Multi-site stimulation of subthalamic nucleus diminishes thalamocortical relay errors in a biophysical network model

Abstract

This paper presents results on a computational study of how multi-site stimulation of the subthalamic nucleus (STN), within the basal ganglia, can improve the fidelity of thalamocortical (TC) relay in a parkinsonian network model. In the absence of stimulation, the network model generates activity featuring synchronized bursting by clusters of neurons in the STN and internal segment of the globus pallidus (GPi), as occurs experimentally in parkinsonian states. This activity yields rhythmic inhibition from GPi to TC neurons, which compromises TC relay of excitatory inputs. We incorporate two types of multi-site STN stimulation into the network model. One stimulation paradigm features coordinated reset pulses that are on for different subintervals of each period at different sites. The other is based on a filtered version of the local field potential recorded from the STN population. Our computational results show that both types of stimulation significantly diminish TC relay errors; the former reduces the rhythmicity of the net GPi input to TC neurons and the latter reduces, but does not eliminate, STN activity. Both types of stimulation represent promising directions for possible therapeutic use with Parkinson’s disease patients.

Introduction

Deep brain stimulation (DBS) is an established clinical intervention for Parkinson’s disease (PD), essential tremor, and dystonia that is now being explored for use in a variety of disorders (reviewed in McIntyre and Hahn, 2010, Wichmann and DeLong, 2006). In the PD case, DBS is implemented by an implanted pulse generator that delivers an ongoing stream of high frequency current pulses. Although this form of therapy has achieved remarkable success, improvements in DBS would be desirable in order to reduce the associated energy use and need for invasive battery changes, help patients with symptoms that do not respond to current DBS paradigms, and allow for individualized optimization of DBS strategies (Deuschl et al., 2006, Feng et al., 2007a, Feng et al., 2007b, Hauptmann et al., 2007, Rodriguez-Oroz et al., 2005, Volkmann, 2004). Efforts to develop such improvements are hampered, however, by a lack of theoretical understanding of the mechanisms through which DBS achieves its clinical efficacy.

Parkinson’s disease (PD) and experimental models of parkinsonism are associated with changes in activity patterns of neurons in the basal ganglia, including increases in synchrony, firing rates, and bursting activity in the subthalamic nucleus (STN) and internal segment of the globus pallidus (GPi) (Bergman et al., 1994, Boraud et al., 1998, Brown et al., 2001, Hurtado et al., 2005, Levy et al., 2003, Magnin et al., 2000, Nini et al., 1995, Raz et al., 2000, Wichmann et al., 1999, Wichmann and Soares, 2006). Since motor outputs from the basal ganglia emanate specifically from the GPi (Alexander et al., 1990, Kelly and Strick, 2004, Middleton and Strick, 2000), it seems likely that changes in GPi activity contribute to the development of parkinsonian motor complications. Furthermore, because motor outputs from GPi target the anterior ventrolateral nucleus of the thalamus (VLa) (DeVito and Anderson, 1982, Kelly and Strick, 2004, Yoshida et al., 1972), which serves to relay signals between cortical areas (Guillery and Sherman, 2002a, Guillery and Sherman, 2002b, Guillery and Sherman, 2002c, Haber, 2003), we and other authors have hypothesized that pathological GPi outputs may induce parkinsonian signs by changing thalamic activity patterns or information processing (Dorval et al., 2010, Dorval et al., 2008, Garcia et al., 2005, Grill et al., 2004, Montgomery and Baker, 2000, Xu et al., 2008) or, in particular, by compromising thalamocortical (TC) relay (Cagnan et al., 2009, Guo et al., 2008, Pirini et al., 2009, Rubin and Terman, 2004). This idea is consistent with the properties of TC neurons, specifically their tendency to fire rebound bursts when exposed to phasic synaptic inhibition, such as that induced by burstiness in the GPi under parkinsonian conditions. According to this viewpoint, DBS achieves its therapeutic efficacy for PD by restoring TC relay fidelity. Importantly, computational models and analysis suggest that returning STN, GPi, and TC activity patterns to their non-parkinsonian states is not necessary for achieving this goal, in as much as a variety of alternative activity patterns can also be associated with successful relay in computational models or alleviation of bradykinesia in human patients with parkinsonism (Dorval et al., 2010, Feng et al., 2007a, Feng et al., 2007b, Guo et al., 2008, Rubin and Terman, 2004).

In this paper, we study a form of STN DBS that has been suggested in the literature as an alternative to standard DBS, namely multi-site STN stimulation with delays between stimulation periods at different stimulation sites (Hauptmann et al., 2007, Hauptmann et al., 2005, Tass, 2003). We consider such stimulation with two types of current injection, one using periodic square pulses and another based on a local field potential signal recorded from the STN population (Hauptmann et al., 2007, Hauptmann et al., 2005, Rosenblum and Pikovsky, 2004b, Tukhlina et al., 2007). To perform our investigation, we introduce these stimulation paradigms into a computational model based on our earlier work (Rubin and Terman, 2004, Terman et al., 2002). This model consists of a small network of conductance-based STN, GPe (external segment of the globus pallidus), GPi, and TC neurons that, by design, generates parkinsonian activity patterns in the absence of stimulation. We simulate the delivery of an excitatory input train to the model TC neurons and compare TC relay performance across simulations without stimulation and simulations with coordinated reset or LFP-based delayed feedback stimulation of various amplitudes and periods. Although the forms of stimulation that we study were introduced previously (Hauptmann et al., 2007, Hauptmann et al., 2005, Tass, 2003), this represents the first work in which they are incorporated into a biophysically-detailed basal ganglia network model and in which their impact on TC relay fidelity is explored. We find that both forms of multi-site stimulation, applied to the STN, regularize GPi outputs and significantly diminish TC relay errors. While both can completely suppress STN activity if introduced with a large enough amplitude, both can also restore TC relay performance without such a drastic effect. Moreover, multi-site delayed feedback stimulation based on the LFP in particular requires relatively small currents (see also Hauptmann et al., 2007, Hauptmann et al., 2005, Tukhlina et al., 2007) and, unlike stimulation with a constant current of similar magnitude, maintains STN responsiveness to its own excitatory inputs (e.g., from the hyperdirect pathway). Hence, these results support the idea that multi-site delayed feedback stimulation of STN merits further consideration as a possible alternative to standard forms of DBS for PD.