Role of cytological analysis of synovial fluid in diagnosis and research.

Synovial joints are generally considered to consist of a number of discrete elements (capsule, cartilaginous articular surfaces, synovial fluid, and synovium'); and synovial fluid is regarded as a body fluid lying within a tissue space bounded by synovium. Because this arrangement is reminiscent of the major body cavities, synovium and synovial fluid are often discussed in similar terms to those used for the lining and contents of other body cavities. By analogy synovial fluid and the changes that it undergoes in disease are often likened to those seen in other body fluids; for instance, by the use of the term joint effusion. This almost subconscious view of the synovial joint has important implications for the way in which joints and joint diseases are perceived and, in particular, how changes in the amount and composition of synovial fluid are evaluated. The synovial joint is so commonly regarded as containing a body cavity that it is perhaps surprising to discover that if its anatomy is studied in detail its structure is not in any way comparable with that of the major body cavities. The fundamental biological difference between the 'joint space' and the body cavities lies in the nature of the interface between the elements lining the 'joint space' (synovium and cartilage) and the synovial fluid. It is accepted that cartilage consists of cells in a solid matrix with no discrete surface cell layer.2 Direct exchange of water, solutes, and certain macromolecules occurs between the synovial fluid and cartilage matrix.3 It is also well recognised that at the ultrastructural level the interface between cartilage and synovial fluid is ill defined.4 Thus at both the physical and chemical level there is no discrete boundary between cartilage and synovial fluid, there being a small but important area of the joint in which components of both are mixed. Although the surface synoviocyte layer often appears as a discrete boundary zone between the liquid synovial fluid and the more solid synovium, electron microscopy shows that synoviocytes are structurally dissimilar to the mesothelial cells of pericardium, peritoneum, or pleura, and morphological4 and epitope expression studies' have shown that in lineage and function synoviocytes and mesothelial cells are quite distinct. Although the synoviocytes that aggregate on the synovial side of the interface between 'solid' synovium and 'liquid' synovial fluid are surrounded by basement membranelike material, this is not a true basal lamina as it lacks the ordered trilaminar arrangement typical of this structure.6 7 Indeed, synoviocytes are not natural lining cells but are rather a mixture of bone marrow derived cells of the monocyte-macrophage lineage and locally derived secretory cells.8 Because in normal joints both the cell layer and the basement membrane-like material are incomplete,9 in places synovial fluid is in direct continuity with the extracellular fluid of the synovium.10 Thus there is not the cell layer and complex basal lamina characteristic of the lining of other body cavities that directly controls the exchange of fluid, ions, and macromolecules between conventional lining tissues and the adjacent tissue space. As a result the synovial intima is a relatively ineffectual barrier to the passage of fluid, macromolecules, and cells.8 The morphological relation between the apparently discrete elements within the joint is, as is indicated above, quite unlike the structure of any conventional body cavity. As might be predicted the synthesis of synovial fluid is also very different from the production of body cavity fluids. Unlike pleural or pericardial fluid, which is a highly selected residuum of predominantly low molecular weight components of tissue fluid, synovial fluid consists of a mixture of a filtrate of plasma and macromolecules produced by cells within adjacent tissue." The difference can be fully explained by the lack in the joint of any form of the cell/basement membrane 'filter' mechanism seen at the interface between fluid and tissue in the body cavities. When comparison is made with other body constituents it is found that in general terms there is no essential difference between the formation of synovial fluid and an extracellular tissue matrix. In view of the way in which it forms and its physical and chemical relationships with surrounding elements it is probably better, therefore, to regard synovial fluid, not as a carefully controlled transudate into a body cavity, but as a peculiarly liquid connective tissue. Departments of Internal Medicine and Pathological Sciences, University ofManchester A J Freemont