Mylène Jansen

Comparing radiographic JSW and MRI cartilage thickness using CT 239 12 Introduction Knee osteoarthritis (OA) is a degenerative joint disease that is characterized by, among other factors, articular cartilage degeneration and subsequent thinning. 1 The gold standard for quantifying cartilage thinning has traditionally been measurements of the joint space width (JSW) on weight-bearing radiographs. 2 The radiographic JSW provides a 2-dimensional projectional estimate of the bone-to-bone distance and thus reflects, to a certain extent, articular cartilage thickness. Radiographic JSW is often required for evaluating the rate of cartilage degeneration/regeneration in clinical trials and, when managed well with a high degree of acquisition standardization, the reliability and reproducibility of JSW measurement techniques are considered to be high. 3–5 Because knee radiographs are generally taken in a weight-bearing position, quality of the cartilage (with respect to deformability of the tissue) may be an important factor in the assessment of radiographic JSW. However, representing only an indirect measure for cartilage thickness, JSW measurements can be influenced significantly by positioning, acquisition errors, focal cartilage degeneration, and changes in other joint tissues. 6,7 The meniscus, in particular, has been shown to substantially impact radiographic JSW measurements. 8,9 A more recent method is the direct measurement of articular cartilage thickness on MRI scans. Using MRI, cartilage tissue itself can be visualized 3-dimensionally. Different quantitative measurements have been described and the average cartilage thickness generally shows high reproducibility. 10,11 However, unlike radiography, MRI images are taken in a non-weight- bearing position. As such, deformability of the cartilage tissue is not taken into account. Yet, it has been shown that knee OA affects the mechanical properties of cartilage, which influences the amount of deformation. 12 Literature comparing both techniques for natural OA progression show moderate to strong correlations cross-sectionally. 13–15 In cross-sectional evaluation, differences in cartilage thickness between individuals are relatively large (millimeters) and as such in favor of finding these relations. However, when looking at longitudinal changes over time, changes are much less pronounced (tenths of millimeters), limiting the measurement window. In these longitudinal studies, no or at best weak correlations were found between the change in radiographic JSW and the change in MRI cartilage thickness. 16–20 This may be the result of the various differences between the techniques described previously: weight-bearing versus non-weight-bearing, bone-to-bone distance versus cartilage thickness, and 2D versus 3D. In the present study we include CT as an imaging technique, as it is performed without weight-bearing, like MRI, but specifically visualizes the bone-to-bone distance, like radiographs. CT is a 3D imaging technique, but is also capable of creating a projectional image