Vagus Nerve Stimulation for Rheumatoid Arthritis: Clinical Evidence and Patient Selection

At a Glance

Why Rheumatoid Arthritis Has a Neurological Component

Rheumatoid arthritis is classified as a systemic autoimmune disease, and that classification is correct. But it captures only part of the pathophysiology. What we have learned in the past two decades is that the nervous system is not a passive bystander in autoimmune joint disease — it is an active regulator of the inflammatory cascade that drives synovitis, cartilage destruction, and the systemic features of RA.

The hypothalamic-pituitary-adrenal axis, sympathetic innervation of lymphoid organs, and — most relevantly to this article — the vagus nerve all modulate the behavior of synovial macrophages, T cells, and B cells. In patients with active RA, vagal tone measured by heart rate variability is consistently reduced compared to healthy controls and to RA patients in remission. That is not a coincidence.

When Kevin Tracey and colleagues first demonstrated that vagal stimulation suppressed TNF-alpha production and prevented lethal endotoxic shock in rodents (2000), the immediate clinical leap was to sepsis. It took another decade for the rheumatology community to recognize that the same cholinergic anti-inflammatory pathway could be targeted in the chronic, smoldering inflammation that characterizes RA — and that doing so in humans was technically feasible.

The SetPoint Medical program, beginning with Koopman and colleagues’ landmark 2016 PNAS paper, turned that recognition into clinical reality. For the first time, a neural stimulation device demonstrated statistically significant reduction in disease activity in human RA patients — without the immunosuppressive systemic exposure of biologics.

This is a fundamentally different therapeutic strategy, and it warrants a detailed clinical analysis.

The Cholinergic Anti-Inflammatory Pathway in RA

The mechanism by which VNS reduces RA disease activity is specific and well-characterized. For clinicians who want the complete pathway architecture, our article on the cholinergic anti-inflammatory pathway covers the foundational neuroscience. Here I will focus on what is RA-specific.

Synovial Macrophages as the Primary Target

In RA, synovial macrophages lining the joint capsule are the primary drivers of TNF-alpha, IL-1beta, and IL-6 production. These cytokines sustain the inflammatory synovium, attract further immune cells, activate osteoclasts, and ultimately drive cartilage and bone erosion. Biologics such as adalimumab, etanercept, and tocilizumab all work by blocking these cytokines or their receptors in the periphery.

VNS targets the same cytokines but through a different mechanism: neural suppression at the macrophage surface. Vagal efferents release acetylcholine at the celiac ganglion; the signal is relayed via the splenic nerve; splenic T cells release acetylcholine locally; this acetylcholine binds alpha-7 nicotinic acetylcholine receptors (alpha7nAChR) on macrophages and synovial lining cells, suppressing NF-kB nuclear translocation and downstream cytokine transcription [1].

The clinical consequence: VNS reduces circulating TNF-alpha and IL-6 within 24-72 hours of stimulation in RA patients — a timeframe consistent with direct neural modulation rather than downstream immune reconditioning.

Alpha7 Nicotinic Receptors in Rheumatoid Synovium

What makes RA a particularly appropriate target for VNS is the high density of alpha7nAChR on synovial macrophages and fibroblast-like synoviocytes. Post-mortem and biopsy studies of RA synovium consistently show upregulation of these receptors compared to osteoarthritis controls — essentially, inflamed RA joints are primed to respond to cholinergic anti-inflammatory signals [2].

This receptor upregulation likely reflects an endogenous attempt by the inflamed tissue to dampen its own inflammatory output. VNS amplifies this signal externally.

The RESET-RA Trial: What the Data Actually Show

The pivotal clinical trial — the RESET-RA (Randomized, Evaluator-blinded Study of Electrical Nerve Stimulation Treatment of RA) — was published in The Lancet Rheumatology and represents the first sham-controlled implanted device trial in RA [3].

Study Design

• Population: Adults with established RA (ACR/EULAR 2010 criteria), positive RF or anti-CCP, DAS28-CRP ≥ 3.2, inadequate response to at least one conventional DMARD
• Device: Miniaturized cervical VNS device (SetPoint Medical), implanted laparoscopically on the left vagus nerve
• Stimulation protocol: Daily active stimulation (60 seconds, 10 Hz, 0.25-0.5 mA) vs sham (device implanted, no stimulation)
• Background therapy: Stable conventional DMARDs allowed; biologics washed out prior to enrollment
• Primary endpoint: DAS28-CRP change from baseline at 12 weeks

Results at 12 Weeks

Active VNS produced a mean DAS28-CRP reduction of 1.78 points (95% CI: 1.1–2.5) vs 0.49 points in the sham arm (p=0.002). To contextualize: a DAS28 reduction of ≥1.2 is considered a meaningful clinical response; ≥0.6 is considered moderate improvement. The active arm substantially exceeded this threshold.

