J.D. Harrison

Quantitative comparisons of cancer induction in humans by internally deposited radionuclides and external radiation.

Purpose: To compare quantitative estimates of lifetime cancer risk in humans for exposures to internally deposited radionuclides and external radiation. To assess the possibility that risks from radionuclide exposures may be underestimated. Materials and methods: Risk estimates following internal exposures can be made for a small number of alpha-particle-emitting nuclides. (1) Lung cancer in underground miners exposed by inhalation to radon-222 gas and its short-lived progeny. Studies of residential 222Rn exposure are generally consistent with predictions from the miner studies. (2) Liver cancer and leukaemia in patients given intravascular injections of Thorotrast, a thorium-232 oxide preparation that concentrates in liver, spleen and bone marrow. (3) Bone cancer in patients given injections of radium-224, and in workers exposed occupationally to 226Ra and 228Ra, mainly by ingestion. (4) Lung cancer in Mayak workers exposed to plutonium-239, mainly by inhalation. Liver and bone cancers were also seen, but the dosimetry is not yet sufficiently good enough to provide quantitative estimates of risks. Comparisons can be made between risk estimates for radiation-induced cancer derived for radionuclide exposure and those derived for the A-bomb survivors, exposed mainly to low-LET (linear energy transfer) external radiation. Data from animal studies, using dogs and rodents, allow comparisons of cancer induction by a range of alpha- and beta-/gamma-emitting radionuclides. They provide information on relative biological effectiveness (RBE), dose-response relationships, dose-rate effects and the location of target cells for different malignancies. Results: For lung and liver cancer, the estimated values of risk per Sv for internal exposure, assuming an RBE for alpha-particles of 20, are reasonably consistent with estimates for external exposure to low-LET radiation. This also applies to bone cancer when risk is calculated on the basis of average bone dose, but consideration of dose to target cells on bone surfaces suggests a low RBE for alpha-particles. Similarly, for leukaemia, the comparison of risks from alpha-irradiation (232Th and progeny) and external radiation suggest a low alpha RBE; this conclusion is supported by animal data. Risk estimates for internal exposure are dependent on the assumptions made in calculating dose. Account is taken of the distribution of radionuclides within tissues and the distribution of target cells for cancer induction. For the lungs and liver, the available human and animal data provide support for current assumptions. However, for bone cancer and leukaemia, it may be that changes are required. Bone cancer risk may be best assessed by calculating dose to a 50 µ;m layer of marrow adjacent to endosteal (inner) bone surfaces rather than to a single 10 µ;m cell layer as currently assumed. Target cells for leukaemia may be concentrated towards the centre of marrow cavities so that the risk of leukaemia from bone-seeking radionuclides, particularly alpha emitters, may be overestimated by the current assumption of uniform distribution of target cells throughout red bone marrow. Conclusions: The lifetime risk estimates considered here for exposure to internally deposited radionuclides and to external radiation are subject to uncertainties, arising from the dosimetric assumptions made, from the quality of cancer incidence and mortality data and from aspects of risk modelling; including variations in baseline rates between populations for some cancer types. Bearing in mind such uncertainties, comparisons of risk estimates for internal emitters and external radiation show good agreement for lung and liver cancers. For leukaemia, the available data suggest that the assumption of an alpha-particle RBE of 20 can result in overestimates of risk. For bone cancer, it also appears that current assumptions will overestimate risks from alpha-particle-emitting nuclides, particularly at low doses.

Induction of osteosarcoma and acute myeloid leukaemia in CBA/H mice by the alpha-emitting nuclides, uranium-233, plutonium-239 and amercium-241

