Appendix B
Risk of Selected Cancers and Nonmalignant Diseases in Uranium Workers

COLORADO PLATEAU URANIUM MILLERS

At the Colorado Plateau uranium mills, uranium was extracted from ores to produce “yellowcake,” a semirefined mixture of diuranates, basic uranyl sulfate, and hydrated uranium oxides that consists of 80-96% uranium (Pinkerton et al. 2004). The processes involved ore handling and preparation, extraction, concentration and purification, and precipitation, drying, and packaging. The uranium ore dusts in the preparation areas consisted principally of insoluble uranium oxides, although a small fraction was soluble uranium compounds. The mills dried the yellowcake at relatively high temperatures (370-538°C) in a process in which it was principally triuranium octaoxide with low solubility. Workers at the mills were also exposed to silica and vanadium in the dusts, and most of the mills recovered vanadium as part of the processing.

Pinkerton et al. (2004) conducted a mortality followup through 1988 of a cohort of 1,485 men who worked at seven uranium mills during 1940-1971 on the Colorado Plateau. Most of the cohort had previously been studied by Waxweiler et al. (1983) and Archer et al. (1973). The men were employed for 1-36 y (median, 3.6 y) in milling operations. Those who had worked in uranium mining were excluded because radon-exposure effects would probably predominate in that group. In addition to obtaining mortality data, the researchers linked the workers to the end-stage renal-disease data maintained by the Health Care Financing Administration (HCFA), which documents people who received renal dialysis or renal transplants.

Pinkerton et al. (2004) observed no statistically significant excess of cancer mortality overall or related to any specific site of interest. The lung-cancer risk was not statistically significant (relative risk [RR], 1.13), and there was a significant negative trend with duration of uranium-milling employment. The negative trend may be due to the “healthy-worker survivor effect” or to differences in smoking habits between short- and long-term employees. It may also be due to the fact that before 1955, when uranium-mill dust concentrations were higher, the workers tended to be short-term employees, whereas those employed after 1955 were exposed to lower concentrations of uranium dust and tended to



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 146
Appendix B Risk of Selected Cancers and Nonmalignant Diseases in Uranium Workers COLORADO PLATEAU URANIUM MILLERS At the Colorado Plateau uranium mills, uranium was extracted from ores to produce “yellowcake,” a semirefined mixture of diuranates, basic uranyl sul- fate, and hydrated uranium oxides that consists of 80-96% uranium (Pinkerton et al. 2004). The processes involved ore handling and preparation, extraction, con- centration and purification, and precipitation, drying, and packaging. The ura- nium ore dusts in the preparation areas consisted principally of insoluble ura- nium oxides, although a small fraction was soluble uranium compounds. The mills dried the yellowcake at relatively high temperatures (370-538oC) in a process in which it was principally triuranium octaoxide with low solubility. Workers at the mills were also exposed to silica and vanadium in the dusts, and most of the mills recovered vanadium as part of the processing. Pinkerton et al. (2004) conducted a mortality followup through 1988 of a cohort of 1,485 men who worked at seven uranium mills during 1940-1971 on the Colorado Plateau. Most of the cohort had previously been studied by Wax- weiler et al. (1983) and Archer et al. (1973). The men were employed for 1-36 y (median, 3.6 y) in milling operations. Those who had worked in uranium mining were excluded because radon-exposure effects would probably predominate in that group. In addition to obtaining mortality data, the researchers linked the workers to the end-stage renal-disease data maintained by the Health Care Fi- nancing Administration (HCFA), which documents people who received renal dialysis or renal transplants. Pinkerton et al. (2004) observed no statistically significant excess of can- cer mortality overall or related to any specific site of interest. The lung-cancer risk was not statistically significant (relative risk [RR], 1.13), and there was a significant negative trend with duration of uranium-milling employment. The negative trend may be due to the “healthy-worker survivor effect” or to differ- ences in smoking habits between short- and long-term employees. It may also be due to the fact that before 1955, when uranium-mill dust concentrations were higher, the workers tended to be short-term employees, whereas those employed after 1955 were exposed to lower concentrations of uranium dust and tended to 146