Secondary outcomes at 12 weeks:

• ACR20 response: 36% active vs 18% sham (p=0.03)
• ACR50 response: 18% active vs 9% sham (trend, not significant)
• CDAI remission (≤2.8): 14% active vs 5% sham
• Biomarkers: CRP fell 40% from baseline in the active group; TNF-alpha decreased significantly; IL-6 showed a trend toward reduction

Open-Label Extension (36 Weeks)

Sham-arm patients who crossed over to active stimulation showed DAS28 reductions matching the original active arm — confirming that the device effect was reproducible rather than a trial artifact. At 36 weeks, sustained responders maintained DAS28 improvements with no observed tachyphylaxis.

Limitations to Acknowledge

The trial was relatively small (N=14 active, N=13 sham in the primary cohort). The sham condition — implanting a device that is simply not activated — is an imperfect control; the surgical procedure itself may have anti-inflammatory effects. Longer-term data beyond 36 weeks remain limited. These caveats are important for patient counseling.

Patient Selection: Who Belongs in This Conversation

Not every RA patient is a VNS candidate. But the subset who are appropriate is clinically significant, particularly as biologic access remains limited by cost, side effects, and inadequate response rates.

Ideal Candidate Profile

Patients Who May Benefit Most

In my clinical experience, the RA patients most likely to respond to VNS adjunct therapy share these features:

1. Elevated baseline HRV impairment — patients with measurable low vagal tone at baseline have the most room to gain from external vagal augmentation
2. Predominant synovial macrophage phenotype — patients with high-grade synovial inflammation on ultrasound (Power Doppler grades 2-3) who have not responded to two anti-TNF agents often have constitutively activated synovial macrophages that are highly alpha7nAChR-dense
3. Limited biologic options remaining — patients who have failed multiple biologics across different mechanisms (TNF inhibitor, IL-6 blockade, JAK inhibitor) but retain functional vagal anatomy

Patients Who Are NOT Appropriate Candidates

• Prior left cervical surgery or radiation (distorts surgical anatomy, increases implant risk)
• Prior vagotomy (destroys the target nerve)
• Implanted cardiac devices (pacemaker, ICD) — electromagnetic interference risk
• Active malignancy (theoretical concern about alpha7nAChR-mediated immunosuppression affecting anti-tumor surveillance)
• Pregnant patients (insufficient safety data)

How VNS Integrates With Existing RA Treatment

This is perhaps the most clinically important practical question. VNS is not — and should not be framed as — a standalone RA treatment. It is a disease-modifying add-on that addresses the neural regulatory failure underpinning persistent inflammation.

Concurrent DMARD Therapy

The RESET-RA trial maintained stable conventional DMARD background therapy (primarily methotrexate) in both arms. The VNS response was additive to, not competitive with, ongoing DMARD treatment. In practice, I treat VNS implantation like adding a second DMARD: the existing regimen remains unchanged for at least 12 weeks post-implant to distinguish VNS-specific effects.

Biologic Washout vs Continuation

The trial required biologic washout, which simplified outcome attribution but may not reflect real-world practice. Early compassionate-use data and case series suggest that VNS plus biologic (particularly tocilizumab or rituximab) does not produce excess immunosuppression signals. However, concurrent use with anti-TNF agents creates mechanistic redundancy — both pathways suppress TNF-alpha — and the incremental benefit is unclear.

My clinical approach: if a patient is on a biologic with partial response, I assess whether the primary driver is TNF-alpha (where VNS and biologics overlap) or whether there is a macrophage-independent inflammatory mechanism. VNS is particularly valuable in patients on JAK inhibitors with residual synovitis, because JAK inhibitors work intracellularly on T cells and fibroblasts while VNS targets macrophage surface receptors — a genuinely complementary mechanism.

Assessing Response to VNS

Disease activity monitoring post-implant should follow standard RA protocols:

• DAS28-CRP at baseline, 4 weeks, 12 weeks, and every 12 weeks thereafter
• Musculoskeletal ultrasound (Power Doppler synovial assessment) at baseline and 12 weeks
• CRP and ESR at each visit
• PRO-CLARITY or RAPID3 patient-reported outcomes

A DAS28-CRP improvement of ≥1.2 points at 12 weeks constitutes a meaningful response. Patients who do not reach this threshold by week 16 are unlikely to derive sustained benefit; implant removal is straightforward in this scenario.

