In the first half of the twentieth century, a group of prominent scientists interested in the genetics of cancer developed inbred strains of mice, some of which were characterized by a high frequency of various neoplasias, including mammary tumors and leukemias. The in-vestigators’ objective was to mate mice from these strains with mice from low-frequency strains, and from the distribution of the characteristic cancer phenotype, they were hoping to understand the genetic factors underlying the susceptibility and resistance to the disease. However, in the course of developing inbred stocks of high cancer frequency, a genetic background was inadvertently created that allowed the multiplication of endogenous, can-cer-causing retroviruses; in turn, this made identification of the viruses possible. A second unforeseen consequence of these crossing and tumor transplantation studies among differ-ent mouse strains was the discovery of the loci associated with the major histocompatibil-ity complex (MHC), a discovery that has shaped and directed the progress of immunology to the present day.
Additional studies and clinical observations led to the development in the 1950s of the concept of immune surveillance, a term coined to describe natural immunological re-sistance against cancer. Although studies conducted with immunodeficient mice have mostly supported the idea that the main purpose of T-cell–mediated cellular immunity is to protect against viral and other infections, the immunosurveillance hypothesis provided an important conceptual framework for the field of tumor immunology. For instance, it is well known that renal transplant patients undergoing treatment with immunosuppressive agents exhibit a significant increase in skin cancers, especially on sites of the body exposed to UV- containing sunlight, suggesting that immunosurveillance does play a role in protecting hu-mans against special types of tumors, including virally induced warts.
The lack of clear evidence that immune surveillance plays a major role in protecting humans from the most common forms of cancer does not imply that human tumors express no tumor antigens nor that they are insensitive to destruction by immunological means. As a matter of fact, an increasing body of experimental evidence indicates that a variety of hu-man cancers express tumor-associated antigens, some of which may serve as targets for cancer immunotherapy. Moreover, additional experimental evidence suggests that tumors may exercise immunosuppressive effects on the host, indirectly confirming that a healthy immune response can help the host in controlling cancer growth and spread.