Introduction

Persistent and nonresolving inflammation has been causally linked to many common disorders. Among these, inflammatory bowel disease (IBD), comprising Crohn’s disease and ulcerative colitis, and rheumatoid arthritis (RA) often affect young and middle-aged adults with severe and life impacting symptoms. IBD and RA affect about 1.4 ( ) and 1.5 ( ) million people in the United States, respectively, with total associated cost estimated between $33 and 70 billion ( ). Standard step-care treatments for both IBD and RA begin with glucocorticoids and immunosuppressive drugs (e.g., azathioprine in IBD and methotrexate in RA) and progress to newer, targeted biological or small molecule therapies such as tumor necrosis factor (TNF) and Janus kinase inhibitors ( ). While these biological therapies have improved disease control for many patients, not all respond (primary failure), and clinical responses are often lost over time (secondary failure) due to generation of antagonizing antibodies or other reasons ( ). In addition, these treatments are very expensive and are associated with severe side effects. Thus, there remains a significant need for alternative therapeutic approaches.

The recent discoveries of neural-immune reflexes have revolutionized our understanding of how the central nervous system (CNS) maintains immunologic homeostasis. One prototypical circuit, the “inflammatory reflex,” utilizes the vagus nerve as a bidirectional conduit, carrying sensory information about peripheral inflammation to the brain (afferent limb) and supplying antiinflammatory signals to immunoactive effector cells in the tissues (efferent limb) ( ). The protective effects of driving the inflammatory reflex, either with drugs targeting molecular components of the circuit or through electrical vagus nerve stimulation (VNS), have been demonstrated in numerous disease models ( ). In standard subacute rodent models of RA and IBD, VNS, using implanted electrodes, reduced the clinical and histological manifestations of disease in the joints or intestines, respectively, and reduced tissue and systemic mediators of inflammation ( ). These studies provided both the rationale and the framework for exploring VNS as a potential nonpharmacologic, bioelectronic treatment for RA and IBD ( ) ( Fig. 127.1 ).

Neurostimulation of the cholinergic antiinflammatory pathway with an implantable medical device: A potential alterative therapeutic approach for rheumatoid arthritis and inflammatory bowel disease.

Signals from the device pass through the efferent vagus nerve and are relayed directly to the intestine, and across the celiac ganglion, traversing the splenic nerve to terminate in the spleen (1). Neurotransmitter release in the end organs results in reduced activation of resident immune cells (1), as well as reduced production of inflammatory mediators and activation of circulating immune cells (2). These effects should result in attenuation of inflammation with improvement in signs and symptoms, and reduction in joint and intestinal mucosal damage (3–4).

From Levine, Y.A., Koopman, F., Faltys, M., Zitnik, R., Tak, P-P., 2014a. Neurostimulation of the cholinergic anti-inflammatory pathway in rheumatoid arthritis and inflammatory bowel disease. Bioelectron. Med. 1, 34–43.

In this chapter, we will discuss the proof-of-concept clinical studies in both RA and IBD. We will describe a novel, miniaturized VNS system that will soon enter studies in RA and IBD, and speculate on ways that this implanted device might be modified in the future to improve usability and function.

The First Clinical Study of Vagus Nerve Stimulation in Patients With Rheumatoid Arthritis

A clinical proof-of-concept study was completed that tested the feasibility of using VNS to treat the sequelae of RA. The investigational study devices used in this study were commercially purchased VNS systems, yet treated as investigational due to their off-label use in patients with RA ( Fig. 127.2 ). The study reported two distinct patient cohorts, an early-stage cohort of patients who have had an insufficient response to the standard first-line oral drug, methotrexate, and a second cohort, including patients who have had an insufficient response to at least two different biologic drugs (e.g., TNF inhibitors and IL-1β receptor antagonists), with at least two different modes of action. This study was designed as a single armed, open-label, proof-of-concept study with a treatment withdrawal period ( ).

The vagus nerve stimulation system used in proof-of-concept studies.

The implanted system is composed of two components; a bipolar lead containing two helical coil electrodes and a helical anchor tether that were wrapped around the cervical vagus nerve, and the pulse generator that is placed within a subcutaneous pocket in the chest wall (A and B). The lead was tunneled subcutaneously from the neck to the pulse generator (A) and inserted into the lead attachment port on the generator.

From Koopman, F.A., Chavan, S.S., Miljko, S., Grazio, S., Sokolovic, S., Schuurman, P.R., et al., 2016a. Vagus nerve stimulation inhibits cytokine production and attenuates disease severity in rheumatoid arthritis. Proc. Natl. Acad. Sci. U.S.A. 113 (29), 8284–8289.

In brief, the devices were implanted, a single intraoperative stimulation was delivered, and the patients healed for approximately 2 weeks. On study day 0, the patients were given a single 60 s train of pulses, with the output current intensity titrated to the maximum tolerated dose. From day 7 to day 42, the patients self-stimulated once to four times daily by magnetically actuating the device. Stimulation trains were delivered at 10 Hz frequency, 0.25 ms pulse width, for 60 s. Following the primary endpoint of the study on the day 42 visit, there was a two-week treatment withdrawal period during which the device was switched off. The device was reactivated on day 56, and patients received stimulation with the same parameters as they had received at the day 42 visit, through the final study visit at day 84 ( ).

