Leukaemia, Chernobyl and Epidemiology

Publisher Website
Google Scholar
Abstract
International Epidemiology Institute, 1550 Research Blvd, Suite 200, Rockville, MD 20850-3127, USA Chernobyl is located in Ukraine, 20 km south of Belarus and near the Russian border. It will be linked indelibly in people's minds as the site of the world's worst nuclear accident in April 1986. Thirty firemen who quelled the smouldering fires of the burning reactor died shortly after excessive exposure to lethal radiation, over 100 000 citizens living in the surrounding area were evacuated, vast areas in the Ukraine, Russia and Belarus were contaminated with radioactive debris spewing from the blazing nuclear core, and hundreds of thousands of workers from all over the former Soviet Union were sent to clean up the contaminated environment and to entomb the extinguished nuclear reactor in a `sarcophagus' (UNSCEAR 1988, NEA 1996, Karaoglou et al 1996, IAEA 1996). The only late health affect ascribed to the reactor accident is a remarkable increase in the incidence (but not mortality) of childhood thyroid cancer among surrounding populations heavily exposed to radioactive iodines, most notably I-131 (half-life a bit over 8 days). The radioactive iodines concentrate to relatively high levels in the thyroid gland after consumption of foodstuffs, but primarily after drinking milk. Childhood leukaemia, however, has not occurred in excess in the most heavily contaminated areas in Belarus (Ivanov et al 1993), Europe (Parkin et al 1996), or Scandinavia (Auvinen et al 1994, Hjalmars et al 1994). The whole body doses from caesium-137 and other radionuclides were apparently much too low to result in a detectable increase over that expected from natural causes (Linet and Boice 1993, Boice and Linet 1994). Until recently, little information has been published on the possible health effects among the clean-up workers, who have also been called liquidators, emergency workers, accident recovery workers. Those sent in 1986 - 87 to within the 30 km zone around the damaged reactor received the highest exposures and number about 300 000. The workers were mainly male, 20 - 39 years of age, and received on average 100 mGy. Few, perhaps only 4%, received over 250 mGy. Biological dosimetry studies of large numbers of Chernobyl clean-up workers from the Baltic countries generally support the physical dose estimates on a population basis (Bigbee et al 1996, 1997, Inskip et al 1997a). Based on extrapolation from high dose atomic bomb survivor data, the relative risk (observed/expected) of cancer mortality from a 100 mGy brief whole body dose would be of the order of 1.05 (UNSCEAR 1994, Pierce et al 1996). The ability for epidemiological studies to detect a risk of cancer following such low exposures is accordingly low and extremely problematic because it would require very large numbers followed carefully for perhaps 50 years (Boice 1997). For leukaemia, however, the situation is different because the risk coefficient is much higher than for solid tumours and any radiation-related excess would be anticipated to occur much earlier and primarily within 10 years of exposure. The relative risk of leukaemia following 100 mGy might be about 1.2 - 1.4 (Preston et al 1994, UNSCEAR 1994). However, because most of the excess, perhaps 80%, would be anticipated within 10 years, the relative risk during the first 10 years of follow up might be of the order of 3 - 4 (Cardis et al 1996), and certainly within the capability of epidemiology to observe. In fact, the predicted excess of leukaemia (excluding chronic lymphocytic leukaemia (CLL) which has never been linked to radiation in any population) occurring within 10 years among 200 000 clean-up workers exposed to 100 mGy might be of the order of 100 - 150 for 100 mGy, compared with a background of about 40 occurring `naturally' (Cardis et al 1996). To date, however, an `epidemic' of leukaemia of this magnitude has not been reported in any series of Chernobyl clean-up workers (Rahu et al 1997, Karaoglou et al 1996, IAEA 1996). The new report by Ivanov and colleagues (this issue) on the 155 000 Russian liquidators adds new understanding to the study of radiogenic leukaemia in the former Soviet Union and new caution in interpreting such studies. While the data are the same as in an earlier publication by the same group (Ivanov et al 1997), the interpretation is diametrically different. In the previous cohort analysis, a significant risk of leukaemia was associated with the mean radiation dose received (excess relative risk, ERR/Gy of 4.3; 95% CI 0.8 - 7.8) which was surprisingly close to that estimated among atomic bomb survivors (Preston et al 1994, UNSCEAR 1994). However, general