Leukaemia Among Uranium Miners Late Effects of Exposure to Uranium Dust L. Tomášek 1, A. Heribanová 2 1 National Radiation Protection Institute, Prague, Šrobárova 48, Czech Republic E-mail: ladislav.tomasek@suro.cz 2 State Office for Nuclear Safety, Prague, Senovážné nám. 9, Czech Republic Abstract. The paper reports recent observations among Czech uranium miners. The study includes two groups of underground miners exposed in periods 1948-63 and 1968-88. A total of 9960 workers have been followed to the end of 1999 representing 267 thousands person-years. By the end of 1999, a total of 29 cases of leukaemia were observed. The standardized mortality ratio is 1.31, 90%CI: 0.88-1.88. In relation to cumulated exposure to radon progeny, no statistical association is seen (p = 0.172), but the trend with duration of underground work is significant (p = 0.015). These contrasting results suggest that the risk of leukaemia is probably related to other components in the underground environment. On the basis of the ICRP respiratory tract and biokinetic models using extrapolation of measurements of long lived radionuclides from uranium dust and external gamma radiation in Czech uranium mines since the 1970s, annual doses to the red bone marrow were estimated in the range 10-30 msv. About 2/3 of the total dose for jobs involving drilling is due to inhalation of uranium and its decay products with ore dust, about 1/4 is due to gamma radiation, and only 5% of the dose is from radon and its progeny. The risk coefficient (excess relative risk per sievert) corresponding to these estimates is ERR/Sv = 7.1 (90%CI: 3.6 13.9). The dose estimates are subject to a considerable uncertainty. Therefore, in the absence of more exact estimates, it is more practical to assess the risk using only duration of exposure modified according to job category. As the red bone marrow estimates for hewers are about two-fold, the modification consists in accounting only half of the year for job categories other than hewer. Using this approach, the ERR per year is 0.15 (90%CI: 0.08 0.27). 1. Introduction The evidence of the risk of leukaemia from ionizing radiation is mainly based on studies of atomic bomb survivors and populations exposed to x rays for diagnostic or therapeutic purposes. There are also an increasing number of occupational studies. In studies of uranium miners, the main objectives were to investigate the risk of lung cancer in relation to exposure to radon progeny and a possible role of modifiers, eg. time since exposure, age, or exposure rate. Mortality from other diseases has been studied in the second step. Results of such analyses have not revealed any significant increase related to radon progeny exposure. Relationship between leukaemia risk and cumulated exposure to radon progeny was not significant in any study of uranium miners, which is not surprising as mortality from leukaemia is about ten times lower than that from lung cancer and therefore numbers of cases in single studies are low. For this limitation, a joint analysis of 11 studies of miners was conducted [1]. The risk of leukaemia (based on 69 cases) was not found significant in the combined study. However, in one study of Czech uranium miners [2], the risk of leukaemia (based on 10 cases) was significantly related to duration of exposure, suggesting that other components in the mining environment other than radon might be involved, eg. long lived radionuclides from uranium dust and external gamma radiation. In contrast to estimates of exposure to radon (measured since 1949) and its progeny, data for other components are scarce. For instance, measurement of long lived alpha and gamma activities begun not early than in the late 1960s. Another issue is related to the estimation of the dose to red bone marrow from all components of ionizing radiation in uranium mines. 2. Methods The cohort study of uranium miners was established by Ševc [3] in 1970 in order to analyze the risk of lung cancer from radon. The details of the studies are given elsewhere. In brief, the cohort involves two groups of uranium miners exposed in two different periods. A total of 4339 miners from the older cohort (S) were employed in the Jáchymov region (West Bohemia) in the period 1948-63 [4]. The other 5621 miners belong to the newer cohort (N) and were employed in the Příbram mines (Central Bohemia) during 1968-88 [5].