OCR for page 146
147 Appendix B have longer employment. Lymphopoietic malignancies are of interest because human data indicate that inhaled insoluble uranium compounds accumulate in the tracheobronchial lymph nodes (Singh et al. 1987). The trend by duration of employment was in the positive direction for lymphopoietic cancers other than leukemia but was not significant. Within the category of nonmalignant respiratory disease, there were ex- cesses of emphysema (standardized mortality ratio [SMR], 1.96; 95% confi- dence interval [CI], 1.21-2.99; n = 21) and pneumoconioses and other respira- tory diseases (SMR, 1.68; 95% CI, 1.26-2.21; n = 52). However, all non- malignant respiratory diseases and emphysema were inversely associated with duration of milling work. There was a suggestion of an excess of deaths from chronic renal failure in the Colorado uranium-miller cohort for the period 1940-1998, when only the primary cause of death was examined (SMR, 1.35; 95% CI, 0.6-2.7; 8 observed, 5.9 expected). In support of a possible increase in risk was the trend with dura- tion of uranium-milling employment (SMRs, 1.27, 1.33, and 1.53 for 1-2, 3-9, and over 10 y of employment, respectively). However, when multiple causes of death as given on a death certificate were examined, there was no excess (SMR, 1.05); and when the incidence of treated end-stage renal disease was evaluated with HCFA’s program data, there again was no excess (standardized incidence rate [SIR], 0.71; 95% CI, 0.26-1.65). A possible explanation for not finding sig- nificant effects is that renal exposures may have been low because the workers were exposed mainly to uranium compounds of low solubility. TENNESSEE EASTMAN CORPORATION, OAK RIDGE, TENNESSEE From June 1943 to May 1947, the Tennessee Eastman Corporation (TEC) plant was engaged in the enrichment of uranium with the electromagnetic sepa- ration process. Polednak and Frome (1981) determined the mortality experience through 1974 of a cohort of 18,869 white men who worked at the TEC Y12 plant. A substantial number of the employees (8,345) worked in the chemical departments (code named alpha and beta to refer to stages in the electromagnetic separation process), where exposures were high. In the alpha departments, ura- nium trioxide was converted to enriched uranium tetrachloride. Those depart- ments, which had the highest uranium concentrations, operated until September 1945. Thereafter, uranium hexafluoride from the Oak Ridge K-25 (gaseous- diffusion) facility was fed to the beta stage of the process. Before late 1945, ura- nium trioxide was received from Mallinckrodt Chemical Works and converted to uranium tetrachloride, but thereafter exposure to the insoluble oxides was partly replaced by exposure to the more soluble uranium hexafluoride and uranyl fluoride. Uranium hexafluoride was converted to the oxides uranium tetraoxide and uranium trioxide and then to “green salt” in the beta departments. Other toxic exposure in the workplace included exposure to phosgene gas, mer- cury, carbon tetrachloride, and trichloroethylene.

