimaging
• Neuropathology
• Clinical Neurology

Background:

• Progressive supranuclear palsy (PSP) is a neurodegenerative disease with Richardson syndrome as its most common clinical presentation.
• Current understanding of PSP pathophysiology, particularly the sequence of brain atrophy, is limited.
• Determining the atrophy sequence is crucial for understanding disease mechanisms and improving clinical trial design.

Purpose of the Study:

• To determine the sequence of brain atrophy in Richardson syndrome using a probabilistic event-based model.
• To validate the model's staging system using longitudinal MRI data and clinical scores.
• To correlate disease progression with clinical severity and duration.

Main Methods:

• Utilized a probabilistic event-based model on cross-sectional structural MRI scans from 341 Richardson syndrome patients and 260 controls.
• Incorporated 12-month follow-up MRI scans and the Progressive Supranuclear Palsy Rating Scale (PSPRS) for longitudinal validation.
• Cross-validated the model's robustness and staging utility.

Main Results:

• The model identified initial atrophy in the brainstem and subcortical regions, progressing to the superior cerebellar peduncle, deep cerebellar nuclei, and then the cortex.
• Cortical atrophy followed an anterior-to-posterior pattern: insula, frontal, temporal, parietal, and occipital lobes.
• Longitudinal data confirmed consistent progression through model stages; clinical severity and disease duration correlated significantly with the model stage (P < 0.01).

Conclusions:

• The study provides a detailed in vivo sequence of brain atrophy in Richardson syndrome, aligning with post-mortem findings.
• The event-based model offers a validated staging system for tracking disease progression.
• This framework can aid in stratifying patients for clinical trials and monitoring disease advancement.