Immunohistological analysis of synovial tissue for differential diagnosis in early arthritis.

OBJECTIVE An early diagnosis in patients presenting with arthritis is important to provide information about prognosis and to initiate treatment. The objective of this study was to determine which markers applied in immunohistological analysis of synovial tissue (ST) specimens could be used to differentiate rheumatoid arthritis (RA) from other forms of arthritis. METHODS Synovial biopsies were obtained by blind needle techniques from 95 patients with early arthritis. After follow-up of at least 2 yr to verify the diagnosis, the patients could be classified as follows: RA (n=36), undifferentiated arthritis (UA; n=21), osteoarthritis (OA; n=17), reactive arthritis (ReA; n=10), ankylosing spondylitis (AS; n=3), psoriatic arthritis (PsA; n=2) and crystal-induced arthritis (CA; n=6). ST sections were analysed by immunohistochemistry using monoclonal antibodies against CD3, CD4, CD8, CD22 (B cells), CD38 (plasma cells), CD68 (macrophages) and CD55 (fibroblast-like synoviocytes). RESULTS Logistic regression analysis revealed that the higher scores for the numbers of CD38+ plasma cells and CD22+ B cells in RA were the best discriminating markers comparing RA to non-RA patients (CD38: P=0.0001; CD22: P<0.05). Polychotomous regression analysis comparing three diagnostic categories (1: RA; 2: UA, ReA, AS and PsA; 3: OA and CA) also identified the score for the number of CD38+ plasma cells (P<0.0001) as well as the numbers of CD68+ macrophages in the synovial sublining (P=0.05) as discriminating markers. CONCLUSION The results suggest that immunohistochemical analysis of ST specimens from early arthritis patients can be used to differentiate RA from non-RA patients. The numbers of plasma cells, B cells and macrophages are especially increased in ST of patients with RA. Future studies in early arthritis patients with clinical features which do not allow an immediate confident diagnosis may clarify the role of this test system in differential diagnosis.