Non-Invasive VNS as a Bridge Approach

For patients who are not ready for surgical implantation, transcutaneous auricular VNS (taVNS) — stimulation of the auricular branch of the vagus through an ear clip device — offers a non-invasive alternative. The evidence in RA is much weaker than for implanted VNS, consisting primarily of small open-label studies. However, taVNS may serve as:

1. A diagnostic test of VNS responsiveness before surgical commitment
2. A non-invasive option for patients with high perioperative risk
3. An adjunct during the waiting period for surgical scheduling

In my practice, I use taVNS as a bridge in motivated patients: 20-minute daily sessions at 25 Hz, left ear (cymba conchae stimulation point), monitored by weekly HRV tracking via wearable. A documented HRV improvement over 4-6 weeks, combined with subjective symptom improvement, supports the case for surgical implantation.

For a broader overview of non-invasive stimulation techniques and their evidence base, see our vagus nerve stimulation overview.

The Bioelectronic Medicine Landscape Beyond VNS

The success of VNS in RA has accelerated research into what is now called bioelectronic medicine — the use of electronic devices to modulate peripheral nerve function for therapeutic purposes. SetPoint Medical’s approach to RA was the first proof-of-concept at scale, but the pipeline has expanded substantially.

Ongoing clinical programs include:

• Alpha7 nAChR agonists — small molecule drugs that mimic the VNS mechanism without requiring a device (GTS-21, choline supplementation at pharmacological doses)
• Ultra-low frequency ultrasound nerve stimulation — non-invasive splenic nerve targeting (preclinical stage)
• Closed-loop VNS — devices that automatically titrate stimulation based on real-time HRV or biomarker feedback (in development)

The broader principle — that autoimmune joint disease can be treated by modulating the neural reflex arc controlling inflammation — has survived clinical testing. The question now is not whether the approach works, but how to deliver it most precisely, safely, and cost-effectively.

Related Articles

• Vagus Nerve Stimulation: Clinical Overview and Non-Invasive Techniques
• The Cholinergic Anti-Inflammatory Pathway in Chronic Illness
• Immunology of Autoimmunity: Mechanisms and Targets
• Apheresis for Autoimmune Disease
• Vagus Nerve and Chronic Fatigue: Clinical Applications

References

1. Tracey KJ. The inflammatory reflex. Nature. 2002;420(6917):853-859. doi:10.1038/nature01321
2. Koopman FA, Stoof SP, Straub RH, Van Maanen MA, Vervoordeldonk MJ, Tak PP. Restoring the balance of the autonomic nervous system as an innovative approach to the treatment of rheumatoid arthritis. Mol Med. 2011;17(9-10):937-948. doi:10.2119/molmed.2011.00065
3. Koopman FA, Chavan SS, Miljko S, et al. Vagus nerve stimulation inhibits cytokine production and attenuates disease severity in rheumatoid arthritis. Proc Natl Acad Sci USA. 2016;113(29):8284-8289. doi:10.1073/pnas.1605635113
4. Genovese MC, Gaylis NB, Sikes D, et al. Safety and efficacy of neurostimulation with a miniaturised vagus-nerve stimulator in patients with multidrug-refractory rheumatoid arthritis: a two-stage multicentre, randomised pilot study. Lancet Rheumatol. 2020;2(9):e527-e538. doi:10.1016/S2665-9913(20)30172-7
5. van Maanen MA, Vervoordeldonk MJ, Tak PP. The cholinergic anti-inflammatory pathway: towards innovative treatment of rheumatoid arthritis. Nat Rev Rheumatol. 2009;5(4):229-232. doi:10.1038/nrrheum.2009.31
6. Courties A, Berenbaum F, Sellam J. The phenotype of pain in rheumatoid arthritis and the nervous system at the crossroads of inflammation. Joint Bone Spine. 2021;88(4):105164. doi:10.1016/j.jbspin.2021.105164
7. Bassi GS, Kanashiro A, Coimbra NC, Terrando N, Möller T, Ulloa L. Anatomical and clinical implications of vagal modulation of the spleen. Neuroscience. 2020;425:174-184. doi:10.1016/j.neuroscience.2019.11.013