RBE of neutrons in cancer mortality among atomic bomb survivors Hiroshima and Nagasaki, 1950-78

A study of the relative biological effectiveness (RBE) of neutrons in cancer mortality in the period 1950-78 in a fixed cohort of 109,000 individuals is reported. The RBE of neutrons has been estimated by fitting several models to the dose-response curves in Hiroshima and Nagasaki for leukemia, lung, breast, and stomach cancers where the dose-response relationship has been confirmed and for cancers of all sites other than leukemia. A simple additive model was used because two other models where mortality is assumed to level off in the high dose group due to cell killing fit closely resembling the simple additive model. Thus, primarily, the L-L model where mortality is assumed to increase linearly with gamma and neutron doses and the Q-L model where mortality increases proportionally to the square of the gamma dose and linearly with the neutron dose were used. Since the observations on cancers of all sites fit both the L-L and Q-L models, the RBE was estimated for both. The RBE based on the kerma dose (and tissue dose) is leukemia 7.3 (14.7), stomach cancer 3.0 (7.4), lung cancer 2.2 (4.7), and breast cancer 4.6 (6.5) using the L-L model. Under the Q-L model, the RBE is dose dependent and decreases with increasing dose. At 100 rad it is 5.1 (5.5), 3.1 (3.4), 8.3 (9.2), and 11.6 (12.7), respectively. The L-Q-L model where mortality increases with the gamma dose and its square, and linearly with the neutron dose, fits better than either the L-L model or the Q-L model for the data on cancer other than leukemia. However, one or the other of the two estimated coefficients associated with the gamma dose are negative. (author)