Supplementary Material for: Anatomical Group-level studies of the Volume of Tissue Activated by Deep Brain Stimulation in Parkinson’s Disease: a model for targeting?

Introduction
Parkinson’s disease (PD) is a neurodegenerative disorder for which deep brain stimulation (DBS) of the subthalamic nucleus (STN) is an established treatment. Despite standardized programming, some patients seem to respond very well to DBS (optimal responders), while others seem to react poorly (poor responders). The objective is to compare the area of tissue activated between optimal and poor responders and determine whether there is a potential optimal stimulation area.

Methods
For 338 PD patients with STN-DBS, four outcome categories on the Movement Disorders Society Unified Parkinson Disease Rating Scale (MDS-UPDRS) motor part were assessed: hemibody, rigidity, bradykinesia, and tremor score for left and right separately. For each outcome category, patients were divided into one of three responder groups, based on their percentage hemibody improvement (optimal responders, >70% improvement; responders, 30-70% improvement; poor responders, <30% improvement). For each of the resulting 12 groups, volumes of tissue activated (VTA) were modeled for every individual electrode based on the stimulation parameters during follow-up assessment. To enable the responder groups comparison, all VTAs were aggregated into a so-called heatmap in normalized space. As we were mainly interested in the difference in VTA location for the optimal and poor responders, only these group heatmaps were visually assessed in reference to the STN. For quantitative sub-analyses the amount of current applied and spread of electrode location was compared.

Results
Considerable overlap between heatmaps of optimal and poor responders within the dorsolateral region of the STN was seen. The amount of current applied and spread of electrode location did not differ.

Conclusions
This study comparing anatomical group-level studies of VTAs of optimal responders with poor responders for STN-DBS in PD did not find an area of optimal stimulation to reduce variability in DBS outcome. However, the heatmap of optimal responders can facilitate easier DBS targeting. To reduce variability in DBS outcome, focus could shift more towards patient-specific anatomy and connectivity levels in order to determine the individual optimal subthalamic area for programming.