Tumor-promoting inflammatory networks in pancreatic neoplasia: Another reason to loatheKras

Abstract

Oncogenic Kras is increasingly appreciated as an instigator of an inflammatory program that facilitates pancreatic cancer. In this issue of Cancer Cell, McAllister et al use both gain-of-function and loss-of-function studies to demonstrate that oncogenic Kras activates an IL-17-dependent pathway that promotes the initiation and progression of preinvasive pancreatic neoplasia.

Pancreatic cancer is well-known for its very poor prognosis. The disease is notoriously refractory to standard therapy, especially for patients who present with metastatic disease. Unfortunately, the “statistics” are getting worse. Pancreatic cancer is one of only two cancers among 21 histotypes for which the death rate in the United States rose from 1990 to 2008 (the other is liver/intrahepatic bile duct cancer). Pancreatic cancer is likely to become the second leading cause of cancer-related death in the United States by 2020 (behind lung cancer, which is declining). Although new combination chemotherapies are able to stabilize many patients who present with metastatic pancreatic cancer (representing a major clinical advance), response rates remain <35%, and patients with long-term complete remissions after treatment with these therapies are rare.

In the context of this unmet clinical need, genetically engineered murine models of pancreatic neoplasia have become an increasingly important tool to garner biological insights that might lead to novel therapeutic strategies. Based largely on the selective expression of oncogenic Kras in the pancreas of immune-competent hosts, these models reproduce key biological aspects of the human disease including its highly desmoplastic and inflammatory tumor microenvironment. These models offer the opportunity of uncovering novel “non-tumor cell autonomous” therapeutic targets in the tumor stroma which can be tested for efficacy in tumor-bearing mice prior to translation to the clinic (Beatty et al, 2011). Because pathology in these models advances from preinvasive disease to invasive disease, agents that prevent or reverse the earliest neoplastic events can also be explored.

In this issue of Cancer Cell, McAllister et al (2014) identify early inflammatory events, dependent on the cytokine IL-17 and oncogenic Kras, that drive pancreatic neoplasia in vivo in the Mist1^CreERT2/+; LSL-Kras^G12D murine genetic model. IL-17 and related cytokines potently evoke tissue inflammation and can trigger autoimmunity by interacting with multiple cell types in various tissues. In cancer, the role of IL-17 (as well as that of Th17 cells, helper T cells that secrete IL-17) is far from clear cut; in some experimental systems, Th17 cells promote tumor formation but in other studies, Th17 cells have clear, direct, and powerful anti-tumor effects (Chen and Oppenheim, 2014). In addition, IL-17 has been shown to mediate a paracrine network that promotes tumor resistance to anti-angiogenic therapy by the induction of GCSF expression and recruitment of immunosuppressive immature myeloid cells in the tumor microenvironment (Chung et al, 2013). Many of these studies, however, focus on implantable tumor models or employ immune-incompetent mice, for which the tumor microenvironment, especially at early stages of neoplastic growth, is not ideally reproduced. In their genetic model, McAllister et al identify two distinct IL-17-secreting lymphoid cell populations, Th17 cells and γδ T cells, recruited to the pancreas in the setting of mutant Kras expression and inflammation. Both cellular subsets are rare in the tissue, outnumbered by myeloid inflammatory cells, and in fact, were not specifically noted in the initial immunophenotyping studies of genetically engineered mice bearing Kras-driven pancreatic intraepithelial neoplastic (PanIN) lesions (Clark et al, 2007). Nevertheless, in both gain-of-function and loss-of-function studies in vivo, McAllister et al (2014) demonstrate that IL-17 helps to drive initiation and progression of pancreatic neoplasia. Interestingly, Th17 cells have been previously reported to promote pancreatic inflammation in the setting of autoimmune diabetes in immune-incompetent mice in which anti-IL-17 treatment reduces insulitic inflammation and disease progression (Martin-Orozco et al, 2009). In human PanIN tissues, McAllister et al (2014) identify many elements of the IL-17 inflammatory fingerprint, including PanIN-infiltrating lymphocytes that express RORγt, the transcription factor classically linked to Th17 cells. These findings underscore the potential relevance of IL-17 in the initiation of human pancreatic cancer.