Purpose : To compare tumour induction in CBA/H mice, principally osteosarcoma and acute myeloid leukaemia, resulting from exposure to the alpha-emitting nuclides, uranium-233, plutonium-239 and americium-241, and to relate differences between the three nuclides to the pattern of dose delivery within tissues. Materials and methods : Each nuclide was administered intraperitoneally in citrate solution to three groups of adult male CBA/H mice at levels of activity which gave estimated life-time average skeletal doses of about 0.25-0.3 Gy, 0.5-1 Gy and 1-2 Gy. Animals were carefully monitored and sacrificed as soon as they showed signs of ill health; tumours were identified by standard histopathological techniques. Results : Statistical modelling by Cox regression showed that, considering all three nuclides together, there was a highly significant increase in risk of death from osteosarcoma or myeloid leukaemia with increasing dose rate. For osteosarcoma, the effect was significantly greater for 239 Pu than 241 Am, while separate analysis for 233 U showed no significant increase with increasing dose rate. For example, the increase in relative risk of death from osteosarcoma for an increase in life-time average dose rate to bone of 1 mGy d -1 was 4.2 (2.7-6.5) for 239 Pu, 2.3 (1.4-3.4) for 241 Am and 1.1 (0.4-3.1) for 233 U. For myeloid leukaemia, there was no significant difference between 239 Pu and 241 Am in the effect of dose rate. The increase in relative risk from myeloid leukaemia for an increase in average dose rate of 1 mGy d -1 was 1.8 (1.1-2.8) for 239 Pu, 2.0 (1.4-2.9) for 241 Am and 1.5 (0.8-2.7) for 233 U. Significant increases in renal and hepatic carcinomas were also recorded in animals exposed to 233 U and 241 Am, respectively. Studies of the distribution of the nuclides within the skeleton, published separately, have shown differences in their retention in individual bones and within bone. The proportions of decays occurring near to endosteal bone surfaces and throughout bone marrow were in the order: 239 Pu > 241 Am > 233 U. Conclusions : For osteosarcoma, the relative effectiveness of the nuclides in terms of average bone dose, in the order 239 Pu > 241 Am > 233 U, is consistent with the proportion of dose delivered near to endosteal surfaces. For myeloid leukaemia, the greater effectiveness of 239 Pu and 241 Am than 233 U is consistent with their accumulation in marrow.PURPOSE To compare tumour induction in CBA/H mice, principally osteosarcoma and acute myeloid leukaemia, resulting from exposure to the alpha-emitting nuclides, uranium-233, plutonium-239 and americium-241, and to relate differences between the three nuclides to the pattern of dose delivery within tissues. MATERIALS AND METHODS Each nuclide was administered intraperitoneally in citrate solution to three groups of adult male CBA/H mice at levels of activity which gave estimated life-time average skeletal doses of about 0.25-0.3 Gy, 0.5-1 Gy and 1-2 Gy. Animals were carefully monitored and sacrificed as soon as they showed signs of ill health; tumours were identified by standard histopathological techniques. RESULTS Statistical modelling by Cox regression showed that, considering all three nuclides together, there was a highly significant increase in risk of death from osteosarcoma or myeloid leukaemia with increasing dose rate. For osteosarcoma, the effect was significantly greater for 239Pu than 241Am, while separate analysis for 233U showed no significant increase with increasing dose rate. For example, the increase in relative risk of death from osteosarcoma for an increase in life-time average dose rate to bone of 1 mGyd(-1) was 4.2 (2.7-6.5) for 239Pu, 2.3 (1.4-3.4) for 241Am and 1.1 (0.4-3.1) for 233U. For myeloid leukaemia, there was no significant difference between 239Pu and 241Am in the effect of dose rate. The increase in relative risk from myeloid leukaemia for an increase in average dose rate of 1 mGyd(-1) was 1.8 (1.1-2.8) for 239Pu, 2.0 (1.4-2.9) for 241Am and 1.5 (0.8-2.7) for 233U. Significant increases in renal and hepatic carcinomas were also recorded in animals exposed to 233U and 241Am, respectively. Studies of the distribution of the nuclides within the skeleton, published separately, have shown differences in their retention in individual bones and within bone. The proportions of decays occurring near to endosteal bone surfaces and throughout bone marrow were in the order: 239Pu> 241Am>233U. CONCLUSIONS For osteosarcoma, the relative effectiveness of the nuclides in terms of average bone dose, in the order 239Pu>241Am>233U, is consistent with the proportion of dose delivered near to endosteal surfaces. For myeloid leukaemia, the greater effectiveness of 239Pu and 241Am than 233U is consistent with their accumulation in marrow.

The gastrointestinal absorption of the actinide elements

The greatest uncertainty in dose estimates for the ingestion of long-lived, alpha-emitting isotopes of the actinide elements is in the values used for their fractional absorption from the gastrointestinal tract (f1 values). Recent years have seen a large increase in the available data on actinide absorption. Human data are reviewed here, together with animal data, to illustrate the effect on absorption of chemical form, incorporation into food materials, fasting and other dietary factors, and age at ingestion. The f1 values recommended by the International Commission on Radiological Protection, by an Expert Group of the Nuclear Energy Agency and by the National Radiological Protection Board are discussed.

