Clinical Outcomes with Chronic Adaptive Versus Conventional
DBS
in Parkinson's Disease: A Pilot Randomized Crossover Trial
Jun Tanimura, Takehiro Yako, Takao Hashimoto Abstract
Background
Currently available adaptive deep brain stimulation (aDBS) modulates stimulation guided by real‐time beta‐band local field potential activity, potentially offering advantages over conventional DBS (cDBS). Clinical evidence under chronic stimulation remains limited.
Objectives
To compare aDBS versus cDBS efficacy in patients with Parkinson's disease (PD) under chronic stimulation.
Methods
This was a double‐blind, randomized crossover trial in nine PD patients with bilateral subthalamic nucleus DBS. Patients underwent consecutive one‐month periods of cDBS and aDBS after a four‐month postoperative interval. aDBS used dual‐threshold algorithm adjusting amplitude based on subthalamic beta‐band LFP power. Primary outcomes were daily ON/OFF durations and dyskinesia duration under optimized medication. Mixed‐effects ANCOVA estimated treatment effects. Bayesian analyses integrated prior evidence to assess probabilities of clinically meaningful differences.
Results
No statistically significant differences emerged between treatments. Directional effects varied: ON duration favored cDBS, whereas troublesome dyskinesia and UPDRS scores favored aDBS. Bayesian analyses indicated low probabilities that these differences reached clinical importance (29.4% probability for cDBS advantage on ON time, 7.5% for aDBS advantage on UPDRS Part III). Moderate‐to‐substantial between‐patient heterogeneity was observed in comparative treatment effects across outcomes.
Conclusions
The current beta‐band LFP‐guided aDBS approach with dual‐threshold algorithm showed efficacy comparable to cDBS in patients with PD. These findings both call for and provide a basis for appropriately powered follow‐up studies to determine whether, for whom, and with which adaptive stimulation strategies aDBS offers greater benefit than cDBS—a key step in the further development of personalized aDBS.