The osteoarthritis process is intricately connected with the shape of joints. On the one hand, changes in shape of the peri-articular bone such as flattening of the femoral head and osteophytosis are part of the etiology ofosteoarthritis (OA). On the other hand, more and more evidence adds to the idea that the shape of the joint itself is responsible for initiating the OA process, especially in the hip. Interesting in this respect is the dynamic nature of bone morphology that is guided by mechanical and hormonal interactions and likely interact with the evolving osteoarthritic disease aspects. Some of those most striking morphological problems as seen in Perthes’ disease, slipped capital femoral epiphysis, and congenital hip dysplasia pose a high risk for early development of hip OA. But also mild non-clinical forms of dysplasia have been shown to be associated with development of OA. Furthermore, there is evidence that other subtle shape variations such as e.g. a decreased anteversion angle of the femoral neck, retroversion of the acetabulum, a deep acetabular socket are all associated to OA. Together, these findings show that both the orientation and the way the femur is placed into the acetabular socket (congruity) are highly relevant for the development of hip OA. Abnormal contact of the proximal femur and acetabulum due to these anatomical variations also referred to as femoral acetabular impingement, likely leads to a non-optimal joint congruity with subsequent high stresses on the labrum, cartilage, and the underlying bone, which eventually triggers the onset of OA. New techniques have been developed that use statistical shape or appearance models to mathematically describe the bone morphology from either 2D (x-ray or dexa) or 3D (CT or MRI) data. A map of the contour of the bone can be made through surface points and semi automated computer programs help to quickly transfer the data in statistical models that enable to analyze large data sets. Predefined measures such as alfa-angle or sphericity can be easily derived from these models. Using principal component analyses the models can also be used to determine (new) independent components of variation, referred to as modes. Each mode represents a specific pattern of variation in shape or appearance that exists within the population studied. Numerically this is expressed as normalized values with 0 denote the mean and positive and negative values denote deviations (in numbers of standard variation) from the mean. We found that shape characteristics (either from modes or predefined measures) are descriptive for the clinical status of OA independent from other radiological measures. In particular parameters that reflect aspect of femoral acetabular impingement are highly predictive for the future development OA.Recently shape aspects were correlated with OA characteristics and tested for the association of the shape with single-nucleotide polymorphisms of OA susceptible genes and investigated. It was found that specific aspect of hip shape correlate with OA characteristics only in carriers of the susceptibility allele (a SNP of DIO2), indicating that this SNP increases the vulnerability of cartilage to non-optimal bone shapes rather than directly influencing the formation of these shapes.The statistical shape method was also applied in relation to the shape of the knee joint and OA, in both radiographs and 3D MRI. Relatively larger femurs and tibias where found when OA is present, in particular tibial width increased with increasing OA severity. Interestingly, increased width was also found very early in the OA process, when no structural OA changes could be seen on radiographs, but cartilage damage was already present on MRI scans.Taken together we have shown now in several analyses that subtle bone shape and appearance variations are associated with OA. Some of these shape variations such as a CAM deformity develop during growth (puberty) and are influenced by biomechanical aspects and more or less static during ageing. Other variations are more dynamic in the ageing process and interact with or are part of the OA development process.
Disclosure of Interest None Declared
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.