Perspectives in Cancer Research The Autoimmune Nature of Cancer1

Four different kinds of data from the 3-methylcholanthrene-induced sarcoma system of the mouse show that the immune system stimulates oncogenesis; i.e., the presence of a tumor-specific immune reaction is a positive aid to tumor development. It seems proper, therefore, to consider cancer, at least in part, an autoimmune disease. The purpose of this paper is to review recent work supporting the thesis that tumor-specific immunity in the MCAMnduced sarcoma system of the mouse usually facilitates rather than inhibits oncogenesis. In particular, four quite different types of evidence show that an intermediate level of antitumor immune reaction is more conducive to oncogenesis than is either a greater or a lesser level. Inasmuch as the level of immune reaction is determined by the interaction of tumor antigens with the immune system of the host, tumors of little ant igen ¡city, as are often produced by very low concentrations of an oncogenic chemical, tend to be dependent upon an unusually responsive host immune mechanism to facilitate their growth. Conversely, the highly antigenic tumors that are often produced by a strong oncogenic stimulus may be facilitated by measures that reduce the level of (but do not abolish) the immune reaction. This thesis has been supported by the following four disparate kinds of data: (a) data from the titration, by various means, of the immune reaction, both in vivo and in vitro; (b) data derived from the peculiar relationship between the immunogenicities of tumors and the latent periods of their inductions in their primary hosts; (c) data concerning the reversal of the rank order of strain susceptibilities to oncogenesis when the dosage of MCA is drastically altered; and (d) data describing the effect of 3-day thymectomy, which produces autoimmune hyperplasias, on oncogenesis with high or low dosages of MCA. Each of these kinds of data is independently supportive of the thesis; together they seem to us to be virtually conclusive. Titration of the Immune Response The first of the titration experiments used "Winn" tests in which tumor cells and spleen cells from specifically immune animals were mixed in varying proportions and the mixtures were implanted s.c. in syngeneic mice. Maximal tumor growth occurred with about a 1/1 ratio and growth at this ratio was significantly better than that with normal spleen cells as con trols. However, at high spleen cell/tumor cell ratios, the im mune spleen cells were inhibitory relative to the effect of normal spleen cells (which were themselves stimulatory to tumor growth) (1). Quite analogous results have been obtained in tests done completely in vitro. Some of the experiments showed the tumor facilitation to be specific and others showed it to be nonspecific (2); it seems that there may be both specific and nonspecific Received 12/6/84; revised 10/17/86; accepted 12/3/86. 1Supported by Grant CA 31837 from the NIH. 'Present address: Department of Pathology, University of Washington, Seattle, WA 98195. 3The abbreviation used is: MCA, 3-methylcholanthrene. components. Judging by these results, it is possible that any immune effector, i.e., macrophage, T-cell, natural killer cell, or pure antibody and complement, may be stimulatory in the proper proportions; the literature supportive of this judgment is suggestive, but far from conclusive (3). It is clear from the work with antibody that one and the same molecular species can either facilitate or inhibit tumor growth depending only upon the concentration (4). Although some species of splenic effector cells may be more stimulatory or inhibitory to tumor growth than are others (5, 6), the observations with antibody do not encourage the belief that the immune reaction is divisible into two parts, one facilitating and one inhibiting. Results very similar to those already described were also obtained in oncogensis experiments. For example, mice were thymectomized as adults and then subjected to 450 R of Xirradiation to reduce their immune capacities to minimal levels. The animals were then immunologically restored to varying degrees by inoculating them i.p. with graded numbers of normal syngeneic spleen cells; they were then exposed to various amounts of MCA. It was found that, at larger dosages of the carcinogen, the most rapid sarcogenesis occurred in animals partially restored with small numbers of cells. Oncogenesis in the presence of a small degree of restoration was significantly faster than that in unrestored animals or in animals restored with larger numbers of cells (7). Similar results were obtained in mice in which the presumptive titration of the immune mechanism was accomplished directly with varying amounts of X-irradiation rather than with spleen cell restoration (8) or in which the immunological defect in nude mice was rectified to various extents by the inoculation of various numbers of thymus or spleen cells (9). In the nude mouse experiments, the induced tumors were skin neoplasms, rather than s.c. sarcomas, showing that the phenomenon was not limited to one histolÃ3gica! type of cancer. There is evidence that mouse mammary tumors may also be subject to immunological facilitation (10). Relationship of Latency to Immunogenicity The immunological facilitation of oncogenesis has been shown by a new consideration of the relationship between tumor immunogenicities (as measured by classical in vivo tests com paring the growths of tumor implants in immunized versus nonimmunized syngeneic mice) and the latencies of the pri mary, untransplanted tumors (11). This long known relation ship is mediated by host immunity (12, 13). Although tumors of higher immunogenicity tend to have had shorter latencies, it has recently been found, among 154 different MCA-induced sarcomas, that tumors with the very shortest latencies were usually of an intermediate level of immunogenicity with rela tively little scatter (11). Why should tumors with the very shortest original latencies tend to have intermediate levels of immunogenicity? The usual explanation of the relationship between latency and immuno genicity is that the immune capacities of the mice change with time after MCA administration; the MCA is a known ¡numi nosuppressant, but one that takes time to act (14). Tumors that