Encyclopedia of Cancer. Springer-Verlag (2015)
Although there is an abundance of evidence for immunoediting in murine models of cancer, the evidence that this occurs in human cancer is largely circumstantial. It has long been known that cancer patients develop immune responses to their own tumors. Indeed, both tumor-specific T cells and antibodies isolated from cancer patients have been harnessed to identify tumor-associated antigens by methods such as SEREX. In addition the ability of a patient to mount a response to their tumors, particularly if there is evidence of tumor infiltration of immune effector cells, is a strong predictor of a favorable prognosis. That the immune system sculpts tumor immunogenicity in tumors that arise in cancer patients is supported by the emergence of tumors in patients undergoing antigen-specific immunotherapy that have downregulated the tumor antigen to which the therapy is directed or components of the cellular machinery that generates the cell surface complex of HLA I and antigen recognized by CTLs. These include LMP proteasome subunits and transporter molecules that chaper- one antigen peptides from the cytosol to the Golgi apparatus for loading onto HLA class I molecules. Sometimes the HLA I molecule itself is lost, usually by mutation of the β2 microglobulin subunit, common to all HLA class I molecular complexes. In some cases, however, lack of HLA class I on the surface of a tumor is due not to mutational events in the HLA class I genes themselves but to downregulation of expression of the protein. This can often be reversed by cytokines, such as IFN-γ, but it is interesting to note that human tumors have been shown to arise that lack the IFN-γ receptor. The phenomenon of HLA class I loss from the surface of tumor cells has been documented in numerous clinical cancer vaccine trials for melanoma, prostate carcinoma, and HER-2/neu positive tumors.
In addition, even in the absence of immunotherapy, HLA class I expression has been shown to be lost or downregulated in all types of tumors especially in patients with advanced disease. Indeed, the frequency of deletion or downregulation of these cell surface molecules has been found to be as high as 15% in primary melanoma lesions and 50% in primary prostate carcinoma lesions. A further piece of evidence for immunoediting in human cancer is that patients that are seriously immunosuppressed, for example, post organ transplant, have a higher incidence of cancer than healthy individuals. Finally, the recent emergence of so-called checkpoint inhibitors of immunoinhibitory molecules such as CTLA-4, PD-1, or PD-L1 to treat cancer, particularly melanoma, by augmenting the natural immune response of the patient against his or her own tumor also supports the existence of immunosurveillance in human cancer.