Researchers Propose Nerve-Targeted Radiofrequency Treatment for Knee Osteoarthritis

A research team has proposed a staged test of pulsed radiofrequency applied directly to subchondral bone as a possible way to treat pain associated with knee osteoarthritis. The approach targets bone marrow lesions seen on magnetic resonance imaging, abnormalities that are associated with pain and structural progression in osteoarthritis but are not addressed by an approved therapy with a demonstrated structural effect. The article, published in Frontiers in Pain Research as a hypothesis-and-theory paper, describes a research program rather than reporting a successful clinical treatment. Its central proposal is that an electrode placed within the trabecular bone of the medial tibial plateau could alter a local circuit involving sensory nerves, osteoclasts and osteoblasts, potentially reducing both pain and the volume of bone marrow lesions.

The biological rationale combines observations from human osteoarthritis, animal models and other experimental settings. Subchondral bone is innervated by peptidergic sensory fibers, and these fibers in human osteoarthritis accompany vascular channels across the osteochondral junction. In mouse models, osteoclasts can induce sensory nerve growth into subchondral bone. Genetic experiments in mice have also linked this innervation to bone remodeling: removing sensory nerves reduced bone volume, while sensory knockout of EP4 or TrkA attenuated load-induced bone formation. Together, these findings support the hypothesis that sensory signaling may help sustain a pathological cycle of nerve growth and bone alteration. The authors emphasize, however, that much of this evidence comes from settings other than human osteoarthritis, so the proposed mechanism remains to be tested in an animal model designed specifically to evaluate the treatment.

Pulsed radiofrequency is described as a way to produce electric fields near the tip of an electrode. According to the hypothesis, these fields can reach hundreds of volts per centimeter within a few millimeters of the tip, at least four orders of magnitude greater than the approximately 0.07 to 2 millivolts per centimeter induced in bone-cell studies of pulsed electromagnetic fields. The comparison does not mean that the radiofrequency dose is already known to be safe or effective in bone. Its electroporation benchmark came from a non-neuronal cell line and requires bone-specific validation, while electric-field dosimetry and electrothermal modeling must first establish what tissue exposure would occur around the proposed electrode placement.

The initial laboratory stage is intended to include bone-specific electrothermal modeling and ex vivo dosimetry. The next stage would use a randomized, blinded, sham-controlled experiment in mice with destabilization of the medial meniscus, a model of knee osteoarthritis, together with functional measurements of neuronal activity and targeted sensory denervation. Large-animal dosimetry would follow before a human feasibility study. The authors envision a subsequent external pilot randomized trial using quantitative volumetric measurement of bone marrow lesions. This sequence separates questions about heat, tissue exposure, biological effects and procedural feasibility, but it also shows that the source provides a plan rather than evidence that the treatment already relieves pain, shrinks lesions or prevents structural progression. Even the limited human precedent remains incomplete: an uncontrolled series of 10 patients with femoral head osteonecrosis demonstrated that intraosseous electrode placement was technically feasible, but it did not report pain scores or adverse-event data and therefore did not establish safety.

A major unresolved issue is the proposed role of the nerve growth factor–TrkA signaling axis. The authors identify this pathway as a possible mediator of the effect, but its behavior is internally paradoxical. Nerve growth factor can appear reparative in necrotic bone while sensitizing sensory neurons in osteoarthritic joints. In an animal model, blocking anti-nerve growth factor treatment relieved osteoarthritis pain but was associated with worse subchondral bone and cartilage outcomes. The proposed program therefore includes prespecified criteria for success, failure and inconclusive results, with the balance between nerve repair and sensitization presented as the central test. The concept offers a technically specific way to investigate a sensory nerve–bone remodeling circuit, but efficacy, dose limits and safety in people with knee osteoarthritis cannot yet be inferred from the hypothesis or from the small feasibility series cited in the paper.

Subject of Research: A staged experimental proposal for using intraosseous pulsed radiofrequency to modulate sensory nerves, bone remodeling and marrow lesions in knee osteoarthritis.

Article Title: Pulsed radiofrequency neuromodulation of subchondral bone in knee osteoarthritis

Article References: Lana, J. F., Jorge, D., Protasio Netto, J., Pontes de Macedo, A., Pires, L., Costa, F., Mosaner, T., Sobreiro, P., Santos, D., Teixeira, A., & Rohof, O. (2026). Pulsed Radiofrequency Neuromodulation of Subchondral Bone in Knee Osteoarthritis. Frontiers in Pain Research, 7, 1968444. https://doi.org/10.3389/fpain.2026.1968444

Image Credits: AI Generated

DOI: 10.3389/fpain.2026.1968444

Keywords: knee osteoarthritis, pulsed radiofrequency, subchondral bone, bone marrow lesions, sensory nerves, nerve growth factor, TrkA signaling, bone remodeling, neuromodulation, mice, Pulsed, radiofrequency