MEDIAL FIXED JSW TRAJECTORIES: PREDICTORS OF TRAJECTORIES AND THEIR ASSOCIATION WITH THE RISK OF KNEE REPLACEMENT

Loss of JSW is a hallmark of knee OA structural progression, and knee replacement (KR) is considered by regulatory agencies as a hard clinical endpoint in knee OA DMOAD clinical trials. There has been limited longitudinal investigation of JSW trajectories and their association with the risk of KR, however. In addition, identification of predictors of worse vs. better JSW loss trajectories may help identify potential targets for therapeutic interventions and may also be useful for selecting more homogenous study populations.

This study aimed to 1) identify distinct JSW trajectories; 2) identify predictors of different rates of change in medial fixed JSW (fJSW) trajectories; and 3) examine the association between JSW trajectories and the risk of KR.

Knees (n = 2862) with available baseline MRI Osteoarthritis Knee Scores (MOAKS) from OAI were included. Observations collected after KR were excluded. Medial fJSW at 25% of femur width (i.e., x=0.250) was assessed longitudinally at baseline and each follow-up visit (i.e., 12, 24, 36, 48, 72, 96, and 120 months) to construct change-from-baseline JSW trajectories. Latent Class Mixed Models (LCMM) were employed to classify knees into two clusters having distinct fJSW trajectories, with (fJSW) change from baseline as the outcome and quadratic year term as the predictor, while jointly identifying predictors of latent class membership within the model. Covariates included demographic, clinical, psychosocial and imaging features. The response covariance matrix was allowed to be different for each latent class. Kaplan-Meier curves were used to illustrate time to KR for each fJSW trajectory cluster, and differences between clusters were assessed using the log-rank test.

Two fJSW trajectory clusters were identified: a faster progressive trajectory (n = 824) and a slower progressive trajectory (n = 2038) (Figure 1). Predictors of membership in the faster progressive trajectory included meniscal extrusion, effusion-synovitis/Hoffa synovitis, cartilage surface/depth damage, larger bone marrow lesion (BML) size, and being overweight/having obesity (i.e., BMI > 25) (Figure 2). The faster progressive trajectory was also associated with a higher risk of KR (p < 2e-16) CONCLUSION: We have identified two distinct JSW trajectories that were associated with MRI-detected structural damage and specific demographic and clinical features. The faster progressive trajectory was associated with an increased risk of KR. Risk factors for the faster trajectory may represent suitable targets for interventions aimed at preventing KR.