McAllister et al (2014) also observe Kras-dependent expression of functional IL-17 receptor (IL-17R) on pancreatic neoplastic epithelium. Based on a genetic lineage tracing allele, IL-17R expression is also evident on pancreatic neoplastic cells undergoing epithelial-mesenchymal transition. Neutralization of IL-17 in vivo leads to a rapid change in the gene expression program of neoplasic cells, including the loss of IL-6 expression and STAT-3 phosphorylation, two key regulators of pancreatic neoplastic progression (McAllister et al, 2014). Importantly, this IL-17-epithelial cell crosstalk is observed only in the presence of oncogenic Kras. Although it remains to be determined if IL-17R signaling triggers a cell-intrinsic effect that promotes oncogenesis, there likely is a paracrine aspect ofIL-17 in the neoplastic microenvironment, culminating in a disease-promoting scenario.

The findings are significant for a number of reasons. First, these results underscore a role for IL-17 as an inflammatory mediator that promotes neoplasia. Second, the results offer novel insights regarding the role of IL-17 in pancreatic neoplasia in particular and Kras-mediated oncogenesis in general. Third, the findings point to a potential translational path to neutralize IL-17 in patients with pancreatic neoplasia, or those at high risk, given the active development of effective IL-17 monoclonal antibodies that have shown stunning promise in other (non-cancer) inflammatory diseases (van den Berg and McInnes, 2013).

Curiously, CD4 T cell depletion in this model (by the use of depleting monoclonal antibody) reproduces aspects of the phenotype observed with IL-17 neutralization, even though only a small fraction of CD4 T cells secrete IL-17 and other IL-17-secreting cells identified (such as γδ T cells) do not express CD4. These observations are consistent with a second report published earlier this year that IL-17 secreting CD4 T cells are found in the pancreas upon induction of mutant Kras and treatment with the inflammatory agent cerulein, and that genetic deletion of CD4 cells abrogates PanIN formation, an effect that requires CD8 T cells (Zhang et al, 2014). Regulatory T cells (Tregs) comprise a far larger percentage of PanIN-infiltrating CD4 T cells than Th17 cells, and therefore, Th17 cells, Tregs, and perhaps CD4 Th2 T cells may cooperate to accelerate pancreatic neoplasia disease. Accumulating evidence shows that Th17 cells and Tregscan stimulate each other in vivo such that therapeutic manipulation of one cell type will impact the other (Chen and Oppenheim, 2014).

Further experiments are needed to evaluate whether IL-17 neutralization is a useful therapeutic strategy in the setting of invasive pancreatic carcinoma as it appears to be, at least experimentally, in the early stages of pancreatic neoplasia. The development of potent, clinical-grade anti-IL17 mAb certainly presents a translational temptation (van den Berg and McInnes, 2013); however, additional studies are needed to ensure that such treatment would not be counterproductive, given the anti-tumor activity of Th17 in certain models (Chen and Oppenheim, 2014). Such strategies are amenable for testing via genetically engineered mice with pancreatic cancer.

Finally, it is of great interest that the IL-17 phenotype reported by McAllister et al (2014) emerges in the setting of oncogenic Kras, which appears capable of orchestrating a tumor-promoting microenvironment beyond well-described tumor-cell autonomous Kras mechanisms. GM-CSF expression by PanIN and invasive pancreatic cancer cells is another example of a pathway downstream from oncogenic Kras that has non-cell autonomous effects, in this case functioning to establish an influx of suppressive myeloid cells that inhibit adaptive immunity (Bayne et al, 2012; Pylayeva-Gupta et al, 2012). Pharmacological inhibition of oncogenic Kras, therefore, might realistically be expected to derail these tumor-promoting non-cell autonomous mechanisms, providing even more incentive (if more were needed) for renewed efforts to drug Kras.