Exploratory Clinicopathological Associations Between Prespecified Clinical–Structural Domains and Synovial Tissue Findings in End-Stage Knee Osteoarthritis

Background: Knee osteoarthritis (OA) is a heterogeneous whole-joint disease. This exploratory clinicopathological study evaluated associations between four prespecified continuous clinical and structural domains and synovial tissue findings in patients undergoing total knee arthroplasty (TKA) for primary end-stage knee OA. Methods: This prospective single-center study included 79 patients undergoing cemented TKA. Standardized inflammatory, metabolic/lipomatous, osteocartilaginous/remodeling, and mechanical domain scores were evaluated against suprapatellar histopathological and intraoperative macroscopic synovial outcomes. Associations were examined using Spearman correlations and age- and sex-adjusted regression models with false-discovery-rate (FDR) correction. Results: Low-grade synovitis was present in 63.3% of knees and high-grade synovitis in 31.6%. Unadjusted analyses showed weak correlations between the inflammatory domain and macroscopic synovitis (ρ = 0.24, p = 0.032) and between the metabolic/lipomatous domain and synovial lipomatosis severity (ρ = 0.23, p = 0.042). Adjusted estimates were β = 0.24 (p = 0.043) for inflammatory domain–Krenn synovitis, β = 0.23 (p = 0.050) for inflammatory domain–macroscopic synovitis, and OR = 1.57 (p = 0.044) for metabolic/lipomatous domain–lipomatosis severity. No associations were observed for the osteocartilaginous/remodeling or mechanical domains. None of the six prespecified associations remained statistically significant after FDR correction. Conclusions: End-stage knee OA showed substantial synovial heterogeneity, but no prespecified domain–tissue association remained significant after multiple-testing correction. The evaluated domains should therefore be considered exploratory clinicostructural constructs rather than validated biological or diagnostic classifications. Further multicenter longitudinal studies integrating imaging, molecular biomarkers, and tissue assessment are needed.