population rates of cancer incidence in Russia were used as the comparison and it was not possible to separate out CLL. This estimate of risk is now seen to have been terribly biased and very misleading. The new case-control analysis within the same cohort of liquidators takes estimated dose into account, assumes a two-year minimum latency for radiogenic leukaemia, and excludes CLL. There was no significant correlation made between leukaemia and any measure of exposure. The erroneous risk estimate in the earlier investigation was because of ascertainment bias, using an inappropriate comparison population, and including a leukaemia subtype that has never been linked to radiation in any human population. Chernobyl clean-up workers receive annual examinations and thus leukaemias are readily detected and recorded. Perhaps even too much, in that 50% of the initial diagnoses of leukaemia were not confirmed (Cardis et al 1996). On the other hand, registration of malignancy in Russia is not complete, and for some unknown reason the proportion of CLL among liquidators is much higher than in the general population of persons of similar age (Parkin et al 1992). Thus the biased radiation risk coefficient was due to increased surveillance among the workers that resulted in relatively complete (or more complete) registration of leukaemias; underreporting of leukaemias in the general population that resulted in an underestimate of leukaemia; and the peculiar excess of CLL among workers which may be due to chance, surveillance, misdiagnoses, or non-radiation factors unique to workers sent to Chernobyl from Russia. Interpretation of these early epidemiological studies of Chernobyl workers must be made with care. Despite the relatively straightforward design and defensible assumptions and analyses, results may nonetheless be in error due to bias and confounding associated with differences in surveillance, completeness of follow up, choice of a comparison group, and choice of leukaemia subtypes to include in the analysis. The new analysis improves greatly on the previous one by making internal comparisons based on the nested case-control design. To eliminate the bias associated with noncomparable comparison with the general population, Ivanov et al (this issue) matched each leukaemia case with a liquidator control with similar characteristics and then evaluated whether the exposures differed and whether risk varied by categories of dose. It did not. At first glance, the 48 leukaemias appear excessive based upon a probable background rate of 3 - 4 non-CLL/100 000 per year anticipated from non-Chernobyl factors for men of these ages. But 14 leukaemias occurred within 2 years and are unlikely to be due to radiation and 10 were CLL which is not radiation related. Thus only 24 leukaemias were available for analysis in years 2 - 8 following the Chernobyl accident. In the absence of any radiation effect, the background number expected would be of the order of 18 - 27 (155 000 workers × 6 yr × (3 - 4)/100 000/yr). However, the data cannot exclude the possibility that radiation has not contributed to a slight increase in risk. The analysis of these 24 leukaemia cases and matched controls did suggest a slight trend with dose, although the trend was not statistically significant and the range of possible risks was broad (ERR/Gy = 1.67; 95% CI 5.9 - 9.2). While there are uncertainties in predicting what the radiation excess might be during these 8 years, the best estimates, without including a dose and dose rate effectiveness factor (DRREF), are of the order of a 3-fold excess or 60 - 80 cases (Cardis et al 1996). Clearly, the observed total number of 24 non-CLL is not consistent with such an excess over background predicted based on models from high dose and dose rate exposures and not applying a DRREF (Cardis et al 1996). The Russian data are also consistent with a small but comprehensive study of Estonia clean-up workers recently published (Tekkel et al 1997, Rahu et al 1997). A nearly complete roster of 4800 workers was compiled from a variety of sources, including military lists. Over 80% had recorded doses and the mean dose was essentially the same as that seen in Russia, 100 mGy. Comprehensive biodosimetry using flow cytometry and the glycophorin A mutational assay of red blood cells, and chromosome aberration analyses using fluorescent in situ hybridisation techniques, confirmed that high exposures were infrequent and unlikely, and that the mean population dose might adequately characterise the exposure, although an even lower mean dose than 100 mGy is possible (Bigbee et al 1996, 1997, Inskip et al 1997a). A thyroid screening study of 2500 workers with ultrasound and fine needle aspiration