First hemisphere implantation reshapes thalamic stimulation maps in Essential Tremor

Deep brain stimulation has become one of the most reliable surgical therapies for severe Essential Tremor, yet the way it quiets the pathological tremor circuit remains one of neuromodulation’s stubborn mysteries. A team of researchers from Switzerland and France has now uncovered a striking asymmetry in how the two hemispheres of the thalamus respond to stimulation during surgery, and their findings suggest that the very act of implanting the first electrode leaves a temporary imprint on the brain that reshapes the stimulation map of the second hemisphere. The study, published in Frontiers in Neurology, analyzed intraoperative data from 23 patients and revealed that the tremor-suppressing region in the second implanted hemisphere was roughly twelve times larger than in the first, a difference the authors attribute to a transitory lesioning effect triggered by the initial implantation.

The research focused on the ventrointermediate nucleus of the thalamus, or Vim, the canonical target for Essential Tremor deep brain stimulation. Chronic high-frequency stimulation of this nucleus and its neighboring subthalamic region modulates a distributed network, notably the cortico-ponto-cerebello-rubro-thalamo-cortical circuit, which has long been implicated in the generation of pathological oscillations. Although the clinical benefit of Vim stimulation is well established, the precise neuronal mechanisms underlying the improvement remain incompletely deciphered, and the contribution of bi-hemispheric modulation is particularly unclear. Ipsilateral effects, in which stimulation of one thalamus influences tremor on the same side of the body, have been reported but are not always apparent, hinting that the two hemispheres do not operate as independent, mirrored modules.

To probe this hemispheric organization, the team developed a data-driven probabilistic mapping approach that linked simulated stimulation volumes, known as Volumes of Tissue Activated, to symptom improvement and side effects for each hemisphere separately. Unlike conventional probabilistic studies that pool data from both sides by mirroring one hemisphere onto the other, this framework preserved the spatial identity of each hemisphere, allowing genuine inter-hemispheric comparisons. The investigators retrospectively analyzed intraoperative stimulation data collected between 2008 and 2017 at the University Hospital Clermont-Ferrand in France, where all 23 patients underwent bilateral Vim lead implantation under local anesthesia. The left hemisphere was systematically explored and implanted first, with the right hemisphere following approximately one hour later, a consistent surgical sequence that provided a natural experimental contrast between a naïve hemisphere and one already carrying an implanted electrode.

The intraoperative exploration itself was remarkably dense. Using rigid semi-microelectrodes with a contact diameter of 0.5 millimeters, the surgeons tested stimulation positions in one-millimeter increments along trajectories averaging 14 millimeters in length, one central track optimized for the Vim and Forel’s field H, and a secondary track two millimeters away, mostly posterior, exploring the ventrocaudal nucleus of the thalamus in search of paresthesia. At each position, the stimulation current was increased from 0.2 to 3 milliamperes in 0.2-milliampere steps, and the physician rated tremor improvement on a five-point Likert scale converted into linear percentage data. In total, 605 stimulation spots were tested, corresponding to an average of twelve sites per hemisphere per patient, and 696 stimulation tests were recorded. Among these, roughly 23 percent elicited side effects, with paresthesia being the most common, followed by dysarthria, oculomotor disturbances, vegetative symptoms, and pyramidal signs.

Each stimulation spot was then categorized according to its therapeutic window, defined as the difference between the minimum current inducing clinical improvement and the threshold at which side effects occur, combined with the presence or absence of side effects at the upper stimulation limit. Six categories emerged, ranging from clinically unusable sites where stimulation immediately induced side effects or produced no effect at all, to the optimal category 3, considered the clinical sweet spot, characterized by a therapeutic window of at least 1.5 milliamperes and no side effects. This categorization framework, combined with the probabilistic volumes, provided a refined, clinically relevant lens for understanding where stimulation produces benefit, where it produces harm, and where the two effects spatially converge or diverge.

The computational pipeline behind the probabilistic maps was equally sophisticated. Patient-specific brain conductivity models were generated from individual anatomical MRI scans, assigning tissue classes of cerebrospinal fluid, gray matter, and white matter with distinct but isotropic conductivities. Electric field simulations were run for each patient and each tested stimulation setting, and the resulting field meshes were thresholded at 0.2 volts per millimeter to obtain the Volumes of Tissue Activated. A total of 696 VTAs were computed, 381 in the left hemisphere and 315 in the right, and these were transformed using a non-linear approach to a group-specific MRI template aligned to the anterior commissure-posterior commissure plane, enabling group-level statistical analysis. The voxels associated with significant improvement were selected using a Bayesian one-sample t-test with a null hypothesis that improvement was less than 75 percent, a threshold chosen as clinically relevant, and a Bayes factor of 10 as the significance criterion, comparable to a p-value of 0.01. The side effect volumes were computed with a binomial test, since the information was binary, and corrected for multiple comparisons using the False Discovery Rate.