Plutonium and americium uptake in rats fed with cumbrian shellfish: implications for estimates of dose to man

Winkles (Littorina littorea) and mussels (Mytilus edulis) collected on the Cumbrian coast contain americium-241 and isotopes of plutonium discharged from the nuclear-fuel reprocessing plant at Sellafield. Shellfish have been fed to rats and measurements made of the gastrointestinal absorption of the actinides. For shellfish collected over a 1-year period from March 1983 to February 1984, the average values for the fractional absorption of plutonium and americium were 9 x 10(-4) and 3 x 10(-4), respectively, for winkles and 1.5 x 10(-3) and 6 x 10(-4), respectively, for mussels. Comparisons with results for winkles collected in December 1981 and mussels collected in July 1982 suggest that there may be considerable seasonal variation in the availability of the actinides for absorption. The results suggest that in calculations of doses to individuals consuming shellfish in west Cumbria, it may be prudent to examine the effect of using the new ICRP gut transfer factor of 1 x 10(-3) for both actinides, in comparison with the value of 5 x 10(-4) recommended previously by NRPB. The use of 1 x 10(-3) would increase the estimate of the committed effective dose equivalent for 1985 intakes, from the value of 0.73 mSv calculated by the Ministry of Agriculture, Fisheries and Food, to 1.29 mSv. However, taking into account up-to-date estimates of the retention of the actinides in liver and bone would reduce this value to 1.07 mSv. If, in addition, allowance is made for the effect of the burial and recycling of actinides in bone, a significant reduction in the dose estimate could result; for example, the use of one proposed dynamic bone model would reduce the value from 1.07 to 0.54 mSv.

Dose coefficients for the embryo and foetus following intakes of radionuclides by the mother

Committee 2 of the International Commission on Radiological Protection (ICRP) has the responsibility for calculating radiation doses from intakes of radionuclides for all age groups in the population. Publication 88 of the ICRP, which has recently been published, describes the development of models used for calculating radiation doses to the embryo and foetus following intakes of radionuclides by the mother. It also gives radiation doses to the offspring for intakes of radionuclides by the mother either before or during pregnancy. The approaches used in the development of the biokinetic and dosimetric models are summarised here together with a comparison of the doses to the offspring with those to the reference adult.

Gastrointestinal Absorption and Retention of Polonium in Adult and Newborn Rats and Guinea Pigs

The gastrointestinal absorption of 210Po was determined by comparing tissue retention after oral and systemic administration. The results indicate an increase in absorption in adult rats for 210Po administered in liver compared with 210Po nitrate with estimated absorption of 5 and 13%, respectively. For 210Po citrate, values of about 7% were obtained in 1-day-old neonate and adult rats while absorption in guinea pigs was estimated to be about 23% in 1-day-old neonates, 17% in 5-day-old neonates, and 9% in adults. Gut retention of ingested 210Po in neonates was high in rats but not guinea pigs. In adult animals, but not neonates, the liver accounted for a greater proportion of 210Po reaching the bloodstream after ingestion than after systemic injection. The significance of these results is discussed in relation to current assumptions made in the calculation of doses from 210Po.

The speciation of plutonium in foodstuffs and its influence on gut uptake

Soluble plutonium complexing agents in foodstuffs have been identified. These have been labelled to high specific activity with 238Pu and their gut uptake determined in rats. The range of values found for naturally-occurring complexes was between 0.03% for oxalate and 0.1% for phytate. However, reasons are advanced for believing that the enhancement of uptake in rats fed the phytate complex would not be expected to occur in man. These results are not taken, therefore, as suggesting that there should be any change in the gut uptake factor of 0.05% currently recommended by the National Radiological Protection Board for plutonium contained in foodstuffs eaten by humans.

Microdistribution and localized dosimetry of the alpha-emitting radionuclides 239Pu, 241Am and 233U in mouse femoral shaft.

