available and relevant data . The US National Council for Radiation Protection and Measurements ( NCRP ) reviewed all data including the most recent finding for the Japanese A-bomb survivors , and studies of workers and patients exposed to radiation . The NCRP verdict was that the LNT model is appropriate for assessing cancer risk for RP purposes . Other recent studies on thyroid cancer following low dose exposures in childhood and on UK / US / France nuclear workers showed significant risks following low-level exposures . The highest risks have been shown to be for exposures in utero and during early childhood , with significant excess risks from radiation at doses down to a few 10s of mSv . However , it is unlikely that epidemiological studies will ever discern statistically significant cancer risks below around 10 mSv – for lower doses we must turn to the biological evidence .

Biological evidence A recent review by the United Nations Scientific Committee on the Effects of Atomic Radiation ( UNSCEAR ) concluded that there is good justification for the use of a nonthreshold model for risk inference for RP purposes , given the present robust knowledge on the role of mutation and chromosomal aberrations in carcinogenesis . That is , ionising radiation can cause DNA damage in single cells that result in stable chromosomal aberrations and mutations which lead to proliferative advantage , clonal expansion , and the possibility of cancer following further mutations . On this basis , even the smallest doses of radiation have a chance of causing such damage , however small that probability . UNSCEAR further concluded that , despite decades of study , any implications from studies of other cellular responses are still not clear .

Conclusions Ionising radiation is a carcinogen for which there is a substantial body of epidemiological evidence and mechanistic understanding from biological studies . Based on current information , the use of a linear nonthreshold dose-response model for cancer causation is the most realistic approach for

Radiation Protection Today Winter 2022

RP purposes and should not be thought of as overly conservative . However , acceptance of the probability of small risks at low doses should not mean that exposures should be minimised . Increased cancer risks at doses of 1 – 10 mSv and below will be a very small proportion of background risks and , depending upon the circumstances of exposure , may be considered negligible . ALARA should be applied with due consideration of what is reasonable .

What is the Linear Non-Threshold ( LNT ) model for radiation-induced stochastic effects ? A number of dose effect curves have been postulated for cancer ( see below ) which include a supralinear model in which the effect per Gy is greater at low doses than high doses , a threshold model in which there are no effects at low doses , and a hormesis model in which low doses are beneficial . The straight-line LNT model is the simplest relationship between excess stochastic effects and radiation dose and its use for RP purposes is supported by cancer epidemiology and biology ( see text ).

Excess Cancer

0

0

Supralinear

Threshold

LNT

Dose

Hormesis

A complete version of this article , including full references , is available via the digital version of Radiation Protection Today ( www . srp-uk . rpt )

33