Engineering Solutions for Vagus Nerve Stimulation to Minimize Invasiveness and Reduce Side Effects

Vagus nerve stimulation (VNS) is an effective treatment for epilepsy, depression, and stroke rehabilitation with ongoing studies of additional clinical applications including rheumatoid arthritis and heart failure. Acute VNS alleviates inflammation caused by infection, and clinical studies of VNS treatment for immune dysregulation are promising. However, more widespread use of VNS is limited by surgical implantation, and risks associated with surgery are deleterious in patients with pre-existing neurocognitive impairment. Non-invasive methods of VNS (e.g., transcutaneous VNS) produce inconsistent results and lack a robust biomarker to confirm nerve stimulation. There is a need for a method of VNS that provides targeted stimulation with reduced invasiveness. As well, side effects limit therapy. Reduced heart rate (HR) during stimulation is associated with therapy for heart failure, but stimulation frequency and amplitude are limited by patient tolerance. An understanding of physiological responses to parameter adjustments would allow control of therapeutic and side effects. The purpose of this dissertation was to develop novel methods of VNS, investigate new applications of VNS for post-operative treatment, and conduct parametric studies to increase the dynamic range between therapeutic effects and side effects.First, we developed a minimally invasive, targeted, percutaneous vagus nerve stimulation approach (pVNS). We stimulated the cervical vagus nerve in mice using an ultrasound-guided needle electrode under sevoflurane anesthesia. The concentric bipolar needle electrode was placed adjacent to the carotid sheath, and nerve stimulation was verified in real-time using bradycardia as a biomarker. Activation of vagal fibers was also confirmed by immunohistochemical quantification of c-Fos and choline acetyltransferase expression in relevant brainstem structures, including the dorsal motor nucleus and nucleus tractus solitarius. Following lipopolysaccharide (LPS) administration, pVNS reduced plasma levels of tumor necrosis factor at 3 h post-injection. pVNS also prevented LPS-induced hippocampal microglial activation as analyzed by changes in Iba-1 immunoreactivity, including cell body enlargement and shortened ramifications. LPS injection reduced memory function at 24 and 48 h but not when pVNS was delivered post-injection. These results provide a novel therapeutic approach using VNS to modulate neuro-immune interactions that affect cognition. Second, we assessed pVNS efficacy in prevention of surgery-induced delirium superimposed on dementia in Alzheimer’s Disease-like (CVN-AD) mice. Orthopedic surgery increased hippocampal microglia activation, amyloid-beta (AB) accumulation, and neuronal loss as measured by histology. At 24 h post-operation, neural pathologies were absent in animals that received post-operative pVNS. We quantified blood-brain barrier (BBB) permeability with intracardial tracer injection to investigate vessel integrity. Following surgery, pVNS reduced tracer concentration in the hippocampus indicating improved BBB integrity. We paired 5x familial autosomal dominant (5xFAD) AD mice and wild type (C57) mice using parabiosis to evaluate the contribution of circulating factors from sterile surgery on post-operative pathologies. AB accumulation and microglia activation increased in 5xFAD mice when the C57 parabiosis partner received orthopedic surgery. Thus, post-operative pVNS protects against surgery-induced increases in dementia pathology driven by systemic inflammation. Finally, we measured the effect of the temporal pattern of VNS on HR (a proxy to therapy) and laryngeal EMG (a side effect) in anesthetized mice. Amplitude, intra-burst frequency, and mean pulse rate (MPR) modulated HR while only MPR modulated EMG. Neither outcome was sensitive to stimulation pattern at clinical frequencies. However, stimulation modulated HR to a greater degree than EMG at an amplitude and frequency above those used clinically. We leveraged collected data to construct computational models of HR and laryngeal muscle activity that reproduced VNS responses for amplitudes and patterns. To model HR, we incorporated a mechanism for frequency-dependent filtering of vagal pulses by cardiac ganglia, and the model overestimated HR changes at a high intra-burst frequency when filtering was removed. The experiment outcomes indicate concurrent increases of stimulation amplitude and MPR modulate HR more than laryngeal EMG, and the model outcomes indicate that ganglia fidelity contributes to stimulation frequency effects on HR. The results impact the field of VNS through the invention of a minimally invasive method of stimulation to modulate neuroinflammation, demonstration of protective effects of VNS on delirium superimposed on dementia, and identification of stimulation adjustments to maximize a proxy for therapy over side effects.