Accounting for Neutron Doses Received by the Japanese Atomic-bomb Survivors

*Harry M Cullings, Radiation Effects Research Foundation 
Albrecht Kellerer, Ludwig-Maximilians University 
Donald A PIerce, Oregon Health Sciences University 

Keywords: neutrons, RBE, risk estimation

In studies of the Japanese atomic-bomb survivors, accounting for doses received from neutrons has long been a concern. Doses Dn due to neutrons were << doses Dg from gamma-rays: for dose to colon, Dn/Dg was < 0.015 at the most proximal survivor distances in Hiroshima (< 0.005 in Nagasaki), declining rapidly with distance. However, account must be taken of neutrons’ considerably larger relative biological effectiveness (R) vs. gamma rays in producing adverse health effects such as cancer and leukemia. The Radiation Effects Research Foundation (RERF) has generally used a constant R of 10 for all neutron doses. In contrast, considerations from radiation biology suggest that R depends on both Dn and Dg due to substantial curvature in the dose-response for gamma rays. We show that a proper formulation of R(Dn,Dg), for a mixed field in which Dn << Dg, can be well-approximated by a simple decreasing function R(Dg) of Dg alone, with two parameters, one governing how variable is R(Dg) and the other being its value at some reference Dg such as 1 Gy. Furthermore, R(Dn,Dg) ~ R(Dg) for such a mixed exposure declines more rapidly with Dg than the R(Dn) that would apply for the neutron exposures in the absence of the gamma-rays. We show the quantitative relationships involved and provide examples of risk estimates from RERF’s Life Span Study cohort, under plausible assumed parameters for R(Dg), such as Rmax = 90 at Dg ~ 0 and R1 = 10 at Dg = 1 gray, or Rmax = 90 and R1 = 20, contrasting these to a constant R = 10. Because Dn decreases more rapidly with distance than Dg, the largest values R apply to vanishingly small Dn, and the effects of the different assumed R functions on Deq = RDn + Dg are not critical.