The committee considered two aspects of infant leukaemia at the third meeting: its biological origin and uncertainties in current radiation ‘risk factors’ for the disease.


Infant leukaemia is very rare (there are only about 32 cases per million live births) and it has only become possible to study its origin in detail in the last few years. It has recently been discovered that in about 80% of cases of infant leukaemia a specific rearrangement of the so-called MLL gene has occurred. Such rearrangements are known to be linked to exposure to particular chemotherapy drugs and to bioflavenoids. This has led to a view that many of these 80% of infant leukaemias may be associated with the mother’s intakes of substances present in pesticides and her intakes of other chemicals found in food and the environment. It appears that external radiation is very unlikely to cause these MLL gene rearrangements, so there is a view that it is unlikely to be implicated to a large extent in the majority of infant leukaemias. Little is known about the origins of the approximately 20% of infant leukaemias that are not linked to MLL gene rearrangements and it is not known how important radiation may be in these cases.

There have been no studies on whether internal irradiation causes MLL gene rearrangements and there is little information on possible synergistic effects of radiation and chemicals on MLL and other genes. Current understanding of molecular mechanisms is not adequate to explain any excesses of infant leukaemias found by epidemiological studies, nor to allow confident determinations of whether such excesses were or were not caused by internal irradiation.


In this context, ‘risk factor’ means an estimate of the probability that radiation exposure of the parents will lead to infant leukaemia. The question considered by the committee was whether the reported observations of excess numbers of infant leukaemias following the Chernobyl accident (see summary of second CERRIE meeting) are consistent with the uncertainties inherent in current risk factors for the disease.

In a paper for the committee it was estimated that the current risk factors for infant leukaemia are uncertain by a factor of about 2-5 either way. A risk factor five times higher than that currently used would be consistent with the reported excess numbers of infant leukaemias in Belarus, but not with the numbers reported in Greece, west Germany, Scotland and Wales, or the US. Risk factors would have to be hundreds of times higher than those currently used in order to be consistent with the reported numbers of infant leukaemias in these latter countries. The position is therefore that either the reported post-Chernobyl excess infant leukaemias were not caused by radiation, or the current risk factors are substantially in error, or these studies based on small numbers have led to misleading generalisations.


Discussion of this topic focused on an epidemiological study published in 1992 on childhood leukaemia in Nordic countries and its relationship to fallout from the atmospheric nuclear weapons tests carried out in the 1950s and 1960s. The study derived risk factors for childhood leukaemia (for exposure of children after birth and their parents prior to birth) that are similar to those currently used. Questions were raised about aspects of the methodology used in the study and about the estimated doses that had been used in deriving the risk factors. It was felt that the uncertainties in these risk factors might be considerably higher than the factor of five (either way) quoted in the study (which was based on statistical precision only). The committee will return to the topic of childhood leukaemia and weapons test fallout at its next meeting, when epidemiological studies for other countries will be considered.


The second event theory is about radiation effects at the cellular level. The theory is that a first radiation ‘hit’ pushes a cell from a quiescent state to a much more sensitive state and the second radiation hit then causes damage. The two hits have to take place within about 2-10 hours of each other for the cell to be in a sensitive state when the second hit occurs. Existing models implicitly assume that, at low doses and dose rates, each cell is hit only once (or that subsequent hits act independently) and that some hits are to cells in a sensitive state and some to cells in a quiescent state. The second event theory predicts that some specific types of internal emitters cause greater biological damage than is taken into account in existing models. These emitters include parent/daughter radionuclide pairs such as strontium-90 and yttrium-90, and tellurium-132 and iodine-132.

Published calculations give different estimates of the factor by which the damage from strontium-90/yttrium-90 is predicted to be greater than that caused by the same dose of external gamma radiation. One set of calculations indicates a factor of about 30 and another a factor of about 1.3. Calculations for the committee gave a factor of about 1.8 for tellurium-132/iodine-132, using the same assumptions as in obtaining the factor of 1.3 for strontium-90/yttrium-90. It seems unlikely that further calculations will resolve these differences of view about the importance of second events. It is possible that there are published data from experiments carried out before the second event theory was postulated that could shed light on the matter. The committee will therefore return to this topic after further examination of the existing literature.

There are three other aspects of the second event theory on which the committee held initial discussions and to which it will be returning. One is the possibility that radioactive particles, especially those containing alpha emitters such as plutonium-239, could cause more damage than is currently thought because of second events. Another is the prediction inherent in the second event theory that within any population of cells there is, at any one time, a sub-population that is very much more sensitive than the rest. The third is the topic of ‘bystander effects’, that is the possibility that cells adjacent to the one receiving the radiation hit are affected by it although they are not hit themselves. If bystander effects are important then the first and second hits may not need to be in the same cell in order to lead to greater damage to tissue than single events.


In existing models for the metabolism of radioactive particles it is assumed that the dominant mechanism for transfer of radionuclides across the placenta to the foetus is dissolution of the particles in the lung or gut and movement of the dissolved radionuclides in blood. Insoluble inhaled particles are assumed to move from the lung to the tracheobronchial lymph nodes, where most remain. Most ingested insoluble particles are assumed to be excreted. Any particles that do enter the mother’s blood are assumed to be removed in her liver, spleen or bone marrow, or to be prevented from reaching the foetus by the placenta or yolk sac. Each of these modelling assumptions is supported by human and animal data.

It has been suggested that small particles (those of diameter a few tens of nanometres or less) that do enter the mother’s blood could be transferred across the placenta to the foetus. The data in the existing literature seem to show that transfer of particles across the placenta decreases with increasing particle size but do not rule out some transfer of small particles. Further data are expected to be published in the near future. The committee will return to this topic in due course, and also consider whether particles reaching the foetus could cause more damage than radionuclides in soluble form.


Uncertainties in risk factors for external radiation are important because current risk factors for internal radiation are, to a large extent, derived from them. Recent analyses have shown that the uncertainty in the present estimates of overall fatal cancer risk from external exposures to low LET radiation (ie gamma, beta and X-rays) is of the order of a factor of two to three either way. The main source of this uncertainty is in the value used for the ‘dose and dose rate effectiveness factor’ (DDREF), which is used to extrapolate from the acute radiation doses received by populations such as the Hiroshima and Nagasaki survivors to the prolonged, low doses that are received in most exposure situations today. Some experts are of the view that no DDREF should be used for many cancers, which would lead to an overall external radiation risk factor towards the top of the current range. The uncertainty range of a factor of two might not take full account of phenomena such as bystander effects and genomic instability (which it has been suggested could lead to higher risks), nor of suggestions that low doses of radiation may have beneficial effects on health (‘hormesis’). The committee will use the current external radiation risk factors and uncertainty ranges as a baseline for further examination of internal radiation risks.

Next (Meeting 4)