A pessimistic view of serologic markers for diagnosis and management of osteoarthritis. Biochemical, immunologic and clinicopathologic barriers.

The diagnosis of osteoarthritis (OA) is usually based on clinical and radiologic findings and is usually made only after the patient presents with joint pain. There is today much interest in development of an immunologic "marker" of OA, to detect subclinical disease and/or monitor therapy. The approach employs measurement of serum or synovial fluid levels of articular cartilage macromolecules, such as proteoglycans or glycosaminoglycans, or fragments of these. The data, however, raise questions about interpretation and utility of such tests. What causes egress of such macromolecules from OA cartilage? Overproduction? Hypercatabolism? Leakage from an excessively permeable matrix? Does the marker reflect the rate of cartilage breakdown? Or of repair? How reliable are the quantitative immunologic methods in tests of sera from patients with OA? Data show, for example, that serum keratan sulfate levels may be influenced by the mode of presentation of the antigen, i.e., single vs multiple chains, and by the degree of sulfation, etc. To what extent might serum levels of a marker reflect release from degenerating but asymptomatic joints, rather than from painful joints? Also, since all putative marker molecules studied to date are widely distributed throughout the connective tissue of the body, they could be released from an asymptomatic degenerating intervertebral disc rather than, e.g., a painful osteoarthritic hip or knee. In the present climate of "marker mania," it should be emphasized that no marker exists today which can confidently be used for diagnosis of subclinical OA or for monitoring therapeutic response.