Effusion and synovial hyperplasia: Distinct constructs in early and late-stage knee OA.

OBJECTIVE: Ultrasound (US) permits the assessment of synovial effusion and synovial thickening (hyperplasia) in knee osteoarthritis (OA). We aimed to measure the associations of these features with radiographic OA severity and patient-reported knee-specific pain using a quantitative US image analysis approach. METHODS: Patients with symptomatic knee OA were included. The Knee Injury and Osteoarthritis Outcome Score (KOOS) pain subscale, Kellgren-Lawrence (KL) grade, and lateral suprapatellar recess B-mode images were acquired for Outcome Measures in Rheumatology (OMERACT) measurement and grading. US images of effusion-synovitis were manually segmented to calculate effusion and hyperplasia area. Correlation was assessed, and associations with KL grade/KOOS pain were modelled using multivariable linear and quadratic regression with and without adjustment for age, sex, and body mass index (BMI). RESULTS: Quantitative measures of effusion and hyperplasia area correlated strongly (ρ 0.83 to 0.86) with OMERACT grades and effusion-synovitis depth. Adjusting for age, sex, and BMI, synovial effusion and hyperplasia area are non-linearly related to radiographic severity, a pattern also observed in models of OMERACT global synovitis and effusion-synovitis depth. KOOS pain was associated with effusion-synovitis depth (β=-0.53) and borderline associated with OMERACT hyperplasia (β=-5.55), but not with OMERACT global synovitis, effusion, nor any quantitative effusion-synovitis area measures. CONCLUSION: Minimal US effusion-synovitis among knee OA patients with late-stage disease may indicate maladaptive remodeling of the synovium instead of reduced inflammation. Discriminating between patients with and without maladaptive synovial remodeling in early and late-stage disease should be a priority, as it may improve our ability to predict joint failure.