OCR for page 146
148 Risks to Military Personnel from Exposure to Depleted Uranium The average uranium concentrations in the air in 1945 by department and activity are shown in Table B-1. Many readings were considerably higher, in- cluding occasional concentrations greater than 10,000-20,000 µg/m3. There were no statistically significant increases in risk of total cancer, lung cancer, bone cancer, renal cancer, lymphoma, or leukemia, nor was there an excess of chronic renal disease or other diseases. Cookfair et al. (1983) later conducted a case-control study at the facility to examine the possible association of uranium exposure and lung cancer in more detail. They had smoking information on about half the 330 cases and their con- trols. They reported no association overall but did find a nominal increase in risk in those who were over age 45 y old at exposure and received a lung dose of at least 200 mGy. URANIUM-MATERIALS FABRICATION PLANT, OAK RIDGE, TENNESSEE In 1947, the TEC uranium-enrichment operations ceased, and the Y12 fa- cility shifted its function to fabrication and assembly of nuclear-weapons mate- rials and the recovery and recycling of nuclear products or materials on behalf of the U.S. government. The production processes involved the conversion of ura- nium hexafluoride to uranium tetrafluoride and its reduction to uranium metal (Loomis and Wolf 1996). Exposure was primarily through inhalation of ura- nium-bearing dusts. Other workplace toxicants included solvents, machine oils, mercury, lead, and beryllium. The Y12 worker cohort consisted of 10,597 men and women employed during 1947-1974; mortality was ascertained through 1990 (Checkoway et al. 1988; Loomis and Wolf 1996). The primary statistics were based on 6,591 white men because the nonwhite contingent was relatively small; there was substantial underascertainment of mortality in the women beause of difficulties in linking records. In the white men, the only cancer that had a statistically significant ex- cess was lung cancer (RR, 1.20; 95% CI, 1.04-1.38; n = 194 deaths). However, there was no consistent pattern in lung-cancer risk by number of years since beginning Y12 employment or by duration of employment (SMRs for less than 5, 5-9, 10-19, 20-29, and 30+ y of employment were 1.23, 1.87, 1.87, 1.11, and 0.73, respectively). A recent reanalysis of the lung-cancer data that used exter- nal-dose measurements and estimated internal doses to the lungs, primarily from uranium dust, did not reveal a consistent relationship for the internal exposures, although the group with the highest external and internal exposures had the highest risk (RR, 2.2; 95% CI, 0.7-6.7; n = 7 lung cancers) (Richardson and Wing 2006). In the 452 nonwhite (99% black) men, there were no statistically signifi- cant excesses of cancer or other mortality, although there was a suggestion of an excess of digestive-system cancers (SMR, 1.6; n = 7). Of the other common causes of death, no excess was seen for lung cancer (SMR, 0.60) or cardiovascu-

OCR for page 146
149 Appendix B TABLE B-1 Average Air Concentrations of Uranium Average Concentration of Uranium Compounds (as Uranium) in Air (µg/m3) Department and Activity Alpha chemistry Sublimation (purification of UCl4) 300 Bottle filling with UCl4 250 Recovery of uranium, conversion to UO3 500 Alpha process, Calutron, first stage 25 Beta chemistry UCl4 production and bottle filling; recovery 50 Beta process (Calutron; second stage) 25 Source: Polednak and Frome 1981. Reprinted with permission; copyright 1981, Lippin- cott Williams & Wilkins. lar disease (SMR, 0.55). In the 1,073 female employees, there was a nominal, nonsignificant excess of breast cancer (SMR, 1.21), a nominal deficit of lung cancer (SMR, 0.78), and a statistically significant deficit of circulatory diseases (SMR, 0.40; 95% CI, 0.2-0.8). There was no excess mortality from either renal cancer or chronic renal disease in the Y12 cohort. Loomis and Wolf (1996) noted, however, that soluble uranium is more nephrotoxic than insoluble uranium and that the Y12 exposure was primarily to insoluble forms, although some people were working with ura- nium hexafluoride. MALLINCKRODT CHEMICAL WORKS, ST. LOUIS, MISSOURI The mortality experience of 2,514 white men employed during 1942-1966 at the Mallinckrodt uranium-processing facility in St. Louis, Missouri, has been assessed through 1993 (Dupree-Ellis et al. 2000). Mallinckrodt processed “pitchblende” uranium ore from the Belgian Congo, which contained up to 70% uranium and high concentrations of radium, so external radiation exposure (mean, 47.8 mSv) was more of an issue with this cohort than with other ura- nium-processing cohorts. Processing methods were crude by modern standards. Daily average uranium-dust concentrations in poorly ventilated processing areas were as high as 5,000-10,000 µg/m3. Silica and sulfuric acid were other hazard- ous substances in the workplace. White men were studied because other race-sex groups were too small. They had worked for a mean of 5.2 y and had a median followup of 36 y. There were no apparent excesses of total cancer, lung cancer, lymphoma, or leukemia. There was a slight suggestion of an excess of renal cancer (SMR, 1.17), and the researchers reported a significant trend in renal-cancer risk with external- radiation dose (excess RR, 10.5/Sv; 90% CI, 0.6-57.4). There was also a sugges-

