Amg 592 Is an Investigational IL-2 Mutein That Induces Highly Selective Expansion of Regulatory T Cells

Background: Low-dose interleukin-2 (IL-2) therapy expands regulatory T cells (Tregs) and provides clinical benefit for inflammatory diseases; however, concomitant increases in conventional effector T cells (Tcon) and natural killer (NK) cells may result in toxicity. AMG 592 is an investigational IL-2 mutein designed for greater Treg selectivity and longer half-life compared with recombinant IL-2 (aldesleukin). In preclinical studies and a phase 1, double-blind, placebo (PBO)-controlled first-in-human (FIH) study, we investigated the safety and tolerability of AMG 592 and its effects on expansion of Treg, Tcon, and NK-cells.

Methods: AMG 592 activity was assessed as in vitro phosphorylated STAT5 (pSTAT5) in primary human peripheral blood mononuclear cells (PBMC). Treg, Tcon, and NK-cell expansion and cytokine production were assessed after culture of pre-stimulated human PBMC with increasing doses of AMG 592 or aldesleukin. The in vivo effects of AMG 592 on body temperature, C-reactive protein (CRP), and peripheral blood Treg, Tcon, and NK numbers were evaluated in cynomolgus monkeys after escalating single SC doses of AMG 592 or 5 consecutive daily SC doses of aldesleukin. In the FIH study, healthy volunteers in 8 dose cohorts received a single SC ascending dose of AMG 592 (n=6 per dose) or placebo (n=2 per dose) and were evaluated for safety and tolerability, pharmacokinetics (PK), pharmacodynamics, and cytokine production for 28 days.

Results: In human PBMC cultures, AMG 592 caused more selective Treg response (pSTAT5 and proliferation) and lower production of pro-inflammatory cytokines than aldesleukin. Dose-dependent expansion of FoxP3+ Tregs in aldesleukin-treated cynomolgus monkeys was accompanied by increased body temperature and CRP; however, AMG 592 caused Treg expansion without significant effects on body temperature or CRP. In the ongoing FIH study, AMG 592 was well tolerated, with no serious adverse events (AEs). The most common AE across treatment arms was grade 1 painless erythema at or near the injection site that resolved without treatment. Preliminary PK results indicate dose-related increases in AMG 592 serum exposure. AMG 592 caused a robust, dose-dependent expansion of Tregs relative to Tcon in all treated individuals. Expanded Tregs had increased levels of CD25 and Foxp3, and were enriched for recent thymic emigrants. At the highest dose, the increase in the Treg:Tcon ratio peaked at day 8 (~4-fold vs baseline) and remained elevated up to day 29. Treg expansion by AMG 592 was highly selective, with no directional change in NK-cell numbers and minimal increase in Tcon. Moreover, there were no increases in the serum pro-inflammatory cytokines IL-6, TNFα, or IFN-γ above the limits of detection.

Conclusion: AMG 592 caused a dose-dependent, selective expansion of Tregs in healthy volunteers. The lack of pro-inflammatory cytokines and reduced markers of inflammation by AMG 592 compared with aldesleukin suggests a wider therapeutic margin. The sustained Treg elevation compared with aldesleukin suggests less frequent dosing may be feasible. Further investigation of the ability of AMG 592 to restore immune homeostasis by Tregs in inflammatory and autoimmune diseases is warranted.