The effect of treatment with a monoclonal antibody against the CD4 antigen present on T helper cells was studied in 10 patients with severe intractable rheumatoid arthritis. In an open trial, monoclonal antibody 16H5 was infused at a dosage of 0.3 mg/kg of body weight on 7 consecutive days. Studies of the kinetics demonstrated a drastic depletion of CD4+ cells, to as low as 25 cells/microliters, 1 hour after the first infusion. The subsequent recovery of the CD4+ cell numbers 24 hours after infusion did not reach initial levels, and after the full 7-day treatment cycle there was a significant reduction of the number of CD4+ cells (mean +/- SD 51 +/- 28%; P less than 0.02). There was a reduced or even inverse CD4:CD8 ratio, which generally persisted 3-4 weeks. Lymphocyte transformation assays demonstrated significantly reduced reactivity in 5 of the 9 patients who completed the 7-day course, whereas 4 individuals exhibited an unexpected elevation in the T cell response to mitogens and common antigens. Parallel laboratory studies showed a significant decrease in the erythrocyte sedimentation rate (P less than 0.05), rheumatoid factor titer (P less than 0.04), and total immunoglobulin values (P less than 0.01), as well as a reduction in C-reactive protein levels, in 7 of the 9 patients. Clinically, there was a significant reduction in the Ritchie articular index (P less than 0.05) and in the number of swollen joints (P less than 0.04). Adverse effects were urticaria in 2 patients, which led to withdrawal of therapy in 1 of them, and chills with fever, suggestive of a lymphokine release syndrome, in another 2 patients. Only low levels of human anti-mouse immunoglobulin antibodies developed (not exceeding 1.7 mg/liter). It was therefore possible to repeat the treatment cycle, achieving still better efficacy, in 4 of the patients (reductions in the Ritchie index and the number of swollen joints P less than 0.02). Our findings indicate that treatment with monoclonal antibodies against the CD4 antigen leads to immunomodulation which results in clinical benefits, at least during initial observation periods (up to 6 months postinfusion). However, it remains to be determined whether long-term remission can be induced with this therapeutic approach. The use of immunosuppressive therapies or repeated antibody treatments will have to be considered.