The results revealed a pronounced asymmetry between the two hemispheres. In the first implanted, left hemisphere, the Probabilistic Volume of Improvement was confined to the infero-posterior part of the Vim, reaching the subthalamic tegmental fields traversed by rubro-thalamo-cortical fibers, while the Probabilistic Volume of Side Effects sat above, posterior, and lateral to it, spanning both the Vim and the ventrocaudal nucleus. In the second implanted, right hemisphere, the improvement and side effect volumes overlapped substantially, intersecting the Vim and the ventrocaudal nucleus in their intermediate parts. The volume of the improvement region in the right hemisphere reached 457 cubic millimeters, about twelve times larger than the 37 cubic millimeters in the left, while the side effect volume followed the opposite trend to a lesser extent, at 388 cubic millimeters in the left versus 285 cubic millimeters in the right. Stimulation spot clusters, identified using a density-based spatial clustering algorithm, confirmed these differences, with optimal positions in the left hemisphere distributed more medially and inferiorly than in the right, and the clusters in the right hemisphere occupying the intermediate and superior parts of the Vim.

The interpretation the authors offer is that the first lead implantation produces a transitory, tangible lesioning effect, a reversible phenomenon linked to the mechanical and electrical perturbation of the tissue during exploration and implantation. This quasi-experimental lesioning of the left hemisphere achieved near-complete tremor suppression within a larger, functionally undifferentiated central region of the right Vim, encompassing both anti-tremor and side effect-related areas. When the surgeons subsequently explored the right hemisphere, the stimulation map they encountered was no longer that of a naïve thalamus but of one already functionally rearranged by the contralateral implantation. The confined area of improvement revealed in the initially explored hemisphere widened the contralateral improvement volume, resulting in the overlap between improvement and side effect regions that characterizes the second hemisphere. This expansion likely accounts for the masking of hemisphere-specific lesion effects when data from both hemispheres are averaged, a practice common in probabilistic mapping studies.

The findings also resonate with pioneering stereotactic work from the 1960s, in which the improvement volume of the first implanted hemisphere corresponds to the camptotomy site, while the overlay of improvement and side effect volumes in the second hemisphere corresponds to both the thalamo-capsular and the relay-nuclei sites. The left improvement volume is clearly related to the cortico-ponto-cerebello-rubro-thalamic pathway, whereas the right combined volume is more related to the cortico-thalamic and thalamo-cortical systems, and both are also related to the cortico-subthalamic fibers. These anatomical relationships fit within the broader circuits implicated in Essential Tremor syndrome, including the cerebellar and thalamic relay systems whose dysfunction generates the pathological oscillations that stimulation suppresses. The colocation of Vim neuron clumps and fiber pathways in the second hemisphere may explain why the improvement and side effect volumes converge there, since the same stimulation current activates both relay neurons and passing fibers.

The clinical implications of this work are potentially far-reaching. The authors suggest that the large probabilistic volume in the second explored hemisphere should be considered when positioning the deep brain stimulation lead, given that the deeper contacts may be effective, mirroring the improvement volume of the first hemisphere once the transitory lesioning effect wears off. During chronic stimulation, both electrodes exert contra- and ipsilateral effects that should incorporate the geometric and functional modifications of stimulation revealed by this study. A possible development could be the application of novel chronic stimulation strategies, such as interleaved interhemispheric multi-energy and multi-contact stimulations, in which stimulation intensity alternates between hemispheres, with one receiving a conventional therapeutic amplitude while the contralateral hemisphere simultaneously receives a lower amplitude, followed by periodic switching. Future prospective studies with randomized first hemisphere implantation could further investigate whether the lesioning effect is truly hemisphere-independent or whether lateralized functional markers exist across the extended tremor circuitry. The retrospective design and limited sample size of 23 patients warrant caution, though the high density of stimulation spots per patient and the robustness of the Bayesian statistical approach mitigate some of these concerns. As the understanding of bi-hemispheric modulation deepens, the decoupling of first and second hemisphere analyses in acute intraoperative effects may become a standard consideration, not only for Essential Tremor but potentially for other movement disorders such as Parkinson’s disease and dystonia, where the interplay between lesioning and stimulation effects could similarly shape the spatial organization of clinical benefit.

Subject of Research: Intraoperative asymmetry of thalamic stimulation mapping in Essential Tremor: specificity of first hemisphere implantation and transitory lesioning effect

Article Title: Intraoperative asymmetry of thalamic stimulation mapping in Essential Tremor: specificity of first hemisphere implantation and transitory lesioning effect

Article References: Bucciarelli, V., Vogel, D., Coste, J., Chaix, R., Hemm, S., & Lemaire, J.-J. (2026). Intraoperative asymmetry of thalamic stimulation mapping in Essential Tremor: specificity of first hemisphere implantation and transitory lesioning effect. Frontiers in Neurology, 17, Article 1893911. https://doi.org/10.3389/fneur.2026.1893911

Image Credits: AI Generated

DOI: 10.3389/fneur.2026.1893911

Keywords: Intraoperative, asymmetry, thalamic, stimulation, mapping, Essential, Tremor, specificity, first, hemisphere, implantation, transitory