A Novel Human Anti-Pd1/Il15 Bi-Functional Protein With Robust Anti-Tumor Activity And Low Systemic Toxicity

Background IL-15 is a key cytokine promoting CD8+ T, NK, and NKT cell proliferation and has demonstrated clinical activity in cancer patients without evidence of any Treg stimulation.1 2 However, its short half-life and systemic toxicity limit its clinical utility. Kadmon has established an IL-15 fusion protein platform to extend the IL-15 serum half-life and direct its action to the tumor microenvironment.3 A major asset of this platform is anti-PD1/IL15 bifunctional protein. To test the bifunctionality hypothesis of this fusion protein in murine models, four different formats of the surrogate bi-functional proteins were engineered by fusing mouse IL-15 to a mouse-human chimeric anti-mouse PD1 antibody (m3A7). We presented earlier that the single IL-15 N-terminal fusion to anti-PD1 antibody (1N-IL-15/m3A7) showed significantly stronger anti-tumor activity in vivo mainly due to the cis-presentation to the PD1 and IL2Rβγ co-expressed on TILs. The cis-presentation potentially maximizes the bi-functionality of PD1 blockade and IL-15 stimulation.4 Here, the therapeutic single IL-15 N-terminal fusion antibodies containing a novel human PD1 antagonist antibody (38B2) and either wild-type IL15 (1N-IL-15/38B2) or mutated 65S-IL15 (65S-1N-IL-15/38B2) have been constructed; their anti-PD1 functions, IL15 stimulation and anti-tumor activities were evaluated. Methods Purified 1N-IL-15/38B2 and 65S-1N-IL-15/38B2 were generated and characterized in vitro.4 The anti-tumor activities were examined in the human-PD-1/PD-L1 transgenic BALB/c mice subcutaneously transplanted with the human-PD-L1 positive CT26 colon carcinoma. The treatment was started when tumors reached 100 mm3 (IP, QW). Results All 1N-IL-15/anti-PD1 fusions showed similar potencies in binding to the soluble and cell expressed human PD1 and blocking the hPDL-1 binding to hPD1. Comparing to wild-type 1N-IL-15/38B2, mutated 65S-1N-IL-15/38B2 showed lower stimulation (\u003e150 folds) in the M07e, CTLL2, mouse spleen cells and hPBMC (mainly CD8T+ T cell) proliferation. When we treated hPDL1-CT26 tumor transplanted hPDL1-hPD1 transgenic mice with 65S-1N-IL-15/38B2 at 6 mg/kg, 80% of tumor growth inhibition (TGI) was achieved with no body-weight loss. Although wild-type 1N-IL-15/38B2 at 3 mg/kg demonstrated similar efficacy, a significant mouse body-weight loss was observed and 1/3 mice died after second injection. No anti-tumor activity was observed for 65S-1N-IL-15 non-target fusion in 6 mg/kg. Conclusions The previous observation of robust anti-tumor activity of surrogate 1N-IL-15/m3A7 in PD1 resistant LL2 model was replicated with the therapeutic bifunctional protein in this study. We also found that lower stimulation 65S-1N-IL-15/38B2 showed strong anti-tumor activity with significant low systemic toxicity; suggesting that the 65S mutation increased the therapeutic window of this bi-functional protein References Goldrath, AW etc. Cytokine requirements for acute and basal homeostatic proliferation of naive and memory CD8+ T cells. J Exp Med 2002; 195:1515–1522. Waldmann TA. Cytokines in Cancer Immunotherapy. Cold Spring Harb Perspect Biol 2018;103. Martomo, S. etc. ESMO 2019 1221P; SITC 2019 P#4854. Polonskaya Z. etc. AACR 2020 #2263