Engaging dystonia networks with subthalamic stimulation.
Konstantin Butenko,Clemens Neudorfer,Till A. Dembek,Barbara Hollunder,Garance M. Meyer,Ningfei Li,Simon Oxenford,Bahne H. Bahners,Bassam Al-Fatly,Roxanne Lofredi,Evan M. Gordon,Nico U.F. Dosenbach,Christos Ganos,Mark Hallett,Philip A. Starr,Jill L. Ostrem,Yiwen Wu,ChenCheng Zhang,Michael D. Fox,Andreas Horn
DOI: https://doi.org/10.1101/2024.05.24.24307896
2024-05-25
Abstract:Deep brain stimulation is a viable and efficacious treatment option for dystonia. While the internal pallidum serves as the primary target, more recently, stimulation of the subthalamic nucleus (STN) has been investigated. However, optimal targeting within this structure and its complex surroundings have not been studied in depth. Indeed, multiple historical targets that have been used for surgical treatment of dystonia are directly adjacent to the STN. Further, multiple types of dystonia exist, and outcomes are variable, suggesting that not all types would profit maximally from the exact same target. Therefore, a thorough investigation of the neural substrates underlying effects on dystonia symptoms is warranted.
Here, we analyze a multi-center cohort of isolated dystonia patients with subthalamic implantations (N = 58) and relate their stimulation sites to improvement of appendicular and cervical symptoms as well as blepharospasm. Stimulation of the ventral oral posterior nucleus of thalamus and surrounding regions was associated with improvement in cervical dystonia, while stimulation of the dorsolateral STN was associated with improvement in limb dystonia and blepharospasm. This dissociation was also evident for structural connectivity, where the cerebellothalamic, corticospinal and pallidosubthalamic tracts were associated with improvement of cervical dystonia, while hyperdirect and subthalamopallidal pathways were associated with alleviation of limb dystonia and blepharospasm. Importantly, a single well-placed electrode may reach the three optimal target sites. On the level of functional networks, improvement of limb dystonia was correlated with connectivity to the corresponding somatotopic regions in primary motor cortex, while alleviation of cervical dystonia was correlated with connectivity to the recently described 'action-mode' network that involves supplementary motor and premotor cortex. Our findings suggest that different types of dystonia symptoms are modulated via distinct networks. Namely, appendicular dystonia and blepharospasm are improved with modulation of the basal ganglia, and, in particular, the subthalamic circuitry, including projections from the primary motor cortex. In contrast, cervical dystonia was more responsive when engaging the cerebello-thalamo-cortical circuit, including direct stimulation of ventral thalamic nuclei.
These findings may inform DBS targeting and image-based programming strategies for patient-specific treatment of dystonia.