Vagus nerve stimulation has become one of the most widely adopted neuromodulation therapies in clinical practice, with more than 130,000 patients carrying implanted devices as of 2024. Yet the human cervical vagus nerve contains over 100,000 fibers, with efferent fibers reaching most visceral organs and afferent fibers returning signals to the brainstem. When an extraneural electrode activates the entire nerve, current spreads indiscriminately, and because larger motor fibers have lower activation thresholds than smaller fibers, stimulation can trigger unintended muscle contractions, cough, dysesthesia, intestinal inflammation, and disturbances of heart and breathing rate. A new review in Bioelectronic Medicine examines why the next generation of bioelectronic therapies depends on shrinking nerve interfaces to fit branches measuring one millimeter or less in diameter, and surveys the materials, fabrication methods, and device concepts that could get them there.
The review begins with the anatomy that constrains every design. Peripheral nerves are cables of fiber bundles called fascicles, each wrapped in a perineurium sheath and bundled together with blood vessels, Schwann cells, and immune cells inside an epineurium. Fibers may be myelinated or unmyelinated, and both diameter and myelination govern conduction velocity and activation threshold. Although larger fibers are easier to activate and produce higher amplitude action potentials that are easier to record, peripheral nerves contain far more small, unmyelinated C-type fibers, generally located in the nerve interior. These fibers act as nociceptors and afferents and play significant roles in regulating functions such as glycemic control, pain transduction, and interoception. Activating and recording from them remains difficult because of their higher stimulation thresholds and their position deep within the nerve.
The historical arc stretches back further than most readers might expect. Ancient Egyptians, Greeks, and Romans used electric fish for headache and pain relief, and Luigi Galvani’s serendipitous observation of a frog nerve contracting near a metal knife and spark established the principle of electrical neural interfacing. Early transcutaneous vagal and cervical sympathetic stimulation for epilepsy in the 1880s was controversial, producing bradycardia and syncope, and the resulting lack of scientific credibility stalled the field until the mid-twentieth century. Clinical applications emerged in the 1960s, including phrenic nerve stimulation for diaphragmatic pacing and trigeminal nerve stimulation for facial pain. The FDA subsequently approved a phrenic nerve stimulator in 1987, a vagus nerve stimulator with a helical cuff electrode for drug-resistant epilepsy in 1997 and for treatment-resistant depression in 2005, and sacral nerve stimulators for urinary urge incontinence, also in 1997.
Today’s approved devices converge on a handful of form factors and materials. Interfaces may connect to an implantable pulse generator through a flexible lead, like a cardiac pacemaker, or integrate electronics and electrodes into a single leadless unit, as in wireless tibial nerve stimulators for overactive bladder. Cuff electrodes wrap tubular insulating structures around the nerve circumference, from multi-turn helical cuffs to cuffs with radially distributed disc electrodes. Insulation is typically medical-grade silicone or polyurethane applied by molding and extrusion; wiring uses coiled stainless steel, MP35N, or platinum alloys; and electrode sites are punched metal foils embedded in silicone rubber, often platinum-iridium or sputtered iridium oxide for stimulation. Crucially, all approved peripheral nerve interfaces are extraneural, and current cuff-style devices are compatible only with nerves larger than about two millimeters in diameter.
The physics of selectivity pushes designers toward the nerve itself. Placing an electrode as close as possible to the signal source reduces power consumption and increases selectivity for both stimulation and recording, but the nerve’s structure and the surrounding tissue impose tradeoffs involving the foreign body reaction and electrode-tissue stability. The review categorizes interfaces by electrode placement: extraneural, outside the epineurium; intraneural, inside the epineurium but outside the fascicles; intrafascicular, within individual fascicles; and regenerative, embedded within a bisected, regenerating nerve. Greater invasiveness correlates with greater achievable selectivity and resolution, but also with greater injury. Smaller nerves demand close electrode positioning, adequate insulation without signal leakage, no constrictive pressure, mechanical compliance matched to nerve tissue, and thin layers that still provide moisture barriers and low trace resistance.
