tional to D^2 (quadratic) at higher doses. Yet, other experts believe that there
is no risk of carcinogenesis up to a certain threshold dose, after which the
curve becomes linear or quadratic (Fig. 115.15). While the linear response
model is preferred for all solid tumors, the linear-quadratic model is more
suitable for leukemia.
Risk Estimates of Excess Cancer
The BETR VII Committee (2005) estimated the excess cancer risk based
on the LNT theory and by extrapolation of high-dose incidences to the low-
dose situations. It is estimated that 1 in 100 individuals is expected to
develop solid tumor or leukemia from a radiation dose of 10 rem (0.1 Sv),
whereas approximately 42 of these 100 persons are expected to develop
these cancers from other causes. Lower doses would produce proportion-
ally lower risks. So, approximately 1 in 1000 individuals would expect to
develop cancer from an exposure of 1 rem (0.01 Sv). Another estimate of
cancer risk by BETR VII is that 1 in 100 persons would likely develop
cancer from a lifetime (70 yr) exposure to low-LET natural background
radiations excluding radon and high-LET radiations. For comparison, the
ICRP estimate of cancer risk for the general population is 0.1 per Sv (0.1%
per rem) for high doses and dose rates, and 0.05 Sv (0.05% per rem) for low
doses and dose rates (ICRP 60, 1991).
Leukemia
Leukemia is one of the most common cancers induced by radiation in
humans, accounting for one in five mortalities from radiocarcinogenesis.
Risk of leukemia varies with age, with younger persons being more prone
252 15. Radiation Biology
Fig. 15.15. Three general shapes of the dose–response curves permit prediction of
different incidences of low-dose radiation effects when the curves are fitted to high-
dose data. (Adapted from Murphy PH. Acceptable risk as a basis for regulation.
Radiographics1991; 11:889–897.)