In this paper the radiation-associated relative risks of
second primary cancer incidence in groups treated for first primary
cancer by
radiotherapy are compared with radiation-associated relative risk estimates in the Japanese atomic bomb survivor
cancer incidence data. For four
cancer sites, namely
lung cancer,
bone cancer,
ovarian cancer and leukaemia, the relative risks in the comparable (age at exposure, time since exposure, sex matched) subsets of the Japanese data are significantly greater than those in the majority of
second cancer studies. Even when the differences between the relative risks in the Japanese atomic bomb survivors and the medical series do not approach conventional levels of statistical significance, relative risks tend to be higher in the Japanese data than in the
second cancer studies. At least for leukaemia, the discrepancy between the Japanese and
second cancer risks can be largely explained by cell-sterilisation effects. There are few indications of modification of radiation-associated
second cancer relative risk among those treated with
adjuvant chemotherapy, nor are there strong indications of modification of radiation-associated relative risk by heritable genetic factors. If anything, there is evidence that
second cancer relative excess risks are lower among those patients with
cancer-prone disorders than among non-susceptible patients. However, the higher underlying
cancer risk in some of these medically exposed populations should also be considered, in particular for those with
cancer-prone conditions, so that the absolute excess risk is sometimes higher than in the Japanese data.