Noninvasive vagal nerve stimulation for gastroenterology pain disorders

Abstract

Abdominal pain continues to be a major challenge and unmet need in clinical practice. Normalization of bidirectional gut-brain signaling has generated much interest as a therapeutic approach to treat chronic abdominal pain. Vagal nerve stimulation (VNS) is emerging as a potential non-pharmacologic strategy for the treatment of abdominal pain. In this review paper, we will summarize the etiologies of chronic pain in gastrointestinal disorders and discuss the rational for VNS as a therapeutic approach to chronic abdominal pain, with particular emphasis in the gammaCore stimulator which allows for noninvasive VNS.

Lay abstract

Abdominal pain is still a major medical problem that needs more study and better treatment options. The connection between the gut and the brain is very important. This connection could reveal new therapies to treat abdominal pain. One such therapy is vagal nerve stimulation (VNS). VNS is showing promise for the treatment of abdominal pain and other conditions. In this review, we will summarize the causes of abdominal pain. We will also explain how VNS could work to treat abdominal pain, with particular emphasis in the gammaCore stimulator.

Practice points.

• Gut-brain signaling has come to the forefront as a target for the treatment of chronic abdominal pain.

• Vagal nerve stimulation (VNS) is emerging as a nonpharmacologic therapeutic strategy for disorders of gut–brain interaction.

• Various devices for noninvasive VNS are already on the market for migraine and headaches and under investigation for disorders of gut–brain interaction.

• Early trials with noninvasive VNS are showing promising clinical responses in functional gastrointestinal disorders.

Worldwide, the third most common patient reported reason for visiting a primary care doctor is abdominal pain and associated abdominal symptoms [1]. In the USA, abdominal pain is the leading reason for visiting emergency departments, accounting for 8.7% of all visits [2]. In 2014 alone, abdominal pain was responsible for over 21 million ambulatory care visits, accounting for over $10 billion USD in healthcare expenditures the following year [3]. Given its pervasiveness and burden to individuals and the healthcare system, improved therapies for abdominal pain are paramount.

Abdominal pain can have different etiologies, sources, and modes of neural transmission. From the viewpoint of a gastroenterologist, pain can be somatosensory, originating from the abdominal wall or visceral, originating from the digestive tract. This review will focus mostly on digestive tract visceral pain, but the principles discussed regarding pathogenesis of chronic pain and the therapeutic potential of vagal nerve stimulation (VNS) are also applicable to other disorders associated with visceral and somatic pain. Visceral pain is common and highly prevalent, occurring in health and disease and in individuals with organic disease, such as inflammatory bowel disease, pancreatitis or peptic ulcer disease; and as well as in those with ‘functional gastrointestinal disorders’ (recently reclassified as disorders of gut–brain interaction) such as irritable bowel syndrome or functional dyspepsia, who have no identifiable biochemical or structural abnormalities. Management of visceral pain in these patients remains a major challenge in Pain, Gastroenterology, and Primary Care practices, and is associated with worse disease severity, quality of life and chronic opiate use. Contributing to this problem are gaps in our understanding of the pathophysiological mechanisms underlying visceral nociception, although much is known.

The gastrointestinal tract predominantly receives dual sensory innervation from spinal and vagus nerve afferent neurons. Further, the digestive tract is intrinsically innervated by afferent enteric nervous system neurons, which are thought to contribute principally in gastrointestinal motility and secretion [4]. The vagus nerve and its branches innervate the digestive tract from the upper esophageal sphincter to the transverse colon [5]. The distal colon and rectum afferents originate from sacral plexus neurons. The subdiaphragmatic vagus is composed predominantly of sensory afferent fibers (80–90%), which are mostly unmyelinated C-fibers [6]. The cell bodies for these vagal afferents are found predominantly in the dorsal brainstem nucleus tractus solitarius, where they synapse to the motor efferent vagal nuclei (dorsal motor vagal nucleus) and ‘higher’ centers of the CNS including the hypothalamus, amygdala, parabrachial nucleus and insular cortex. Spinal afferent neurons found in the dorsal root ganglia of the spinal cord ascend via the dorsal horn to the thalamus and then project on to a number of brain regions involved in pain perception. Multiple types of gut sensory neurons have been described based on the location of their receptive fields (serosa, muscular layer, mucosa, etc.) as well as their major stimuli: chemical, mechanical, thermal, etc. Intrinsic enteric neurons normally sense luminal stimuli, but are not involved in pain, as the information is hardly transmitted to the central nervous system [7]. Nociceptive neurons have been described innervating the gut from both the vagus and spinal cord [8,9]. However, splanchnic nerves carrying spinal cord afferent fibers are mainly thought to be responsible for pain processing in the gut, whereas the vagal sensory pathways mainly mediate nonpainful sensations. Nociceptive information may involve the vagal pathways directly, but most evidence points toward an indirect role for pain modulation in the CNS [10].

