Predictive Oncology & Intervention Strategies
Molecular Basis of Oncogenesis & Cancer Control
February 7 - 10, 2004Hotel WestminsterNice, France

The three E's of cancer immunoediting

H Ikeda MDa, A Bruce PhDa, MP Gil PhDa, G Dunna, KCF Sheehan PhDa, PhD Shankarana, LJ Old MDb, RD Schreiber PhDa

aWashington University School of Medicine, Washington, MO, St. Louis and, bLudwig Institute for Cancer Research at Memorial Sloan-Kettering Cancer Center, New York, NY, United States

Previous work from our laboratories has shown that mice lacking lymphocytes and/or responsiveness to IFN-gamma (IFNg) form more chemically induced and spontaneous tumors compared to their wild type counterparts. The increased tumor formation in IFNg insensitive mice is due, at least in part, to a failure of the developing tumor to respond to endogenously produced IFNg thus resulting in a tumor that expresses reduced immunogenicity and which can escape immune detection/elimination. Thus, IFNg and lymphocytes collaborate to protect the host against primary tumor development. Chemically induced tumors that develop in immunodeficient mice are, as a group, more immunogenic than tumors that form in immunocompetent hosts. Therefore, the immune system not only eliminates the most immunogenic tumors as they form but also selects for tumor variants with reduced immunogenicity that paradoxically have a better chance of surviving in an immunocompetent host. We have termed this process “Cancer Immunoediting” and envisage that it occurs as a result of three processes: Elimination, Equilibrium, and Escape (the three E's of Cancer Immunoediting). Immunosurveillance occurs during the Elimination process while the Darwinian selection of tumor variants that may have survived the elimination phase occurs during the Equilibrium process. This in turn can ultimately lead to Escape and the appearance and expansion of clinically apparent tumors. Current work has focused on exploring whether there are defined molecular differences between immunologically edited tumors from wild type mice and un-sculpted tumors derived from genetically matched immunodeficient mice. One approach we have employed has been to compare gene expression profiles in progressively growing tumors derived from wild type mice to that of highly immunogenic tumors derived from RAG2-/- mice that are rejected when transplanted into wild type hosts. This analysis has revealed that a limited set of gene sequences 197/36,486 are differentially expressed either in unedited or edited tumors. Some of these have obvious immunological importance and will be discussed further. Thus, we have now obtained molecular evidence for the Cancer Immunoediting process and may have identified some its key molecular targets.

Paper presented at the International Symposium on Predictive Oncology and Intervention Strategies; Nice, France; February 7 - 10, 2004; in keynote session 902 (Keynote).