Technical Report No. 4-89
Radiation-related posterior lenticular opacities in Hiroshima and Nagasaki atomic bomb survivors based on T65DR and DS86 dosimetry systems
Otake M, Schull WJEditor’s note: A publication based on this report was published in Radiat Res 121:3-13, 1990.
Summary
This paper investigates the quantitative relationship of ionizing radiation to the occurrence of posterior lenticular opacities among the survivors of the atomic-bombings of Hiroshima and Nagasaki, as suggested by the DS86 dosimetry system. DS86 doses are available for 1,983 (93.4%) of the 2,124 A-bomb survivors analyzed in 1982. The DS86 kerma neutron component for Hiroshima survivors is much less than its comparable T65DR component, but is still 4.2-fold higher (0.38 Gy at 6 Gy) than that in Nagasaki (0.09 Gy at 6 Gy). Thus, if the eye is especially sensitive to neutrons, some useful information on neutron effects may yet be discernible, particularly in Hiroshima. The dose-response relationship has been evaluated as a function of the separately estimated gamma-ray and neutron doses. Among several different dose-response models with and without two thresholds, we have selected the one with the smallest chi2 or the largest log likelihood value associated with the goodness of fit. The best fit is a linear gamma-linear neutron relationship which assumes different thresholds for the two types of radiation.
In the DS86 system, both gamma-ray and neutron regression coefficients for the best-fitting model are positive and highly significant for the estimated energy deposited in the eye, here termed the eye organ dose. The DS86 gamma regression coefficient is almost the same as that associated with the T65DR gamma kerma, the ratio of the two coefficients being 1.1 (95% confidence limits: 0.5-2.3) for DS86 kerma in the individual data. If the risks based on the DS86 eye organ dose and DS86 kerma are compared, the ratio is 1.3 (0.6-2.8). However, the risk estimates associated with neutron exposure are 6.4-fold (2.2-19.2) higher for the DS86 kerma than for the T65DR kerma and 1.6-fold (0.5-5.2) higher for the DS86 eye organ dose than for the DS86 kerma.
The relative biological effectiveness (RBE) values based on the individual gamma and neutron components of the DS86 eye organ dose are estimated to be 32.4 + 0.73/(Dnu – 0.06) > 0 with the 95% confidence limits ranging from 11.8 to 88.8 + 1.39/(Dnu – 0.06) > 0, where Dnu is the neutron dose in gray. When such a threshold for the neutron dose is used, the RBE estimates are 105 at 0.01 Gy when Dnu is 0.07 Gy, 40 at 0.10 Gy when Dnu is 0.16 Gy, 36 at 0.20 Gy when Dnu is 0.26 Gy, 35 at 0.30 Gy when Dnu is 0.36 Gy and so on. The RBE value with the 95% lower bound suggests the constant to be 12. It should be noted that we cannot estimate the RBE when Dnu is less than or equal to 0.06 Gy based on the restriction of (Dnu – 0.06) > 0.
In any case, these values strongly suggest that the neutron component could be more important for the eyes than for other sites of the body. If we take into consideration the 95% lower bound of the neutron threshold including zero, we estimate the RBE values as 32.4 + 0.73/Dnu with a range from 11.8 to 88.8 + 1.39/Dnu. Finally, it is interesting to observe that a linear-quadratic gamma and linear neutron model with two thresholds, which fits the data less well, produces very similar estimates of the two thresholds as the linear gamma-linear neutron-response model. In this model, however, the regression coefficient is not significantly associated with the quadratic gamma response.
This paper investigates the quantitative relationship of ionizing radiation to the occurrence of posterior lenticular opacities among the survivors of the atomic-bombings of Hiroshima and Nagasaki, as suggested by the DS86 dosimetry system. DS86 doses are available for 1,983 (93.4%) of the 2,124 A-bomb survivors analyzed in 1982. The DS86 kerma neutron component for Hiroshima survivors is much less than its comparable T65DR component, but is still 4.2-fold higher (0.38 Gy at 6 Gy) than that in Nagasaki (0.09 Gy at 6 Gy). Thus, if the eye is especially sensitive to neutrons, some useful information on neutron effects may yet be discernible, particularly in Hiroshima. The dose-response relationship has been evaluated as a function of the separately estimated gamma-ray and neutron doses. Among several different dose-response models with and without two thresholds, we have selected the one with the smallest chi2 or the largest log likelihood value associated with the goodness of fit. The best fit is a linear gamma-linear neutron relationship which assumes different thresholds for the two types of radiation.
In the DS86 system, both gamma-ray and neutron regression coefficients for the best-fitting model are positive and highly significant for the estimated energy deposited in the eye, here termed the eye organ dose. The DS86 gamma regression coefficient is almost the same as that associated with the T65DR gamma kerma, the ratio of the two coefficients being 1.1 (95% confidence limits: 0.5-2.3) for DS86 kerma in the individual data. If the risks based on the DS86 eye organ dose and DS86 kerma are compared, the ratio is 1.3 (0.6-2.8). However, the risk estimates associated with neutron exposure are 6.4-fold (2.2-19.2) higher for the DS86 kerma than for the T65DR kerma and 1.6-fold (0.5-5.2) higher for the DS86 eye organ dose than for the DS86 kerma.
The relative biological effectiveness (RBE) values based on the individual gamma and neutron components of the DS86 eye organ dose are estimated to be 32.4 + 0.73/(Dnu – 0.06) > 0 with the 95% confidence limits ranging from 11.8 to 88.8 + 1.39/(Dnu – 0.06) > 0, where Dnu is the neutron dose in gray. When such a threshold for the neutron dose is used, the RBE estimates are 105 at 0.01 Gy when Dnu is 0.07 Gy, 40 at 0.10 Gy when Dnu is 0.16 Gy, 36 at 0.20 Gy when Dnu is 0.26 Gy, 35 at 0.30 Gy when Dnu is 0.36 Gy and so on. The RBE value with the 95% lower bound suggests the constant to be 12. It should be noted that we cannot estimate the RBE when Dnu is less than or equal to 0.06 Gy based on the restriction of (Dnu – 0.06) > 0.
In any case, these values strongly suggest that the neutron component could be more important for the eyes than for other sites of the body. If we take into consideration the 95% lower bound of the neutron threshold including zero, we estimate the RBE values as 32.4 + 0.73/Dnu with a range from 11.8 to 88.8 + 1.39/Dnu. Finally, it is interesting to observe that a linear-quadratic gamma and linear neutron model with two thresholds, which fits the data less well, produces very similar estimates of the two thresholds as the linear gamma-linear neutron-response model. In this model, however, the regression coefficient is not significantly associated with the quadratic gamma response.