Adaptive Neurostimulation to Restore Sleep in Parkinson’s Disease

Description

Although STN-DBS is routinely used to treat PD motor symptoms, several studies have reported that STN-DBS also provides benefit for sleep dysregulation through normalization of sleep architecture. In our previous work, using local field potentials (LFP) recorded from STN DBS electrodes implanted for the treatment of PD, unique spectral patterns in STN oscillatory activity were identified that correlated with distinct sleep cycles, offering insight into sleep dysregulation. These findings were used to construct an Artificial Neural Network (ANN) that can accurately predict sleep stage. Building on this work with the use of new DBS battery technology that allows exploration of potential biomarkers and prototyping of closed-loop algorithms, the investigators will test the hypothesis that STN-a highly interconnected node within the basal ganglia- contributes to the regulation and disruption of human sleep behavior and can be manipulated for therapeutic advantage.

This is the first part, Aim 1, of a two-part study. Investigators will enroll 20 subjects for Aim 1 of this study and 20 subjects for Aim 2, with 10 subjects enrolled at each clinical site for each aim (University of Nebraska Medical Center and Stanford University Medical Campus). In Aim 1, subjects will undergo standard-of-care STN DBS lead implantation surgery for the treatment of PD. They will return 3 weeks later to the in-patient Sleep Lab for 3 nights of STN LFP recordings with concurrent PSG, EMG, EOG, actigraphy, and video-EEG. The first two nights of recording will be used to establish a physiological sleep baseline for each patient. The third night of recording will involve sub-clinical thresholds of stimulation in all subjects, in an effort to favorably alter sleep-stage duration, so that NREM and REM-3 are prolonged. As a secondary outcome, subjects will be asked to complete a sleep questionnaire for all three nights, sleep during which stimulation occurred will be compared to the preceding two nights. Data collected during all three nights of recordings will be used to predict sleep stage identity from the LFPs recorded within STN, with the ground truth for each sleep stage provided by sleep-expert evaluated PSG. These data will also be used to identify the optimal sub-clinical threshold current amplitude and sleep-stage timing for adaptive stimulation to improve sleep. The stimulation algorithm developed in Aim 1 will be implemented in the second part of the study, Aim 2, to provide adaptive stimulation to subjects during nighttime sleep, over the course of 3 weeks of in-home sleep.