Can low-dose radiotherapy reduce knee osteoarthritis symptoms?

At four months, responder rates were 70.3% with 3 Gy versus 41.7% with sham (p=0.021); the 0.3-Gy rate was 58.3% (p=0.168 versus sham).

Original responder rates at 1, 4, 8 and 12 months and 95% confidence intervals versus sham; the 3-Gy difference was not significant at 12 months.
Kim, Ro, Wang et al., BMC Medicine, 2026 · DOI: 10.1186/s12916-026-05287-9 · Figure 2.
Evidence & Original Source Image use terms

Original CONSORT diagram of allocation and analysis populations.
Kim, Ro, Wang et al., BMC Medicine, 2026 · DOI: 10.1186/s12916-026-05287-9 · Figure 1.
Evidence & Original Source Image use terms

Original WOMAC, VAS and PGA trajectories; continuous secondary outcomes did not show significant between-group differences.
Kim, Ro, Wang et al., BMC Medicine, 2026 · DOI: 10.1186/s12916-026-05287-9 · Figure 3.
Evidence & Original Source Image use terms

Participants and Methods

A multicenter randomized sham-controlled trial compared total doses of 0.3 Gy and 3 Gy, each in six fractions, with sham treatment. The primary endpoint was the four-month OMERACT–OARSI responder rate in a modified intention-to-treat analysis. The trial randomized 114 patients aged at least 50 years with primary knee OA and baseline walking pain of 50–90/100 into three groups.

Main Findings

At four months, responder rates were 70.3% with 3 Gy versus 41.7% with sham (p=0.021); the 0.3-Gy rate was 58.3% (p=0.168 versus sham). The 3-Gy responder advantage persisted at eight months but was not statistically significant at twelve months. Continuous clinical outcomes, including WOMAC and VAS, and MRI measures showed no significant between-group differences. No treatment-related toxicity was reported.

The point of reading on surgery

The study assesses a nonoperative symptom-treatment option for selected patients; it does not establish structural restoration or avoidance of arthroplasty.

Limitations

Short-term responder benefit does not establish cartilage regeneration, long-term disease modification or a standard treatment for all patients. The absence of reported treatment-related toxicity does not exclude rare or long-term harms. This article-in-press version may receive editorial changes.

Let's put it briefly.

What did this study confirm?

At four months, responder rates were 70.3% with 3 Gy versus 41.7% with sham (p=0.021); the 0.3-Gy rate was 58.3% (p=0.168 versus sham).

Can it be applied directly to individual patients' treatment decisions?

Short-term responder benefit does not establish cartilage regeneration, long-term disease modification or a standard treatment for all patients.

Keywords

knee osteoarthritis
randomized trial
low-dose radiotherapy
symptom response

Abstract · Original text

Byoung Hyuck Kim, Du Hyun Ro, Joon Ho Wang, Dae-Hee Lee, Kichul Shin, Min Jung Kim, Tae Woo Kim, Moon Jong Chang, Dong Hyun Kim, Mira Han, Ji Hyun Lee, Ji Hee Kang, Ji Young Kim, Eun-Hee Hong, Seong-Jun Cho, Hyuk-Soo Han, Hak Jae Kim, Won Park. Single course low-dose radiation therapy in knee osteoarthritis (LoRD-KNeA): a randomized controlled trial. BMC Medicine. 2026. doi:10.1186/s12916-026-05287-9.

Excerpt from the original abstract

Low-dose radiotherapy (LDRT) has been used as a non-pharmacological treatment for osteoarthritis (OA), but robust evidence from randomized, sham-controlled trials remains limited.…

Read abstract at source Terms of use

The original paper.

Single course low-dose radiation therapy in knee osteoarthritis (LoRD-KNeA): a randomized controlled trial

DOI · 10.1186/s12916-026-05287-9 Based data

Reviewed academic journal articles