Classification of radiation effects for dose limitation purposes: history, current situation and future prospects

Abstract

Radiation exposure causes cancer and non-cancer health effects, each of which differs greatly in the shape of the dose-response curve, latency, persistency, recurrence, curability, fatality and impact on quality of life. In recent decades, for dose limitation purposes, the International Commission on Radiological Protection has divided such diverse effects into tissue reactions (formerly termed non-stochastic and deterministic effects) and stochastic effects. On the one hand, effective dose limits aim to reduce the risks of stochastic effects (cancer/heritable effects) and are based on the detriment-adjusted nominal risk coefficients, assuming a linear-non-threshold dose response and a dose and dose rate effectiveness factor of 2. On the other hand, equivalent dose limits aim to avoid tissue reactions (vision-impairing cataracts and cosmetically unacceptable non-cancer skin changes) and are based on a threshold dose. However, the boundary between these two categories is becoming vague. Thus, we review the changes in radiation effect classification, dose limitation concepts, and the definition of detriment and threshold. Then, the current situation is overviewed focusing on (i) stochastic effects with a threshold, (ii) tissue reactions without a threshold, (iii) target organs/tissues for circulatory disease, (iv) dose levels for limitation of cancer risks vs prevention of non-life-threatening tissue reactions vs prevention of life-threatening tissue reactions, (v) mortality or incidence of thyroid cancer, and (vi) the detriment for tissue reactions. For future discussion, one approach is suggested that classifies radiation effects according to whether effects are life threatening, and radiobiological research needs are also briefly discussed.

Keywords: detriment; life-threatening effects; non-life-threatening effects; stochastic effects; threshold; tissue reactions.

Fig 1.

A diagram schematizing the original definition, current situation and our proposal for radiation effect classification and associated radiation protection (RP) endpoints. Orange- and green-colored solid lines indicate radiation effects of which risk is managed in terms of mortality and incidence, respectively. Yellow- and blue-colored solid areas represent radiation effect categories. The yellow-colored dotted line shows radiation effects categorized in tissue reactions but with a linear-non-threshold (LNT)-like dose response, whereas the blue-colored dotted line shows radiation effects categorized in stochastic effects but with a threshold-like dose response. A purple-colored dotted line further emphasizes radiation effects that obscure the boundary between stochastic effects and tissue reactions. Radiation effects were originally classified into stochastic effects and non-stochastic or deterministic effects (presently called tissue reactions). Stochastic effects comprise fatal cancer, non-fatal cancer (e.g. thyroid cancer and non-melanoma skin cancer) and fatal hereditary effects with an LNT dose–response model, whilst a threshold-like dose–response relationship has been known for several cancers such as skin cancer and bone cancer. Non-stochastic or deterministic effects include radiation effects with a threshold-type dose–response model (e.g. vision-impairing cataracts and cosmetically unacceptable non-cancer skin changes). For RP purposes, mortality and incidence were endpoints employed to manage the risk of stochastic effects and non-stochastic or deterministic effects, respectively, whereas an ‘emergency dose level’ for the thyroid was assigned based on its cancer incidence. At present, these have not been changed basically. However, a tissue reaction category now includes life-threatening circulatory disease, though its dose limit has not been set. Moreover, for cataracts and circulatory disease, epidemiological evidence now suggests an LNT dose response, and the RP system assumes no dose-rate effect, making the boundary between stochastic effects and tissue reactions fuzzy. Such a situation may be improved if radiation effects are classified into life-threatening disease and non-life-threatening disease, for which the risk management endpoints are mortality and incidence, respectively. Taken together, the necessity to limit heritable effects should also be reconsidered because human evidence has thus far shown no detectable transgenerational effects. This approach may be useful for future discussion.