Analyses are based on the relative risk (RR) model with a linear dependence on cumulated exposure (W) in the form RR = 1 + b W. (1) The relative risk was separately analyzed in relation to cumulated exposure to radon progeny (WLM 1 ), radon gas (kbq m -3 y), and duration of exposure (y). Due to low numbers, analyses related to leukaemia subtypes and other malignancies are based on simple relative risk model based on hewer equivalent duration in two categories. 3. Results A total of 9960 workers have been followed to the end of 1999 representing 267 thousands personyears. With the exclusion of lung cancer, there were practically no other significant malignancies in this study related to radon exposure. In later periods of follow-up, however, some increase of haematopoetic cancers was observed, including 29 cases of leukaemia, mostly myeoloid and chronic lymphatic (Table I). Table I. Leukaemia subtypes in Czech U miners (1954-1999) ICD-9 a ICD-10 a Leukaemia subtypes Observed % % CZ a 204.0 C91.0 Acute lymphatic 2 7 8 204.1 C91.1 Chronic lymphatic 7 24 24 205 C92 Myeloid 16 55 52 207 C94 Other types 4 14 16 Total 29 100 100 a International Classification of Diseases b Czech population proportions 1997-99 The ratio of the observed number to the nationally expected number is 1.31, 90%CI: 0.88-1.88 and most of cases were observed 20 and more years since first exposure (Table II). The relative risk for cases observed after 20 years since first exposure is 1.9 (90%CI: 1.2-2.8) in comparison to period 0-19 years. Table II. Leukaemia cases by time since first exposure Years since Observed first exposure PY a cases O/E b 95%CI c xx-19 164 053 6 0.90 0.33-1.95 20-xx 102 622 23 1.89 1.20-2.84 Total 266 675 29 1.31 0.88-1.88 a person-years b standardized mortality ratios c confidence interval The mortality from leukaemia mortality was separately analyzed in relation to cumulated exposure to radon progeny, radon gas, and duration of exposure (Table III). Table III. Excess relative risk per unit exposure UE a ERR/UE b 90%CI p Rn progeny WLM 0.0073-0.0154-0.0616 0.698 Rn gas kbqm -3 y 0.0029-0.0020-0.0158 0.432 Modified duration c y 0.15 0.08-0.27 0.013 a UE = unit of exposure b Excess relative risk per unit exposure c years of exposure, non-drilling professions accounted as half year 1 WLM is a unit of cumulated exposure equivalent to one working level for a working month (170 hours) 1WLM = 0.0035 J h m -3 2
When the risk is related to cumulated exposure to radon progeny or radon gas, no statistical association is seen (Table III), but the trend with duration of underground work is significant (p=0.013). These contrasting results suggest that the risk of leukaemia is probably related to other components present in the underground environment. In comparison to estimated doses to the lungs, which are based on extensive radon measurements commencing already in 1949, doses to the red bone marrow can only be estimated with some uncertainty. One such estimate was suggested by Jacobi and Roth [6] in a report on risk of extra-pulmonary cancers from radiation exposure among workers in German uranium mining company. On the basis of the ICRP respiratory tract model [7] and biokinetic model [8], they showed that for miners involved in drilling, the largest contribution to the dose is due to exposure to uranium dust and its progeny (57%), whereas the contribution from radon and its progeny is minor (35%). Extrapolating this indirect dosimetric approach and results of measurements of alpha activities in the dust and external gamma exposure in Czech uranium mines conducted since the 1970s, annual doses to the red bone marrow can be estimated as follows: Table IV: Annual estimated equivalent doses to the red bone marrow 1950-59 1960-69 1970-89 drilling jobs 34 msv 23 msv 12 msv other jobs 10 msv 7 msv 4 msv FIG. 1: Relative risk of leukaemia (RR) in dependence to estimated dose (msv) The risk coefficient (excess relative risk per sievert) corresponding to these estimates in the present study of Czech miners is ERR/Sv=7.1,90%CI: 3.6 13.9 (FIG. 1). The dose estimates, however, are subject to a considerable uncertainty depending among others on dust content, ore mineralization, or type of drilling. Therefore, it is more practical to assess the risk using only duration of exposure that is modified according to job category. As the red bone marrow estimates for hewers are about two-fold, the modification can consist in accounting only half of the year for job categories other than hewer. Due to low numbers of cases, risk for leukaemia subtypes can only be analyzed in a simple dichotomic design (Table V). Table V: Leukaemia subtypes in dependence to modified duration of exposure O O/E 90%CI <10y >10y RR 1 90%CI noncll 22 1.27 0.85-1.83 16 6 3.06 1.56-5.98 CLL 2 7 3.89 1.82-7.30 4 3 4.93 1.91-12.8 all types 29 1.46 1.04-2.00 20 9 3.50 2.02-6.06 1 risk relative to modified duration of exposure <10years 2 chronic lymphatic leukaemia 3
