Background/Purpose: For some rheumatologic diseases (e.g. lupus), separate scores evaluate cumulative damage and disease activity.  No such strategy exists for osteoarthritis (OA).  The prevailing approach to evaluate OA focuses on individual structural features or developing composite scores.  Our group aimed to define an MRI composite score that accounted common features of knee OA: articular cartilage damage, bone marrow lesions (BMLs), and effusion-synovitis volume. However, we struggled to find such a composite score.  We had greater success when we considered OA from the perspective of measuring a cumulative damage score and a disease activity score.  The cumulative damage score – based on articular cartilage – measures damage over time and does not wax and wane. In contrast, the disease activity score – based on BMLs and effusion-synovitis – fluctuates over days or weeks. We aimed to evaluate the construct validity of these two scores.  

Methods: A convenience sample of 197 participants in the Osteoarthritis Initiative with complete clinical, radiographic, and MRI data at baseline and 24-months was selected. We generated quantitative measures of articular cartilage using the Double-Echo Steady State sequence, and BML and effusion-synovitis volumes using the Intermediate Weighted Fat Suppressed sequences. We assessed BMLs and cartilage damage index at the medial and lateral patella, tibia, and femur. A single effusion-synovitis volume was assessed. All MRI measures were standardized to bone width, and the 2-year difference was standardized so all measurements were on the same scale. The cumulative damage score was calculated by summing the change for each of the 6 locations with cartilage measures. Similarly, a disease activity score was calculated by summing the change in effusion-synovitis volume and the BML volumes for 6 regions. To evaluate construct validity we used logistic regression to estimate odds ratios (OR) for the outcomes of Kellgren Lawrence (KL) progression, joint-space width (JSW) progression (change > the median change) and worsening in WOMAC pain.

Results: Our sample was 54% female, 93% with KL grade 2 or 3, mean age of 61 years and mean body mass index of 30.1 kg/m². Mean WOMAC pain at baseline was 5.0. Worsening cumulative damage score was associated with KL progression (OR=1.52, 95% CI=1.11 to 2.08) and JSW progression (OR=1.67, 95% CI= 1.22 to 2.33), but not with WOMAC pain progression (Table). Conversely, the disease activity score was associated with WOMAC pain (OR=1.67, 95% CI= 1.14 to 2.45), but not KL or JSW progression.

Conclusion: The cumulative damage score (based on cartilage damage) had good construct validity for structural outcomes, while the disease activity score (based on BML and effusion-synovitis volumes) was associated with pain. This suggests that separate scores for cumulative damage and disease activity may have important utility in studying OA and provide critical insights into the disease.