Non‐invasive, Spatiotemporal Characterization of Muscle Activation Patterns from Vagus Nerve Stimulation in Human Subjects

Vagus nerve stimulation (VNS) is used for treatment of epilepsy in over 100,000 patients worldwide and is a potential treatment for inflammatory disorders such as rheumatoid arthritis, diabetes, and sepsis. VNS-evoked compound nerve action potentials (CNAPs) may have applications as a biomarker of VNS treatment efficacy, but transcutaneous measurement of this activity is obscured by laryngeal muscle artifacts during cervical VNS. Our objective is to develop a more precise understanding of muscle activation patterns that could improve recording and analysis protocols for isolating the vagus nerve (VN) CNAP from EMG artifacts. We hypothesize that analyzing multi-electrode array (MEA) surface recordings overlying the VN will allow us to characterize this muscle artifact.Twelve patients undergoing VNS therapy for epilepsy were enrolled in a clinical study at Indiana University School of Medicine (IRB #2006075899). The anatomical dimensions of the neck were recorded for seven of the twelve subjects. A custom-made MEA with a grid of 32 soft foam electrodes was placed on the skin overlying the VN on each side of the neck with the short axis parallel to the sternocleidomastoid muscle and the midpoint at the level of the laryngeal prominence. Surface potentials were recorded for approximately 20 minutes at the patient's established device settings.VNS-evoked potentials were visualized in nine subjects (three subjects were excluded due to insufficient data). Two probable muscle artifacts were identified, defined as non-propagating features in the mean response to n > 600 pulses of VNS. The first had an onset latency of ~1-3 ms, and the second had a latency of ~7-10 ms. Both artifacts appeared primarily in the electrodes overlying the laryngeal muscles. In one subject, propagating activity at an approximate velocity of 70-110 m/s was visualized between 1-2 ms latency.The short latency of the first muscle artifact suggests stimulus signal leakage activating the superior laryngeal nerve. The propagating activity visualized in one subject may show the activity of the external branch of the superior laryngeal nerve, which stimulates the cricothyroid muscle to contract, creating the first muscle artifact. The second artifact with longer latency is likely caused by the recurrent laryngeal nerve, which causes contraction of intrinsic laryngeal muscles such as the posterior cricoarytenoids. Previous analyses may have mischaracterized the first muscle artifact as nerve activity [Usami, K., Kawai, K., Sonoo, M., and Saito, N. (2013). Brain Stimul. 6(4), 615-623]. The use of MEA-based recordings clarifies our understanding of the VN's response to VNS, which may lead to better treatment efficacy and the eventual development of personalized VNS therapies for epilepsy and a range of inflammatory disorders.