The influence of cell killing on radiation oncogenesis: possible implications for high-let risk assessment at medium...

As the radiation dose increases, the prevalence of radiation-induced cancer in vivo, and transformation frequency per exposed cell in vitro, reach a maximum and often decrease. The most commonly considered mechanism to account for these effects is one in which, as the dose increases, cell killing becomes an important factor and ultimately controls the rate of oncogenesis. However, this model is by no means proven. We investigate the biophysical consequences of such a mechanism and show that it leads to the testable prediction that, at high LET, the maximum incidence of oncogenesis that can be achieved (the maximum oncogenic potential) is independent of LET. Data contradicting this prediction could therefore be taken as arguing against the model; what relevant data are available do not contradict this prediction. If the model is correct (or even if it is not, but the dose-oncogenesis relation follows this trend for other reasons), then it follows that the conventional risk assessment system, in which risk increases linearly with dose, would be inappropriate at medium doses (> 0.5 Gy) of high-LET (> or = 30 ke V/microns) radiation. In this dose-LET region, a risk assessment system based on the maximum oncogenic potential might be a) more credible than a system in which risk increases linearly with dose; b) appropriately conservative; c) independent of dose rate; and d) simple and practical to use.