TIME VARIATIONS IN THE RISK OF RADIATION-INDUCED CANCER

The models of Armitage and Doll (ref. 1) and Moolgavkar, Venzon and Knudson (ref. 2) can be shown to predict that, asymptotically, the excess relative risk (ERR) following an instantaneous radiation dose decreases at least as fast as [time after exposure]"1 for some positive j (ref. 3). The purpose of this paper is to tie together these theoretical predictions with what is observed in various irradiated groups, and draw some conclusions for population cancer risks.

An analysis of the Japanese LSS cohort (with the DS86 dosimetry) together with information from the latest follow-ups of the New York tinea capitis children, the Oxford Childhood Cancer Research Group cohort and the Rochester thymus irradiated children finds significant (and consistent) decreases in the ERR for solid cancers among those exposed in childhood. These analyses show that the ERR for solid cancers decreases with a speed of between [time after exposure]~2•0 and [time after exposure]~3•2 (depending on the year after which this reduction might take effect) (ref. 4). The implications of reductions in ERR on calculations of population cancer risk estimates are significant; if the range of observed reductions in ERR are assumed to operate 40 or more years after exposure, UK population risk estimates would be reduced by 30%-45% compared to those based on a constant relative risk model (ref. 4).