Risk of solid cancer in low dose-rate radiation epidemiological studies and the dose-rate effectiveness factor

Abstract

Purpose: Estimated radiation risks used for radiation protection purposes have been based primarily on the Life Span Study (LSS) of atomic bomb survivors who received brief exposures at high dose rates, many with high doses. Information is needed regarding radiation risks from low dose-rate (LDR) exposures to low linear-energy-transfer (low-LET) radiation. We conducted a meta-analysis of LDR epidemiologic studies that provide dose-response estimates of total solid cancer risk in adulthood in comparison to corresponding LSS risks, in order to estimate a dose rate effectiveness factor (DREF).

Materials and methods: We identified 22 LDR studies with dose-response risk estimates for solid cancer after minimizing information overlap. For each study, a parallel risk estimate was derived from the LSS risk model using matching values for sex, mean ages at first exposure and attained age, targeted cancer types, and accounting for type of dosimetric assessment. For each LDR study, a ratio of the excess relative risk per Gy (ERR Gy^?1) to the matching LSS ERR risk estimate (LDR/LSS) was calculated, and a meta-analysis of the risk ratios was conducted. The reciprocal of the resultant risk ratio provided an estimate of the DREF.

Results: The meta-analysis showed a LDR/LSS risk ratio of 0.36 (95% confidence interval [CI] 0.14, 0.57) for the 19 studies of solid cancer mortality and 0.33 (95% CI 0.13, 0.54) when three cohorts with only incidence data also were added, implying a DREF with values around 3, but statistically compatible with 2. However, the analyses were highly dominated by the Mayak worker study. When the Mayak study was excluded the LDR/LSS risk ratios increased: 1.12 (95% CI 0.40, 1.84) for mortality and 0.54 (95% CI 0.09, 0.99) for mortality?+?incidence, implying a lower DREF in the range of 1–2. Meta-analyses that included only cohorts in which the mean dose was <100 mGy yielded a risk ratio of 1.06 (95% CI 0.30, 1.83) for solid cancer mortality and 0.58 (95% CI 0.10, 1.06) for mortality?+?incidence data.

Conclusions: The interpretation of a best estimate for a value of the DREF depends on the appropriateness of including the Mayak study. This study indicates a range of uncertainty in the value of DREF between 1 and about 2 after protracted radiation exposure. The LDR data provide direct evidence regarding risk from exposures at low dose rates as an important complement to the LSS risk estimates used for radiation protection purposes.