Light Therapy Aids Spinal Cord Recovery in Rats. Now Scientists Want to Try It in People.

When a vertebral fracture or a bullet severs the spinal cord, the damage is only half done at the moment of impact. Bone, disc, or projectile cuts tissue, and motor, sensory, and autonomic function drop at the scene. Over the following weeks and months, a second biochemical wave of swelling, inflammation, and disrupted blood flow kills nerve cells the original trauma spared. Current treatment is largely limited to surgical stabilization, anti-inflammatory drugs, and rehabilitation. A 2026 narrative review in Photobiomodulation, Photomedicine, and Laser Surgery argues the field's most promising next move is to act on that second wave, using light. (Mary Ann Liebert review, 2026)

That route is called photobiomodulation, or PBM. It is a low-level laser or red- and near-infrared light therapy: nonionizing, non-thermal, and absorbed by naturally occurring light-sensitive molecules in tissue, called chromophores. The cells that absorb it are thought to release more of the molecule the body uses to store and deliver energy, adenosine triphosphate. More ATP, in the working theory, gives damaged neurons enough fuel to survive the secondary cascade. The catch is that nobody has shown this works in a human spine yet.

The evidence is rodent-strong and human-thin. A 2020 systematic review and meta-analysis pooled the animal data, and a separate preclinical study tested both implantable and skin-surface PBM delivery in rat spinal cord injury. Across multiple dosing parameters, the rodent work is consistent enough that the 2026 review calls functional recovery "extensively demonstrated" in animals. What is missing is the next layer: large-animal efficacy, large-animal safety, and a clinical-grade delivery system a surgeon can actually use. The review's own required next steps are those three things, in that order.

One human trial is now in the field. A multicenter randomized controlled trial in acute cervical spinal cord injury tested PBM delivered intraoperatively. Once the cord was exposed for surgical stabilization, a diffusing fiber placed about one centimeter above the tissue beamed 810-nanometer light at 300 milliwatts for 30 minutes a day, for seven consecutive days, in patients classified ASIA grade C (incomplete injury with some motor function preserved below the level of damage). The co-primary outcomes are tracked at three, six, and twelve months. Its dosimetry is the working answer to the "what would a clinical-grade system even look like" question the review raises.

Readers weighing the field's optimism should also weigh the authors behind it. The 2026 review's conflict-of-interest disclosure lists a provisional patent on photobiomodulation technology for traumatic spinal cord injury. That does not invalidate the science, but it does mean the framing of PBM as a "highly effective therapy" comes from a group with a financial stake in that framing landing. Independent large-animal replication is what would change that equation.

The regulatory road map, at least, is not blank. The FDA's 510(k) guidance for Photobiomodulation Devices already defines the clearance pathway for a PBM device. PBM is a treatment modality, not a single product: the device that ends up in front of the FDA will need to demonstrate dosimetric equivalence and safety for the specific spinal indication. That pathway was not built for spinal cord injury; it was built for the broader category. A spinal-specific submission would be new work.

A useful comparison point sits next door. Onward Medical's ARC-EX system, a transcutaneous spinal cord stimulator and not a light device, received FDA de novo clearance in December 2024 for chronic spinal cord injury. ARC-EX targets a different patient (months or years out, not the acute hospital window) and a different mechanism (electrical, not photonic), so it is not a head-to-head benchmark. It is, however, the proof that a device-based SCI therapy can clear the FDA, and the gauge for what the next PBM sponsor will be measured against.

The rodent wins are real. The patient wins are still unwritten. What closes the gap first will be a larger multicenter study using the 810 nm, 300 mW, 30-minute protocol from the acute cervical trial, then a large-animal model that reproduces the rodent signal, and finally a spinal-specific 510(k) submission under the existing PBM guidance.