ESTIMATES OF NEUTRON RELATIVE BIOLOGICAL EFFECTIVENESS DERIVED FROM THE JAPANESE ATOMIC-BOMB SURVIVORS

Purpose: To investigate neutron relative biological effectiveness. Mat erials and methods: The latest Japanese atomic bomb survivor cancer in cidence and mortality datasets with the current (DS86) dosimetry syste m are analysed using generalized relative risk models and generalized absolute risk models, both with and without recently indicated adjustm ents to the Hiroshima DS86 neutron dose estimates. Results: Without ad justments to the Hiroshima neutron doses, the best estimate of neutron relative biological effectiveness for all tumours in the incidence da ta is 63.3 (95% CI<0-275.3) when a generalized relative risk model is used; when a generalized absolute risk model is used in the incidence data the best estimate is 53.5 (95% CI<0-201.0); when a generalized re lative risk model is used in the mortality data, the best estimate is 287.7 (95% CI 38.0 - > 10(3)). When likely adjustments are made to the Hiroshima neutron doses the best estimate of neutron relative biologi cal effectiveness in the incidence data using a generalized relative r isk model is 15.1 (95% CI <0-51.4); when a generalized absolute risk m odel is used in the incidence data the best estimate is 9.0 (95% CI<0- 32.9); when a generalized relative risk model is used in the mortality data the best estimate is 55.1 (95% CI 9.5-280.3). Although there are no significant differences between groupings of the solid tumour site s in their estimated neutron relative biological effectiveness, there are indications that the neutron relative biological effectiveness of solid tumours is lower than that of leukaemia, whether or not adjustme nts are made to the Hiroshima neutron dose estimates. Uncertainties in the likely adjustments to the DS86 Hiroshima neutron and gamma dose e stimates as well as uncertainties in the modelling of excess risk in t he two cities (Hiroshima and Nagasaki) imply that these findings shoul d be treated with caution. Conclusions: Likely adjustments to the Hiro shima neutron dose estimates imply a substantial increase in informati on on neutron relative biological effectiveness. Whether or not adjust ments are made to the Hiroshima neutron doses, there are indications o f inconsistency between the estimates of neutron relative biological e ffectiveness for solid tumours and leukaemia. Dosimetric and modelling uncertainties mean that these findings should be treated with caution .