OCR for page 146
150 Risks to Military Personnel from Exposure to Depleted Uranium tive excess of chronic nephritis. They reported nominal excesses of 30% or more for cancers of the esophagus, rectum, and brain and for multiple myeloma, al- though these excesses were not statistically significant. FERNALD FEED-MATERIALS PRODUCTION CENTER, OHIO Mortality was ascertained through 1990 in a cohort of 4,014 workers em- ployed at the Fernald (Ohio) uranium-processing facility during 1951-1989 (85% were hired before 1960; Ritz 1999). Fernald workers processed uranium- ore concentrate and low-grade enriched uranium into fabricated uranium metal products. Most uranium exposure was to relatively insoluble compounds. Urine samples were routinely obtained to monitor uranium exposure. Radiation doses from both external and internal radiation sources were obtained; 33% of the co- hort was estimated to have cumulative internal lung doses exceeding 50 mSv; only 9% of the cohort was estimated to have external doses exceedeing that level. Many workers also had exposure to solvents, especially trichloroethylene, and cutting oils; there was also exposure to tributyl phosphate, ammonium hy- droxide, and hydrogen fluoride. Smoking information was available on 17% of the workers. The workers were followed for an average of 31 y since hiring. There was no excess of total cancers or of lung, renal, brain, or testicular cancers. There was a suggestive but nonsignificant excess of hepatic cancer (SMR, 1.62) and a statistically significant excess of lymphomas (SMR, 1.81; 95% CI, 1.0-3.0). There was no excess of circulatory disease, nonmalignant respiratory disease, or hepatic cirrhosis. Chronic renal disease was not examined. LUNG-CANCER CASE-CONTROL STUDY Dupree et al. (1995) conducted a case-control study of lung-cancer risk in workers with uranium-dust exposure at four processing facilities (TEC/Y12, Y12, Mallinckrodt, and Fernald). The primary advantages of their study over the studies of lung cancer at the separate facilities (above) are the larger number of cases from the pooling across facilities, that they were able to perform a more detailed exposure assessment of the 787 lung-cancer cases and their 1:1 matched controls than was possible in the large cohorts, and that they were able to extract data to control, at least roughly, for socioeconomic status (pay codes) and smok- ing (ever vs never). Smoking information was available on only 48% of cases and 39% of controls. The smoking frequency among controls was high (75%) and was higher in the people with lung cancer (91%), as expected. No associa- tion was found between lung cancer and the lung dose resulting from internal exposures to uranium dust. Compared with a baseline group of those with inter- nal lung exposures of less than 0.5 mGy, the odds ratios for those with estimated doses of 0.5, 2.5, 5, 25, 50, and 250+ mGy were 1.0, 0.6, 0.9, 0.8, 0.6, and 2.1 (95% CI, 0.2-21), respectively.

OCR for page 146
151 Appendix B PORTSMOUTH URANIUM-ENRICHMENT FACILITY This plant used a gaseous-diffusion process to enrich uranium up to 98% 235 U. Plant operations began in 1954. The 5,773 white men employed through 1982 were chosen for study by the National Institute for Occupational Safety and Health (Brown and Bloom 1987). The median length of employment at the plant was about 5 y. Potential exposure at the plant included exposure to uranyl fluoride and uranium hexafluoride. Because the uranium compounds were very soluble, urinary measurements of uranium were used. The exposures were rela- tively low: 94% of reported urinary measurements were below the limits of de- tection, 5.1% were 10-50 µg/L, and 0.6% were 50-200 µg/L. Exposures at the plant included exposure to hydrogen fluoride and technetium-99. The study found no excess mortality from total cancers or lung cancer. There was no gradient of lung-cancer risk with amount of exposure. There was also no greater cancer mortality in those with longer employment. Although there was a nominal, nonsignificant increase in gastric and lymphohematopoietic malignancies, a subcohort selected for the greatest potential uranium exposure had reduced risks of these malignancies. Likewise, there was no excess mortal- ity from chronic renal disease (SMR, 0.54), nonmalignant neurologic disease (SMR, 0.40; 95% CI, 0.21-0.68), or chronic respiratory disease (SMR, 0.46). SAVANNAH RIVER NUCLEAR-FUELS PRODUCTION FACILITY Operations at this plant included uranium and thorium processing; nu- clear-fuel fabrication; nuclear-reactor operation, overhauling, modification, maintenance, and refueling; and nuclear-fuel reprocessing. There was a potential for varied internal exposure, including exposure to tritium, uranium, fission products, iodine, activation products, and several transuranics. A detailed analy- sis of the cause of death of 9,400 past employees showed deficits of all cancers, lung cancers, digestive cancers, brain cancers, and several types of nonmalig- nant disease. There was a nominal deficit of renal cancer and a significant deficit of chronic nephritis (2/7.42; SMR, 0.27; 95% CI, 0.03-0.97). The low mortality ratios were most likely due in part to a healthy-worker effect. The one sugges- tive finding was an excess of leukemia in hourly workers first employed before 1956. However, among the 14 people with leukemia in this group, only six had an indication of any kind of internal exposure, so it seems doubtful that uranium played a role in the excess. No information was provided on subgroups with exposure to uranium or other specific agents, so this study provides little infor- mation on the health effects of uranium exposure. LINDE AIR PRODUCTS From 1943 to 1949, the Linde Air Products Ceramics Plant near Buffalo, New York, was a uranium-processing facility. It mainly converted Belgian

