The efficacy of immunotherapy is dependent on a variety of factors and cells in the tumor microenvironment, and understanding these complex interactions is critical to understanding and improving the efficacy of treatment. To study some of these interactions, Garris and Arlauckas et al. used real-time intravital imaging, RNA sequencing, and a series of pathway manipulations to study crosstalk within the tumor microenvironment, and the role that it plays in the the efficacy of anti-PD-1 immunotherapy.
To begin, Garris and Arlauckas et al. used intravital imaging to track IFNγ- and IL-12-producing cells in mice bearing MC38 colon carcinoma during treatment with anti-PD-1 immunotherapy. As expected, anti-PD-1 increased the number of IFNγ-producing CD8+ T cells. It also increased the frequency of IL-12-producing cells, which were likely dendritic cells (DCs) based on their branched morphology. After treatment with anti-PD-1, these DC-like cells were observed deeper in the tumor and were found to be more motile. RNA sequencing confirmed that these cells were in fact a subset of DCs, related to the well-described cDC1 subset of DCs, though interestingly, they did not express CD103.
Selective depletion of DCs abrogated the antitumor effect of PD-1 blockade, as did neutralization of IL-12, confirming that cytokine support by dendritic cells was required for therapeutic efficacy. After ruling out direct activation of DCs by anti-PD-1, Garris and Arlauckas et al. used intravital imaging to track IL-12 and found that when CD8+ T cells were depleted, or when IFNγ was blocked, production of IL-12 was reduced in the TME, which was mediated by both reduced IL-12 production by DCs and a reduced number of IL-12-producing DCs. Together this suggested that dendritic cells are activated by sensing of IFNγ, which is produced by T cells in direct response to anti-PD-1.
While PD-1 blockade induced production of IFNγ and IL-12 and proliferation of CD8+ T cells within the TME, similar responses were not observed in draining lymph nodes, suggesting that pre-existing tumor-infiltrating T cells were the primary drivers of antitumor response in this model. To determine how IL-12 might impact this effector population, the researchers intratumorally administered IL-12 to MC38-bearing mice and found that IL-12 increased the frequency of infiltrating IFNγ+ cells and led to robust tumor regression. CD8+ T cells isolated from MC38 tumors also increased IFNγ production upon exposure to IL-12 ex vivo, indicating that IL-12 plays a direct role in activating T cells and maximizing IFNγ production.
To better understand the effect of IL-12 in humans, Garris and Arlauckas et al. examined skin lesion biopsies from 19 melanoma patients before and after administration of an IL-12-expressing DNA plasmid and found that intratumoral expression of IL-12 enhanced the expression of a cytolytic gene signature, which was further associated with therapeutic responses. To confirm that IL-12 was acting directly on infiltrating T cells in the human setting, the researchers isolated CD8+ T cells from 6 solid tumor samples samples (various tumor types) and found that exposure to IL-12 ex vivo increased IFNγ production in 5 of 6 samples, confirming that crosstalk observed in mice is likely consistent with human cancer.
Because IL-12 appears to be a critical component in inducing a strong IFNγ response, Garris and Arlauckas et al. explored whether agonizing IL-12-producing cells could be used to enhance responses to anti-PD-1 immunotherapy. To this end, the researchers tested the use of either agonistic CD40 monoclonal antibodies or a small molecule inhibitor of cellular inhibitor of apoptosis protein (cIAP) to trigger the non-canonical NF-κB pathway, which was found to be enriched in IL-12-producing DCs. Both drugs directly enhanced the production of IL-12 (shown using intravital imaging for CD40 agonist and in vitro for the small molecule inhibitor of cIAP), increased the frequency of IL-12-producing DCs in vivo, and showed incomplete antitumor efficacy in MC38-bearing mice. Combination treatment using CD40 agonist and anti-PD-1 outperformed either monotherapy and led to complete, durable antitumor responses in most MC38 tumor-bearing mice. Similar results were observed in the less immunogenic B16F10 melanoma model, with combination treatment leading to tumor rejection in 50% of mice, dependent on IL-12.
Overall, Garris and Arlauckas et al. show that anti-PD-1 directly activates T cell production of IFNγ, which goes on to induce IL-12 production by a subset of dendritic cells. This increase in IL-12 further licensed and maximized effector T cell responses and was essential for the efficacy of anti-PD-1 immunotherapy, challenging the simple model that direct activation of resident antitumor T cells by anti-PD-1 is sufficient for antitumor effects. In preclinical studies, agonizing the NF-κB pathway enhanced IL-12 production and improved the efficacy of PD-1 checkpoint blockade. Data from patient samples suggest that the crosstalk observed in mice is likely maintained in humans, suggesting that the mechanisms observed here may be relevant in ongoing and future clinical trials.
by Lauren Hitchings