Evidence builds for long-term ineffectiveness of steroid shots for knee OA

SAN DIEGO – Real-world, nontrial research confirms the findings of a high-profile study released earlier in 2017: Corticosteroid shots are ineffective in the long term for knee osteoarthritis. In fact, researchers found a greater likelihood of a worsening condition in knees treated with the injections.

“Our findings are consistent with the latest randomized, controlled trial,” said study coauthor Jie Wei, MD, of Central South University in Changsha, China. She spoke in a plenary presentation about the study findings at the annual meeting of the American College of Rheumatology.

The use of corticosteroids for knee OA is a controversial topic. As Dr. Wei noted, there has been wide disagreement among medical societies about whether the treatment is useful in the long term for patients with pain flare-ups.

^{Thinkstock/Zinkevych}

For the randomized, controlled study released in 2017, researchers tracked 140 patients aged 45 and older with inflammation of the synovial membrane. They were randomly assigned to injections of intra-articular triamcinolone or a placebo.

After 2 years of injections every 12 weeks, there was no difference in reported pain between the intervention and control groups. Also, those who received injections lost more cartilage (JAMA. 2017 May 16;317[19]:1967-75).

Researchers launched the new study to seek insight through a real-life cohort. They examined findings from the Osteoarthritis Initiative, a longitudinal study of 4,796 patients aged 45-79 at four U.S. clinics with knee OA or high risk of knee OA. Patients underwent annual examinations at baseline and annually for 4 years.

In an adjusted marginal structural analysis, knee replacement or worsening of Kellgren Lawrence grade at the tibial femoral joint was more likely in 149 injection knees than 2,191 noninjection knees (odds ratio, 5.74; 95% confidence interval, 2.01-16.42).

Knee replacement or joint space width worsening at the tibial femoral joint was also more likely in 120 injection knees than 2,112 noninjection knees (OR, 1.64; 95% CI, 0.91-2.93).

In another analysis, researchers tracked 134 injection knees (58 whose OA progressed) and 498 noninjection knees (132 whose OA progressed) for up to 8 years. After adjustment, the injection knees were more likely to have progressed (hazard ratio, 1.60; 95% CI, 1.21-2.12,).

“Several explanations may account for our study findings,” Dr. Wei said. One possibility, she said, is that corticosteroids may hurt chondrocytes by, among other things, inducing apoptosis and synovial membrane inflammation.

It’s also possible, she said, that patients may feel pain relief after injections and subsequently boost the risk of OA progression by increasing their physical activity.

“We need to know what types of physical activities may increase the OA progression,” she said. “Did patients who received steroid injection indeed increase this type of physical activity compared to subjects without steroid injection?”

Dr. Wei noted the study’s limitations, including the fact that patients who received injections had more pain at baseline, potentially indicating they had worse structural lesions that are more susceptible to progression.

The study authors reported no relevant disclosures. The National Natural Science Foundation of China funded the study. The Osteoarthritis Initiative is a partnership between the National Institutes of Health and Merck, Novartis, GlaxoSmithKline, and Pfizer.

SOURCE: Lei G et al. ACR 2017 abstract 1788