Comparison of radiation-related cancer risks from the atomic bomb survivors with studies of pediatric medical radiation exposure.

Medical exposures are the largest source of human-made ionizing radiation received by the general population. Cancer risk assessment for medical radiation is often based on the Life Span Study (LSS) of the Japanese atomic bomb survivors. Various assumptions are required to transfer radiation-related cancer risk estimates from the LSS to medical radiation exposures, including dose and fractionation effects. We compared organ-specific cancer risk coefficients from pediatric medical radiation studies (age, <22 years) with the LSS, controlling for average age at exposure, attained age, or time since exposure. We compared 21 studies (including 5 pooled analyses) of brain, breast, thyroid cancer, and leukemia (the most radiosensitive cancers in children), including 6 low-dose (mean < 100 mGy), 7 moderate-dose (mean = 100 mGy to -<2 Gy) and 8 high-dose (mean = 2+ Gy). The high-dose studies all had lower dose-response estimates than the LSS (ratio range = 1.3-37), as did most of the moderate-dose studies (ratio range = 0.4-9.4). In contrast, the dose-response estimates for the low-dose studies were all higher than the LSS (ratio range = 0.1-0.7). These results do not provide strong support for a dose reduction effectiveness factor of 2 for risk assessment for low-dose medical radiation exposures using the LSS. Whilst there is a clear reduction in risk from high-dose fractionated exposures compared with the LSS, the wide variation in ratios makes it difficult currently to quantify these effects.