Interpretation and Utility of Novel Composite Structural Endpoints of Cumulative Damage and Disease Activity in Knee Osteoarthritis

The United States Food and Drug Administration (FDA) recognizes osteoarthritis as a serious disease with an unmet need for therapies that slow, stop, or reverse joint damage. A challenge to testing new therapies is the lack of a comprehensive definition of disease progression that incorporates multiple structural changes. Another challenge is not knowing how much change in a structural measure would reflect meaningful benefit or worsening to a patient. Magnetic resonance (MR) imaging has great potential to address these critical gaps. However, no-one has developed an MR-based composite outcome that reflects multiple structural aspects (“whole-knee”) of knee osteoarthritis progression and is sensitive to change. This remains a critical technological obstacle to testing and developing new therapies. We recently tackled these barriers and found that structural measures of knee osteoarthritis progression can be summarized as: 1) cumulative damage (quantitative articular cartilage damage throughout a knee; relates to radiographic progression), and 2) disease activity (quantitative bone marrow lesions and effusion-synovitis volumes; relates to pain progression). The critical next steps are to demonstrate these outcomes are valid and useful predictive biomarkers of KOA progression by determining 1) how much change in each measure translates to indicators of clinically meaningful improvement or worsening and 2) the discriminative performance of each composite measure. We will accomplish this by assessing annual MR images in 3 nested case-control samples to determine the magnitude of change in cumulative damage and disease activity that predict changes in patient-reported outcomes (Aim 2) and walking speed (Aim 3), and knee replacement (Aim 4) across biological factors (e.g., sex). To achieve our goal, we will quantify 1- or 2-year change in articular cartilage as well as bone marrow lesions and effusion-synovitis volume on MR images for 5,270 knee visits. We will then determine the amount of change in cumulative damage or disease activity that relates to these outcomes during the same time period. We will also test how much change in these measures differentiates adults who will receive a knee replacement. Finally, we will test whether cumulative damage and disease activity will reliably predict changes in patient-centered outcomes across sex, body mass index, radiographic severity, and duration of observation period. This study will support an application for approval of these structural endpoints as predictive biomarkers for KOA progression, which could greatly improve the chance of success for clinical trials testing disease-modifying therapies for knee osteoarthritis.