BAIT-628

Despite broad anti-cancer efficacy as Toll-Like Receptor (TLR) 7/8 agonists, imidazoquinolines remain limited in use via systemic administration or in situ vaccination therapies due to inflammatory toxicity. One approach to address this challenge involves better targeting the action of imidazoquinolines by caging them as glycoconjugate prodrugs. Within cancer cells, imidazoquinoline glycoconjugates are activated by hydrolases prior to efflux by ABC transport proteins, where they then elicit tumoricidal effects from the assistance of bystander immune cells, such as tumor-infiltrating lymphocytes and associated macrophages, in local proximity. While this concept of Bystander-Assisted ImmunoTherapy (BAIT) has been established at a molecular level in vitro, tolerability or efficacy of BAIT has not been reported in vivo. Here, we evaluate the MTD and tumor growth delay efficacy of a lead BAIT prodrug (BAIT628) in a male C57BL/6 mouse TRAMP-C2 prostate cancer model to further establish this methodology. Overall, we find that systemic BAIT628 is well tolerated at over 5-fold the dose-limiting inflammatory toxicity of the parent imidazoquinoline (up to 5 mg/mouse/day I.P. for 10 days). Analyzing serum cytokines reveals that IL-10 production, elicited by the mannoside caging group, likely contributes to the enhanced MTD. Using BAIT628 as an in situ vaccination immunotherapy (seven times over 3 weeks) resulted in significant tumor growth delay and increased survival, both alone and in combination with a murinized α-PD-L1 checkpoint blockade. The tumor histology of tumor-infiltrating immune cell subsets (CD4+, CD8+, CD11c+) reveals significant increases in CD11c+ populations, consistent with TLR7/8 agonism. Overall, BAIT628 is well tolerated and exhibits significant efficacy in the TRAMP-C2 model. These results demonstrate how the BAIT approach can optimize imidazoquinolines for in vivo tolerability and subsequent efficacy as cancer immunotherapeutics.