Identification of a novel resistant mechanism for agonistic costimulatory therapy

Recent approval of YERVOY™ (ipilimumab, anti-CTLA-4) and Keytruda (pembrolizumab, anti-PD-1) by the US FDA marks the validation of immunotherapy for advanced metastatic melanoma patients, which has led to a renaissance in melanoma therapeutics using immunostimulatory monoclonal antibodies (mAbs) that enhance immune responses. In addition to blocking immune-inhibitory molecules (e.g. CTLA-4 and PD-1), activating immune co-stimulatory pathways (e.g. OX40, CD137 and CD40) to potentiate antitumor immune responses is a promising approach (1, 2). OX40 (TNFRSF4, CD134), a member of the tumor necrosis factor receptor (TNFR) superfamily is expressed on activated both CD4+ and CD8+ T cells (3). As OX40 engagement can promote T cell proliferation, survival, effector function and memory (4-7), agonistic mAbs and/or soluble forms of OX40L have been used successfully in a variety of preclinical tumor models and clinical trials (1). As a monotherapy, OX40 engagement has been unable to provide adequate antitumor immunity in established and more clinically relevant poorly immunogenic tumors, while being effective in eradicating primarily small immunogenic tumors (1), calling for identification of resistance mechanisms. It is conceivable that an immunosuppressive tumor microenvironment results in resistance to agonistic anti-OX40 therapy. CD73-mediated adenosinergic effects are now considered one of the most important immunosuppressive pathways in the tumor (8- 13). We and others have demonstrated the pivotal role of tumor and host CD73-mediated adenosinergic effects on tumor growth and metastasis (14-17), and up-regulation of CD73 expression has been also observed in a broad spectrum of solid cancers including melanoma (8). We further verified the antitumor effect of anti-CD73 administration in a more clinically relevant melanoma model, i.e. the Braf-activated/Pten-deficient mouse model (18, 19). To this end, we examined the inhibitory role of CD73 in the context of therapies targeting co-stimulatory molecule OX40 and set out to determine whether CD73-mediated adenosinergic effects are a crucial resistant mechanism for agonistic anti-OX40 mAb therapy. Our preliminary data showed that CD73-deificient mice treated with agonistic anti-OX40 mAb completely rejected B16 melanoma growth when compared with WT mice, highlighting the importance of an inhibitory role for host-derived CD73. To explore the therapeutic relevance, we combined anti-OX40 and anti-CD73 mAb to treat mice with established poorly immunogenic B16 melanoma. This combination therapy induced tumor regression associated with enhanced antitumor CD8+ T cell responses and decreased number of both peripheral and tumor-infiltrating Treg cells leading to a favorable T effector/ T reg cell ratio. In contrast, neither anti-OX40 nor anti-CD73 monotherapy is effective to control the growth of established melanoma. Along these lines, we hypothesize that CD73-mediated adenosinergic effects are a novel crucial resistant mechanism for agonistic anti-OX40 mAb therapy.