Preoperative Glymphatic Asymmetry Is Associated with Motor Outcomes Following STN-DBS in Parkinson's Disease.
BACKGROUND Interindividual variability in motor outcomes after subthalamic nucleus deep brain stimulation (STN-DBS) in Parkinson's disease (PD) remains incompletely understood. Glymphatic-related imaging abnormalities have been reported in PD, but their relevance to neuromodulation response is unclear. OBJECTIVES To investigate whether preoperative glymphatic function, assessed using the Diffusion Tensor Imaging along the Perivascular Space (DTI-ALPS) index, is associated with motor outcomes after STN-DBS and whether hemispheric asymmetry provides additional explanatory value. METHODS We retrospectively evaluated 74 advanced PD patients undergoing bilateral STN-DBS with ≥6 months follow-up. Preoperative assessment included 3 T MRI-derived ALPS measures, Fazekas score, perivascular space burden, MDS-UPDRS-III, levodopa responsiveness, and metabolic markers. Primary outcome was percentage motor improvement at follow-up. Multivariable regression models assessed ALPS associations after adjustment for levodopa responsiveness and measured clinical covariates. RESULTS Lower ALPS asymmetry index (AI) was independently associated with greater percentage motor improvement (standardized β = -0.508, p < 0.001), while greater preoperative levodopa responsiveness was also associated with improvement (β = 0.740, p < 0.001). Lower ALPS-AI was additionally associated with axial motor improvement and responder status at both ≥50% and ≥ 33% thresholds. All four principal levodopa-adjusted ALPS-AI associations remained significant after Benjamini-Hochberg correction (all pFDR≤0.0069). Mean ALPS index, Fazekas score, and perivascular space burden were not associated with motor outcomes. CONCLUSIONS Preoperative glymphatic asymmetry was associated with motor outcomes after STN-DBS independently of levodopa responsiveness and measured clinical covariates. ALPS-derived asymmetry may capture outcome variability not reflected by conventional imaging, although prospective validation and evaluation alongside lead localization, stimulation parameters, and programming factors are required.