Optimizing Adaptive DBS for Parkinson's Disease

University of California, Davis35 enrolled

Overview

Parkinson's disease (PD) affects more than 10 million people worldwide and causes progressive motor symptoms such as slowness, stiffness and tremor. For patients whose symptoms are no longer adequately controlled with medication, deep brain stimulation (DBS) can substantially improve motor function. More recently, adaptive DBS (aDBS) has become available. Unlike conventional DBS, which delivers continuous stimulation, aDBS automatically adjusts stimulation in response to brain activity recorded by the implanted device. Current clinical aDBS programming is based primarily on brief recordings obtained during clinic visits. However, Parkinson's symptoms and the underlying brain signals change throughout the day in response to medication, daily activities and other factors. As a result, recordings collected during a single clinic visit may not fully capture the neural activity that best reflects a patient's symptoms in everyday life. The purpose of this study is to determine whether incorporating long-term brain recordings collected during daily life improves adaptive DBS programming and clinical outcomes. Participants will first undergo in-clinic testing to identify brain signals associated with their symptoms. Brain activity and symptoms will then be monitored during everyday life using the sensing capabilities of the implanted DBS device. Participants with suitable brain signals will enter a randomized, blinded crossover study comparing three stimulation approaches: conventional continuous DBS, adaptive DBS programmed using the current clinic-based approach and adaptive DBS programmed using both clinic and at-home recordings. The study will compare the effects of these approaches on motor fluctuations and quality of life. It will also determine how frequently different brain signals occur in people with Parkinson's disease and how well they reflect motor symptoms, providing information that may improve future adaptive DBS therapies.

Adaptive deep brain stimulation (aDBS) automatically adjusts stimulation based on brain activity recorded by an implanted deep brain stimulation (DBS) system. Current clinical programming of aDBS relies on recordings of brain activity collected during brief clinic visits. However, Parkinson's disease symptoms and the underlying brain activity fluctuate throughout the day in response to medication, daily activities and other factors. It is unknown whether recordings collected during routine clinic visits adequately capture the brain signals needed to optimize adaptive DBS for everyday life. Commercially available sensing-enabled DBS systems now allow chronic recording of brain activity during everyday life. These recordings provide an opportunity to characterize patient-specific brain signals under real-world conditions and determine whether incorporating this information improves adaptive DBS programming and clinical outcomes. This study addresses two primary questions: 1. Does more comprehensive in-clinic characterization of brain signals across medication states improve adaptive DBS programming compared with the current clinic-based approach? 2. Does incorporating chronic at-home recordings of brain activity further improve adaptive DBS programming and clinical outcomes? The study consists of two sequential parts. Part 1: Brain signal characterization and adaptive DBS programming All enrolled participants will undergo comprehensive in-clinic testing across medication and stimulation states using a commercially available sensing-enabled DBS system. Brain activity will subsequently be recorded during everyday life using the device's chronic sensing capabilities together with participant-reported symptoms. These data will be used to identify patient-specific brain signals associated with medication state and motor symptoms and to develop individualized, data-driven adaptive DBS settings. Participants with brain signals that track their motor symptoms will be eligible for Part 2. Part 2: Randomized crossover comparison of adaptive DBS programming strategies Eligible participants will enter a randomized, blinded, within-subject crossover study comparing three stimulation approaches: 1. Conventional DBS with a fixed amplitude (cDBS) 2. Adaptive DBS programmed using the current clinic-based optimization procedure (aDBS-Standard) 3. Adaptive DBS programmed using an extended optimization procedure that combines comprehensive in-clinic characterization with chronic at-home recordings of brain activity (aDBS-Extended) During the crossover phase, participants will wear a smartwatch to continuously monitor motor function and complete daily symptom and quality-of-life assessments using a smartphone application while continuing their usual daily activities. These measures will be used to compare the three stimulation approaches under real-world conditions. This study will determine whether incorporating comprehensive in-clinic characterization and chronic at-home recordings of brain activity into adaptive DBS programming improves clinical outcomes compared with current programming methods. It will also characterize the prevalence and clinical relevance of brain signals used for adaptive DBS and provide information to guide future programming strategies for Parkinson's disease.

Interventions

aDBS-Extended (Medtronic Percept™)DEVICE

All parameters are derived from Steps 2-4. Amplitude modulation is based on individualized biomarker thresholds established through in-clinic and at-home testing, within the participant's effective stimulation amplitude range. Delivered via implanted Medtronic Percept™ DBS system using FDA-approved settings.

aDBS-Standard (Medtronic Percept™)DEVICE

aDBS-Standard, with stimulation parameters optimized by an independent clinician using the current standard clinical programming procedure, with iterative programming across multiple visits as needed.

Continuous DBS (cDBS) - Active Control (Medtronic Percept™)DEVICE

Participants receive their clinically optimized continuous DBS settings without adaptive modulation. Adaptive DBS parameters are configured to maintain blinding, but stimulation amplitude remains fixed at the standard cDBS level. Active stimulation is delivered throughout the study via the implanted Medtronic Percept™ DBS system.

