Tumour-specific antigens: their possible significance in the etiology and treatment of malignant disease.

In the past decade appreciable evidence has accumulated from clinical observation to support the concept that host factors may influence the development and progression of lluman cancer. While some tumours grow without apparent restraint, others such as carcinoma of the breast and prostate may be inhibited by appropriate endocrine modification (Huggins 1967). Further evidence suggestive of the importance of host factors in human neoplasia derives from the following phenomena: the spontaneous regression oftumours (Everson 1964): the regression of metastases following treatment of the primary tumour and conversely by the sudden appearance of tumours after long latent periods, sometimes coinciding with immunosuppressive therapy (Woodruff 1964, Wissler 1964); the failure of circulating tumour cells and autotransplants to form metastases (Grace 1964, Nadier and Moore 1965, Southam I 964, 1965) : the favourable response of certain tumours (for example, placental choriocarcinoma and Burkitt lymphoma) to chemotherapy (Math#{233}, Dausset, Hervet, Amiel, Colombani and Brule 1964; Fass, Herberman and Ziegler 1970); the histological pattern of reaction of lymph nodes draining the tumour (Black and Speer 1958, Hamlin 1968); and the association of some tumours with abnormal immunity mechanisms (Fairley 1969). Presently available evidence implicates the immune response as a regulatory factor in these types of neoplasms. This response is effected by lymphoid cells in two ways: indirectly by the production of classical antibody (humoral immunity), or directly without antibody as intermediary (cell-mediated immunity); and its induction depends on the presence of distinctive tumour antigens not present in, or foreign to, normal host tissues. These antigens are the logical target for immunotherapeutic attack. The ultimate aim of research in tumour immunology is thus the development of clinical methods for immunisation. It is already apparent, however, that the study of the immunological characteristics of animal and human tumours will also contribute substantially to knowledge of the causation of malignant diseaseparticularly with respect to the role of viruses-and will determine the validity of the concept that cell-mediated immunity provides a natural anti-cancer surveillance system. In this paper the evidence for the existence of tumour-specific antigens in animal and human tumours is discussed; in addition, possible etiological and therapeutic implications arising from experimental studies are reviewed with particular reference to osteosarcoma.