OCR for page 146
152 Risks to Military Personnel from Exposure to Depleted Uranium Congo pitchblende and domestic uranium ores to uranium tetrafluoride; the rela- tively insoluble uranium oxide, uranium trioxide, and uranium dioxide were intermediate products. There was also exposure to cutting oils, welding fumes, organic solvents, and asbestos. The vital status of the 995 white men employed during 1943-1949 was determined through 1979 (Dupree et al. 1987). The re- sults showed no excess of total cancers, lung cancers, or lymphopoietic malig- nancies. Teta and Ott (1988) extended the followup for an additional 2 y. They reported excesses of soft-tissue sarcomas and lung cancers, although the lung- cancer excess was not associated with length of employment. Hepatic cirrhosis was also found to be increased, perhaps because of the use of carbon tetrachlo- ride at the plant. UNITED NUCLEAR CORPORATION, CONNECTICUT The United Nuclear Corporation fabricated nuclear fuels at a plant in Connecticut for many years; the 4,106 workers during 1956-1978 were followed for both mortality and cancer incidence through 1979 by Hadjimichael et al. (1983). A subset of 2,613 men called industrial employees sustained most of the uranium exposure. Their cancer incidence showed no statistically significant increases in risk. There was a statistically significant excess of nonmalignant respiratory diseases, but it was based on small numbers. ATOMIC WEAPONS ESTABLISHMENT, UNITED KINGDOM Several studies of uranium workers in the United Kingdom have been conducted. Beral et al. (1988) examined mortality among 22,552 workers with various exposures employed at the UK Atomic Weapons Establishment during 1951-1982. A total of 3,044 were monitored for uranium exposure. Most of the SMRs for the uranium-exposed group were unremarkable, but there was a statis- tically significant excess of prostatic cancer (SMR, 2.81; 95% CI, 1.14-5.84; 6 cases observed) and a marginally increased rate of renal cancer (SMR, 4.30; CI, 0.89-12.6; 3 cases observed). It should be noted that many workers were ex- posed to more than one radionuclide, so the apparent increases in risk may be associated partly with other exposure. In particular, the excess prostatic cancers were accounted for by workers who had had multiple radionuclide exposures, and two of the three people with renal cancer had also been monitored for polo- nium. BRITISH NUCLEAR FUELS CAPENHURST PLANT McGeoghegan and Binks (2000b) documented mortality and cancer inci- dence in a cohort of 12,540 employees at the Capenhurst plant of British Nu- clear Fuels or its predecessors during 1946-1995. The plant mainly performed 235 U enrichment, beginning in 1953. The gaseous-diffusion process was used

