Towards connectome-guided optimization of deep brain stimulation for gait dysfunction

Despite the success of deep brain stimulation (DBS) in treating many symptoms of Parkinson Disease (PD), treatment of gait dysfunction has remained challenging. Recent work suggests that gait dysfunction may require activation of different electrode contacts connected to different brain circuits than those used to treat other PD symptoms. In this study we developed an algorithm to select DBS contacts and parameters based on the location of implanted electrodes that maximizes overlap of the modeled stimulation volume with brain circuitry associated with gait dysfunction. We trained and tested this algorithm using two independent cohorts of Parkinson STN DBS patients who had previously undergone clinical optimization of their DBS settings (training n = 44; test n = 100). We found that the gait-optimized stimulation volumes differed greatly from the clinically selected stimulation volumes, with use of different electrode contacts in over 85% of patients. Increased similarity between the gait-optimized stimulation volume and clinical stimulation volume was correlated with gait improvement after DBS. To illustrate how this information might be used clinically, we reprogrammed six PD patients from their clinical settings to the gait-optimized settings, and found all patients reported subjective gait improvement. These findings suggest that the optimal DBS settings for gait differ from clinically selected DBS settings and that a connectome-based algorithm might help guide DBS reprogramming to improve gait function.