One hundred forty-five synovial biopsy specimens were obtained from 30 procedures performed on the knee joints of 29 patients with rheumatoid arthritis. All patients had clinically active rheumatoid arthritis and none had received slow-acting disease-modifying drugs or intraarticular corticosteroids. Scores were assigned to each biopsy specimen for each of 6 histologic features to quantify variation within each joint. In the majority of knee joint biopsies, there was considerable clustering of scores for all histologic features. Thus, on a scale of 0-10, 82% of the scores for synoviocyte hyperplasia were within 1 point of the median score for a given joint. Similarly, between 69% and 85% of the scores for the remaining features (fibrosis, vessel proliferation, perivascular infiltrates, focal aggregates, and diffuse infiltrates of lymphocytes) were within 1 point of the median values. Multiple biopsies were obtained at arthroscopy in 8 patients. Tissue was selected from areas of apparent maximal and minimal involvement, to enhance the likelihood of regional histologic variation. Of the scores for synoviocyte hyperplasia, 91 % were within 1 point of the median values for a given joint, and of the scores for the remaining 5 features, 72-94% fell within 1 point of the median values. In addition, highly significant statistical correlations of the intensity of synovial lining layer hyperplasia, vessel proliferation, mononuclear cell infiltration, fibrosis, and clinical measurements of synovitis were observed.

No abstract