Relationship between lateral joint space narrowing and change in knee cartilage thickness: data from the OA initiative

Abstract

s / Osteoarthritis and Cartilage 20 (2012) S54–S296 S212 64y, mean BMI 30) were compared to 122 right non-exposed control knees (47 males, 75 females, age 55y, BMI 24) in this analysis. The non-exposed age-gender-height adjusted distribution of the femur curvatures had a small overlap with the TKR femur curvature (95% Sensitivity, 95% specificity AUC1⁄40.99). The false discovery adjusted SPM analysis indicated that 94% of the TKR knees had significant abnormal flattening of the femur bone (Figure 1). The largest prevalence of bone flattening was located at the central medial condyle, and on average affected 25% of the medial condyle area whereas only 9% of the medial tibia was affected by significant bone abnormalities (9%) (Table 1). Figure 1. Statistical Parametric Mapping (SMP) analysis of bone abnormalities. Left, prevalence map of TKR bone abnormalities. Right, the non-exposed vs. TKR ROC curve of the femur curvature. Table 1: ROI analysis. Population means (Standard deviations) are shown for the TKR and non-exposed groups with corresponding Adjusted ROC values. The relative amount of abnormal bone shape is shown in the last column. Conclusions:Bone curvature of end stage OA knees differs significantly in specific locations from non-OA knees with the most extensive differences located in the central medial femur. Therefore, the presence of large bone shape abnormalities in the central medial femurmay be a possible imaging biomarker to predict the need for knee replacement. Mean change (mm) in the lateral (LFTC) and medial (MFTC) femorotibial compartme Unicompartimental lateral JSN: No JSN (grade 0) Any later FLASH sample: LFTC -17 -91