The knee is the joint most often treated with low-dose radiation therapy (LDRT) for osteoarthritis and the one with the strongest clinical evidence base. The 2026 American Radium Society Appropriate Use Criteria (ARS AUC) synthesized the modern evidence and concluded that 60–90% of patients with knee OA respond to treatment. This page explains how the treatment works for the knee, what the trials show, who qualifies by Kellgren-Lawrence grade, and how it compares to your other options.
Knee osteoarthritis pain is driven partly by synovial inflammation — an activated immune response inside and around the joint that releases cytokines and inflammatory mediators causing pain and swelling. Low-dose radiation modulates this inflammatory environment at the cellular level, calming the signals that drive chronic pain. The mechanism is anti-inflammatory, not structural. LDRT does not regrow cartilage, reverse joint space narrowing, or remove bone spurs. Its goal is to reduce pain and improve function — often enough that patients can move more comfortably, reduce their reliance on NSAIDs or injections, and in some cases defer or avoid surgery. For a general explanation of how the treatment works, see our treatment overview.
What changes: inflammatory behavior. What does not: cartilage and bone are not rebuilt.
When people hear radiation, they often imagine cancer treatment and cell killing. Low-dose radiotherapy for arthritis uses a much smaller dose and a different biological goal.
In an inflamed joint, the synovial lining and its tiny blood vessels help recruit immune cells. Those cells release cytokines and other signals that can keep the tissue irritated and amplify pain.
At low doses, radiation appears to change that local conversation. Experimental studies show less immune-cell adhesion to the vessel wall, less migration into tissue, and shifts in macrophage polarization and signaling. The full pathway is still being worked out.
It does not rebuild cartilage or erase osteoarthritis on an X-ray. Instead, over the following weeks, the synovium may become less reactive and send fewer pain-amplifying signals.
The simplest way to remember it is this: low-dose radiotherapy is not trying to destroy the joint. It is trying to turn down local inflammation so the joint can hurt less.
LDRT is not a first-line treatment. The ARS AUC describes the typical appropriate candidate as someone who has completed a course of physical therapy, tried oral or topical NSAIDs, and received at least one corticosteroid or hyaluronic acid injection without adequate sustained relief — and who either is not yet ready for or appropriate for knee replacement. See the full candidacy criteria for details.
LoRD-KNeA (ASTRO 2025 conference abstract): A sham-controlled randomized trial conducted at three Korean academic centers enrolling 114 patients with symptomatic knee OA. Patients received either 3 Gy in six fractions of 0.5 Gy (the modern standard dose) or a sham procedure identical in every detail except radiation delivery. At 4 months: responder rate 70% (LDRT) vs 42% (sham), p = 0.014. This is the largest sham-controlled RCT using the current dose protocol and the primary contemporary evidence supporting the ARS AUC's "appropriate" rating for knee OA.
Fazilat-Panah et al. (2025): A double-blind, sham-controlled randomized trial, same 3 Gy in six fractions of 0.5 Gy protocol. All primary and secondary outcomes — pain score, function, performance status, and analgesic consumption — were statistically significant in favor of LDRT (p < 0.01 across all measures), with benefit sustained through the 6-month follow-up period. Independently replicates the LoRD-KNeA result with a different investigator team and patient population. Full evidence summary at the evidence page.
Makarova et al. — 10-year follow-up: 292 patients with early knee OA randomized to LDRT (4.5 Gy total in 0.45 Gy fractions) plus standard pharmacotherapy (NSAIDs and SYSADOA) versus pharmacotherapy alone. Both arms received identical background pharmacotherapy — LDRT was the only treatment difference. At 10 years: 9.5% of LDRT patients became disabled vs 17.8% of controls. Radiographic deterioration of ≥2 KL grades: 31.3% LDRT vs 44.9% control at 10 years. MRI sub-study showed preserved cartilage thickness and less bone marrow edema in the LDRT group. A 2023 analysis of the same cohort found a 67% reduction in total knee arthroplasty over 9 years in the LDRT group versus controls — the strongest long-term data on LDRT's potential to delay or avoid surgery.
