While there is no question that substantial doses of radiation provide a considerable risk of car- cinogenesis to the human being, estimating and predicting the risk of low doses of radiation is extremely difficult. It is made more difficult by the fact that there is no single unifying model that describes the dose-response effect when dealing with doses less than 100 rad. Moreover, because of the large number of subjects and the long time elapsing between irradiation and the appearance of cancer (for instance), studies on risk estimates are usually complicated by uncer- tainties about the size of the irradiation dose to affected individuals and the possible influence of factors other than radiation. For these reasons, estimates of risk are based on extrapolations from assumptions about dose-response relationships.
Some of the most extensive conclusions about the relative risk of ionizing radiation were developed from studies of survivors of the atomic bomb blasts at Hiroshima and Nagasaki. Fig- ure 12.4 shows the relative risk for cancer at one grade of exposure for survivors from 1950 through 1985 (Shimizu et al., 1990). The overall incidence of cancer in Hiroshima and Nagasaki between 1958 and 1987 increased more than 40% in males and some 15% in females (Goodman et al., 1994). The rate of liver cancer has increased dramatically among males in these cohorts during the past 20 years, with a twofold increase in incidence in the past 10 years alone. How-
Figure 12.4 Relative risk for development of individual types of cancer at one Gy exposure with 90% confidence intervals for survivors of the atom bomb blasts at Hiroshima and Nagasaki from the years 1950 through 1985. (Adapted from Shimizu et al., 1990, with permission of the authors and publisher.)
ever, this finding undoubtedly involves a number of risk factors other than the radiation expo- sure. Interestingly, in children of atomic bomb survivors who were less than 20 years old between 1946 and 1982, the risk of cancer did not increase significantly as the parental gonadal dose increased (Yoshimoto, 1990).
The absolute risk from exposure to ionizing radiation for the development of neoplastic disease has always been considered as linear, with the absence of any threshold effect (Figure
12.3; Chapter 3). Recently, this concept was given substantial experimental reinforcement by the report by Hei et al. (1997) that the traversal through a cell by a single α particle has a low prob- ability of being lethal to the cell, with over 80% survival following such an exposure, but the frequency of gene mutations is enhanced more than twofold over the background in these sur- viving cells. This clearly indicates that a single photon or particle can have mutational and po- tentially carcinogenic effects on a single cell (Hei et al., 1997). While it appears that almost all models of radiobiological cellular reactions conform approximately to a linear-quadratic dose- response relation at low, protracted doses (Sachs et al., 1997), an analysis of the dose-response curves for rates of several different types of cancer, especially leukemia, in survivors of the atom bomb blast indicated a U-shaped dose-response curve. Such a curve for leukemia deaths is seen in Figure 12.5 and follows the findings by Kondo (1990). While several reports have questioned these findings and conclusions (Little and Muirhead, 1997; Pierce and Vaeth, 1991), this horme- sis (biphasic dose-response relationship) effect has been reported for some chemical carcinogens (Teeguarden et al., 1998; Parsons, 1994; Rossi, 1981). No obvious mechanism for such a “pro- tective” effect of low doses of ionizing radiation is obvious; however, Makinodan and James (1990) have found that low doses of ionizing radiation potentiate responses in the immune sys- tem, which could in part explain this apparent protective effect of carcinogens.
The federal government and the National Academy of Sciences of the United States have been extremely concerned about the estimation of risks to ionizing radiation. Several reports of the National Research Council’s Committee on the Biological Effects of Ionizing Radiations (BEIR) have been concerned with this subject in particular. The latest BEIR V report has validated previous estimates of radiation effects and risks with some claim that ionizing radiation is more hazardous than had been previously assumed. Despite the extensive efforts of many, the BEIR V report “illustrates that the data currently available about radiation effects are still too meager to define quantitative risk estimates at low exposures with absolute certainty” (Hendee, 1992).
Figure 12.5 Dose-response relation of leukemia deaths among atom bomb survivors. (Adapted from Hattori, 1994, with permission of the author and publisher.)