OCR for page 146
153 Appendix B exclusively until 1977, when the changeover to the gas-centrifuge process was begun; in 1983, it was completed. Workers at a small tritium-production plant in the complex also were included in the cohort. The mean cumulative external radiation dose to the 3,244 radiation workers was 9.85 mSv. Radiation doses from internally deposited radionuclides were not reported. The mean followup for mortality was 26.7 y. Cancer incidence was ascertained for 1971-1991. The SMRs for the cohort were generally less than unity except for weak suggestions of increases in risk of lymphoma and of brain and central nervous system (CNS) malignancies. The findings on cancer incidence were qualitatively similar to the cancer-mortality data except that there was a small deficit of inci- dent lymphoma (SIR, 0.59) and no indication of an increase in brain or CNS malignancies (SIR, 1.03). Of nonmalignant diseases, there was a nominal but not statistically significant excess of chronic renal failure (4 cases observed, 2.2 expected). BRITISH NUCLEAR FUELS SPRINGFIELDS PLANT The other UKstudy was of the Springfields plant of British Nuclear Fuels (McGeoghegan and Binks 2000a). The main activities at the plant were ura- nium-fuel fabrication and uranium hexafluoride production. The plant received yellowcake containing about 75% uranium. The yellowcake was dissolved in nitric acid to produce uranyl nitrate, which was purified and concentrated and then converted to uranium tetrafluoride by a process that included reacting at a high temperature with air and then with hydrogen fluoride. Some uranium tetra- fluoride was converted to uranium metal, and some was reacted with fluorine to produce uranium hexafluoride, which was then sent for enrichment before being converted to uranium oxide fuel. The Springfields study consisted of 19,454 workers during 1946-1955, of whom 13,960 were radiation workers. The mean external radiation dose was 22.8 mSv when one method of calculation of pre-1953 doses was used and 20.5 mSv when an another method was used. Doses from internal radionuclides were not available. On the average, the workers were followed for mortality for 24.6 y. Cancer incidence was examined for 1971-1991. It is notable that this was the largest study of uranium workers, in whom there were 971 cancer deaths. There were no indications of excess cancer mortality or incidence in the radiation workers compared with the general population, and there was no excess of chronic renal failure. NONSPECIFIC STUDIES OF URANIUM AND OTHER RADIONUCLIDE EXPOSURES Two epidemiologic studies have examined mortality in workers exposed to uranium or other radionuclides. Carpenter et al. (1998) examined cancer- mortality risk in UK workers exposed to plutonium, tritium, and other radionu-

OCR for page 146
154 Risks to Military Personnel from Exposure to Depleted Uranium clides, including uranium, but they did not report any results specific to uranium exposure. Similarly, a study of cancer mortality was conducted at Rocketdyne (Ritz et al. 2000), where workers were exposed primarily to uranium and mixed fission products but also to strontium, plutonium, and other radionuclides. No analyses of a specifically uranium-exposed group were reported. FLORIDA PHOSPHATE-PRODUCTION WORKERS Several studies of phosphate-fertilizer production workers have been con- ducted. The earliest reported an excess of lung cancer in blacks (SMR, 1.82; 95% CI, 0.7-4.0; 5 cases observed, 2.74 expected) at a fertilizer-production plant (Stayner et al. 1985). The risk was found in those with 20 y or more of employ- ment and 20 y or more of observation (SMR, 12.5; 95% CI, 1.5-45; 2 cases ob- served, 0.16 expected), but no excess was observed in whites (SMR, 0.85; CI, 0.4-2.9; 5 cases observed). Phosphate-fertilizer production involves likely expo- sure to chromium, arsenic, and radon daughters in addition to uranium. A larger study was based on 18,466 white and 4,546 nonwhite men at 16 fertilizer plants in Florida (Checkoway et al. 1985, 1996). The SMRs did not differ appreciably between the white and nonwhite cohorts, so the committee combined the two cohorts in its assessment. The men had worked in the industry for an average of 9 y and were followed for mortality for a median of 22 y. The findings were unremarkable with two exceptions: there were statistically significant deficits of hepatic cancer and lymphoma, but there was an excess of lung cancer (SMR, 1.18; CI, 1.1-1.3). In white men, there were significant lung-cancer excesses for less than 5 y of employment (SMR, 1.23; CI, 1.05-1.4) and for 30 y or more (SMR, 1.94; CI, 1.13-3.1). The former may reflect smoking patterns in short- term workers, and the latter might reflect exposure to uranium or other contami- nants.