In sum, the primary neural pathways leading to pain perception from the digestive tract to the brain include afferents from the spinal cord ascending via the dorsal horn to higher brain centers, as well as vagal afferents from the nucleus tractus solitarius. In this review, we summarize the etiologies of chronic pain in gastrointestinal disorders and discuss the rational for VNS as a therapeutic approach to chronic abdominal pain, with particular emphasis on the gammaCore stimulator which allows for noninvasive VNS (nVNS).

Gastrointestinal chronic pain disorders

In health, the majority of interoceptive information from the gastrointestinal tract reaching the brain is not consciously perceived, but serves primarily as input to autonomic reflex pathways to maintain gut homeostasis or adapt its functions. In patients with chronic functional abdominal pain syndromes, conscious perception of interoceptive information from the GI tract can occur in the form of chronic pain [11].

The genesis of chronic abdominal pain is dependent on two major related phenomena, neural plasticity and visceral hypersensitivity. Nearly all chronic gastrointestinal disorders exhibit a disease-stage-dependent, structural and functional neuroplasticity involving changes in neurochemical coding and synapse/innervation density [12]. These changes may lead to pathologic decreased sensory thresholds, which result in the generation of pain or other gastrointestinal related symptoms in response to what would otherwise be normal physiologic stimuli. This condition is referred to as visceral hypersensitivity [13,14]. This sensitization can occur in the primary afferent neurons, referred to peripheral sensitization and/or in the second/third order neurons in the central nervous system, known as central sensitization (Figure 1) [15]. In addition to sensitization of sensory pathways, there is a parallel phenomenon involving loss of modulatory mechanisms in visceral nociceptive pathways. This is typically associated with overactivation of the major stress response systems [9]: the hypothalamic-pituitary-adrenal axis and the efferent sympathetic branch of the autonomic nervous system [16], which impact digestive function extensively.

Figure 1. . Mechanisms underlying pain sensitization in the gut.

DMV: Dorsal motor nucleus of the vagus; DRG: Dorsal root ganglia; NTS: Nucleus tractus solitarius.

Peripheral sensitization is thought to be caused principally by inflammation. Part of a normal physiological response to acute organ inflammation, is increased pain perception, which leads to the appropriate behavioral modifications conducive to healing. This sensory response is mediated by inflammatory cues acting on sensory neurons (such as histamine, prostaglandins, protons, reactive oxygen species and cytokines, generated from activated tissue resident immune cells such as mast cells, macrophages and eosinophils) and neurogenic cues exerting positive feedback and further activation of immune cells (such as substance-P, calcitonin gene-related peptide and neuropeptide K, produced by activated nociceptive neurons) [12,17]. Inflammation is increasingly recognized as playing an important role in gastrointestinal pain disorders [18]. Peripheral sensitization can occur either in the context of chronic inflammatory disorders, such as inflammatory bowel disease or chronic pancreatitis; or after isolated acute bouts of inflammation, such as postinfectious irritable bowel syndrome, where the inflammatory cues resolve, but the afferent pain pathway has undergone lasting dysfunctional plasticity [19].

Central sensitization also occurs in chronic gastrointestinal disorders [20] and may involve multiple CNS regions that regulate the sensory-discriminative (spinothalamic tracts, thalamus and primary/secondary sensory cortices), as well as the affective-emotional, behavioral-motor responses and cognitive components of pain processing (thalamus, limbic system, amygdala, cingulate cortex, midbrain, spinoreticular, spinomesencephalic tracts). Aberrant pain modulation of afferent visceral signaling via inhibitory pathways involving the brainstem and descending corticolimbic pathways may also be affected [9]. Gut sensory information transmitted via the vagus nerve to the brainstem plays an important role in the indirect central inhibitory modulation of pain, by activating regions such as the periaqueductal grey and rostroventral medulla that control the descending inhibitory pathways that can gate nociceptive information at the spinal level [21]. Thus, chronic abdominal pain is likely the result of aberrant signaling in the bidirectional gut-brain axis involving both mechanisms of peripheral and central sensitization leading to visceral hypersensitivity and hyperalgesia.

