Time sequence of cancer occurrence. Implications in low level radiation risk assessment

Abstract

A total of 8 229 C57 Black mice of both sexes were randomly assigned to various groups. In some groups, mice aged 33 ± 3 days were submitted to either sham, neutron or cobalt external radiation at 32, 45, 63, 88 and 123 mGy or at 18, 25, 36 and 51 cGy dose levels, respectively. In other groups, mice either at birth or weaning, were injected with tritiated thymidine or tritiated water, or were given tritiated water as drinking water for the entire lifespan. The main purpose of the experiment was to investigate the low dose-response relationship of cancer induction, especially leukemogenesis and to evaluate the excess risk, using actuarial age-specific rates. Following neutron or cobalt exposure, the phenotypic occurrence of lymphocytic lymphomas was earlier in appearance and higher in yield during the first decades of lifespan in irradiated groups versus matching controls, whereas such occurrence was markedly lower in yield at a later age. Under parallel experimental conditions, induction of reticulum cell lymphomas, however, was uniformly enhanced throughout the entire lifespan. Induction rates of all tumors (reticular and solid) pooled were significantly increased, and more so following cobalt than neutron irradiation. In mice injected with tritiated thymidine, the overall tumor incidence was increased monotonically throughout the lifespan. In mice exposed to tritiated water, the incidence of lymphocytic lymphomas was markedly increased throughout the lifespan, whereas no such effect was observed for reticulum cell tumors. In the light of tumor data analysis, it appears that selection of a particular type of tumor as a dependent variable for doseresponse assessment cannot disregard the primary modulation of the whole tumor spectrum. In C57 Black murine leukemogenesis, the shape and structure of the doseresponse regression curve over the entire lifespan does not fit the linear-quadratic model. It is, however, theorized that our data are consistent with the two-mutation clonal expansion model, assuming creation of initiated cells which are added to the pool of spontaneously occurring initiated cells and implying that the excess risk is initially high at early age and then decreases with increasing age following exposure. It is concluded that murine radiocarcinogenesis investigation may contribute to improving the assessment as well as understanding the underlying mechanisms of low level radiation hazards.