TITLE:
Correlation of Disease Severity to Patient Synovial Fluid Characteristics and Development of a Simulated Synovial Fluid in Knee Osteoarthritis
AUTHORS:
Annie C. Bowles-Welch, Hazel Y. Stevens, Rebecca S. Schneider, Linda E. Kippner, Angela C. Jimenez, Theresa Kotanchek, Thanh N. Doan, David A. Frey Rubio, Carolyn Yeago, Hicham Drissi, Andrés J. García, Krishnendu Roy
KEYWORDS:
Osteoarthritis, Synovial Fluid, Potency Assays, Machine Learning
JOURNAL NAME:
Open Journal of Orthopedics,
Vol.16 No.3,
March
31,
2026
ABSTRACT: Knee osteoarthritis (OA) is a common degenerative disease resulting from pathological changes to the joint. Structural changes are radiographically identified to assess the severity of knee OA by the Kellgren-Lawrence (KL) scoring system; however, various risk and lifestyle factors contribute to the disease pathology and progression which complicate the assessment of OA severity and effective treatments. Growing evidence suggests that disease severity is closely tied to changes to the synovial fluid (SF) composition in OA-afflicted knees, which can modulate local cells. Thus, SF contains critical information about knee OA that can offer insight into the disease milieu. Herein, we characterized human OA patient-derived SF (pdSF) by measuring its molecular, physiological, and mechanical properties. Machine learning (ML) was used to correlate these data with either the corresponding KL scores or a grade of hyaluronic acid degradation based on a novel method as measures of OA. Results from ML models identified top variables in the pdSF as critical attributes, or potential predictors, of OA severity. Moreover, pdSF characterization informed the development of an OA-simulated SF (simSF) which can be used as a surrogate when access to pdSF is limited. Using the simSF to mimic an OA environment, we developed in vitro assays to create: 1) a research tool to evaluate cell therapies (e.g., mesenchymal stromal cells (MSCs)), given their exposure to SF when injected intra-articularly, and 2) a pharmacological screening tool by incorporating local immune cells of the knee (e.g., macrophages). As a research tool, this OA-targeted potency assay can be used to elucidate donor-specific secretory responses by the MSCs, and quantifiable outcomes demonstrated that simSF elicited similar cellular responses compared to pdSF. We also demonstrated macrophage-specific secretory responses to simSF compared to pdSF that recapitulate OA-induced changes to these innate immune cells of the knee. This approach provides a tool for screening drug candidates directed toward macrophage-based mechanisms. Together, this study uncovered in-depth characteristics of pdSF, identified critical attributes of knee OA in pdSF that were correlative to OA severity, and developed a simSF product for in vitro testing that offers new opportunities for exploring knee OA disease and therapeutics.