FV 5 Striato-pallidal connectivity in oscillatory activity of patients with dystonia

Abstract

Introduction: Dystonia is a network disorder in which aberrant signaling maybe associated with an imbalance between direct and indirect pathway activity. Invasive recordings from the internal pallidum (Globus pallidus internus, GPi), the major output structure of the basal ganglia, have described exaggerated oscillatory activity in the low frequency (LF) band (4-12 Hz). Oscillatory activity of the human striatum, as its major input structure, and its role in the generation and propagation of pathological LF activity is not well characterized. We hypothesize that i) the basal ganglia are functionally coupled within the LF band, ii) this connectivity is driven by the striatum and iii) LF activity could either be generated or amplified within the direct pathway. Patients & Methods: Rest recordings of local field potentials were performed in 11 patients undergoing pallidal deep brain stimulation (DBS) for cervical, segmental or generalized dystonia in the 1-7 days following surgery while DBS leads were externalized. The lead model used was DB-2201-45Bc, with 8 circular contacts spanning over 15.5 mm. DBS-leads were localized using the open-source software Lead-DBS and bipolar recordings from adjacent electrodes were assigned accordingly to either the putamen, Globus oallidus externus (GPe) or GPi. Power spectra were calculated for recordings from each structure, as well as the imaginary part of coherence (iCOH) as measure of functional connectivity and granger causality to assess directionality of coupling between structures in the LF band. Results: In all patients, one contact pair per hemisphere was localized in each structure of interest, namely the putamen, GPe and GPi. LF peaks could be detected in all structures of both hemispheres in all patients. LF- activity was coupled between all structures of one hemisphere (putamen-GPi, putamen-GPe, GPi-GPe) as assessed by iCOH and higher than coherence in shuffled data. Specifically the extent of coherence between putamen and GPi showed a positive correlation with motor symptom severity (R=0.63, P=.03). Directionality analyses were less consistent, with a trend towards striatum leading LF activity of internal and external pallidum that did not reach significance (P>0.05). Conclusion: We showed for the first time, that i) LF-activity can also be detected in the striatum of dystonia patients and ii) that striatum, GPe and GPi are coupled in this frequency band. However, only functional coupling in the LF-band between striatum and GPi significantly correlated with motor symptoms. This provides further evidence for the hypothesis that dystonia might be associated with aberrant signaling within the direct pathway. Regarding the directionality of LF-activity within the BG-loop, our preliminary results are less consistent and further analyses are needed. Taken together, these findings give insight to the pathophysiology of dystonia and the striatum as a potential target for neuromodulation.