Biochemical markers of bone and cartilage in osteoarthritis

Osteoarthritis (OA), the most prevalent joint disease, is characterized by abnormal and degraded cartilage, inflamed and/or thickened synovial tissue, and alterations of bone metabolism (sclerosis, cysts, osteophytes). The extracellular matrix of joint tissues is composed primarily of collagens including type I (bone and synovium), type II (cartilage) and type III (synovium) associated with proteoglycans (e.g. aggrecan in cartilage) and other glycoproteins. Potential biological markers for OA include matrix components (and/or their breakdown products), cytokines, and proteases [1].

Although bone turnover markers may reflect the focal abnormalities of subchondral bone metabolism in OA, circulating and urinary levels are more likely to reflect the overall skeletal turnover, which may be influenced by a variety of conditions including age, menopausal status, osteoporosis and other bone diseases. Major efforts have recently been made to develop more specific markers, especially those reflecting cartilage type II collagen synthesis (PIINP and PIICP) and destruction and synovial activity that plays a pivotal role in the pathogenesis of OA. Antibodies recognizing different type II collagen fragments have been developed and include those directed against the neo-epitopes generated by the collagenases, and against C-telopeptide breakdown fragments (CTX-II). We have characterized a pyridinoline derivative, glucosyl-galactosyl-pyridinoline (Glc-Gal-PYD), which is found in large amounts in human synovium and in very low levels in the cartilage and other soft tissues. Urinary Glc-Gal-PYD excretion has been found to be increased in knee OA and hip OA, and the levels show stronger correlations with symptoms of OA than do markers for bone and cartilage [2].

Recent prospective studies have shown that baseline levels of some cartilage markers are associated with a more rapid joint damage. In a study of 52 patients with knee OA, we found that low serum levels of PIIANP or high urinary CTX-II excretion were associated with faster joint destruction as evaluated over a 1-year period, either by plain radiographs or by arthroscopy [3]. Combining these two biological markers to obtain an index of uncoupling of type II collagen synthesis and breakdown was more effective in predicting cartilage destruction than measurements of a single marker. The predictive value of CTX-II has been confirmed in a larger study of volunteers followed prospectively for more than 6 years, where progression was assessed by joint space narrowing at the knees and hips [4]. In the ECHODIAH cohort, including 333 patients with hip OA followed prospectively for 3 years, we found that among a panel of 10 different biochemical markers of bone, cartilage and synovium activity, increased levels of urinary CTX-II and hyaluronic acid were strongly predictive of radiological progression independent of disease activity and radiological parameters [5]. One of the main issues that currently impair efficient development of structure OA modifying therapies is the low sensitivity of the plain X-ray, requiring long-term studies to show a significant difference between placebo-treated and active-drug-treated patients. In the COBRA study of patients with early rheumatoid arthritis, we found that the combined therapy (sulphasalazine + methotrexate + prednisone) produced a significant decrease in urinary CTX-II that can be observed within 3 months of therapy. The magnitude of the changes of CTX-II at 3 months were predictive – independent of changes in disease activity – of the changes in radiological scores after 5 years, indicating that the higher the decrease of CTX-II at 3 months, the lower will be the progression of joint damage at 5 years [6]. Such data are currently missing in OA, mainly because of the absence of treatments that have definitively been shown to reduce progression.

Recent work on biological markers for OA has made valuable progress; in particular, with the development of specific markers for the production and breakdown of type II collagen and of synovial membrane activity. Although, they are unlikely to be clinically useful for the identification of patients with OA, an increasing body of evidence suggests that combination of some biological markers will be useful adjuncts to radiography and magnetic resonance imaging to predict progression of OA. The rapid responsiveness of biological markers should prove valuable in the clinical development of drugs for preventing joint destruction and in monitoring their treatment effects in patients with OA.