A number of mechanisms of action of TNF inhibitors have been identified to date. These include deactivation of the pro-inflammatory cytokine cascade at the site of inflammation; reduction in mediators of joint destruction; diminished recruitment of inflammatory cells from the blood to the rheumatoid joint and diminished synovial vascularity. A role in modulation of apoptosis remains controversial.
The first formal proof that TNF- regulates other pro-inflammatory cytokines in vivo was the observation that there is a rapid reduction in serum IL-6 concentrations, closely followed by falling serum CRP, following administration of infliximab (Elliott et al. 1994; Charles et al. 1999). Although IL-1 concentrations are often below the limit of detection in the peripheral blood of rheumatoid arthritis patients, where it is detectable, down-regulation has been reported in a proportion of patients (Lorenz et al. 1996). Similarly, in a small study of repeat synovial biopsies obtained before and 2 weeks after a single infusion of 10 mg per kilogram of infliximab, computerized image analysis of sections stained for cytokine-producing cells demonstrated a reduction in synovial IL-1a and IL-10 in a subgroup (Ulfgren et al. 2000). It is clear that the benefits of anti-TNF therapy are not mediated by upregula-tion of endogenous pro-inflammatory cytokine inhibitors, since circulating IL-1ra and soluble TNF receptor levels fall after infliximab infusion (Charles et al. 1999).
It is thought that a major mechanism of action of TNF inhibitors is likely to be mediated by modulation of inflammatory cell traffic. A dose-dependent rise in peripheral blood lymphocyte counts is observed following infliximab infusion, with a maximum rise within 24 h of treatment (Paleolog et al. 1996). This is mediated by modulation of both arms of the inflammatory cell recruitment cascade. Thus there is reduced histo-logical expression of synovial cytokine-induced vascular adhesion molecules, such as E-selectin and VCAM-1, following anti-TNF treatment (Tak et al. 1996) and a significant dose-dependent reduction in soluble serum E-selectin and ICAM-1 concentrations (Paleolog et al.
1996), as well as significantly diminished immunohi-stological expression of the chemokines IL-8 and MCP-1, with a trend to reduction in a number of other chemokines (Taylor et al. 2000).
Further indirect evidence to suggest TNF blockade reduces inflammatory cell recruitment to the joint is based on the observation that infliximab therapy is associated with a reduction in numbers of synovial tissue macrophages and lymphocytes (Tak et al. 1996; Taylor et al. 2000). However, the definitive confirmation that TNF- blockade reduces leucocyte traffic to inflamed joints was obtained in an open-label clinical trial demonstrating a 40-50% decrease in retention of autologous indium-111 labelled granulocytes in the hands, wrists, and knees 2 weeks after infliximab treatment (Taylor et al. 2000). There is a reduction in the marginating granulocyte pool after infliximab treatment, an observation that would normally be associated with a rise in peripheral blood granulocyte counts (Taylor et al. 1999). However, in contrast to peripheral blood lymphocyte counts, numbers of peripheral blood granulocytes decrease after infliximab with maximal changes within 24 h. The reason for this is that myeloid cell production is reduced secondary to down-regulation of GM-CSF as a consequence of TNF blockade. Because of the short circulating half-life of the granulocyte, of approximately 8 h, a diminished rate of cell production dominates the peripheral blood picture.
One factor contributing to the very rapid reduction in joint swelling observed by both patients and physicians after anti-TNF therapy is likely to be a reduction in tissue oedema and capillary leak, mediated by vascular endothelial growth factor (VEGF), a cytokine implicated in new blood vessel formation and found to be elevated in the serum of rheumatoid arthritis patients (Paleolog et al. 1998). Serum concentrations of VEGF show a dose-dependent reduction following infliximab infusions, but without normalization. There is also reduction in synovial vascular density and in particular a reduction in angiogenesis, as assessed by diminished expression of vessels expressed the aV|33 integrin (Taylor 2005). Further evidence for a reduction in synovial vascularity following TNF blockade is the observation that the vascular signal on quantitative power Doppler imaging is significantly reduced following infliximab therapy (Taylor et al. 2004, 2006).
Relatively early in the history of clinical trials of TNF blockade, marked reductions in circulating concentra tions of the precursors of the matrix metalloproteinase enzymes MMP-1 and MMP-3 were reported (Brennan et al. 1997) as well as a significant reduction in synovial tissue expression of matrix metalloproteinases (Catrina et al. 2002). Similarly, serum levels of osteoprotege-rin (OPG) and soluble receptor activator of nuclear factor kB ligands (sRANKL), both of which are elevated in rheumatoid compared to normal sera, are normalized following infliximab therapy without influencing the OPG:sRankL ratio (Ziolkowska et al. 2002). These observations predicted the disease-modifying capability of anti-TNF inhibitors, which is now established.
One hypothesis for the failure of the p75 TNF receptor fusion protein etanercept to give clinical benefits in trials of Crohn's disease, in contrast to the marked benefits demonstrated with monoclonal antibodies to TNF-a, is that the antibodies may cause an increase in apoptosis of lamina propria and peripheral T cells through caspase-dependent mechanisms (Van den Brande et al. 2003). However, this topic remains controversial and the relevance of modulation of apoptosis as the mechanism of action of TNF inhibitors in rheumatoid arthritis is unclear. In one study, decreased synovial cellularity was reported as early as 48 h after inflixi-mab administration, but with no corresponding evidence of apoptosis (Smeets et al. 2003). In another study, however, apoptosis in synovial macrophages was reported to be induced by both etanercept and inflixi-mab, with a corresponding increase in active caspase-3 expression. No increase in lymphocyte apoptosis was observed, however (Catrina et al. 2005). The relevance of these interesting observations to the mode of action of TNF inhibitors in rheumatoid arthritis is not clear.
Was this article helpful?