DOSE-RESPONSE RELATIONSHIP OF NEUTRONS AND GAMMA- RAYS TO LEUKEMIA INCIDENCE AMONG ATOMIC-BOMB SURVIVORS IN HIROSHIMA AND NAGASAKI BY TYPE OF LEUKEMIA, 1950-1971

The incidence of leukemia during 1950-1971 in a fixed mortality sample of atomic bomb survivors in Hiroshima and Nagasaki was analyzed as a function of neutron and γ kerma and marrow doses. Two dose-response models were tested for acute leukemia, chronic granulocyte leukemia, and all types of leukemia, respectively. Each model postulates that the leukemia incidence depends upon the sum of separate risks imposed by γ and neutron doses. In Model I the risk from both types of radiation is assumed to be directly proportional to the respective doses, while Model II assumes that whereas the risk from neutrons is directly proportional to the dose, the risk from γ rays is proportional to dose-squared. The analysis demonstrated that the dose response of the two types of leukemia differed by type of radiation. The data suggested that the response of acute leukemia was best explained by Model II, while the response of chronic granulocytic leukemia depended almost linearly upon neutron dose alone, because the regression coefficients associated with γ radiation for both Models I and II were not significant. The relative biological effectiveness (RBE) of neutrons in relation to γ rays for incidence of acute leukemia was estimated to be approximately 30/ (Dn)(1/2) [95% confidence limits; 17/(Dn)(1/2) ~ 54/(Dn)(1/2)] for kerma and 32/(Dn)(1/2) [95% confidence limits; 18/(Dn)(1/2) ~ 58/(Dn)(1/2)] for marrow dose (Dn = neutron dose). If acute and chronic granulocytic leukemias are considered together as all types of leukemia, Model II appears to fit the data slightly better than Model I, but neither model is statistically rejected by the data.