The existence of lymphoid follicles in the synovium of patients with inflammatory arthritis, including the feature we now recognize as a germinal center of proliferating B cells, has been known since the nineteenth century (1). It has also long been recognized that these structures are present in only a minority of cases and are not specific to rheumatoid disease. Most of the lymphocytes seen in synovium are T cells (2), although this analysis does not include B cell–derived plasma cells, which can be very numerous. Perhaps surprisingly, we are still faced with the question: are these B cell–containing follicles telling us something important about the disease and how to treat it, or are they a secondary curiosity? We have probably learned enough about inflammatory arthritis to know that the answer will lie somewhere in between and that every patient may be slightly different. Nevertheless, careful studies, such as the one by Thurlings and colleagues reported in this issue of Arthritis & Rheumatism (3), may reveal broad truths that may at least guide useful treatment strategies. As is often the case, the study by Thurlings et al is most interesting for what was not found. It confirms the historical belief that organized collections of B cells occur in rather few rheumatoid arthritis (RA) synovia. Moreover, it shows that, at least in terms of measures in the Disease Activity Score 28-joint assessment (4), such as the numbers of swollen and tender joints, amount of bone damage, and presence of rheumatoid nodules, lymphoid neogenesis does not seem to determine disease phenotype. It is difficult to escape the conclusion that B cells in joints may not be that important. Should we be surprised? Probably not, since the role of B cells is not usually at a site of inflammation itself. The findings of van Boxel and Paget (2) (T cell predominance) apply to more or less all inflammation. B cells normally contribute to inflammation either by generating plasma cells, which, in lymphoid tissue or at sites of inflammation, secrete antibodies capable of activating inflammatory mechanisms, or by presenting antigen to T cells in draining lymph nodes so that primed T cells can recirculate to inflamed tissue. Starting with antibody production, arguably the most interesting finding reported by Thurlings and colleagues is that the presence of organized B cell collections in synovium was not correlated with the presence of detectable autoantibodies (rheumatoid factor and antibodies against citrullinated proteins). The implication is that if autoantibodies contribute to inflammation through the interaction of immune complexes with macrophage Fc receptors and complement, then in many cases the B cells responsible are to be found largely or wholly in lymph nodes or spleen. This is consistent with the body of data indicating that autoantibody production frequently precedes any clinical sign of cellular infiltration of synovium by up to several years (5). (Interestingly, preclinical autoantibody production can be associated with a rise in C-reactive protein [CRP] levels, suggesting that immune complexes may drive an acute-phase response, perhaps by Kupffer cell activation, without causing inflammation.) It does not, of course, imply that therapies targeting B cells should not be designed to reach synovial B cell collections as well as those elsewhere. But it does tend to imply that this is not the primary objective. In terms of antigen presentation, the prima facie Jonathan C. W. Edwards, MD, Maria J. Leandro, MD, PhD: University College London, London, UK. Dr. Edwards has received honoraria (less than $10,000) from Roche. Dr. Leandro has received consulting fees, speaking fees, and honoraria (more than $10,000) from Roche. Address correspondence and reprint requests to Maria J. Leandro, MD, PhD, Room 317, Windeyer Building, 46 Cleveland Street, London W1T 4JF, UK. E-mail: maria.leandro@ucl.ac.uk. Submitted for publication January 11, 2008; accepted in revised form February 29, 2008. Arthritis & Rheumatism
Read full abstract