Convergent mapping of a tremor treatment network

Abstract

Tremor occurs in various forms across diverse neurological disorders, including Parkinson's disease and essential tremor. While clinically heterogeneous, converging evidence suggests a shared brain network may underlie tremor across conditions. Here, we empirically define such a network using four modalities: lesion locations, atrophy patterns, EMG-fMRI, and deep brain stimulation outcomes. We show that network connectivity robustly explains clinical outcomes in independent cohorts undergoing deep brain stimulation of the subthalamic nucleus for Parkinson's disease and the ventral intermediate nucleus for essential tremor. Maps from each cohort accounted for outcomes in the respective other, supporting a disorder-independent tremor network. A multimodal agreement map revealed consistent substrates in the primary motor cortex and motor cerebellum. To validate the network, we test its predictive power in a third, independent cohort treated with pallidal stimulation for Parkinson's disease. Our findings define a robust, cross-condition tremor network that may guide both invasive and noninvasive neuromodulation strategies.

Fig. 1. Lesion-derived tremor-relief map connectivity correlates with clinical improvement in deep brain stimulation (DBS) cohorts.

A The panel illustrates 11 published cases of patients who experienced tremor alleviation following beneficial lesions due to stroke. Using lesion network mapping, a tremor-relief network was identified (bottom). Colors represent the probability (p) of voxels associated with tremor relief. B Electrode localizations from subjects in the current study, with blue representing VIM patients and orange representing STN patients. The bottom plot shows a positive Spearman correlation (two-sided) between relative clinical improvement (measured as ratios of the maximum achievable score for UPDRS-III tremor scores or FTM) following DBS and each DBS electrode’s connectivity to the lesion network map. A cohort regressor was applied to control for potential confounding effects related to differences between patient groups. No adjustment for multiple comparisons was applied. The shaded area indicates the 95% confidence interval. The red circle highlights an example of a volume of activated tissue (VTA) with low connectivity, shown alongside an example with high connectivity (green circle) to the lesion network map in (C). Transversal and coronal slices from the BigBrain atlas were used. Cerebellar flatmaps were created using the SUIT toolbox. DBS Deep Brain Stimulation, FTM Fahn-Tolosa-Marin Tremor Rating Scale, GPi Globus pallidus internus, STN Subthalamic nucleus, UPDRS-III Unified Parkinson’s Disease Rating Scale, part III, VIM Ventral intermediate nucleus of the thalamus.

Fig. 2. fMRI derived tremor activity map connectivity correlates with clinical improvement in deep brain stimulation (DBS) cohorts.

A This panel illustrates tremor-amplitude-related activity within the cerebello-thalamo-cortical circuit, derived from individual concurrent EMG-fMRI data in 22 cases. Colors represent z-scores (z). Cerebellar flatmaps were created using the SUIT toolbox. B The plot shows a positive Spearman correlation (two-sided) between relative clinical improvement (measured as ratios of the maximum achievable score for UPDRS-III tremor scores or FTM) following DBS and each DBS electrode’s connectivity to the fMRI-derived tremor map. Orange and blue dots represent electrodes from the STN and VIM cohorts, respectively. A cohort regressor was applied to control for potential confounding effects related to differences between patient groups. No adjustment for multiple comparisons was applied. The shaded area indicates the 95% confidence interval. Illustrations were created using Affinity Designer 2 (Serif Ltd., Nottingham, UK). DBS Deep Brain Stimulation, FTM Fahn-Tolosa-Marin Tremor Rating Scale, STN Subthalamic nucleus, UPDRS-III Unified Parkinson’s Disease Rating Scale, part III, VIM Ventral intermediate nucleus of the thalamus.

Fig. 3. Cross-validation and permutation analysis between DBS cohorts.

A Correlation maps for both the STN-DBS cohort (top left) and the VIM-DBS cohort (bottom right). These maps contain group-level correlation coefficients consisting of each patient’s hemisphere connectivity profile with their clinical tremor improvement. The connectivity of VIM-DBS patients to the STN connectivity profile correlated positively with clinical improvement (top) in a two-sided Spearman correlation, and vice versa for STN patients with VIM profiles (bottom). No adjustment for multiple comparisons was applied. The shaded area indicates the 95% confidence interval. Cerebellar flatmaps were created using the SUIT toolbox. B Results of a permutation analysis for VIM connectivity maps predicting STN-DBS outcomes (left), STN connectivity maps predicting VIM-DBS outcomes (right), and both combined (middle). DBS Deep Brain Stimulation, FTM Fahn-Tolosa-Marin Tremor Rating Scale, STN Subthalamic nucleus, UPDRS-III Unified Parkinson’s Disease Rating Scale, part III, VIM Ventral intermediate nucleus of the thalamus.

Fig. 4. DBS Agreement map.

A An agreement map was calculated by retaining connections that were positively or negatively associated with optimal clinical outcome in both the STN and the VIM cohorts, respectively. Absolute values of retained voxels were multiplied and standardized (z-scored). Cerebellar flatmaps were created using the SUIT toolbox. B Spearman correlation (two-sided) between connectivity of DBS sites to the agreement map and clinical improvements (R = 0.41). This analysis was circular and meant to demonstrate the degree of fit between model and data, i.e., the maximal possible amount of variance that could be explained by this agreement map in the data used to create it. Due to circularity, we avoid reporting the p-value of this correlation. However, correlations remained significant when subjecting the analysis to fivefold cross-validation, which circumvented circularity (rho(135) = 0.20, p = 0.018, 95% CI [0.04, 0.36]). No adjustment for multiple comparisons was applied. The shaded area indicates the 95% confidence interval. DBS Deep Brain Stimulation, STN Subthalamic nucleus, VIM Ventral intermediate nucleus of the thalamus.

Fig. 5. Multimodally-informed convergent tremor map anticipates clinical outcomes in out-of-sample GPi cohort.

A The lesion-derived map, EMG-fMRI-derived map, VIM-STN (agreement) map and an ET-specific atrophy map were superimposed to create the displayed multimodally-informed convergent tremor map, with z-scores visualized. Cerebellar flatmaps were created using the SUIT toolbox. B In an out-of-sample cohort of 31 analyzed hemispheres from PD patients with GPi-DBS, lead localizations were conducted using Lead-DBS (localizations shown in the top right image). Spearman correlation (two-sided) between connectivity of DBS sites to the convergent tremor map and MDS-UPDRS-III tremor improvements. No adjustment for multiple comparisons was applied. The shaded area indicates the 95% confidence interval. ET Essential tremor, DBS Deep Brain Stimulation, GPi Globus pallidus internus, STN Subthalamic nucleus, UPDRS-III Unified Parkinson’s Disease Rating Scale, part III, PD Parkinson’s disease, VIM Ventral intermediate nucleus of the thalamus.