VNS for gastrointestinal disorders

Normalization of bidirectional gut-brain signaling has generated much interest as a therapeutic approach to treat chronic abdominal pain [22]. Given that VNS for the treatment of refractory epilepsy has been in practice for decades, its broader use for other disorders involving vagal signaling has gained much interest. In fact, since its introduction for epilepsy in 1990 and subsequent use for depression [23], over 7000 reports have been published in regard to its various clinical uses and mechanisms of action. Most recently, migraines and other headaches have become approved indications for VNS. Mounting evidence for its therapeutic potential in inflammatory/pain disorders, and the introduction of nVNS modalities have further increased the appeal of this nonpharmacologic therapy for a broad array of disorders [24]. Autonomic vagal dysfunction has indeed been observed in functional and inflammatory digestive disorders [16,25]. Given its extensive innervation of the gastrointestinal tract and its predominant role in parasympathetic regulation of inflammation and motility, the vagus nerve may be utilized as a powerful target for the treatment of gastrointestinal dysfunction and associated pain [25,26].

VNS therapy can broadly be divided into two categories based on route of administration: invasive and noninvasive. Invasive VNS was the first modality to be introduced [27], and consists of placement of helical electrodes on the surgically-exposed cervical vagal nerve. These electrodes are connected to a stimulating generator most commonly positioned in the ipsilateral infraclavicular pocket. Various implantable systems have been in clinical use and there are many more in research development [28]. Although not considered formally VNS, another invasive modality of VNS worth mentioning here is gastric electric stimulation, which has been in clinical use for medically refractory gastroparesis (a motility disorder of gut–brain interaction). Here, two electrodes are implanted surgically into the stomach smooth muscle layer and are then connected to a stimulating generator positioned in the abdominal wall. Although initially designed to increase gastric motility, gastric electric stimulation impacted mostly symptoms of nausea, vomiting and to a lesser extent, abdominal pain, without significant changes in gastric emptying [29]. Subsequent studies have suggested that many of these symptomatic improvements are mediated through decreased visceral hypersensitivity via increases in efferent vagal output and neurostimulation ascending from the stomach to the brain via vagal afferents [30–34].

nVNS is administered via transcutaneous electrodes, and can be further grouped by location. The cervical vagus on the neck and the auricular branch of the vagus in the earlobe are the most common targets of nVNS, and there are multiple companies and devices in clinical trials and on the market [23,35]. ElectroCore has been at the forefront of developing a cervical transcutaneous stimulator called gammaCore. The device (Figure 2) uses two stainless steel round discs that function as skin contact surfaces that deliver a low-voltage electrical signal (5 kHz sine waves each lasting 200 ms, repeated once every 40 ms or 25 Hz) to the cervical vagus nerve on the neck. The device delivers a programmable number of stimulation cycles, each lasting 120 s and is placed medially to the sternocleidomastoid muscle along the cervical branch of the vagus nerve. The intensity of the stimulation is gradually increased until a mild but stable contraction of the ipsilateral orbicularis oris muscle is obtained. The highest voltage produced is 24 V and the highest intensity is 60 mA. A major advantage of nVNS is that patients avoid surgery and are easily able to titrate therapy (intensity, frequency, etc.) to accommodate their needs. It is important to note that although safety trials have been very encouraging, the long-term effects of the chronic use of gammaCore have not been evaluated. Per the manufacturer’s information site, safety and efficacy of gammaCore have not been evaluated in patients with carotid atherosclerosis surgical injury to the vagus, children, pregnant women, or patients with clinically significant hypertension, hypotension, bradycardia or tachycardia; and therefore, use in these populations should be under closely monitored clinical research protocols. Further, nVNS with gammaCore is contraindicated in patients with an active implantable medical device (i.e., pacemaker, hearing aid implant or any implanted electronic device) or metallic device (i.e., stent, bone plate or bone screw) implanted at or near their neck.

Figure 2. . gammaCore noninvasive vagal nerve stimulator.

Copyright electroCore, Inc. All rights reserved. Used with the permission of electroCore.

Animal experiments suggest that nVNS acts via stimulation of vagal afferents. In healthy human volunteers, gammaCore nVNS (25 Hz, 6–24 V) 2 min transcutaneous stimulation over left/right cervical vagus nerve showed vagal somatosensory evoked potentials similar to those found with implanted electrodes or stimulation of auricular nerve in the outer ear [36]. These gammaCore-evoked potentials increased in amplitude with stimulation intensity and disappeared when the stimulator was positioned over neck sternocleidomastoid muscles, suggesting that they are not muscle artifacts. Currently, gammaCore is approved in the USA for the treatment of migraines and cluster headaches and in Europe for asthma and airway reactivity, primary headache, gastrointestinal disorder, anxiety, depression and seizure disorder, with multiple ongoing trials for other indications.

