Description

1. Background and Rationale Parkinson's disease (PD) is a common neurodegenerative disorder affecting 2% of individuals over 65 years. Beyond classical motor symptoms like bradykinesia, rigidity, and tremor, PD is increasingly recognized for its debilitating non-motor symptoms (NMS), which include depression, anxiety, fatigue, apathy, and bradyphrenia. These NMS fluctuate alongside motor symptoms, particularly in later disease stages when dopaminergic treatments become pulsatile. Neuropsychiatric fluctuations (NMF) significantly reduce patients' quality of life, yet their pathophysiology is not well understood.

While EEG studies have characterized motor fluctuations and impulse control disorders through frequency-specific changes, few studies have systematically examined the EEG or intracranial electrophysiological substrates of NMF. STN-DBS offers the opportunity to record local field potentials (LFPs) directly from the subthalamic nucleus (STN). Beta oscillations in STN have been consistently associated with akinesia, while recent studies suggest that alpha and gamma band activities may relate to non-motor and neuropsychiatric features.

Technological advances, such as Medtronic's PERCEPT™ PC DBS system, now enable chronic LFP recordings postoperatively. Coupling these with high-density EEG enables real-time exploration of cortical-subcortical interactions. Our pilot data show distinct EEG dynamics between motor and neuropsychiatric fluctuations following levodopa challenge, supporting a differential spatio-spectral signature for these symptoms. 2. Objectives

Primary Objective:

To characterize the temporal dynamics of cortical EEG and STN LFP oscillatory activity and coherence during the acute levodopa response, and their correlation with motor and non-motor fluctuations in PD.

Secondary Objectives:

To determine the influence of STN-DBS on EEG/LFP biomarkers. To investigate motor and non-motor temporal profiles after levodopa administration.

To characterize phase-coupling changes between STN and cortical networks. To use machine learning to identify spectral EEG/LFP biomarkers predicting ON/OFF states.

To explore anatomical correlates of LFP signals using post-op imaging and electrode localization. 3. Design and Methodology

Design

Prospective, observational, exploratory two-center study (HUG Geneva and CHUV Lausanne).

Participants

30 patients with idiopathic PD (per UK Brain Bank Criteria), in the fluctuations stage, treated or candidates for STN-DBS, on dopaminergic therapy, and without dementia.

Timeline

Total duration: \~36 months

24 months for recruitment and data collection 12 months for data analysis and dissemination

Study Visits:

Visit 1: Baseline clinical and neuropsychological assessment in ON-drug/ON-DBS condition.

Visit 2: Modified levodopa challenge with 5 successive conditions:

OFF-drug/ON-DBS OFF-drug/OFF-DBS Transition period post-levodopa (serial assessments) ON-drug/OFF-DBS ON-drug/ON-DBS

For each condition: 10-min resting-state EEG, 5-min STN-LFP, and brief clinical assessment.

Levodopa Administration:

150% of usual morning dose (Madopar® LIQ), to ensure supramaximal response.

Neurophysiological Recording:

High-density EEG (256 channels, Philips EGI) STN-LFP using Medtronic PERCEPT™ system Synchronization via EMG trigger pulse

Clinical Assessments:

MDS-UPDRS Part III (motor) Neuropsychiatric Fluctuation Scale (tablet-based) Stop-Signal Task (SSRT) Bradyphrenia scale

4\. Inclusion/Exclusion Criteria

Inclusion

Idiopathic PD diagnosis DBS candidate or post-DBS (within 4-8 weeks) Motor or non-motor fluctuations MOCA \>24/30 On stable dopaminergic therapy

Exclusion

Age \>80 Dementia, active psychosis, unstable disease Severe OFF symptoms precluding testing Legal guardianship or participation in other drug studies

6\. Impact and Perspectives This study addresses an unmet need in PD: the poor understanding and characterization of neuropsychiatric fluctuations. By correlating motor and non-motor symptom dynamics with LFP and EEG signatures, it will provide pathophysiological insights and biomarkers to guide future adaptive DBS (aDBS). It will also inform clinical management of NMFs and open avenues for biomarker-driven neuromodulation strategies.

The collaboration between two major Swiss centers ensures feasibility and access to a representative PD cohort. Use of standardized procedures and machine learning ensures translatability to larger future trials.

Ultimately, this project aims to bridge the neurophysiological understanding of motor and non-motor circuits in PD, enhancing the personalization of neuromodulatory treatments.