Anatomical determinants of long-term outcomes in essential tremor deep brain stimulation: Dissociating therapeutic and adverse effects

Background This study was undertaken to identify anatomical determinants of long-term therapeutic and adverse outcomes following bilateral ventral intermediate nucleus (VIM) deep brain stimulation (DBS) in essential tremor (ET), and whether targeting strategies could dissociate beneficial from detrimental effects. Methods Twenty-one patients with ET and DBS were assessed at four years post-implantation. Clinical measures included tremor improvement, tolerance, and cerebellar side effects. Volumes of tissue activated were used to identify significant voxels associated with dependent variable improvement (hot-spots) and worsening (cold-spots) via leave-nothing-out permutation testing, and weighted overlap with atlas structures was calculated. Weighted center-of-mass (COM) distances to atlas structures were computed from z-scored unthresholded hot-spots and cold-spots. Hot and cold COM spatial patterns were compared across dependent variables using principal component analysis. Results Bilateral analyses identified significant hot-spots for tolerance and left-sided tremor improvement and significant cold-spots for cerebellar side effects and tolerance. Tremor improvement, tolerance, and cerebellar side effects were associated with decussating and non-decussating dentato-rubro-thalamic tract stimulation, while VIM grey-matter stimulation was associated with lower tremor improvement but not side effects. Tolerance and cerebellar side effects clustered distinctly from tremor improvement, suggesting different subregional recruitment patterns. Conclusion Long-term therapeutic and adverse outcomes share a common anatomical substrate: the cerebellothalamic pathways. VIM grey matter stimulation may offer a more favorable long-term balance than direct white matter stimulation. Nevertheless, the anatomical overlap of therapeutic and adverse substrates presents fundamental targeting challenges for conventional continuous stimulation, and optimizing long-term outcomes may require novel stimulation paradigms that maintain therapeutic efficacy while minimizing chronic circuit maladaptation.