revealed no increase in thyroid nodular disease or cancer that could be related to dose (Inskip et al 1997a). Cause specific mortality was determined through linkage with national death registry data and cause specific cancer incidence was determined through linkage with national cancer registry data. No leukaemias were reported against only 1 expected based on population rates, and relative risks higher than about 3 could be excluded with 95% confidence. The only significant finding was a 50% increase in death due to suicide. Future analyses are planned with the 12 000 workers from the other two Baltic countries, Latvia and Lithuania (Inskip et al 1997b). What then can be concluded about the risk of cancer among Chernobyl clean-up workers some 10 years after the accident? (1) No excess of leukaemia has been demonstrated to date. No excess of thyroid cancer or other solid cancers has been linked to Chernobyl radiation among clean-up workers. (2) It is likely that few workers received high doses (over 250 - 500 mGy). Thus large populations will need to be studied epidemiologically to discern convincingly any radiation effect for leukaemia. The likelihood of detecting a risk of solid cancers, which would not be anticipated to start until 10 years after exposure, is exceedingly small if it holds true that leukaemia is not measurably increased among liquidators. This excludes, however, the possibility that ingestion or inhalation of significant amounts of fission fragments occurred. (3) There is the need, then, for additional study of the leukaemia experience to date, including combined series from Russia, Ukraine, Belarus and the Baltics if at all possible. Care will have to be taken with regard to dose estimation, cancer surveillance among the liquidators, underreporting among the general population, excluding the first 2 years of observation, and excluding CLL from the analyses. (4) The risk of leukaemia predicted from models based on studies of high dose and high dose rate exposures (atomic bomb survivors and medical series) are not consistent, it appears, with the observed numbers of leukaemias seen among liquidators. Why might this be so? (a) The Chernobyl doses may be lower than estimated. (b) Leukaemia cases among liquidators might have been missed. (c) The risk model may be incorrect (a DRREF of 2 may be appropriate and the time response may differ following low dose and low dose rate exposures). (d) Risk from low doses delivered at low rates may be much less than predicted from higher doses at higher dose rates, possibly because of the opportunity for cellular damage to be repaired. [It may be interesting to debate whether sending 300 000 persons to Chernobyl (mean, 100 mGy) will eventually result in fewer leukaemias and cancers than if 30 000 persons had been sent (with 1000 mGy, mean)]. (e) The time for radiation-induced leukaemias to develop after exposures in 1986 - 87 might be much longer than suggested from studies of populations exposed to high doses delivered at high dose rates. Observations to date have several implications if borne out. A low excess of leukaemia implies that the possible excess of solid tumours later in life will be very low and outside the realm of epidemiological detection. Thus future studies should aim to combine series to have any hope of detecting a leukaemia effect, and focus should be on the groups likely to have received the highest doses, such as the 4% estimated to have received > 250 mGy, the 5000 soldiers who were on the roof shortly after the accident, and the 1200 helicopter pilots dumping material on the reactor to contain the fire. Further, other accident-related, rather than radiation-related, problems may deserve more attention, in particular psychological effects. For example, the only significant finding among Estonian clean-up workers was a 50% increase in suicides (Rahu et al 1997), indicating, perhaps, the severe stress associated with the accident experience. Finally, with the publication of sound epidemiologic studies on the range of risks, it appears that the health consequences may be much less than perceived by most of the world's population. Other than thyroid cancer in children, where the numbers of incident cases are large but the number of deaths is apparently small, not all and perhaps only a few cancers and leukaemias might eventually be attributed to the radiation received during the clean-up activities among the many hundreds of thousands workers involved in civilisation's most remarkable environmental decontamination undertaking. Acknowledgment The author wishes to acknowledge helpful discussions with Drs Dale Preston and Kiyo Mabuchi of the Radiation Effects Research Foundation, Hiroshima, Japan. References