Publications

Documents of the NRPB: Volume 14 , No. 1

Risk of Leukaemia and Related Malignancies following Radiation Exposure: Estimates for the UK Population: Report of an Advisory Group on Ionising Radiation


Author(s):

n/a

Summary:

This report reflects the understanding and evaluation of the current scientific evidence as presented and referenced in the full report.

The summary and conclusions and the recommendations for research of the report are given below. The membership of the Subgroup on Epidemiology responsible for preparing the report is provided in the PDF of the report.

Price:

£25.75
ISBN 0-85951-518-4

Order

PDF download:

The full document
(PDF, 7.04 MB, 123 pages)

Documents of the NRPB 14/1 (PDF, 7 MB)
     

Please note that the paragraph numbering from the full published document has been retained in the extracts below.

Summary and conclusions

  1. It has been known for many years that exposure to ionising radiation can lead to the development of leukaemia. Information from epidemiological studies of groups exposed to radiation has been reviewed on a number of occasions by national and international bodies, and this report updates earlier evaluations by NRPB in 1988 and 1993. In addition, data on related malignancies, namely non-Hodgkin lymphoma (NHL), Hodgkin disease and multiple myeloma, have been reviewed here.
  2. Haematological malignancies have common cell lineages that link myeloid and lymphoid diseases with different clinical conditions. Individual malignancies require careful definition, and some - such as NHL - comprise different entities that can be difficult to separate. Changes in disease classification over the past few decades may have affected some epidemiological findings. Nevertheless, even with the use of fairly crude definitions, there are differences between these conditions, both in terms of age and time trends, and in factors that affect the risk of the disease developing. Some of these conditions, such as acute myeloid leukaemia (AML), chronic myeloid leukaemia (CML), chronic lymphocytic leukaemia (CLL) and multiple myeloma, are much more common at older than at younger ages. In contrast, acute lymphoblastic leukaemia (ALL) occurs mainly in childhood. The incidence of some of these diseases has varied over time, in part due to diagnostic changes - for example, declines in AML and CML have been mirrored by an increase in pre-leukaemic states such as myelodysplasia. Also, there are wide variations throughout the world in the incidence of most of these conditions. Many potential risk factors have been investigated in epidemiological studies. However, apart from ionising radiation, only a few of these risk factors have been established, eg benzene for myeloid leukaemia and chronic immunosuppression for NHL.
  3. Many epidemiological studies have been conducted on leukaemia and related diseases in groups exposed to ionising radiation. However, differences in design mean that these studies are not equally informative. In order to make substantive inferences from a study, the potential for bias or confounding should be low and the statistical power should be high. In particular, findings from cohort and case-control studies are generally more reliable than those from geographical correlation studies. A review of the strengths and limitations of various studies indicates that particular weight should be placed on the Life Span Study (LSS) of Japanese atomic-bomb survivors. This cohort study involves a population of all ages and both sexes who received a wide range of whole-body doses, which have been estimated on an individual basis. Furthermore, as well as a mortality follow-up, data on the incidence of leukaemia and other haematopoietic diseases in the atomic-bomb survivors have been collected over several decades. Studies of medical irradiation can also be informative, although they can be more restricted than the atomic-bomb survivors in the ranges of ages and doses, and individual organ-specific doses are not always available.
  4. There is a substantial amount of information on leukaemia risks following radiation exposure. This reflects the high relative increase in risk for leukaemia compared with that for other cancer types and the temporal pattern in risk, with many of the excess leukaemias occurring within about the first two decades following exposure, particularly among those irradiated at young ages. There are some differences between the Japanese atomic-bomb survivors and some large studies of medically exposed groups in estimates of both the magnitude of the radiation risk and the shape of the dose-response for external low LET exposure, with the atomic-bomb data tending to show higher risks. These differences might be explained by variation in the uniformity of exposure to the bone marrow and in the temporal pattern of exposure, as well as by differences in the pattern of risk among leukaemia subtypes. In contrast to most solid cancers, there is clear evidence of non-linearity in the dose-response for leukaemia, with a slope at lower doses that is about half of that at higher doses. Analyses based on large groups of radiation workers have suggested a raised leukaemia risk, and are compatible with risk estimates ranging from about zero to around four times the risk estimated at low doses from the Japanese atomic-bomb survivors. Studies of in utero irradiation from prenatal x-rays have shown a raised risk of childhood leukaemia in childhood, although there is some uncertainty in the associated risk estimate, which is likely to be at least a factor of two to three.
  5. Studies on internal exposures to low LET radiation do not indicate raised risks of leukaemia, although these data are sparse and the statistical precision of the findings is low, in part because of the generally small doses. Whilst comparisons are not always straightforward, data from - for example - a study of childhood leukaemia and weapons fallout in the Nordic countries appear to be consistent with those from the Japanese atomic-bomb survivors. There is currently no convincing evidence of an increased risk of leukaemia due to environmental exposures associated with the Chernobyl nuclear plant accident, although investigations are continuing. Dose-related increases in leukaemia risk have been seen among patients with large exposures to high LET radiation arising from injections of Thorotrast, a diagnostic x-ray contrast medium. It is difficult to compare leukaemia risks in these high LET studies with those in the Japanese atomic-bomb study, but the findings are likely to be consistent with existing radiation weighting factors for high LET radiation, as well as with a range of other values. There is less evidence for raised risks among patients injected with 224 Ra, and studies with individual assessments of radon exposure, either in mines or in homes, have generally not shown increased risks. Small numbers limit the interpretation of studies of plutonium workers.
  6. In contrast to leukaemia (other than CLL), epidemiological studies of NHL, Hodgkin disease and multiple myeloma do not point strongly to a link with ionising radiation. For many studies the small numbers of cases and absence of dose-response analyses preclude detailed inferences, particularly in the case of internal low or high LET exposures. However, data from the Japanese atomic-bomb survivors and from the most informative medical studies do not indicate clear associations with radiation exposure. The epidemiological findings are mixed in some instances - for example, the indications of raised levels of mortality from multiple myeloma, but not myeloma incidence in the atomic-bomb survivors and other groups. Whilst the interpretation of these findings is not always clear, overall the evidence for associations between radiation and these diseases appears to be weak.
  7. Whilst it is not possible to derive quantitative estimates of any radiation-related risk for NHL, Hodgkin disease or multiple myeloma, leukaemia risk estimates specific to a UK population have been derived in this report. These are based on relative risk models fitted to data on leukaemia mortality and incidence in the Japanese atomic-bomb survivors, which have been applied to baseline rates in the UK. For mortality from leukaemia (other than CLL), the lifetime risk in a UK population of all ages and both sexes following low dose exposure is estimated to be about 0.5-0.6% Sv -1. This is consistent with the risk factor of 0.5% Sv -1 derived in the UNSCEAR 2000 report, and the value of 0.6% Sv -1 obtained in the previous evaluation by NRPB in 1993. For leukaemia incidence, the corresponding lifetime risk estimate following low dose exposure is estimated to be about 0.6-0.7% Sv -1, reflecting in part the relatively high baseline rates and survival for ALL in childhood. The uncertainty in these risk estimates may be of the order of a factor of two, higher or lower. Furthermore, the uncertainty in leukaemia risks for exposure at specific ages would be somewhat greater than that for a population of all ages.
  8. Some other issues concerning ionising radiation and leukaemia risk have been considered in this report. Suggestions that doses to the bone marrow may be increased in the vicinity of power frequency electromagnetic fields, as a consequence of enhanced deposition of radon decay products, are not borne out by dosimetric calculations. Furthermore, epidemiological studies have not tended to find raised childhood leukaemia risks with increased levels of radon in homes or with greater proximity to power lines. Various studies have examined the risks of childhood cancer in relation to preconception radiation. Overall, these studies provide no evidence to suggest that preconception irradiation alone is associated with an increased risk of leukaemia or other childhood cancers.

