[Cancer between infection and heredity: genes, viruses and mice at National Cancer Institute (1937-1977)].

After World War II, in the United States,viral explanation of cancer replaced a vision of the disease emphasizing genetic factors. From the mid 1950s onwards, experimental oncologists favored the notion that cancer was initiated by infectious agents passed from one generation to the next. In order to analyze this displacement, the present paper focuses on the part played by new experimental systems, i.e. mice showing tumors induced by viruses. Since animal models are agencies which "represent" human diseases, and mediate between different social worlds, their uses often result in opposing views. Mouse models thus provided tractable resources which favored the alternation between heredity and infection. The paper describes the emergence, in the late 1930s, at the Jackson Memorial Laboratory, of an agent enhancing the formation of mammary tumors in mice. This laboratory was then involved in the production of marketable inbred mice as well as in research concerned with genetic factors that may cause cancer. After World War II, loose theories and conflicting results helped turn the agent into a virus. At the National Cancer Institute, the virus was associated with a whole range of particles causing leukemia in mice. Owing to the Virus Cancer Program, the value of mouse tumor viruses increased in the late 1960s. This research effort then aimed at finding human tumor viruses, and at crafting cancer vaccines. It was modeled after the experience of the NCI chemotherapy program stemming from war research. In addition to the fact that biomedical research became a state enterprise, the study emphasizes three parameters. First, loose practices--both theoretical and experimental--helped manage the variability of animal models. Secondly, the standardization and mass production of animals and reagents encouraged the stabilization of research programs. Thirdly, private biotechnology companies working under NCI contracts implemented preclinical work, and mediated between virology laboratories and clinical settings.