Chronic pain is a debilitating neurological disorder which affects hundreds of millions of people worldwide. Neurostimulation therapies, such as spinal cord stimulation (SCS) and dorsal root ganglion stimulation (DRGS), are non-addictive alternatives for managing chronic neuropathic pain that is refractory to conventional medical management. SCS and DRGS apply sequences of brief electrical impulses to neural tissue. However, not all patients receiving these therapies obtain adequate pain relief, and patient outcomes are not improving despite decades of clinical experience and advancements in stimulation technology. This dissertation addresses two crucial knowledge gaps limiting the success of neurostimulation therapies: 1) we do not understand the physiological mechanisms of electrical stimulation-induced pain relief, and 2) we do not understand the sources of variability affecting the neural response to stimulation.

The first portion of this thesis examined the mechanisms of action of DRGS. We developed statistical models of neural element (i.e., cell bodies, axons) locations in histological samples of human dorsal root ganglia (DRG) tissue. Next, we employed a histologically informed field-cable modeling approach to study the neural response to DRGS. We coupled a finite element method model of the potential distribution generated by DRGS to multi-compartment cable models of DRG neurons to simulate which types of sensory neurons are activated by therapeutic DRGS. Our data suggest that clinical DRGS directly activates the subset of sensory neurons that code non-painful touch sensations, which may trigger pain-inhibition neural networks in the spinal cord dorsal horn.

The second portion of this thesis investigated how biological variability at different scales (e.g., single cells, patient anatomy) affected the neural response to stimulation. We implemented a Markov Chain Monte Carlo (MCMC) method to parametrize populations of neurons with heterogeneous ion channel expression profiles. We incorporated this approach in our field-cable model of DRGS and showed that variability in ion channel expression can affect the stimulation amplitude required to generate activity in target neurons. We further applied this population-modeling approach to investigate how pathology induced changes in ion channel expression can affect the behavior of neural circuits governing sensory transmission. Finally, we developed a framework for constructing patient-specific field-cable models of patients receiving SCS. This framework captured the effect of key anatomical details (e.g., the amount of cerebrospinal fluid between a patient’s SCS electrode array and the spinal cord) on neural activation during stimulation. Furthermore, this patient-specific modeling framework allows the comparison of model predictions of neural activation during SCS with clinical data, such as patient-reported outcomes (e.g., pain relief).

The results of this dissertation suggest that DRGS may share mechanisms of action with other neurostimulation therapies for pain management, such as SCS. This dissertation also developed frameworks for studying the effect of biological variability on the nervous system’s response to electrical stimulation. To develop safe and effective therapies for neurological disorders, it is crucial to understand both the physiological mechanisms of symptom relief, and how the neural response to therapy may vary across cells, circuits, and patients. This dissertation provides novel insights on both aspects as they relate to neurostimulation for chronic pain.

Date: Friday, March 25, 2022
Time: 12:00 PM EST
Zoom link: https://umich.zoom.us/j/93382361344 (password: neuron)
Chair: Dr. Scott F. Lempka