Recommendations for research

  1. Data on the risks of leukaemia and related diseases from studies of populations exposed to high doses and high dose rates are largely complete. Although findings from further follow-up of these groups will appear from time to time, it is unlikely that these updated results will differ substantially from those currently available. However, re-analyses or combined analyses of these data may be useful, particularly in the development of mechanistic models for radiation-induced leukaemia.
  2. It is important to obtain further data from existing studies of groups exposed to low doses and low dose rates, as a check on the methods currently employed to extrapolate findings at high doses and high dose rates to low doses or low dose rates. Studies of large cohorts of workers with individual dose information (eg the National Registry for Radiation Workers) are likely to be particularly valuable, and continued follow-up and possible expansion of such cohorts is desirable. Pooled analyses of such studies may also be helpful.
  3. Work to obtain a better understanding of the effects of in utero irradiation would be desirable.
  4. Some of the inconsistency in the findings for non-leukaemia conditions might be due to radiogenic subtypes that are not yet defined. Re-analyses of these data would be useful once diagnostic advances are made.
  5. Work underway should be noted. For example, separate projects are in progress to combine data from various countries on cancer among radiation workers in the nuclear industry and among aircraft crew. These combined analyses will provide statistically more powerful information than can be obtained from any one of the contributing studies.
     
    ACKNOWLEDGEMENTS
     
    This report makes use of data obtained from the Radiation Effects Research Foundation (RERF) in Hiroshima, Japan. RERF is a private foundation, funded equally by the Japanese Ministry of Health and Welfare and the US Department of Energy through the US National Academy of Sciences. The conclusions in this report do not necessarily reflect the scientific judgements of RERF or its funding agencies.



Last reviewed: 1 September 2008