Objective To evaluate microscopic measures of inflammation in rheumatoid arthritis synovial tissue samples selected at arthroscopy compared with those obtained blindly by needle biopsy from the suprapatellar pouch (SPP) of the same joint. Methods Samples were selected at knee arthroscopy from the SPP and the lateral and medial gutters. Immediately following arthroscopy, a biopsy needle was inserted through the same portal into the SPP by a second investigator, and 3 further samples were obtained blindly. Using standard immunohistologic methods, all samples were analyzed by a single investigator without knowledge of the original tissue location and biopsy technique. Following staining with anti‐CD3 and anti‐CD68 monoclonal antibodies, T lymphocyte and macrophage infiltration were measured by quantitative analysis. Results Synovial tissues from 14 patients were analyzed. In comparing microscopic measures of inflammation using the 2 procedures, mean scores of lining cell depth and the percentage of CD68+ cells in the lining layer correlated positively (τ = 0.59, P = 0.003 and τ = 0.73, P = 0.0003, respectively). In the sublining layer, CD3+ cell counts also correlated significantly (τ = 0.71, P = 0.0004). Sublining CD68+ cell counts did not correlate. This was explained by the observation that CD68+ cell infiltration in areas adjacent to articular cartilage was significantly greater than in the SPP (P = 0.01), suggesting preferential trafficking to this site by macrophages, but not but T lymphocytes. Macroscopic appearance at arthroscopy did not predict microscopic features. Conclusion Most microscopic measures of inflammation in synovial tissue samples obtained blindly from the SPP were similar to those determined in samples selected at arthroscopy. However, measurements in samples from the SPP may underestimate the intensity of macrophage infiltration in areas more adjacent to cartilage. These observations have important implications for future study of macrophage function in synovial tissue.

No abstract