A consistent two-mutation model of lung cancer for different data sets of radon-exposed rats

A two-mutation carcinogenesis model, formulated in terms of biologically motivated equations for mutation and expansion steps, has been applied in a mechanistic modelling of the lung cancer incidence in two large data sets of rats exposed to radon, both separately and jointly. Results indicate that (1) the equations employed are able to provide an accurate description of the separate data sets, (2) the parameters in the equations take on similar values for both data sets, and (3) it is possible to construct a consistent and well-fitting solution for the joint data set. It proved not to be necessary to take into account the effect of uranium ore dust, administered to part of the data or the different rat strains of the data sets. The joint solution provides a firm basis to investigate the effects of exposure, exposure rate and age at exposure on cumulative incidence, excess relative risk and excess absolute risk. For the same total exposure, cumulative incidence reaches a maximum for exposure rates between 1 and 10 WLM per day. The so-called inverse-exposure-rate effect acts for higher exposure rates. The influence of age at exposure, however, seems to be even more pronounced. Exposure at a young age leads to considerably higher incidences than exposure at a later age. Parameters derived in this study compare fairly well with those derived for uranium miners, suggesting that a consistent model description for the induction of lung cancer by radon in rats and humans may be possible.