. "3 Radiation-Induced Cancer: Mechanisms, Quantitative Experimental Studies and the Role of Genetic Factors." Health Risks from Exposure to Low Levels of Ionizing Radiation: BEIR VII Phase 2. Washington, DC: The National Academies Press, 2006.
The following HTML text is provided to enhance online
readability. Many aspects of typography translate only awkwardly to HTML.
Please use the page image
as the authoritative form to ensure accuracy.
Health Risks from Exposure to Low Levels of Ionizing Radiation: Beir VII Phase 2
By dividing the tumorigenic effectiveness per unit dose of acute exposures using the high-dose data and the low-dose-rate exposures, effectiveness ratios were obtained. These ratios have been termed dose and dose-rate effectiveness factors. Since the data from which these ratios are obtained result from comparing high- and low-dose-rate effects, these ratios are literally dose-rate effectiveness factors (DREF). However, since the actual dose-response for most radiation-induced tumors following single acute exposures was found to be linear quadratic, it can be seen from Figure 10-1 that this procedure would tend to overestimate effects for low single acute radiation doses (in the dose range where the response is predominantly linear) as well as for low-dose-rate exposures over a wide range of total doses. Since the ratio should be equally valid for estimating effects at low dose rates (the DREF) and for low single doses, the term dose and dose rate effectiveness factor (DDREF) has commonly been used. This would not be the case if the dose-response following acute exposures is not linear quadratic.
The derivation and application of DDREF must be performed with caution. Tumors for which there is mechanistic knowledge that they are unlikely to be applicable to radiation carcinogenesis in human populations should not be considered. On this basis, quantitative data on dose-rate effects for thymic lymphomas and for ovarian tumors, which have been shown to be highly sensitive to dose-rate effects, should not be used. Likewise, caution should also be exercised when considering data for the induction of pituitary tumors in RFM female mice because of potential effects associated with the sensitivity of the mouse ovary and the subsequent disruption of pituitary and ovarian hormone functions. This leaves a limited data set upon which to base DDREF calculations, which includes data for myeloid leukemia and a few solid tumors including Harderian gland (for which there is no comparable tissue in humans), lung adenocarcinomas, and mammary tumors. Data for myeloid leukemia are available for two mouse strains and from at least three independent studies. All of the data support a reduced effect when comparing high- and low-dose-rate exposures over the 0–3 Gy dose range. Calculation of DDREF values using the procedures described above yields estimates on the order of 2 to 6, with most values in the range of 4–5. For lung adenocarcinomas and Harderian gland tumors, DDREF values of approximately 3 have been calculated over the 0–2 Gy dose range. For mammary tumors, all of the data suggest a DDREF value of less than 2 and closer to a value of 1 when effects of high-dose-rate and low-dose-rate exposures are compared in this 0–2 Gy dose range. Thus, it appears that myeloid leukemia is probably more sensitive to dose-rate effects than are solid tumors.
It should also be pointed out that these values are based on extrapolation of data from acute doses of 2–3 Gy and that extrapolating data from lower doses would result in lower estimates. The impact of dose range must be considered when applying DDREF factors to human risk estimates for which there are good data at and below 1 Gy. Chapters 10 and 12 describe the use of animal data in developing a specific judgment on the value of DDREF to be used in BEIR VII cancer risk estimates.
Human and animal data relating to adaptive responses to radiation and cancer risk have been reviewed by UNSCEAR (1994). That review concluded that the presence of an adaptive response for cancer risk was not readily evident from the results of animal studies and that, for reasons of statistical power, no clear statements were possible from epidemiologic investigations. Since 1994 a number of further animal studies have reported evidence suggestive of some form of adaptive response in the development of certain tumors.
Ishii and colleagues (1996) reported a decreased incidence of thymic lymphoma in AKR mice following chronic fractionated low doses of X-rays. As described in this chapter, the atypical involvement of cell killing in the etiology of murine thymic lymphoma makes interpretation of all data for this tumor type most difficult. On this basis, no great weight can be placed on the data of Ishii and others (1996). Of potentially greater relevance are the adaptive response data on the induction of AML in CBA mice and the development of osteosarcoma or lymphoma in Trp 53-deficient mice.
In studies with CBA mice (Mitchel and others 1999), prior exposure to low-dose-rate radiation was shown to change the tumorigenic response of animals receiving a second dose at a higher dose rate delivered one day later. Somewhat surprisingly, the principal effect of the priming dose was not to reduce the lifetime risk of AML but rather to increase tumor latency. Similar delaying effects on tumor latency but not lifetime risk of a low (10 mGy) acute priming dose were subsequently reported for spontaneous development of osteosarcoma and lymphoma in Trp 53 heterozygotes. The effects of a 100 mGy priming dose differed for osteosarcoma (decreased latency) and lymphoma (increased latency), a result that is suggestive of a mechanism that is dependent on dose and tumor type. These studies are difficult to interpret, particularly since the priming dose appears to influence tumor development rather than initiation.
This result runs counter to expectations from cellular data on adaptive responses (see Chapter 2), which emphasize the potential importance of adaptive DNA damage response processes. To explain the apparent effects of a priming dose on tumor latency it would be necessary to postulate the existence of low-dose-induced physiological signals that have a lifetime of many months. Mitchel and others (2003) suggest that these signals might act via the inhibition of genomic instability, which would then tend to slow tumor development. However given the great uncertainties on the in vivo activity of radiation-associated genomic instability already noted in this chapter, the adaptive mechanism suggested by Mitchel and others (2003) is regarded as being highly speculative.