Predicting Parkinson's disease trajectory using clinical and neuroimaging baseline measures

Area of Science:

• Neuroimaging
• Biomarkers
• Neurodegenerative Diseases

Background:

• Parkinson's Disease (PD) progression prediction is crucial for treatment development and patient prognosis.
• Current prognostic tools lack precision, necessitating novel biomarkers.
• Resting-state functional magnetic resonance imaging (rs-fMRI) offers potential for non-invasive PD assessment.

Purpose of the Study:

• To develop and validate predictive biomarkers for Parkinson's Disease progression using rs-fMRI.
• To identify specific brain regions and imaging measures that correlate with current and future PD severity.
• To assess the utility of these biomarkers in clinical trial stratification and patient counseling.

Main Methods:

• Regional homogeneity (ReHo) and fractional amplitude of low frequency fluctuations (fALFF) were extracted from rs-fMRI data of 82 PD patients.
• Machine learning models were trained to predict baseline and longitudinal Movement Disorder Society Unified Parkinson's Disease Rating Scale (MDS-UPDRS) scores.
• Model performance was rigorously validated using nested cross-validation, an external dataset, and leave-one-site-out cross-validation.

Main Results:

• Predictive models explained significant variance in current (up to 30.4%) and future PD severity (up to 55.8% at 1 year, 47.1% at 2 years).
• High positive and negative predictive values (up to 79% and 80%) were achieved for distinguishing high and low PD severity.
• Increased ReHo and fALFF in default motor network regions correlated with lower current and future PD severity.

Conclusions:

• rs-fMRI derived ReHo and fALFF serve as accurate prognostic neuroimaging biomarkers for Parkinson's Disease.
• These biomarkers can enhance patient stratification for neuroprotective treatment trials.
• The findings support the clinical utility of these imaging measures for improved PD patient management and counseling.