Immunological control of breast cancer: discussion.

The inclusion of breast cancer in this symposium on virus-associated tumors in man presumes that human breast cancer is induced by a virus or is at least strongly associated with a virus. At the moment, there appear to be 2 lines of evidence for this. The first is simply an analogy to the extensively studied mouse mammary tumor system, where a virus has been directly demonstrated. Much of the data presented at this meeting has related to this model system, with the impl ic i t expectation that some or all of the information will be relevant to human breast cancer. The other line of evidence is quite recent and involves the accumulating data indicating biochemical and immunological cross-reactions of human breast cancer with MTV 2 or with Mason-Pfizer monkey virus, and the presence of B-type particles in human milk and at least 1 tissue culture line derived from human breast cancer. There are several points that came up dur ing this meeting that should be emphasized, particularly with regard to the possible impl icat ions for human disease. In the mouse tumor systems, several laboratories are f inding some relationship between tumor-associated antigens and MTV-associated antigens (3, 12). Such observations need to be explored further. For example, what is the relationship between ML antigen and gp52 of MTV. Further, is the serologically defined ML antigen the same as the detected by assays of cell-mediated immuni ty? These questions should be answered by further puri f icat ion of the antigens and studies of cross-reactivity. It wil l be of particular interest to determine whether the gp52 is the major transplantat ion antigen associated with MTV, i.e., whether immunizat ion with this antigen can protect mice against mammary tumor growth (12). During the course of the discussion, it was pointed out that some of the extracts now used, e.g . , 3 M KCI extracts of tumors, are quite crude and can contain a variety of proteins and antigens, some of which may not even be truly soluble (12, 13). Hosts respond differently to viral antigens depending on whether, when, and how they have been exposed to the virus (3, 12). Different MTV's (at least with different pathogenic spectra) may have antigenic differences. The dynamic state and rapid turnover of some of these antigens have been emphasized (13), and the actual amount of viral-related antigens in different mouse mammary tumors has been shown to be variable and having an inverse relation-