Targeting the IL-12/IL-23 axis: An alternative approach to removing tumor induced immune suppression.

Combination immune checkpoint blockade has demonstrated significant clinical responses in cancers infiltrated by T cells. Many tumors contain high proportions of myeloid cells and these can secrete immunosuppressive cytokines like IL-23. Our data suggest the clinical potential of using anti-CD40 (push) and anti-IL-23 mAbs (pull) to tip the IL-12/23 balance in established tumors and act as an alternative combination cancer immunotherapy.

Article

Cancer immunotherapy has come of age with predictions that immune-based approaches to cancer treatment will increase in use very significantly in the next 10–15 y. Over the past 5 y, the class of immunotherapies that target T cell checkpoint receptors have achieved remarkable success and generated tremendous enthusiasm among oncologists. These include the FDA-approved α-CTLA-4 monoclonal antibody (mAb) and the α-PD-1:PD-1 ligand mAbs. These studies demonstrate that in many cancers, the immune system is not inert, but rather actively suppressed, and that an endogenous anti-tumor response can be reactivated following checkpoint blockade. Although T cell checkpoint blockade has been validated as a means to relieve tumor-induced immune suppression in cancers such as melanoma and non-small cell lung carcinoma (NSCLC)(20–50% response rates), one envisages that multiple suppressive pathways will operate in any given tumor microenvironment. Thus, targeting one pathway may not release full endogenous anti-tumor immunity. Certainly, deeper and more rapid responses were observed in advanced melanoma patients given concurrent α-PD-1 and α-CTLA-4 compared with monotherapy alone. Nevertheless, some cancers (e.g., prostate, pancreatic adenocarcinoma, gastric and esophageal adenocarcinoma did not show objective responses and even in melanoma, there are a large number of patients where checkpoint blockade in combination fails.

Myeloid derived suppressor cells (MDSC) and/or tumor-associated macrophages (TAM) are another major group of immune cells known to suppress anti-tumor immunity. They are prevalent in many human tumors and inhibit T effector function through the production of immunosuppressive cytokines and factors such as IL-10 and TGF-β or indoleamine 2, 3-dioxygenase (IDO) and arginase. Recently, our group and others discovered that IL-23, produced by myeloid cells, is a novel cytokine that promoted tumor growth and development by suppressing T cell and NK cell effector function. Given human tumors are extremely heterogeneous with respect to their proportion of lymphoid and myeloid cells, co-blockade of immunosuppressive pathways mediated by T and myeloid cells may result in a greater release of endogenous anti-tumor immunity. In addition, combination approaches that aim to remove immunosuppression and activate immune cells are currently being examined in clinical trials.

Recently, we demonstrated that the combination of agonistic anti-CD40 mAbs to drive IL-12 production, and anti-IL-23 mAbs to counter the tumor promoting effects of IL-23, had greater anti-tumor activity than either agent alone. This increased anti-tumor efficacy was observed in several experimental and spontaneous lung metastases models as well as in models of de novo carcinogenesis. The combination effects were dependent on host IL-12, perforin, IFNγ, NK and/or T cells and independent of host B cells, IL-17A, and IFNαβ sensitivity. Thus, we have demonstrated that a combination of agonistic CD40 antibody to activate Th1 immune response and anti-IL-23 mAbs to neutralize myeloid immunosuppression of T and/or NK cells is an alternate approach for cancer immunotherapy (Fig. 1). Although the mechanism of action of agonistic anti-CD40 mAbs was thought to be through licensing of antigen presenting cells to activate effector T cells; in recent years, studies have demonstrated that a T-cell independent response is also activated. Using a genetic mouse model of pancreatic ductal adenocarcinoma, agonistic CD40 mAb induced tumor regression in the absence of T cells. This was mediated by activating CD40-expressing macrophages to become tumoricidal or by converting M2 tumor promoting macrophages into M1 phenotype. Indeed many cancer subtypes are enriched in myeloid infiltrates including pancreatic adenocarcinoma, colorectal carcinoma, breast and prostate cancer. In addition, significant upregulation of IL-23p19 mRNA expression has also been detected in human cancers that are myeloid rich, including colorectal carcinoma, lung and breast cancer with one report suggesting that circulating IL-23 was increased in breast cancer patients and negatively correlated with survival. Although we showed that the combination effects in our study were dependent on host NK and/or T cells in the tumor models we used, we envisage this combination may also be effective in patients whose cancer display rich myeloid infiltrates and upregulated IL-23 (Fig. 1). In this setting, neutralization of IL-23 in the tumor microenvironment may allow the agonistic activity of anti-CD40 to be fully maximized. Other cancer types that may benefit from this combination are sarcomas, a disease not well explored immunologically and immunotherapeutically, yet surveillance studies from our group and others suggests that they are highly immunogenic and a good target for immunotherapy. In a de novo mouse model of soft tissue sarcoma, combination anti-CD40/IL-23 therapy of established tumors significantly suppressed tumor growth.

Figure 1. An alternative strategy for alleviating tumor induced immunosuppression by targeting the IL-12/IL-23 axis of inflammation. Cancer patients have heterogeneous infiltrates of T and/or myeloid cells and varying levels of IL-23 in their tumors. Targeting the IL-12/23 axis using anti-CD40 and anti-IL-23 may be an alternative approach to treat cancer patients whose tumors are T cell rich but do not respond to combination checkpoint blockade. Similarly, in cancer types that are myeloid rich, anti-CD40 and anti-IL-23 therapy may induce better anti-tumor response than checkpoint receptor blockade.

The source of IL-23 and how it mediates its suppression are key questions that remain to be addressed. Interestingly, in the experimental lung metastases tumor models used in our study, we observed that intracellular IL-23 production was specifically restricted to MHC-IIhiCD11c+CD11b+ cells. We are currently characterizing the phenotype of these cells and determining if they are also present in subcutaneous tumor models where neutralization of IL-23 was therapeutic. Given that IL-12 and IL-23 share a common receptor subunit, IL-12Rβ1, it was assumed that expression of the IL-12 and IL-23 specific receptors would be expressed on cells where IL-12Rβ1 was present; neutralization of IL-23 therefore allowed the activation of a Th1 signaling pathway. However, a recent paper has demonstrated a dichotomous pattern of expression for IL-12 and IL-23 receptors in both mouse and human, suggesting that immune cells involved in anti-tumor responses may be quite distinct from those that are tumor promoting.

Overall, our data suggest the clinical potential of using anti-CD40 (push) and anti-IL-23 mAbs (pull) to tip the IL-12/23 balance in established tumors.