Spatiotemporal Approaches to Increase the Efficacy of Spinal Cord Stimulation

Abstract

Spinal cord stimulation (SCS) is a surgically implanted therapy for chronic pain that delivers electrical stimulation to the spinal cord. Despite significant technological and clinical improvements, the therapeutic success of SCS has plateaued (North et al., 1993; Taylor et al., 2014), in part due to incomplete understanding of how changing stimulation parameters (i.e. amplitude, pulse duration, and timing) affect the neuronal circuitry that modulates pain perception. The work in this dissertation uses computational modeling and in vivo neural recordings to understand how dorsal horn circuitry changes in response to neuropathic pain and predict optimal stimulation parameters for SCS. Computational models are important tools for studying and predicting neural circuit responses and the first part of this dissertation concerns the development of novel computational models. The models are subsequently used to predict responses to neuropathic pain, and the models quantified distinct shifts in responses observed in experimental recordings, demonstrating their validity as a tool for understanding mechanisms of action of SCS. Multiple single unit recordings in the dorsal horn replicated the predictions made using the computational models and predicted that correlations between neurons could be used as a biomarker of neuropathic pain and stimulation efficacy. Model-based design uncovered multifrequency stimulation parameters and temporal patterns of stimulation that optimized neural responses and present promising avenues for improving clinical efficacy. We also used the computational model to predict the mechanism of action of a novel modality of SCS and validated our predicted mechanism of action through experimental recordings. Overall, this thesis work improved our understanding of dorsal horn circuits and the mechanisms of action underlying multiple modalities of SCS, developed new strategies for optimizing stimulation parameters, and demonstrated the effectiveness of optimized stimulation parameters in preclinical models.