Mücke et al. (multi-center retrospective): An analysis of 4,544 patients with knee osteoarthritis treated at German centers — the largest published cohort for this indication. At 3 months post-treatment, approximately 60% of patients reported meaningful pain reduction. At 1 year, approximately 40% maintained durable benefit. This series provides the best available real-world population-level response rate estimate outside a randomized trial.
U.S. Series (Koneru; Scripps): Early U.S. institutional experience with LDRT for knee OA used the same 6-fraction, 0.5 Gy/fraction protocol now endorsed in the ARS AUC. These series confirmed that the European results are replicable in American patients with knee OA and contributed to the growing evidence base that informed the 2026 ARS AUC development process.
Mahler 2019 and Minten 2018: Both randomized controlled trials found no significant benefit from LDRT for knee OA. Both used an older 1 Gy/fraction protocol — twice the dose per session of the current standard. The ARS AUC identifies the 1 Gy/fraction protocol as a specific methodological limitation. Trials using ≤0.5 Gy/fraction have generally shown benefit; the two trials using 1 Gy/fraction were null. The negative results of Mahler 2019 and Minten 2018 are not evidence against LDRT at the current dose; they are evidence against an abandoned dose protocol.
The Kellgren-Lawrence (KL) scale grades osteoarthritis severity on X-ray from 0 (normal) to 4 (severe bone-on-bone). The ARS AUC considers LDRT appropriate for KL grades 1 through 3 as the primary range.
Full candidacy criteria including age thresholds, prior treatment requirements, and contraindications are on the candidacy page.
vs. Corticosteroid injections: Cortisone works within days and lasts 4–12 weeks. LDRT takes 4–12 weeks to reach full effect but may last 6–24 months. The two are not mutually exclusive — a cortisone injection can bridge to an LDRT course while the radiation effect develops.
vs. Total knee arthroplasty: TKA is the definitive treatment for severe end-stage knee OA in appropriate surgical candidates. LDRT does not replace it. The Makarova 9-year data showing a 67% reduction in arthroplasty risk is compelling but should be interpreted cautiously — it comes from one early-OA cohort. What the data do support is that LDRT may meaningfully delay the point at which surgery becomes necessary in some patients.
A standard LDRT course for knee OA is 6 sessions of 0.5 Gy each (3 Gy total), delivered two to three times per week over approximately two weeks. Each session takes 10–20 minutes in the department; beam-on time is 2–5 minutes. No anesthesia, no needles, no IV lines, no recovery period. Most patients drive themselves. For radiation safety information and detailed side effects and timing, see the dedicated pages.
Most patients who respond to LDRT notice improvement 4–12 weeks after completing the course, not immediately after the final session. The anti-inflammatory effect develops gradually. In published series, peak benefit occurs around 3 months post-treatment. A 3-month follow-up appointment is typically used to assess whether a clinically meaningful response has occurred. Not feeling better immediately after the final session does not mean treatment has not worked.
When LDRT helps, relief typically lasts 6–24 months from a single course. Re-treatment after relapse is described in the European literature and is considered acceptable in appropriate patients because the cumulative dose remains very low.
Use our provider directory to find a radiation oncology center offering LDRT for knee osteoarthritis near you.
See also: All joints treated with LDRT | Related large joints: Hip | Shoulder
Medically reviewed by: Robert Warren Floyd, M.D., Ph.D. — Resident Physician, Radiation Oncology, The University of Texas MD Anderson Cancer Center. Last reviewed: September 8, 2026. Independent personal project — not affiliated with or endorsed by MD Anderson Cancer Center. Funding & independence disclosure.
This content is for education only and is not medical advice. Always consult a qualified healthcare provider for diagnosis and treatment decisions.