Predictive model of spread of Parkinson's pathology using network diffusion.

Area of Science:

• Neuroscience
• Computational Biology
• Pathology

Background:

• Neurodegenerative disorders like Parkinson's disease (PD) show evidence of prion-like mechanisms.
• Alpha-synuclein aggregation and spread are implicated in PD pathogenesis.

Purpose of the Study:

• To adapt and apply a network diffusion model (NDM) to quantitatively predict the trans-neuronal spread of alpha-synuclein in PD.
• To validate the NDM using neuropathological, neuroimaging, and clinical data from Parkinson's patients.
• To compare the NDM with alternative spread models and investigate the role of genetic factors.

Main Methods:

• Developed and tailored a quantitative network diffusion model (NDM) for PD.
• Modeled alpha-synuclein spread from the substantia nigra (SN).
• Validated the model using MRI deformation data from 232 PD patients and compared it with random and distance-based spread models.

Main Results:

• The substantia nigra (SN) was identified as the most likely initial seeding region for PD pathology.
• The connectivity-based NDM provided a better fit than distance-based or random spread models.
• The model's predicted temporal sequence of affected regions closely matched Braak stages III-VI, creating a 'computational Braak' staging system.
• Network processes, rather than regional genetic expression, were stronger predictors of regional atrophy in PD.

Conclusions:

• The NDM successfully models the prion-like spread of PD pathology, originating from the SN.
• Network connectivity is a more significant driver of regional vulnerability in PD than genetic factors.
• The NDM offers a promising tool for PD diagnosis, prognosis, and staging, validated by in vivo human imaging data.