Evidence supporting nVNS for abdominal pain disorders

The overall impact of any VNS modality on chronic abdominal pain should, in theory, be associated with its potential to restore gut-brain signaling homeostasis and normalize endogenous pain neuroregulatory processes such as peripheral and central sensitization (Figure 1). nVNS with gammaCore has been suggested to modulate central descending pathways for pain control, as shown in a study with healthy volunteers that experienced significantly increased thresholds to nociceptive withdrawal reflex of the lower limb 5 and 30 min after nVNS [37]. In a prior small study with healthy volunteers, it was demonstrated that auricular nVNS reduced somatic pain sensitivity and enhanced gastric motility [38]. To date, auricular devices have yielded the most robust (randomized, blinded, sham-controlled) clinical data supporting nVNS therapy for gastrointestinal pain disorders [39,40]. However, further studies are needed to explore the full potential of VNS modalities for gut visceral pain disorders.

nVNS for gastrointestinal disorders associated with abdominal pain

Two open-label single-center pilot studies using short-term (3–4 week) gammaCore nVNS have been published regarding its utility in gastroparesis, a disorder of gut–brain interaction marked by delayed gastric emptying, chronic nausea, vomiting and abdominal pain [41]. The first study by Paulon et al. in the UK found a 43% response rate in symptom improvement after 3–6 week of nVNS in medically refractory gastroparesis patients, with a 43% relative symptom improvement in bloating and abdominal pain [42]. The second study, which we conducted in the USA, showed a similar 40% response rate in gastroparesis symptom improvement, associated with a significant amelioration of abdominal pain after 4–6 weeks of nVNS [43]. The nVNS responder group also noted therapy-associated improvement using the patient reported outcomes measurement information system (PROMIS) Pain Interference scale, which was not the case in the non-responder group [43]. These pilot studies highlight the therapeutic potential of nVNS in disorders of gut–brain interaction as extending well-beyond pain improvement. In fact, randomized, sham-controlled trials in adolescents with irritable bowel syndrome [39] or chronic abdominal pain-related functional gastrointestinal disorders [40] showed that 4 weeks of auricular nVNS was associated with a treatment-specific improvement in maximal pain scores and composite pain scores, which was sustained for an extended period of time.

Importantly and broadly speaking, VNS has been established as a safe, well tolerated and effective treatment of inflammatory pain disorders with associated dysautonomia. VNS has potent anti-inflammatory effects [44], improving animal models of colitis [45,46], postoperative ileus [47,48] and inducing a clinical and biological remission in small cohorts of patients with active Crohn’s disease [49] and rheumatoid arthritis [50]. It is important to note that analgesia was not a primary end point in these studies. The therapeutic effect of VNS in these inflammatory disorders is thought to be based on stimulation of the cholinergic anti-inflammatory reflex pathway via vagal efferents [51,52], and engagement of the hypothalamic-pituitary-adrenal axis by vagal afferents [53]. Specifically in humans, nVNS with gammaCore can downregulate cytokine secretion in lipopolysaccharide-stimulated whole blood cultures [54] and has a subtle anti TNF-α effect in healthy volunteers [55]. In our own hands, we found that four weeks of nVNS ameliorated various inflammatory markers in a cohort of gastroparesis patients (manuscript in preparation). Therefore, in addition to modulation of central descending pain control pathways, nVNS may also ameliorate inflammation-mediated peripheral sensitization via enhanced anti-inflammatory vagal efferent signaling.

Concluding remarks

Electroceuticals hold great potential to normalize, in a safe, nonpharmacologic manner, the endogenous regulatory pathways and reflexes controlling gastrointestinal function and perception of pain. Therefore, nVNS and similar modalities should continue to be explored in larger, randomized, blinded, placebo-controlled trials. It will also be of interest to complement clinical trials with translational studies focused on understanding the exact mechanisms of action of nVNS, such as modulation of central descending pathways involving the amygdala and spinal cord, decreasing inflammation-related peripheral sensitization or increasing thresholds of primary visceral afferents in the gut-brain signaling axis. Importantly, it will be critical that approved nVNS devices showing positive clinical responses are made accessible to the patients who will benefit the most from them.