Purpose: To analyse the temporal change in microdistribution of 239Pu, 241Am and 233U in mouse femur and to compare the calculated radiation doses with regions of the bone marrow thought to contain target cells for osteosarcoma and leukaemia with relative risk for those diseases. Materials and methods: Neutron-induced and alpha-track autoradiographs were prepared from femora of the CBA/H mouse that had been injected with 40kBqkg 1 radionuclide between 1 and 448 days previously. Computer-based image analysis of the autoradiographs was performed and dosimetric methods applied to obtain radiation dose-rates to different regions of the marrow cavity. Results: Initially each radionuclide deposited on endosteal and periosteal bone surfaces; 241Am was additionally deposited on vascular canal surfaces. Redistribution resulted in 233U being incorporated into bone, while 239Pu and 241Am showed transfer into both bone volume and marrow. Accumulation in the central marrow peaked at 112-224 days post-injection, but subsequently was cleared by 448 days. Cumulative doses to both osteosarcomagenic and myeloid leukaemogenic target cell regions showed the trend 239 Pu> 241 Am> 233 U. Conclusions: Calculation of cumulative doses to a 10-mum layer of marrow adjacent to bone surfaces appears to be a suitable predictor for risk of osteosarcoma. Risks of myeloid leukaemia in the mouse are better predicted by considering the central marrow as the target region rather than average dose to all marrow.PURPOSE To analyse the temporal change in microdistribution of 239Pu, 241Am and 233U in mouse femur and to compare the calculated radiation doses with regions of the bone marrow thought to contain target cells for osteosarcoma and leukaemia with relative risk for those diseases. MATERIALS AND METHODS Neutron-induced and alpha-track autoradiographs were prepared from femora of the CBA/H mouse that had been injected with 40 kBq kg(-1) radionuclide between 1 and 448 days previously. Computer-based image analysis of the autoradiographs was performed and dosimetric methods applied to obtain radiation dose-rates to different regions of the marrow cavity. RESULTS Initially each radionuclide deposited on endosteal and periosteal bone surfaces; 241Am was additionally deposited on vascular canal surfaces. Redistribution resulted in 233U being incorporated into bone, while 239Pu and 241Am showed transfer into both bone volume and marrow. Accumulation in the central marrow peaked at 112-224 days post-injection, but subsequently was cleared by 448 days. Cumulative doses to both osteosarcomagenic and myeloid leukaemogenic target cell regions showed the trend 239Pu > 241Am > 233U. CONCLUSIONS Calculation of cumulative doses to a 10-microm layer of marrow adjacent to bone surfaces appears to be a suitable predictor for risk of osteosarcoma. Risks of myeloid leukaemia in the mouse are better predicted by considering the central marrow as the target region rather than average dose to all marrow.

The gastrointestinal absorption of plutonium and americium in rats and guinea pigs after ingestion of dusts from the former nuclear weapons site at Maralinga: implications for human exposure

The gastrointestinal absorption of plutonium and americium present in dusts from the Maralinga test sites in South Australia has been measured as an input to dose assessments. The materials studied were from three different areas, designated Taranaki (Q380II), TM100 (10/2) and N. Plume (26). The three dusts were fed to groups of rats, mixed with their normal food. The Taranaki and TM100 dusts were also fed to guinea pigs, as a suspension in water. Expressed as fractional absorption from the gastrointestinal tract (f1), the values obtained for plutonium were 2 x 10(-6) and 8 x 10(-6) for Taranaki dust in rats and guinea pigs, respectively, 3 x 10(-6) and 10(-5) for TM100 dust in rats and guinea pigs, respectively, and 2 x 10(-6) for N. Plume dust in rats. The f1 values obtained for americium were 3 x 10(-6) and 2 x 10(-5) for Taranaki dust in rats and guinea pigs, respectively, 10(-5) and 5 x 10(-5) for Taranaki dust in rats and guinea pigs, respectively, and 10(-5) for N. Plume dust in rats. On the basis of these results, rounded f1 values used in the dose assessments were 10(-5) for plutonium and 10(-4) for americium, applying to all intakes of dust. These values compare with the International Commission on Radiological Protection recommendations of 10(-3) for unspecified chemical forms of both elements, 10(-4) for plutonium nitrate and 10(-5) for plutonium oxides. The effect of changes in f1 values on doses from 239Pu and 241Am is considered.

The Wound Clearance and Distribution of Plutonium, Americium and Curium in Rodents

Wound contamination was simulated by the intramuscular injections of either 239Pu, 241Am or 244Cm nitrate in the hamster and by the subcutaneous injection of either 239Pu or 241Am nitrate in the rat. The actinides moved from the hamster muscle at approximately the same rate after injection of 370 Bq of each radionuclide (approximate 80 per cent cleared after 6 months). Similarly, the rates of translocation of 239Pu and 241Am in the rat were the same after injection of 370 Bq of each radionuclide (approximate 80 per cent cleared after 1 month). The clearance of 239Pu in the rat was slower after injection of 14.8 kBq and more rapid after injection of 20 Bq (approximately 40 per cent and 90 per cent moved after 1 month, respectively) while the movement of 241Am showed no dependence on the administered dose over the same range. The mixing of 239Pu and 241Am prior to intramuscular injection appeared to enhance the clearance of both radionuclides. The levels of accumulation of each actinide in the skeleton and liver of both species showed that they reached the circulation predominantly in soluble form. Some uptake of Pu and Am in regional lymph nodes was also observed, indicating that lymphatic clearance of polymeric material also took place.