Materials innovation is central to miniaturization. Polydimethylsiloxane, the workhorse silicone of early interfaces, offers a tunable Young’s modulus but low tensile strength, so thicker layers are needed to avoid breakage, making it hard to use below one millimeter. Parylene C, deposited by chemical vapor deposition as a conformal pinhole-free film, can be thermoformed after microfabrication into three-dimensional shapes, enabling small-diameter helices down to 0.25 millimeters and nerve cuff structures. Polyimide, spin-coated in layers of roughly 5 to 80 micrometers, can also be thermoformed. Liquid crystal polymers show superior resistance to water absorption and stronger interlayer adhesion than polyimide or Parylene C under heated in vitro conditions. Shape memory polymers can start stiff for implantation and then soften at body temperature to match tissue properties, an approach demonstrated on rat pelvic nerves roughly 200 micrometers in diameter.
On the conductive side, thin film metals such as gold, platinum, and titanium deposited in nanometer-scale thicknesses provide micron-scale patterning precision, with microcracked films tolerating greater mechanical strain than continuous ones. Liquid metals like eutectic gallium-indium survive deformation but cannot be scaled to fine features and are unsuitable at the electrode interface because gallium oxidizes in aqueous environments. Carbon nanotube and graphene coatings offer large surface area, though their potential cytotoxicity still requires investigation.
The most striking experimental devices attack the small-nerve problem from several directions. Wrapped microwires achieved stable in vivo performance for up to 60 days on the mouse cervical vagus nerve. Self-closing cuffs pre-formed by thermoforming have reached diameters as low as 86 micrometers on the rat vagus nerve, and a thermoformed Parylene C cuff implanted on the locust N5 nerve for two weeks caused no observable impairment in insect activity. Actuating cuffs close themselves using electrochemically activated polypyrrole, swelling hydrogels, or shape memory polymers that transform near body temperature. Microchannel devices with channels as narrow as 40 micrometers have recorded activity from unmyelinated C-fibers in teased branches of the rat vagus nerve. On the invasive end, Utah slanted electrode arrays have been pneumatically inserted into rat sciatic and cat pudendal nerves, carbon fiber arrays have penetrated the rat cervical vagus nerve, and transverse intrafascicular electrode arrays have been placed in the median and ulnar nerves of hand amputees in clinical trials to provide sensory feedback for prosthesis control.
Regenerative interfaces take a still different route, integrating electrodes into transected nerves and letting fibers regrow through sieve pores, microchannels, or scaffolds. Because regeneration takes two to three weeks with stabilization requiring roughly six weeks, applications are more limited. Wireless integration is also advancing: millimeter-scale systems including an ultrasonically powered cuff stimulator of 1.7 cubic millimeters, a near-field communication cuff tested chronically in the canine cervical vagus nerve, and bioresorbable stimulators designed to be metabolized by the body after temporary use.
The authors are careful about what stands between these laboratory demonstrations and the clinic. No microfabricated peripheral nerve interface has yet reached an FDA-approved device. Surgical difficulty rises as target nerves shrink, requiring preoperative localization with magnetic resonance neurography and ultrasound, and intrafascicular and regenerative approaches place heavy demands on dexterity. Engineering hurdles compound the problem: the roughly 27,000 axons of a rat sciatic nerve vastly exceed the electrode counts achievable per unit area, shrinking device footprints raises impedance and crosstalk, multilayer thin films risk delamination and fatigue under chronic motion, and the foreign body reaction can encapsulate electrodes, raising impedance and accelerating failure. Translation, the review concludes, will require cross-disciplinary collaboration across materials, fabrication, surgical technique, and tissue integration, with a few prospects already progressing in animal and early human studies toward interfaces capable of stimulating and recording nerves of one millimeter or less.
Subject of Research: Electrode-based peripheral nerve interfaces for small-diameter nerves in bioelectronic medicine
Article Title: Scaling beyond the vagus nerve: historical and contemporary progress on electrode-based small-diameter peripheral nerve interfaces
Article References: Li, M., Ly, C., Wen, J., Rustogi, P., & Meng, E. (2026). Scaling beyond the vagus nerve: historical and contemporary progress on electrode-based small-diameter peripheral nerve interfaces. Bioelectronic Medicine, 12(1), Article 6. https://doi.org/10.1186/s42234-025-00199-0
Image Credits: AI Generated
DOI: 10.1186/s42234-025-00199-0
Keywords: peripheral nerve interfaces, vagus nerve stimulation, bioelectronic medicine, cuff electrodes, microfabrication, Parylene C, shape memory polymers, electrode coatings, intraneural electrodes, regenerative interfaces, FDA-approved devices, foreign body response
News Source: Denise Maddox. (October 29, 2026). Beyond the vagus nerve: the push for electrodes small enough to reach tiny peripheral nerves. Scienmag.
