Manipulating and Optimising Brain Rhythms for Enhancement of Sleep

This is a feasibility study that will take place in two UK sites: the John Radcliffe Hospital (Oxford) and the Surrey Sleep Research Centre, University of Surrey. The identification, consent and screening of patients will take place in Oxford, and the sleep assessments will take place in Surrey. A parallel study will also be conducted at US sites (Mayo Clinic, Rochester MA, and UCSF, CA) under the same funding but this will be reviewed by their local IRB. De-identified data may be shared between sites for analyses purposes. This will be done securely in an encrypted fashion while participants will be made fully aware of this.

Participants will be identified during their routine clinic visits with the Functional Neurosurgery team (John Radcliffe Hospital) and invited to participate in the study. Screening of interested potential participants will take into account pre-operative assessment data. Testing will ensue (within a variable period from screening, but within the aforementioned study timeline), over two visits to the Surrey Sleep Research Centre, each consisting of two nights. The maximum duration between consent and first study visit will be three months, while there will be a minimum of two weeks between each study visit to allow for preliminary analyses. Informed consent and baseline sleep recordings will be obtained over the first night of the first visit to Surrey, with the remaining nights consisting of stimulation trials. Questionnaires, daytime vigilance testing, autonomic parameters and cortisol levels will be collected. No long-term follow-up is planned at this stage.

With regards to data collection processes, the population of patients visiting the Surrey site will have already been implanted with a DBS system (such as an Activa PC/PC+S (Medtronic) or an Abbot/St Jude device). The investigators will leverage rules from the current AASM guidelines, technologies and methods for manual and automated sleep scoring using standard scalp polysomnography as well as novel methods we previously developed for sleep scoring using intracranial electrophysiology. To avoid contamination of the electrophysiology monitoring amplifiers by the stimulation protocol, a combination of bipolar stimulation and sampling-stimulate rate interactions will be used. For example, we will apply know-how from the Activa PC+S work to place any residual artefacts into bands of little physiological interest (DC, 50/60Hz, etc.). A preliminary assessment of amplifier performance will provide assurance that we will be able to maintain sleep staging accuracy over the course of our experiments.

For perturbations, the CE-marked clinician programmer will be used (especially at the open-loop stage) or a research variant thereof. In this study, high-frequency (HF) is specified at a stimulation range from 41Hz to 250Hz, while low frequency (LF) will be defined as 2-40Hz. All perturbations will be within the clinical range used and approved for DBS patients. A combination of subjective and objective measurements of sleep quality and efficiency to measure the impact of any sleep perturbations will be used.

The following objective metrics will be derived from visual/manual polysomnography (gold standard) \& automated algorithm sleep staging:

1. Total sleep time.
2. Number of sleep cycles (switches of non-rapid eye movement (NREM) and rapid eye movement (REM) sleep).
3. Initial sleep and REM latencies (Time from beginning of study to the first stage of sleep, and from beginning of sleep to first REM sleep onset).
4. Duration of N1, N2, N3, and REM stages, and the duration of NREM (sum of N1,N2,N3)stages.
5. Sleep efficiency (100*total sleep time/time in bed).
6. Number of awakenings during night, including arousals per hour and duration of time spent awake after initial sleep onset (i.e., wake after sleep onset time)
7. Apnoea-hypopnea and respiratory disturbance indices (the frequency of stop or reduced breathing episodes, and respiratory arousals per hour)
8. Periodic limb movement indices (the frequency of periodic leg movements of sleep per hour)
9. Measures of arousal during sleep
10. behavioural macrostate observations such as eye opening or blinking that can be assessed in video recordings
11. microarousal observations defined exclusively by cortical fast frequency shifts (ie, >14 Hz rhythms) lasting 3 seconds or longer, with or without added autonomic measures of tachycardia.
12. Spectral composition of the sleep stage specific EEG over the 0.25-32 Hz frequency range Salivary samples to characterize awakening levels of cortisol will be obtained in the Surrey Sleep Research Centre at the end of each night.

Questionnaires will be administered to patients such as those listed below (not all necessary):

Upon enrolment/prior to the participant's visit:

• Pittsburgh Sleep Quality Index (PSQI) (Buysse et al 1989)
• Movement Disorder Society-Sponsored Revision of the Unified Parkinson's Disease Rating Scale (MDS-UPDRS) (last amendment 2019, original article Goetz et al 2008)
• Parkinson's Disease Sleep Scale (PDSS) (Trenkwalder et al. 2011)
• Patient sleep diary (bedtime, awakening time, arousals/sleep disturbances, perceived sleep quality)

During the visits to the Surrey Sleep Research Centre:

• Karolinska Sleepiness Scale (KSS) (Akerstedt and Gillberg, 1990)
• Samn Perelli Fatigue Scale (Samn and Perelli, 1982)
• Profile of Mood States Scale (POMS) (McNair, Lorr and Doppleman, 1971)

With regards to collecting data relevant to cognitive performance, we will focus on assessments of sleep inertia as measured by performance characteristics assessed by tasks listed below:

• Digit Symbol Substitution Test (DSST) (as in Boyle et al. 2012)
• Psychomotor Vigilance Task (PVT) (as in Santhi et al. 2013)
• N-back task (Santhi et al. 2013)
• Karolinska Drowsiness Test (KDT) (Akerstedt and Gillbert, 1990)