Eligibility

Sex: ALLMin age: 18 Years

Medical Language ↔ Plain English

Inclusion Criteria: * Subjects will be 18 and over * Clinical diagnosis of idiopathic Parkinson's disease (PD), consistent with MDS diagnostic criteria * Motor fluctuations in response to medication * Bilateral subthalamic nucleus (STN) DBS using a commercial sensing-enabled device * Stable cDBS settings for ≥3 months prior to enrollment * Stable antiparkinsonian medication regimen for ≥4 weeks prior to enrollment * Capacity to provide informed consent * Ability and willingness to complete study visits and at-home monitoring * For Part 2: Presence of an adequate STN neural signal (e.g., stable LFP feature suitable for biomarker extraction) during screening * Ability to speak and understand English sufficiently to provide informed consent and complete study procedures and assessments without an interpreter. Exclusion Criteria: * Atypical or secondary parkinsonism (e.g., multiple system atrophy, progressive supranuclear palsy, vascular parkinsonism) * Prior brain surgery other than STN DBS * Clinically significant cognitive impairment or dementia, defined as MoCA \< 24 or equivalent, or lacking capacity to provide informed consent * Active psychiatric illness that could compromise safety or participation (e.g., uncontrolled depression, psychosis, severe anxiety disorder) * History of suicidality or suicide attempt within the past year * Clinically unstable medical conditions (e.g., uncontrolled hypertension, advanced cardiac or pulmonary disease) that would increase study risk * Current substance abuse or dependence * Ongoing participation in another interventional trial that could confound outcomes * Pregnancy or plans to become pregnant during the study period * Inability or unwillingness to comply with study procedures (e.g., at-home monitoring, study visits, data collection)

Outcomes

Primary Outcomes

Prevalence and Spectral Properties of STN Local Field Potential Peaks

Prevalence and spectral properties of STN local field potential (LFP) peaks (beta: 13-30 Hz; gamma: 60-90 Hz), and their modulation by dopaminergic medication state.

Time frame: Neural recordings obtained during 1 in-clinic session. For participants in Part 2, selected neural biomarker will be recorded for 14 days, and symptom and medication logs completed on 3 of those days. Events are recorded for predefined symptoms.

Percent Time With Most Bothersome Motor Symptom

Percent time of the most bothersome motor symptom, assessed via electronic self-ratings (Oehrn et al., Nature Medicine, 2024). Primary outcome measures will be derived from nightly self-reports, including time spent with the most bothersome motor symptom (hours/day) and time awake (hours/day). The outcome is calculated as the percentage of waking hours spent with the most bothersome motor symptom (hours with symptom ÷ waking hours × 100).

Time frame: Daily assessment over each 1-week treatment block; repeated 8 times per condition over approximately 6 months.

Severity of Most Bothersome Motor Symptom

Self-reported severity of the most bothersome motor symptom. Symptom severity (0-10 scale). Scores range from 0 (no symptoms) to 10 (worst possible symptoms); higher scores indicate worse outcome.

Time frame: Daily assessment over each 1-week treatment block; repeated 8 times per condition over approximately 6 months

Quality of Life (EQ-5D-5L)

Quality of life will be assessed using the EuroQol 5-Dimension 5-Level questionnaire (EQ-5D-5L), including the EQ Visual Analogue Scale (EQ VAS; range 0-100), with higher scores indicating better perceived health.

Time frame: Daily assessment over each 1-week treatment block; repeated 8 times per condition over approximately 6 months

Secondary Outcomes

Medication-Related Modulation of STN Local Field Potential Spectral Power

Subthalamic nucleus (STN) local field potential spectral power will be compared between ON- and OFF-medication states using cluster-based permutation testing across frequencies. Data will be pooled across movement and rest conditions and across two stimulation levels to identify medication-related spectral differences that are independent of movement state and stimulation level. For significant frequency clusters, the magnitude of the medication-related difference will be quantified using Cohen's d. The frequency or frequency range with the largest Cohen's d will be selected as the patient-specific neural biomarker for subsequent at-home analyses.

Time frame: During the in-clinic testing session

Wearable-Derived Motor Symptoms Across DBS Conditions

During testing of the DBS algorithms, the effect on motor symptoms will be objectively assessed using a validated wrist-worn movement sensor during each treatment condition. A single wearable-derived motor measure will be selected for each participant based on their predefined most bothersome motor symptom (bradykinesia, dyskinesia, or tremor). Tremor, bradykinesia, and dyskinesia are quantified using validated proprietary algorithm-derived scores (reported in device-defined score units) calculated for consecutive 2-minute epochs. Differences in the selected wearable-derived motor measure across conventional DBS, standard adaptive DBS, and extended adaptive DBS will be evaluated at the group level using a linear mixed-effects model, with participant included as a random effect.

Time frame: Continuous wearable monitoring during each 1-week treatment block, repeated 8 times per condition over approximately 6 months.

Regression of Wearable-Derived Motor Measures on Patient-Specific STN Neural Biomarker Activity

The patient-specific subthalamic nucleus (STN) neural biomarker identified during in-clinic testing will be recorded during daily life together with motor symptoms objectively assessed using a validated wrist-worn movement sensor. Tremor, bradykinesia, and dyskinesia are quantified using validated proprietary algorithm-derived scores (reported in device-defined score units) calculated for consecutive 2-minute epochs. Regression analyses will quantify the relationships between patient-specific STN neural biomarker activity and wearable-derived measures of tremor, bradykinesia, and dyskinesia, with the regression coefficients reported as the outcomes.

Time frame: At-home neural and wearable recordings over 14 days.