Auvinen A, Hakama M, Arvela H, Hakulinen T, Rahola T, Suomela M, Soderman B and Rytömaa T 1994 Fallout from Chernobyl and incidence of childhood leukaemia in Finland 1976 - 92 Br. Med. J. 309 151 - 7 Bigbee W L, Jensen R H, Veidebaum T, Tekkel M, Rahu M, Stengrevics A, Auvinen A, Hakulinen T, Servomaa K, Rytömaa T, Obrams G I and Boice J D Jr 1997 Biodosimetry of Chernobyl clean-up workers from Estonia and Latvia using the glycophorin A in vivo somatic cell mutation assay Radiat. Res. 147 215 - 24 Bigbee W L, Jensen R H, Rahu M, Tekkel M, Veidebaum T, Stengrevics A, Auvinen A, Hakulinen T, Obrams I and Boice J D Jr 1996 Glycophorin A biodosimetry in Chernobyl clean-up workers from the Baltic countries Br. Med. J. 312 1078 - 9 Boice J D Jr (ed) 1997 Radiation epidemiology: past and present Implications of New Data on Radiation Cancer Risk vol 18 (Bethesda, MD: NCRP) pp 7 - 28 Boice J D Jr and Linet M 1994 Chernobyl, childhood cancer and chromosome 21 (Editorial) Br. Med. J. 309 139 - 40 Cardis E, Anspaugh L, Ivanov V K, Likhtarev I A, Mabuchi K, Okeanov A E and Prisyazhniuk A E 1996 Estimated long term health effects of the Chernobyl accident Proc. Int. Conf. on One Decade After Chernobyl: Summing Up the Consequences of the Accident STI/PUB/1001 (Vienna: IAEA) pp 241 - 79 Hjalmars U, Kulldorff M and Gustafsson G, for the Swedish Child Leukaemia Group 1994 Fallout from the Chernobyl accident and risk of acute childhood leukaemia in Sweden Br. Med. J. 309 154 - 7 Inskip P D, Hartshorne M F, Tekkel M, Rahu M, Veidebaum T, Auvinen A, Crooks L A, Littlefield L G, McFee A F, Salomaa S et al 1997a Thyroid nodularity and cancer among Chernobyl clean-up workers from Estonia Radiat. Res. 147 225 - 35 Inskip P D, Tekkel M, Rahu M, Veidebaum T, Hakulinen T, Auvinen A, Rytömaa T, Servomaa K, Obrams G I, Stengrevics A et al 1997b Studies of leukaemia and thyroid disease among Chernobyl clean-up workers from the Baltics NCRP Proc. 18 123 - 41 International Atomic Energy Agency, European Commission and World Health Organization 1996 Proc. Int. Conf. on One Decade After Chernobyl: Summing Up the Consequences of the Accident STI/PUB/1001 (Vienna: IAEA) Ivanov E P, Tolochko G, Lazarev V S and Shuvaeva L 1993 Childhood leukaemia after Chernobyl Nature 365 702 Ivanov V K, Tsyb A F, Gorsky A L, Maksyutov M A, Rastopehin E M, Konogorov A P, Korelo A M, Biryukov A P and Matyash V A 1997 Leukaemia and thyroid cancer in emergency workers of the Chernobyl accident: estimation of radiation risks (1986 - 1995) Radiat. Environ. Biophys. 36 9 - 16 Ivanov V K, Tsyb A F, Konogorov A P, Rastopchin E M and Khait S E 1997 Case-control analysis of leukaemia among Chernobyl accident emergency workers residing in the Russian Federation 1986 - 1993 J. Radiol. Prot. 17 137 Karaoglou A et al (eds) 1996 The Radiological Consequences of the Chernobyl Accident EUR 16544EN (Brussels: European Commission) Linet M and Boice J D Jr 1993 Radiation from Chernobyl and risk of childhood leukaemia (Commentary) Eur. J. Cancer 29 1 - 3 NEA 1996 Committee on Radiation Protection and Public Health Chernobyl - Ten Years On: Radiological and Health Impact (Paris: OECD) Parkin D M, Clayton D, Black R J, Masuyer E, Friedl H P, Ivanov E, Sinnaeve J, Tzvetansky C G, Geryk E, Storm H H et al 1996 Childhood leukaemia in Europe after Chernobyl 5-year follow-up Br. J. Cancer 73 1006 - 12 Parkin D M et al 1992 Cancer Incidence in Five Continents vol VI IARC Publ No 120 (Lyon: IARC) Pierce D A, Shimizu Y, Preston D L, Vaeth M and Mabuchi K 1996 Studies of the mortality of A-bomb survivors. Report 12, Part 1. Cancer 1950 - 1990 Radiat. Res. 146 1 - 27 Preston D L, Kusumi S, Tomonaga M, Izumi S, Ron E, Kuramoto A, Kamada N, Dohy H, Matsuo T et al 1994 Cancer incidence in atomic bomb survivors. Part III: leukaemia, lymphoma, and multiple myeloma 1950 - 87 Radiat. Res. 137 S68 - S97 Rahu M, Tekkel M, Veidebaum T, Pukkala E, Hakulinen T, Auvinen A, Rytömaa T, Inskip P D and Boice J D Jr 1997 The Estonian study of Chernobyl clean-up workers: III. Incidence of cancer and mortality Radiat. Res. 147 653 - 7 Tekkel M, Rahu M, Veidebaum T, Hakulinen T, Auvinen A, Rytömaa T, Inskip P D and Boice J D Jr 1997 Estonian study of Chernobyl clean-up workers: design and questionnaire data. Radiat. Res. 147 641 - 52 United Nations Scientific Committee on the Effects of Atomic Radiation 1988 Sources, Effects and Risks of Ionizing Radiation 1988 Report to the General Assembly, with annexes. Publ E.88.IX.7 (New York: United Nations) United Nations Scientific Committee on the Effects of Atomic Radiation 1994 Sources and Effects of Ionizing Radiation 1994 Report to the General Assembly, with scientific annexes. Publ E.94.IX.11 (New York: United Nations)