• lymphocyte responses
• infectious arthritis

In a previous issue of the Annals, Sibilia and Limbach reviewed the microbiology of “infectious arthritis” and described various ways in which the agents might be related to the arthritis.¹ One approach, for which there is considerable support, was, however, not discussed.

In 1980 synovial lymphocytes were found to respond maximally to stimulation by either chlamydia or ureaplasma antigens in cases of sexually transmitted reactive arthritis.² In 1985 synovial responses were reported in eight cases of enteric and 12 cases of sexually transmitted reactive arthritis.³ Responses to the relevant antigens of each category differentiated the enteric from the sexually transmitted cases. Additionally, peripheral blood lymphocytes in all eight enteric cases and in eight of the 12 sexually transmitted cases responded negligibly or only minimally to the antigens that gave significant synovial responses. In 1991 a review of 12 cases of enteric reactive arthritis showed that the maximal synovial responses to the relevant enteric antigen in 10 cases of salmonella, shigella, or yersinia reactive arthritis would unequivocally differentiate them from the responses in two cases of campylobacter reactive arthritis; the results also indicated that some cross reactivity occurred within the salmonella, shigella, and yersinia group.⁴

These observations on the responses of synovial lymphocytes to the causative antigen in reactive arthritis have been confirmed in several countries between 1989 and 1994.^5–7 However, the data from some studies have shown that the stimulation indices from the responses in the Vancouver experience are higher and more specific than those of other laboratories and one laboratory has downgraded the importance of this approach.⁸ Technical differences between laboratory procedures are hard to define, but the use of fresh as opposed to stored frozen lymphocytes for the tests differentiates the Vancouver laboratory from several others. An early investigation of lymphocyte responses performed in the Vancouver laboratory in the late 1970s showed that the use of stored liquid nitrogen frozen lymphocytes negated or greatly reduced the response to antigenic stimulation, although the lymphocytes still responded to phytohaemagglutinin (PHA). In consequence, fresh lymphocytes were always employed subsequently. The assumption was made that freezing caused loss of associated antigen-processing macrophages, which are required for antigen responses, but are not needed for PHA and other mitogen responses. It is noteworthy that the study referred to above, in which synovial responses were considered unhelpful, did use stored frozen synovial mononuclear cells. Another study found a lack of correlation between the detection of Chlamydia trachomatis DNA in synovial fluid and the presence of an antichlamydial immune response,⁹ but again, frozen synovial mononuclear cell samples were employed.

It is now clear that DNA or RNA from a variety of micro-organisms can be found within the synovium of arthritic, but also of normal and degenerative, joints. To incriminate intrasynovial organisms as a cause of a patient’s arthritis is difficult. The Vancouver experience of a 12 year study of 360 patients with many types of arthritis has indicated that the response of synovial lymphocytes to microbiological antigen stimulation can provide such incriminating evidence, as summarised in 1996.¹⁰ To be an effective investigative tool however, the synovial lymphocytes should be employed fresh and not frozen, and to validate the significance of findings multiple microbiological antigens should be used for each test so that specific stimulation can be identified. If a micro-organism’s DNA or RNA is associated with a statistical maximal intrasynovial lymphoid response to antigens of that agent, then a presumption of pathogenicity is logical, though not proved.