Document resource
Introduction Deep brain stimulation (DBS) for dystonia is effective but programming optimisation can take months. Local field potentials (LFPs) recorded by the Medtronic Percept device may provide biomarkers to guide stimulation. This study will prospectively evaluate whether chronic LFP profiles correlate with clinical outcomes and can inform DBS programming strategies.Methods and analysis LFP-DYT is a single-centre, multi-phase observational study at Newcastle upon Tyne National Health Service (NHS) Foundation Trust. An internal pilot (Cohort 1) will refine recording workflows, followed by Cohort 2 (traditional programming with LFP recordings) and Cohort 3 (LFP-informed programming). 20–25 adults with primary dystonia undergoing globus pallidus internus DBS will be recruited. The study combines chronic LFP sensing with neurophysiology (electromyography, electroencephalography), motor inhibition testing (stop-signal reaction time), patient-reported outcomes and wearable sensor monitoring (STAT-ON) to provide a comprehensive multi-modal assessment framework. Primary outcome: reproducibility of alpha–theta frequency LFP peaks and concordance with optimal stimulation site. Secondary outcomes include stimulation and medication effects on LFP profiles, clinical improvement (Toronto Western Spasmodic Torticollis Rating Scale-2 (TWSTRS-2), Burke-Fahn-Marsden Dystonia Rating Scale (BFMDRS)) and beta-band activity as a marker of stimulation-related bradykinesia. Analyses will be descriptive and exploratory. Feasibility outcomes, including recruitment rates, retention and data completeness, will inform design and power calculations for future multi-centre trials.Ethics and dissemination The study has NHS Research Ethics Committee approval from the East Midlands—Derby Research Ethics Committee (REC reference: 24/EM/0246; IRAS ID: 337426). All participants will provide informed consent. Data will be pseudonymised and stored on secure NHS servers. Results will be disseminated via peer-reviewed publications, conferences and participant summaries. De-identified data and analysis code will be available on reasonable request.Trial registration number NCT07309133.