Increased variance in second electrode accuracy during deep brain stimulation and its relationship to pneumocephalus, brain shift, and clinical outcomes: a retrospective cohort study

ABSTRACT Introduction Accurate placement of deep brain stimulation electrodes within the intended target is believed to be a key variable related to outcomes. However, methods to verify electrode location are not universally established. Research Question The aim of this study was to determine the feasibility of post-op lead localisation in clinical practice and its utility to audit our own DBS accuracy. Material and Methods A retrospective cohort study was performed of a consecutive series of patients with Parkinson’s disease who underwent deep brain stimulation of either the globus pallidus internus (GPi) or subthalamic nucleus (STN) between 2016 and 2019. Image processing was performed using the Lead-DBS toolbox. Institutional ethical approval was granted as a review of service. Results In total 38 participants met the inclusion criteria. Electrode localisation was completed in 79%. Clinical outcomes included improvement in UPDRS III of 46% and PDQ39 of 32%. Overall electrode accuracy was 0.22 +/- 0.4mm for all electrodes to the main nucleus with 9 (12%) outliers but only 3 (4%) electrodes out with 2mm from the intended target. Accuracy was worse for the second electrode implanted and in the GPi but was not affected by pneumocephalus or brain shift. Neither clinical outcomes nor the volume of activated tissue was affected by electrode accuracy. Discussion and Conclusions A neuroimaging approach to electrode localisation allows qualitative appraisal of targeting accuracy and is feasible with routine clinical data. Such methods are complimentary to traditional co-ordinate approaches and lend themselves to developing large, collaborative, and quantitative projects. HIGHLIGHTS Overall electrode accuracy was 0.22 +/- 0.4mm with only 3 (4%) electrodes out with 2mm from the intended target Accuracy was significantly worse in the GPi versus the STN and on the second side implanted Inaccuracy occurred in the X (lateral) plane but was not related to pneumocephalus or brain shift