Search

Objective: Anti-tumor necrosis factor alpha (anti-TNFalpha) therapy is very effective in rheumatoid arthritis (RA), whereas depleting anti-CD4 therapy is relatively ineffective. To explain the differences in efficacy between these 2 therapies, we used an animal model of RA to compare their effects on different aspects of the disease process., Methods: Mice with collagen-induced arthritis were treated with depleting anti-CD4 monoclonal antibodies (mAb), anti-TNFalpha mAb, or phosphate buffered saline. Another group was given a combination of anti-TNFalpha plus anti-CD4. The treatments were compared for their ability to down-regulate the expression of proinflammatory cytokines and adhesion molecules, reduce the cellularity of the joint, and inhibit Th1 activity., Results: Anti-TNFalpha significantly reduced the numbers of cells expressing TNFalpha, interleukin-1beta (IL-1beta), very late activation antigen 4 (VLA-4), vascular cell adhesion molecule 1 (VCAM-1), and numbers of CD4+ T cells and macrophages in the joint. Anti-CD4 treatment led to a small reduction in the expression of TNFalpha, IL-1beta, VLA-4, and VCAM-1, but this did not reach statistical significance. Depleting anti-CD4 was also surprisingly ineffective in eliminating CD4+ T cells from the joint. Anti-TNFalpha therapy was also more effective than anti-CD4 in reducing Thl activity, as assessed by the production of interferon-gamma in lymph node cell cultures. There was a synergistic relationship between anti-TNFalpha and anti-CD4 in the reduction of histologic score and inhibition of TNFalpha/IL-1beta expression in the joints., Conclusion: The efficacy of the 3 treatments correlated with their ability to modulate the expression of inflammatory cytokines and adhesion molecules in the joint, reduce the cellularity of the joint, and inhibit Th1 activity. This kind of analysis may prove useful in the testing of novel therapies for RA.

Rolipram is a type IV phosphodiesterase inhibitor that suppresses inflammation and TNF-alpha production. As anti-TNF-alpha therapy is effective in rheumatoid arthritis, we investigated the effect of rolipram on collagen-induced arthritis (CIA), a murine model of rheumatoid arthritis. Rolipram was administered after the onset of clinical arthritis at doses of 0.5, 3, 5, or 10 mg/kg twice daily, with a dose-dependent therapeutic effect on clinical severity and joint erosion. Immunohistochemical analysis of joints of rolipram-treated mice revealed 67% reduction in TNF-alpha-expressing cells compared with control arthritic mice. In vitro studies using bone marrow-derived macrophages confirmed that rolipram directly suppressed TNF-alpha and IL-12 production following stimulation with IFN-gamma and LPS. The effect of rolipram on T cell activity was studied by measuring Th1/Th2 cytokine production by collagen-stimulated draining lymph node cells from arthritic mice treated in vivo with rolipram. Rolipram reduced IFN-gamma production and increased IL-10, indicating that rolipram down-regulated the ongoing Th1 response to type II collagen. Finally, the effect on CIA of combination therapy was studied using rolipram plus either anti-TNF-alpha or anti-CD4 mAbs. Rolipram plus anti-TNF-alpha was not therapeutically additive, whereas rolipram plus anti-CD4 mAb was clearly additive. This result indicates that the therapeutic effects of rolipram overlap with TNF-alpha blockade, but are complementary to anti-CD4 treatment. It is therefore proposed that a major mechanism of action of rolipram in CIA is suppression of TNF-alpha activity. These findings suggest that type IV phosphodiesterase inhibitors may be effective in pathologic conditions, such as RA, with overexpression of TNF-alpha.

Arthritis spontaneously develops in mice expressing a human TNF-alpha transgene modified with the 3' untranslated region of beta-globin. We have backcrossed these mice onto the arthritis-susceptible DBA/1 background and found an acceleration of the onset of arthritis with successive generations of interbreeding. Bioactive TNF-alpha in primary synovial membrane cell cultures was significantly higher in the DBA/1 transgenic mice than in transgenic mice on the original background. Elevated levels of human TNF-alpha were accompanied by increases in synovial cell expression of murine IL-1beta and IL-6, but murine granulocyte-macrophage CSF, IFN-gamma, and IL-4 could not be detected. Interestingly, the anti-inflammatory cytokine IL-10 could be detected, but levels were not modulated by expression of the transgene. Analysis of the synovial membrane cell composition revealed that >50% of synovial cells were CD45-negative cells, presumably fibroblasts and endothelial cells, and the majority of CD45-expressing cells were neutrophils. Peritoneal macrophages and lymphocytes from the spleen, bone marrow, and lymph nodes required LPS stimulation to produce human TNF-alpha, indicating that, when activated, cells of these lineages were capable of expressing the transgene; however, few were found in synovial tissues. In contrast, fibroblasts derived from synovial tissue spontaneously released human TNF-alpha, and using immunohistochemical techniques, this cytokine was localized to fibroblast-like cells and chondrocytes. We propose that arthritis in DBA/1 human TNF-alpha transgenic mice is driven in part through the spontaneous expression of transgene by connective tissue cells, and there is little evidence of the participation of lymphocytes in this model.

Anti-CD4 treatment is reported to prevent collagen-induced arthritis if administered before the onset of clinical disease but has relatively little effect on established arthritis. In contrast, we have recently shown that anti-tumor necrosis factor alpha/beta (TNF) treatment reduces the severity of established arthritis. We now study the effect of combined administration of anti-CD4 monoclonal antibody (YTS 191.1.2/YTA 3.1.2) and anti-TNF monoclonal antibody (TN3-19.12) in established arthritis. Anti-CD4 treatment caused some reduction in paw-swelling but did not significantly prevent joint erosion. A suboptimal dose of anti-TNF alone had no significant effect on arthritis. In contrast, anti-CD4 plus suboptimal anti-TNF significantly reduced paw-swelling, limb involvement, and joint erosion. As previously reported, an optimal dose of anti-TNF alone inhibited paw-swelling, limb involvement, and joint erosion. However, optimal anti-TNF combined with anti-CD4 caused significantly greater reductions in paw-swelling and joint erosion than those achieved by optimal anti-TNF alone. Coadministration of anti-CD4 was also effective in preventing an antibody response to the hamster anti-TNF antibody, which may have implications for long-term therapy in human disease. Thus anti-CD4 acts synergistically with anti-TNF in ameliorating established collagen-induced arthritis and this combined therapeutic approach may provide effective long-term control of rheumatoid arthritis.