Biological consequences of exposure of cells to ionising radiation include cell death, gene mutations, chromosome aberrations and neoplastic transformations. These effects have been conventionally attributed to DNA damage that produces irreversible changes during DNA replication or when the cell attempts to repair the damage. The characteristics of the effects are that they are fixed at the time of irradiation, or very shortly afterwards, and that mutations are passed on to all the cell’s progeny, which show the same lesions as the irradiated cells.

It is now known that there is another class of effect that arises in the progeny of irradiated cells. These include an enhanced cell death rate (‘lethal mutations’ or ‘delayed reproductive death’) and delayed appearance of new mutations that are different to those observed in the irradiated cells. This class of effect is regarded as a consequence of destabilisation of the genome and is termed ‘radiation-induced genomic instability’. Such effects are a factor of 100 to 1,000 more common than immediate genetic changes to cells following irradiation.

Instabilities are seen in cancerous cells but at present it is not known what role genomic instability plays in causing cancer. It is possible that radiation-induced instabilities cause later mutations that are expressed as cancers. But it is not clear to what extent genomic instabilities result in cell death, rather than mutation, and if cell death is the usual outcome there would be little contribution to cancer induction.

The likelihood is that radiation-induced cancer arises from both mutations at the time of irradiation and genomic instabilities. At present there is too little experimental evidence available to form a judgement on the balance between these two, especially at low doses.


Bystander effects are those that occur in cells that are not themselves hit by radiation but are close to cells that are. There are two basic types of bystander effects. One type is where there is contact between irradiated and unirradiated cells and gap-junction communication takes place between them. The other type is when cells are not in contact but the irradiated cell releases a soluble ‘signal’ substance that travels to unirradiated cells. The consequences in the unirradiated cells can be similar to those of genomic instability as well as of immediate DNA damage, ie cell death, mutation or chromosome aberrations.

In experimental systems, bystander effects seem to occur at fairly low radiation doses, and both gap-junction and cell-signal mechanisms seem to show similar dose responses. If the effects occur in people at low doses then this could call into question the linear relationship that is used to extrapolate risk estimates derived from epidemiological data for populations exposed at high doses and dose rates to the low doses and dose rates that are of most interest for routine exposures. One view is that the result could be either an increase or a decrease in current risk estimates. Another view is that, taking into account other evidence, risk estimates could only increase. Yet another view is that current risk estimates take some account of bystander effects and that any future changes could not be large. Without more insights into the mechanisms involved in bystander effects it is difficult to choose between these views.


The committee continued its work on this topic by considering UK and US epidemiological studies of childhood leukaemia. Atmospheric testing of nuclear weapons was at its peak in the late 1950s and early 1960s and if there was an increase in childhood leukaemia we would expect to see it during this period and into the 1970s. Analysis of UK data for this time is complicated by lack of reliable registering of incidences of childhood leukaemia, increasing accuracy of diagnosis and increasingly successful treatment. There was also a rise in the overall UK and US levels of childhood leukaemia over this time that is similar to that seen in many countries and that seems to be linked to socio-economic development.

The UK data shows that risks of mortality from childhood leukaemia rose steadily in England and Wales from about 1920 to about 1960 and then fell steadily. One study showed that there was a peak in registrations in the 1970s that was compatible with the small effect that would be expected using current risk estimates. Another study of areas of high and low weapons test fallout in Great Britain was inconclusive.

One study for the whole of the US supported current risk estimates but had used doses that were known to be overestimates. A case control study for Utah, using detailed dose reconstruction, indicated that current risk estimates are not significantly in error. This study showed no association between childhood leukaemia and radiation exposure of the foetus in the womb but did show an association for radiation exposure of young children.

Next (Meeting 5)