Ultrasound assessment of response to intra-articular therapy in osteoarthritis of the knee

Objective. Assessment of the synovium in patients with knee OA is of great potential value for clinical trials. Ultrasonography could provide this but few data exist on its ability to assess synovial response to therapies. The aim of this study was to examine whether US can detect synovial response to IA corticosteroid (IACS) therapy and to explore associations between synovial characteristics and symptoms. Methods. A total of 35 people with ACR radiographic knee OA were included, including those who required an injection of 80 mg of IA methylprednisolone. All participants completed a visual analogue scale for pain and underwent US of the knee at baseline, 1 and 4 weeks. Minimum clinically important improvement (MCII) in pain was >= 20 mm. Results. One week of data were available for 33 patients (19 received IACS and 14 others). Synovial thickness (ST) decreased in 16 IACS patients and 2 others [mean between-group difference 4.7 mm (95% CI 1.1, 8.2), P = 0.012]. Absolute reduction was not associated with absolute reduction in pain (r = 0.20, P = 0.289), but decreased ST was substantively associated with reduction in pain greater than or equal to the MCII (52.9% vs 23.1%, P = 0.098, phi = 0.30). The power Doppler score decreased in 13 IACS patients and 3 others {median change in IACS patients -1.0 [interquartile range (IQR) -5.0-0.0], others 0.0 [-0.3-1.3], P = 0.004}. Absolute changes in pain and power Doppler score were weakly associated (rho = 0.36, P = 0.049) and a decreased power Doppler score was associated with reduction in pain greater than or equal to the MCII (64.3% vs 18.8%, P = 0.011, phi = 0.46). Conclusion. Ultrasonography can detect short-term synovial response in knee OA. In particular, power Doppler score may be both responsive to and associated with pain, warranting further investigation.