Neoplastic transformation following split doses of X rays

Most data for radiation-induced malignancies in the human relate to a large single dose of radiation received in an acute exposure. An interpolation is required in order to arrive at cancer risk estimates for low doses or low dose-rates which are relevant to human medical or environmental exposure (Brown, 1977; Upton, 1977). Both the BEIR (BEIR, 1972) and UNSCEAR (UNSCEAR, 1977) committees have used a linear interpolation from high to low doses, which assumes that the risk per rad is independent of dose or dose-rate. Risk estimates calculated in this way are assumed to be “conservative”, since it has been reported that the biological consequences of a given dose are reduced when that dose is divided into a number of fractions and delivered over a prolonged period of time (Hornsey and Silini, 1962; Elkind and Whitmore, 1967; Belli et al., 1967). Our experiments indicate that this is not always true in the case of neoplastic transformations produced in vitro.