Vaccines based on self (tumor) antigens: Accomplishments and challenges.

CS03-02

Tumors are recognized by the immune system through molecules that distinguish tumor cells from the surrounding normal cells or from similar tissues throughout the body. These molecules, known as tumor antigens, are numerous and varied, both in their origin and chemical composition. Thus, proteins, carbohydrates, peptides, lipopeptides, glycopeptides, phosphopeptides, can all be differentially expressed by cancer cells due to their different metabolic properties as well as altered genetic profiles. Furthermore, as cancers have different etiologies, molecules derived from some etiologic agents, such as for example oncoviruses, can serve as tumor antigens. Many tumor antigens were originally defined either by mouse monoclonal antibodies that distinguish tumor cells from normal cells, or by cancer patients' antibodies and T cells that specifically recognized their tumor cells. Identification of tumor antigens recognized by the adaptive immunity provided proof that the immune system is involved in the surveillance of tumor growth, as originally proposed by Burnett but until recently difficult to prove experimentally , and raised questions about the quality and the duration of that response and the apparent ability of tumors to evade immune destruction. Most recently, it has become clear that many tumor antigens interact not only with antigen specific receptors on T and B cells, but also with various pattern recognition receptors on the cells of the innate immune system , which can determine the type of an adaptive anti-tumor immune response that is generated. We have identified and characterized two human tumor antigens, MUC1 and cyclin B1, and explored their interactions with the human immune system. These two molecules belong to the category of shared tumor antigens, being expressed by many human tumors. Furthermore they are in the category of self/tumor antigens, being expressed on the normal tissues from which the tumors originate, but overexpressed or aberrantly expressed on tumors. Both of these antigens are targets for immunosurveillance, and antibodies and T cells specific for MUC1 and cyclin B1 can be found in patients with MUC1+ and cyclin B1+ tumors. Patients who at diagnosis show evidence of anti-MUC1 or anti-cyclin B1 immunity have better prognosis. We will present data in support of this immunity being generated early in the patient's life as part of normal immunosurveillance of premalignant lesions. Depending on the conditions under which the immune responses were generated and on the general immune competence of the individual, that immunity may be very strong and life-long. On the other hand, these immune responses may be weak and short-lived, thus allowing early tumors to evade immune destruction. There are several immunotherapy approaches used currently to boost patient's anti-tumor immune responses. Some "passive immunotherapy" approaches, such as monoclonal antibody treatments or adoptive transfer or tumor specific T cells, have been clearly effective. Others, such as cancer vaccines, have been erroneously applied in the setting of advanced disease and have had only very marginal success . We will present data in support of administration of cancer vaccines either early in the disease process to prevent cancer progression, or prior to the onset of disease in individuals at high risk for cancer. We will show more specifically that immune responses generated against MUC1 and cyclin B1 vaccines are safe and effective against spontaneous tumors in transgenic animal models and discuss ongoing clinical trials in the therapeutic and the preventive settings.