Direct Imaging Guided Lateral Trajectory for Globus pallidus internus Deep Brain Stimulation in Dystonia: Long-term Outcomes, Sweetspot Localization, and Associated Pathways.

Abstract

Introduction: Globus pallidus internus (GPi) deep brain stimulation (DBS) is an established treatment for dystonia; however, clinical outcomes for the technique remain variable despite technically accurate implantation. This study examined whether direct imaging guided lateral trajectory targeting can reproducibly position the stimulation field nearer a therapeutic zone in GPi DBS to yield durable clinical benefit in the long-term.

Methods: We retrospectively analyzed patients with idiopathic isolated or hereditary isolated dystonia treated with bilateral GPi DBS at a single center. Clinical outcomes were assessed longitudinally using the Burke Fahn Marsden Dystonia Rating Scale (BFMDRS). Targeting relied on fused stereotactic computed tomography with T2-weighted and susceptibility weighted magnetic resonance imaging. Final field modeling was anchored to the most recent clinically stable follow-up to represent the chronic therapeutic state. Native space morphometry, postoperative lead reconstruction, volume of tissue activation modeling, voxel wise sweetspot mapping, boundary focused overlap analyses, and normative connectomic analyses were performed to verify the results.

Results: Mean preoperative BFMDRS scores improved from 30.76 ± 31.16 to 10.56 ± 13.23 (Wilcoxon signed rank p < 0.001), corresponding to a mean patient level improvement of 67.0% ± 19.8% after a mean follow-up of 10.04 ± 6.01 years, with individual follow-up extending to 19 years. Direct imaging guided targeting produced a reproducible 48° to 49° lateral trajectory within the posteroventral (PV) GPi. Broadly overlapping sweetspot clusters localized near the dorsolateral GPi-GPe (Globus pallidus externus) border. Greater GPi-GPe border overlap volume showed a positive association with greater clinical improvement (Spearman ρ = 0.43, p = 0.025), and normative fiber filtering identified ansa lenticularis, pallidothalamic sensorimotor pathway, associated with clinical benefit (Spearman ρ = 0.86; p = 2 × 10⁻⁴), with the strongest named tract match to the dominant pallidothalamic system.

Conclusion: Direct imaging guided lateral trajectory targeting was associated with durable clinical benefit and a reproducible stimulation pattern near the GPi-GPe interface. Sweetspot mapping and normative fiber analysis support this approach as a practical refinement of posteroventral GPi targeting rather than a change in target nucleus.