Cytokine Traps: Novel Anti-Cytokine Blockers
Protein-based cytokine inhibitors have been demonstrated to be very efficacious in the treatment of rheumatoid arthritis (RA), in particular antagonists of tumor necrosis factor-a (TNF-a). These consist of a soluble TNF receptor (etanercept;Enbrel®) and two monoclonal antibodies against TNF (infliximab or Remicade®; adalimumab or Humira®). Interleukin-1 is another cytokine that has potent proinflamamtory properties in RA. However, the only available inhibitor of interleukin-1, the naturally occurring IL-1 receptor antagonist (anakinra; Kineret®), must be administered daily and in relatively high dose due to receptor occupancy issues and rapid clearance in vivo.
IL-1 (like IL-4 and IL-6) uses a complex multi-component receptor system in which IL-1 binds with low affinity to the first receptor component (IL-1receptor type I; IL-1RI) and then recruits a second receptor component (IL-1R accessory protein; IL-1RAcP). The complex formed by these two receptor domains (IL1-RI and IL-1RAcP) binds IL-1 with substantially higher affinity than either component alone.
Methods: Economides AN et al (Nature Medicine 9:47-52, 2003) reasoned that a soluble receptor incorporating both receptor components would likely bind cytokine with high affinity and serve as a potentially potent therapeutic antagonist. They termed these molecules cytokine traps. The trap consists of two identical molecules composed of inline (sequential) fusion of each receptor extracellualr domain followed by the human IgG1 Fc, covalently linked by disulfide bonds at the Fc portion of each molecule. The investigators tested IL-6, IL-4 and IL-1 traps in this study both in in vitro and in vivo models. Only the results for the IL-1 trap will be summarized.
Results: IL-1 trap inhibited IL-1 (and IL-1a) induced IL-6 production in the human MRC5 fibroblastic cell line with an IC50 of ~2 pM, while the IC50 of IL-1ra (kineret) was 70 pM. This affinity of IL-1 trap is also approximately 100-fold higher than those reported for soluble IL-1RI and IL-1RII (1000 pM and 500 pM, respectively). The dissociation constant of IL-1 trap was tested directly using an ELISA assay, and the Kd was similar to that derived from the biologic assay (Kd 1.5 pM). IL-1 trap also binds IL-1ra but binds IL-1 about 23-fold more tightly than IL-1ra. IL-1RI alone binds IL-1ra more tightly than IL-1b. The addition of the IL-1RAcP component to create the IL-1 trap reserved the relative affinities so that trap binds IL-1 more tightly than IL-1ra. The terminal half-life of IL-1 trap in primates was 67 hours.
In vivo in mice, IL-1 trap was highly effective in suppressing IL-1 induced elevation in IL-6 levels, whereas a 15-fold higher dose of IL-1ra was ineffective in this regard. IL-1 trap was also highly effective in suppressing inflammation and joint erosions in mice with collagen-induced arthritis.
Conclusion: Cytokine traps (inline fusion proteins) represent a novel approach to the design of highly potent cytokine antagonists. Their low picomolar affinity is comparable to that of monoclonal antibodies and may be higher than the affinity of naturally occurring receptors on cells. In vitro and in vivo animal studies indicate a prolonged half-life and impressive efficacy and potency.
Editorial Comments: This methodology offers a novel, exciting approach to the design of more effective, more potent antagonists for IL-1. In fact, a preliminary dose ranging study of IL-1 trap in human RA was reported at the 2001 ACR meetings wherein significant reductions of CRP were observed at the highest dose. Although clinical improvement was also observed, this was an open label trial without placebo control. A randomized, placebo-controlled phase II trial is currently underway, and the results of this trial should be quite interesting. If the agent proves safe and efficacious, it is likely to replace IL-1ra (kineret). Given its high potency and prolonged half-life, it will be critical to determine whether treatment with IL-1 trap is associated with a higher risk of infection.