Background There is a lack of adequate outcome measures of treatment response in Knee Osteoarthritis (KOA). The role of Ultrasonography (US) in the management of inflammatory articular and periarticular diseases is well established, but it has been less extensively used in OA.
Objectives To develop a new high-resolution US protocol for the assessment of structural and inflammatory changes in primary KOA and to evaluate its inter-observer reliability.
Methods A scanning protocol was developed by consensus among 3 senior Rheumatologists with similar high levels of experience in musculoskeletal US. The protocol includes both inflammatory elementary lesions (synovitis, power Doppler, Baker's cyst) and structural damage lesions (medial collateral ligament bulging, meniscal lesions and osteophytes) measured as dichotomous variables (present/absent). Cartilage thickness (in millimeters) is assessed at the notch and 10mm towards the lateral and medial condyles (mean value of 3 measurements). The OMERACT definitions for elementary lesions were used whenever available. In addition, global scores were calculated for joint inflammation (range: 0–3) and structural damage (range 0–11). Patients with primary KOA according to the American College of Rheumatology classification criteria underwent Knee US assessment in the same day by the 3 experts using the same device. For dichotomous parameters, Inter-observer agreement was estimated by the free-marginal multirater Kappa and the proportion of agreement (Agt). For continuous parameters two-way mixed effects models were used and absolute-agreement Intraclass Correlation Coefficients (ICC) was measured between individual ratings and global scores.
Results In total, 7 patients were included (14 knees), corresponding to 42 US assessments by the 3 readers. Patients were 6 females and 1 male, with a mean age of 58.6 years. Concerning inflammatory lesions, agreement was good for synovitis (κ=0.714; Agt=85.7%) and Baker cyst (κ=0.714;Agt=85.7%). All power Doppler analysis were negative, thus achieving excellent agreement (κ=1; Agt=100%). As a result, agreement was good for the inflammatory score (ICC: 0.774; 95% CI). Results were not as homogenous for structural damage lesions. In one hand, agreement was excellent for medial meniscus cyst (κ=0.810; Agt=90.5%), good for lateral and medial femoral osteophytes [κ=0.619; Agt=80.9%) and κ=0.714; Agt=85.7%) respectively] and for lateral meniscus extrusion and cyst κ=0.619 and 0.619; Agt=81.0%). On the other hand, it was poor for lateral meniscus fracture κ=0.143; Agt=57.1%). For cartilage thickness, agreement was good for notch and medial condyle (ICC=0.652 and 0.619). Finally, the overall structural score agreement was moderate (ICC: 0.484).
Conclusions The proposed protocol has shown to be reliable in patients with KOA, therefore a promising tool to be used both in clinical practice and for research purposes. Long-term reproducibility and responsiveness to change is needed to become a valid outcome measure, thus filling in an important gap in OA assessment.
Disclosure of Interest None declared