4. Discussion Results on leukaemia risk among uranium miners are scarce. In a report from 11 miner studies [1], 1995), no association to radon exposure (based on 69 cases) was observed. However, the results of follow-up by 1990 among Czech uranium miners suggest some increase with duration of exposure [2]. Generally, the evaluation of leukaemia risk is limited by low numbers of cases. In the male population, leukaemia incidence is lower by one order of magnitude in comparison to lung cancer and is manifested mainly in age groups over 55. Therefore, more evidence can be expected in a longer follow-up or by combining several studies. In radiation induced leukaemia, chronic lymphatic leukaemia (CLL) is usually excluded, as in most studies no radiogenic association was observed for this subtype. In the present study all subtypes vere combined. However, the relative risk in the simple dichotomic model was higher for CLL than for other subtypes combined, although the difference between the relative risks was not significant. Although the assessment of the leukaemia risk is based on an approximate estimation of dose, the estimated value of excess relative risk per Sv is consistent with findings of the LSS study, where the risk coefficient in linear model below 4Sv is 8.0 per 1 Sv. However, the evaluation of occupational exposure should be based on more objective data, eg. duration of exposure. As there are differences in dose estimates for drilling and other activities, duration of exposure should be modified according to the job. In the absence of detailed data, the estimation of 50% for other than hewer activities is an appropriate simplification. Due to estimated retention of radionuclides from uranium dust in bones and red bone marrow, it was also analyzed if other relevant malignancies were elevated in the cohort of uranium miners. Due to low number of cases, we compared only modified duration of exposure in categories below and over 5 years. The relative risk of non-hodgkin lymphoma (16 cases) was elevated (RR=1.3, 90%CI: 0.6-3.0), but not significantly, and the relative risk of bone cancer (5 cases) was similar (RR=2.4, 90%CI: 0.5-10.8). These findings, although based on small numbers and simple models, seem to be in line with leukaemia risk. Risk of leukaemia solely from radon was not observed in the present study of uranium miners, despite considerably high concentrations in mines in early years. As estimated by Kendall [9], doses to the red bone marrow from radon gas are higher in comparison to doses from radon progeny. In view of this differences, we analyzed the risk of leukaemia in one Czech cohort study of 12000 inhabitants exposed to high concentrations of radon [10]. The relative risk based on 27 cases of leukaemia was elevated RR=1.4 (90%CI: 0.7-2.8) for residents in exposure category over 400Bq/m 3 in comparison to those living below 400Bq/m 3. These findings, however, need to be confirmed by further observations. 5. Conclusions Results of leukaemia study in a cohort of nearly 10 000 uranium miners based on 29 observed cases show significant association to chronic exposure. The largest contribution is from the uranium dust. Although there are estimates to the red bone marrow, it seems to be more practical to use a model based on modified duration of exposure, where non-hewer exposures are accounted as 50%. The evaluation of leukaemia subtypes is limited by low numbers. More accurate results can be expected in an extended follow-up or by pooling several studies. Acknowledements The present research was supported by the Internal Grant Agency of the Czech Ministry of Health (IGA NJ/6768-3). 4
References 1. Darby, S. C., Whitley, E., Howe, G. R., Hutchings, S. J., Kusiak, R. A., Lubin, J. H., Morrison, H. I., Tirmarche, M., Tomášek, L., Radford, E. P., Roscoe, R. J., Samet, J. M., Yao, S. X., Radon and cancers other than lung cancer in underground miners: a collaborative analysis of 11 studies. J. Natl. Cancer Inst., 87:378-384, (1995). 2. Tomášek, L., Darby, S. C., Swerdlow, A. J., Plaček, V., Kunz, E., Radon exposure and cancers other than lung cancer among uranium miners in West Bohemia. Lancet, 341:919-923, (1993). 3. Ševc, J., Plaček, V., Jeřábek, J., in Proceedings of the 4th Conference on Radiation Hygiene, Jasná pod Chopkom, 1971, edited by T. Trnovec and P. Gaál (Purkyně Medical Research and Postgraduate Institute Press, Hradec Králové, 1971) Part II, p. 315-326. 4. Tomášek, L., Darby, S. C., Fearn, T., Swerdlow, A. J., Plaček, V., Kunz, E., Pattern of lung cancer mortality among uranium miners in West Bohemia with varying rates of exposure to radon and its progeny. Radiat. Res., 137:251-261, (1994). 5. Plaček, V., Tomášek, L., Heribanová, A., Kunz, E., Lung cancer and exposure to radon under present mining conditions. Pracov. Lék., 49:14-20, (1997) (in Czech). 6. Jacobi, W., Roth, P., Risk and probability of causation of non-lung cancers from occuppational exposure among employees of the former Wismut company, GSF Report 4/95, Institut für Strahlenschutz, GSF, Neuherberg (1995) (in German). 7. International Commission on Radiological Protection, Human respiratory tract model. Publication 66. Annals of the ICRP, 24, No. 1-4, Pergamon Press, Oxford and New York (1994). 8. International Commission on Radiological Protection, Dose coefficients for intakes of radionuclides by workers. Publication 68. Annals of the ICRP, 24, No. 4, Pergamon Press, Oxford and New York (1995). 9. Kendall, G. M., Smith, T. J., Doses to organs and tissues from radon and its decay products. J. Radiol. Prot., 22:389-406, (2002). 10. Tomášek, L., Müller, T., Kunz, E., Heribanová, A., Matzner, J., Plaček, V., Burian, I., Holeček, J., Study of lung cancer and residential radon in the Czech Republic. Centr. Eur. J. Publ. Health, 9:150-153, (2001). 5