The primary study endpoint was the mean change in a standard composite RA study endpoint, the disease activity score 28 (DAS28), between baseline and day 42 visit. The DAS28 is a validated and scaled composite score composed of swollen joints and tender joints counts (maximum of 28), a patient defined, visual analog score (VAS), a global health assessment score, and serum C-reactive protein (CRP) levels ( ). The data showed a highly significant reduction in the mean DAS28 at day 42. The mean DAS28 worsened significantly during the withdrawal period between days 42 and 54 and was reduced once again at day 84 ( Fig. 127.3 ). American College of Rheumatology (ACR) rates of 20%, 50%, and 70% response, secondary clinical endpoints commonly used to measure effectiveness of a therapy in RA clinical trials ( ), were reported as 70.6%, 42.2%, and 11.8% respectively. A significant and DAS28-dependent decrease in IL-6 was reported, as well as significant decreases in whole blood LPS-induced TNF release. There were no serious adverse events (SAEs) or infections and the device was well-tolerated ( ). The study extends the evidence of chronic VNS reducing cytokine production and arthritic symptomatology from animals to humans and provides a clinical proof-of-concept for VNS devices in treating RA.

Vagus nerve stimulation decreases clinical disease score in rheumatoid arthritis.

Mean change from baseline in 28 joint C-reactive protein (CRP)-based disease activity score (DAS28-CRP) by study visit day for the combined cohort is shown. Significance of differences in mean change by paired t-test between visits is shown ( ⁺ p < .001 vs. day −21, ^# p < .001 vs. day 42).

From Koopman, F.A., Chavan, S.S., Miljko, S., Grazio, S., Sokolovic, S., Schuurman, P.R., et al., 2016a. Vagus nerve stimulation inhibits cytokine production and attenuates disease severity in rheumatoid arthritis. Proc. Natl. Acad. Sci. U.S.A. 113 (29), 8284–8289.

The First Clinical Studies of Vagus Nerve Stimulation in Patients With Inflammatory Bowel Disease

While there is broad evidence of the therapeutic potential of VNS from animal models of disease, evidence of efficacy of VNS in clinical IBD has only begun to emerge. An initial proof-of-concept study was reported in by Bonaz et al. that tested chronic VNS in patients with Crohn’s disease. The study, as of publication date, had recruited seven patients with active disease that had insufficient response to standard of care treatments, but were naïve to biological therapies such as TNF inhibitors. Patients were screened and had baseline clinical and biomarker assessments. The device, a commercially available pulse generator and lead ( Fig. 127.2 ), was implanted and stimulation was delivered at 10 Hz frequency, at 0.50 ms pulse width, and a continuous duty cycle of 30 s on, followed by 5 min off. The current output was gradually increased to 1.25 mA for all subjects and stimulation continued for 6 months.

The primary endpoint of the study was change in the Crohn’s Disease Activity Index (CDAI) from baseline to the final reported visit after 6 months of stimulation. The CDAI is a validated and scaled composite score that includes the number of liquid stools, abdominal pain, patient general well-being, extraintestinal complications, the use of antidiarrheal drugs, the presence of an abdominal mass, and changes in hematocrit and weight from an established “normal.” Five (5) of seven (7) patients enrolled in the study had significant improvement in CDAI from baseline, with all five achieving endoscopic remission after 6 months of treatment. An exploratory endpoint reported was the normalization of heart rate variability (HRV) components. As will be described in further detail later in this chapter, HRV components have been associated with aspects of autonomic neural activity, including sympathetic and vagal tone. In their analysis of the HRV data, the authors focus on markers of the sympatho-vagal balance (low frequency/high frequency) rather than on the high frequency component (vagal) alone. They report that VNS has normalized this balance in all five responding patients, bringing the sympatho-vagal balance closer to unity, whether by increasing or decreasing this ratio. There were no SAEs or infections and the device was well-tolerated ( ). These results provide the first evidence that VNS may be a feasible means to treat active Crohn’s disease.

The SetPoint Medical group and colleagues have also begun a clinical proof-of-concept study in Crohn’s disease. In contrast to the Bonaz study, later stage patients, those with active disease and insufficient response to TNF inhibitors and vedolizumab (α4β7 inhibitor), are recruited. Two distinct enrollment criteria are applied: in one cohort, the subjects are enrolled and implanted after previous, biologic treatments have been washed out for at minimum 3 months and VNS is studied as a monotherapy. In the second cohort, the subjects are enrolled and implanted while on a stable regimen of a biological background therapy to which an insufficient response is apparent; thus, VNS is studied as an adjunctive therapy. In addition, a much lower duty cycle of VNS is applied to these patients than is reported in the Bonaz study. The results of this study should increase our understanding of which Crohn’s disease patient population would be best treated by VNS therapy.

Summary of Clinical Safety Data

VNS devices have been used for decades in patients with refractory epilepsy, and more recently, depression ( ). Implanted in more than 100,000 patients, VNS devices have been well-tolerated and have not been associated with immunosuppression or long-term complications ( ). Implantation procedure–related events are usually self-limiting and patients rapidly accommodate to stimulation-related events, such as hoarseness due to depolarization of motor fibers in the recurrent laryngeal nerve. In the RA treatment study described earlier, the reported adverse events (AEs) were mild to moderate in severity, and included transient hoarseness, postoperative hoarseness from neuropraxis, and transient, intraoperative bradycardia during surgery ( ). These events were comparable in type and frequency to prior studies of VNS therapy reported in epilepsy patients. None of the patients developed opportunistic infections. Similarly, Bonaz et al. report that the most common AEs of the stimulation were voice alterations, hoarseness, and throat pain, all minor and time-limited to the length of the brief stimulation period ( ). The safety profile of neuromodulation devices in chronic inflammatory disease will be further evaluated in larger study populations to determine the risk–benefit ratio as compared to the toxicity and side effects of pharmacological and targeted therapies.