The predicted clinical success of anti-TNF therapy was based on several experimental observations. The first of these was the expression of TNF-a and its receptors in rheumatoid arthritis synovial tissue (Maini and Taylor 2000). Secondly, evidence from in vitro experiments employing dissociated synovial cell cultures pointed to TNF- as a regulator of many other pro-inflammatory cytokines (Brennan et al. 1989; Butler et al. 1995; Haworth et al. 1991). Thirdly, a number of independent in vivo studies demonstrated that blockade of bioactive TNF in murine collagen-induced arthritis can ameliorate clinical symptoms and prevent joint destruction in established disease (Williams et al. 2000a). Finally, in a murine model, over-expression of a human TNF-a transgene modified at its three prime ends to prevent degradation of its MRNA is associated with the development of a destructive form of polyarthritis 4-6 weeks after birth. This can be prevented by administration of a monoclonal antibody with specificity for human TNF (Keffer et al. 1991).
An extensive range of pro-inflammatory cytokines can be detected at the protein level in human synovial samples, regardless of differences in donor disease duration, severity, or even drug therapy, and has been confirmed in studies from a number of laboratories. These findings imply that there is prolonged cytokine expression in the rheumatoid joint, contrasting with the transient production induced by mitogenic stimulation. This hypothesis was confirmed following the observation that pro-inflammatory cytokines are produced over several days in dissociated rheumatoid arthritis synovial membrane cell cultures in the absence of extrinsic stimulation (Buchan et al. 1988). This finding suggested the presence of one or more soluble factors regulating prolonged cytokine synthesis within the rheumatoid synovial membrane cultures. These dissociated cell cultures comprise a heterogeneous population of cells producing numerous cyto-kine and other non-cytokine molecular messengers. A key observation using this cell culture system was that addition of anti-TNF antibodies strikingly reduced the production of other pro-inflammatory cytokines including IL-1, GM-CSF, IL-6, and IL-8 (Brennan et al. 1989). Furthermore, using the same rheumatoid arthritis synovial cell culture system, blockade of IL-1 by means of the IL-1 receptor antagonists results in reduced IL-6 and IL-8 production but not that of TNF-a (Butler et al. 1995). These observations led to the formulation of the hypothesis that TNF-a occupies a dominant position at the apex of a pro-inflammatory cyto-kine network. At the time of these experimental findings, in the early 1990s, TNF-a had also been described as a pleiotropic cytokine with biological properties that included enhanced synovial proliferation, production of prostaglandins and metalloproteinases (Dayer et al. 1985), as well as regulation of other pro-inflammatory cytokines. For all these reasons, TNF- was considered to represent a potential therapeutic target in rheumatoid arthritis.
The first data in vivo to support the hypothesis that TNF- is a good therapeutic target for inflammatory arthritis came from studies of murine collagen-induced arthritis. Monoclonal anti-TNF antibodies or soluble TNF receptor-Fc fusion proteins, administered either during the induction phase of arthritis or, more importantly, in the established phase of disease after the onset of symptoms, were able to ameliorate clinical features and significantly inhibit joint destruction (Williams et al. 1992; Piguet et al. 1992). Further unequivocal validation ofTNF-a as a therapeutic target came following the administration of biologic agents to patients with rheumatoid arthritis.
Was this article helpful?