Abstract
To explore whether pretreatment features of synovial tissue in patients with gonarthritis could predict the clinical effect of radiation synovectomy with yttrium-90 (90Y) and glucocorticoids or with intra-articular glucocorticoids alone. A synovial biopsy was carried out blindly 2 weeks before treatment in 66 patients with persistent gonarthritis, who were randomised to treatment either with 90Y and triamcinolone or with placebo and triamcinolone. Immunohistochemistry was used to detect T cells, macrophages, B cells, plasma cells, fibroblast-like synoviocytes, adhesion molecules and pro-inflammatory cytokines. Stained sections were evaluated by digital image analysis. Individual patient improvement was expressed using a composite change index (CCI; range 0-12). Successful treatment was defined as CCI > or = 6 after 6 months. Patients with rheumatoid arthritis, psoriatic arthritis, undifferentiated arthritis and other causes of gonarthritis were included. The overall response rate was 47%. Clinical efficacy in both therapeutic groups was similar and not dependent on diagnosis. No significant differences were noted between baseline microscopic features of synovial tissue inflammation in patients with rheumatoid arthritis and in those with non-rheumatoid arthritis (ie, all diagnoses other than rheumatoid arthritis). The number of macrophages in the synovial sublining was significantly higher in responders than in non-responders (p = 0.002), independent of treatment group and diagnosis. The clinical effect was positively correlated with pretreatment total macrophage numbers (r = 0.28; p = 0.03), sublining macrophage numbers (r = 0.34; p = 0.005) and vascular cell adhesion molecule 1 expression (r = 0.25; p = 0.04). The observations support the view that intra-articular treatment either with 90Y and glucocorticoids or with glucocorticoids alone is especially successful in patients with marked synovial inflammation.