Supplementary material
The BEIR VII [21] report presents methods to estimate the lifetime attributable risk (LAR) of cancer incidence as well as cancer mortality, using excess relative risk (ERR) models and excess absolute risk (EAR) models. EAR models describe the excess risk as the difference in the total risk and the background risk, i.e. EAR is the rate of disease in an exposed population λE minus the rate in an unexposed population, λU.
EAR=λE-λU (1)
ERR models express the excess risk relative to the background risk. The concept of ERR is based on the relative risk (RR), which is the rate of disease in an exposed population divided by the rate in an unexposed population. ERR equals the relative risk minus one.
RR=λEλU (2)
ERR=RR-1 (3)
The potential increase in cancer risk attributable to exposure of ionizing radiation is often estimated using risk projection models derived from the Life Span Study (LSS). LSS provides data on about 120 000 survivors of the atomic bombs in Japan, has a long follow-up period, and includes both sexes and all ages at exposure. Despite uncertainties in the estimated doses to the exposed population in Nagasaki and Hiroshima, data from atomic bomb survivors provide evidence of statistically significant increase of cancer incidence from whole-body exposure of 100 mSv or more and for most cancer types the dose-response is best described as linear [21]. However, there is no general consensus on how to estimate the risks associated with low doses (<100 mSv) such as those of single PET/CT scans, mostly due to the lack of consensus regarding the dose-risk relationship in this low-dose range (see footnote). All analyses of populations exposed to low doses of radiation suffer from large uncertainties such as statistical variability, limitations of sample size and impact of confounding factors.
The BEIR VII risk models have been generated using linear extrapolation of atomic bomb survivor data in the higher dose-range to assess the risk of solid cancer induction from low-dose exposures. Subsequently, a mathematical model has been chosen to describe the risk function: for solid cancer induction, a combined exponential and power function depending on dose D, attained age a, age at exposure e and gender s was used. The function has three unknown parameters (βS, γ and η) which are estimated by fitting the model to the extrapolated LSS-data and provided in Table 12-2 in [21].
ERRD,e,a=βS∙D∙expγ∙e⋆∙a60η (4)
EARD,e,a=βS∙D∙expγ∙e⋆∙a60η (5)
In BEIR VI, ERR and EAR only depend on age at exposure for exposure ages under 30, which is accounted for by using e⋆=e-3010 in the calculations.
As the models are based on data from the Japanese population, with very different baseline risks for most cancer diagnoses compared to the populations in Europe or USA, the calculated risk must be adjusted to better describe the population of interest. In relative risk models, it is assumed that the excess risk is proportional to the baseline risk so that the ERR is same for all populations. In absolute risk models, the excess risk is independent of the baseline risk, which means that the EAR is the same. For most organs, the committee behind BEIR VII proposes a weighted mean, with weight of 0.7 and 0.3 for the estimate obtained using ERR and EAR, respectively. Finally, the weighted mean is divided by a dose and dose rate effectiveness factor (DDREF) of 1.5. For more information about the risk estimations, see [21].
BEIR VII assumes a latency of 5 years for solid cancer induction. In our models, we assume a maximum attained age of 100 years old. Thus, the LAR of cancer incidence due to exposing individuals of gender s, at age e, to a certain dose D, is calculated by equation 6 below (which is equivalent to equation 7):
LARD,e,a=0.7e+5100ERRD,e,a∙λIna∙Sa,e+0.3e+5100EARD,e,a∙λIna∙Sa,eDDREF
(6)
LARD,e,a=0.7e+5100LARERR_per_yearD,e,a+0.3e+5100LAREAR_per_yearD,e,aDDREF
(7)
References to sex-specific incidence rate (λIn) and sex-specific conditional survival of cancer patients Sa,e are given in the manuscript text.
Footnote
Some claim that there are no risks associated with low doses of radiation, that the LNT-model is inappropriate and directly misleading in the low-dose range and that there may be beneficial effects (i.e. hormesis) from low doses of ionizing radiation [a, b]. Others argue that there is direct evidence from epidemiologic studies that exposure from 2 or 3 CT scans are non-negligible and enough to render an increased risk of cancer, especially for children [c, d, e]. Data on biologic mechanisms have been interpreted to give no general support for the idea of neither a low-dose threshold nor hormesis [f], and two major expert panel reviews have presented their support of the LNT-model [g, h]. As stated by Brenner et al, it is likely that the LNT model underestimates the risks for some radiation-induced cancers and overestimates the risks for others [i].
[a] Tubiana M, Feinendegen LE, Yang C et al. The linear no-threshold relationship is inconsistent with radiation bilogic and experimental data. Radiology. 2009, 251, 13-22.
[b] Hooker AM, Bhat M, Day TK et al. The linear no-threshold model does not hold for low-dose ionizing radiation. Radiation Research. 2004, No 162, 447-452.
[c] Brenner DJ, Elliston CD, Hall EJ and Berdon WE. Estimated Risks of Radiation-Induced Fatal Cancer from Pediatric CT. American Journal of Roentgenology. 2001, No 176, 289-296.
[d] Berrington de González A, Mahesh M, Kim KP et al. Projected Cancer Risks from Computed Tomographic Scans Performed in the United States in 2007. Archives of Internal Medicine. 2009, Vol. 169, No 22, 2071-2077.
[e] Prasad KN, Cole WC and Hasse GM. Health risks of low dose ionizing radiation in humans: A review. Experimental Biology and Medicine. 2004, 229, 378-382.
[f] Little MP, Wakeford R, Tawn EJ et al. Risks associated with low doses and low dose rates of ionizing radiation: Why linearity may be (almost) the best we can do. Radiology. 2009, Vol. 251, No 1, 6-12.
[g] United Nations Scientific Committee on the Effects of Atomic Radiation. UNSCEAR 2000 Report Vol. 1: Sources and Effects of Ionizing Radiation. Annex D: Medical Radiation Exposures.
[h] The International Commission on Radiological Protection. The 2005 recommendations of the International Commission on Radiological Protection. 2005.
[i] Brenner DJ, Doll R, Goodhead DT et al. Cancer risks attributable to low doses of ionizing radiation: Assessing what we really know. Proceedings of the National Academy of Sciences of the United States of America (PNAS). 2003, Vol- 100, No 24, 13761-13766.