Functional and Structural Connectivity Correlates of Axial Symptom Outcomes After Pallidal Deep Brain Stimulation in Parkinson's Disease.

Background/Objectives: Deep brain stimulation (DBS) of the globus pallidus interna (GPi) is a safe and established therapy for management of refractory motor fluctuations and dyskinesia in Parkinson's disease (PD). However, the relationship between stimulation site connectivity and improvement of axial gait symptoms remains poorly understood, particularly when stimulating in the GPi. This study investigated functional and structural connectivity patterns specifically associated with axial symptom outcomes following bilateral GPi-DBS, and, as a secondary exploratory analysis, examined whether Volumes of tissue activated (VTAs)-based connectivity related to overall UPDRS-III change. Methods: We retrospectively analyzed 19 PD patients who underwent bilateral GPi-DBS at the University of Florida (2002-2017). Unified Parkinson's Disease Rating Scale (UPDRS-III) axial gait subscores were assessed at baseline and 36-month follow-up. VTAs were reconstructed using Lead-DBS and coregistered to Montreal Neurological Institute (MNI) space. Structural connectivity was evaluated with diffusion tractography, and functional connectivity was estimated using normative resting-state fMRI datasets. Correlations between VTA connectivity and clinical improvement were examined using Spearman correlation and voxelwise analyses. Results: Patients with axial improvement in motor scales demonstrated specific VTA connectivity to sensorimotor and supplementary motor networks, particularly lobule V and lobules I-IV of the cerebellum. These associations were specific to axial gait subscores. In contrast, worsening axial gait symptoms correlated with connectivity to cerebellar Crus II, cerebellum VIII, calcarine cortex, and thalamus (p < 0.05). Total UPDRS-III scores did not show a significant positive correlation with supplementary motor area or primary motor cortex connectivity; a non-significant trend was observed for VTA-M1 connectivity (R = 0.41, p = 0.078). Worsening total motor scores were associated with cerebellar Crus II and frontal-parietal networks. These findings suggest that distinct connectivity patterns underlie differential trajectories in axial and global motor outcomes following GPi-DBS. Conclusions: Distinct connectivity profiles might underlie axial gait symptom outcomes following GPi-DBS. Connectivity to motor and sensorimotor pathways supports improvement, whereas involvement of Crus II and occipital networks predicts worsening. Additional studies to confirm and expand on these findings are needed, but our results highlight the value of connectomic mapping for refining patient-specific targeting and developing future programming strategies.