Morphologic brain network predicts levodopa responsiveness in Parkinson disease.

Area of Science:

• Neuroscience
• Medical Imaging
• Neurology

Background:

• The levodopa challenge test (LCT) is standard for Parkinson disease (PD) evaluation and predicting deep brain stimulation (DBS) outcomes.
• Current guidelines suggest DBS for PD patients with >33% UPDRS-III score improvement on LCT.
• LCT effectiveness is limited by influencing factors and only reflects immediate dopamine response.

Purpose of the Study:

• To investigate the relationship between LCT outcomes and brain imaging features in PD patients.
• To determine if brain imaging can identify suitable candidates for DBS treatment.
• To assess the utility of morphologic brain networks in predicting levodopa responsiveness.

Main Methods:

• 38 PD patients were categorized into low (PD-LCT-L) and high (PD-LCT-H) levodopa improvement groups based on UPDRS-III scores.
• Neural networks were reconstructed from T1-weighted MRI using Jensen-Shannon divergence.
• Multiple kernel support vector machine technique was employed to analyze network differences.

Main Results:

• No significant differences in global or local morphologic brain network metrics were found between PD-LCT-L and PD-LCT-H groups.
• 43 consensus connections were affected in both groups.
• PD-LCT-L patients showed decreased connections, with increased connectivity in the dorsolateral superior frontal gyrus and left cuneus.

Conclusions:

• Morphologic brain network assessment is a valuable method for predicting levodopa responsiveness in PD.
• This imaging approach can aid in selecting appropriate candidates for DBS.
• Brain imaging provides a more objective measure than LCT alone for DBS candidacy.