Modeling DBS Tolerability in Tourette Syndrome: Unilateral and Bilateral Effects Characterization

Abstract

Deep Brain Stimulation (DBS) can reduce symptoms in medically refractory Tourette syndrome (TS), but programming optimization is constrained by stimulation-induced side-effects. This study uses thresholding data from 8 participants with TS to model DBS tolerability across targets centromedian thalamus (CM); anterior globus pallidus internus (aGPi), hemispheres, and contact levels, to quantify unilateral and bilateral side-effects. Laterality was evaluated in a unilateral (n=8) and a bilateral (n=4) subset using stimulation-threshold comparisons. Stimulation amplitudes increased incrementally until the first side-effect, and trials without side-effects were treated as ‘censored’ during thresholding.  This study modeled unilateral amplitude-to-side-effect using target, hemisphere, and 4-contact levels as fixed effects and a gamma frailty term to capture subject heterogeneity. The aGPi showed higher relative tolerability than the CM (hazard ratio = 0.6), with contact-level dependence and substantial between-subject variability. Directional mode testing was heavily censored at the 2-mA ceiling, preventing robust modeling. Matched analysis indicated lower bilateral thresholds, particularly at deeper contacts, with wide 95% confidence intervals. These findings support subject-specific tolerability profiling to inform more data-driven DBS programming practices and motivate larger cohort studies.