Longitudinal survey for high density mineralised protrusions identified by 3D dess MRI in patients of the osteoarthritis initiative

Abstract

Purpose: The structural and physiological integrity of the bone-cartilage interface is essential to maintaining joint health. Phenomena that disrupt this boundary can compromise whole joint function and are a major factor in the development of osteoarthritis (OA), particularly in the knee. Early diagnosis of such maladies may offer a predictive biomarker of joint destruction. Microcracks in the subchondral bone plate including the articular calcified cartilage (ACC) are in some cases repaired with a high density mineralised infill material. This material may also extend from the mineralising front into hyaline articular cartilage (HAC) as a high density mineralised protrusion (HDMP). Such features are likely to impair biomechanical performance of surrounding tissues. Their formation can occur in the absence of symptomatic OA and they are potentially a primary cause of OA. Recent reports have confirmed the presence of HDMPs in human ex- and in-vivo knee joints and their detectability with clinical MRI. Here we present the first longitudinal MRI study prospecting for HDMPs in the human knee. Methods: Twenty-one patients were selected from the osteoarthritis initiative (OAI) progression cohort for whom bilateral knee MRI data were available at five consecutive annual visits. Sagitally-acquired dual echo steady state MRI images were assessed for the presence of HDMP-like features at baseline. Incidences were recorded and categorised to one of six regions according to bone and laterality: medial femoral condyle (MFC), lateral femoral condyle (LFC), intercondylar fossa of femur (ICF), medial tibial condyle (MTC), lateral tibial condyle (LTC) and intercondylar ridge of tibia (ICT). Imaging data from follow-up visits were used to monitor the development of features over time. Results: Seventeen patients were found to possess MRI features identified as HDMPs in at least one knee with incidence being unilateral in 5 and bilateral in 12 cases. A total of 58 features were found, ranging from 1 to 5 in a single joint. Features were found in all six regions of the knee joint but distribution between regions was inequitable with 73% being located somewhere on the femur (ICF = 36%; LFC = 25%; MFC = 12%) and only 27% on the tibia (LTC = 23%; MTC = 2%; ICT = 2%). Of those HDMPs identified at baseline, the percentage still present at each subsequent annual visit was 96, 90, 85 and 81, respectively. Morphology of those still visible varied greatly; some remained similar in appearance while others altered substantially, for example, changing in size and shape. Some appeared to have fragmented, forming clusters within HAC. Figure 1 shows an example of a HDMP located on the articular surface of the medial femoral condyle. The material extends from the subchondral bone surface initially. It can be seen breaking away and moving gradually through HAC over time. Conclusions: Observation of features in regions of high mechanical stress and noting their disappearance over time is suggestive of a role in arthropathy. Fine particles of fragmented HDMP may abrade HAC and damage surrounding tissue. The potential for mineral features to fragment may account for apparent clustering of features over time, or they may be discrete incidences. Longevity of features throughout the study duration suggests fragmentation is slow and not guaranteed. Disappearance may indicate destruction or may be caused by average effects. The lower proportion of HDMPs in tibiae may indicate the bone is less susceptible to HDMP formation, or they may be shorter lived. The appearances of HDMPs in clinical scans were consistent with reported ex-vivo studies. Clinical DESS yields high contrast between signal intensities of mineral (low yield) and HAC (high yield) and the features described should be considered a potential imaging biomarker for the prediction of joint destruction in some forms of OA.