Infant leukaemia after the Chernobyl accident
 
A temporary increase in the incidence of infant leukaemia in Greece was reported by Petridou et al., which was attributed to in utero exposure to ionising radiation resulting from the Chernobyl accident. We performed a similar analysis based on the data of the German Childhood Cancer Registry in order to check whether the observation could be confirmed by means of independent data. Applying the same definitions as Petridou et al., we also observed an increased incidence of infant leukaemia in a cohort of children born after the Chernobyl accident. More detailed analyses, regarding areas with different contamination levels and dose rate gradients over time after the accident, showed, however, no clear trend with regard to exposure. It would therefore appear less likely that the observed effect was caused by exposure to ionising radiation due to the Chernobyl accident.

STEINER, M., BURKART, W., GROSCHE, B., KALETSCH, U., MICHAELIS, J.
Trends in infant leukaemia in West Germany in relation to in utero exposure due to Chernobyl accident. Radiat Environ Biophys 37, 87-93, 1998.

In a letter to Nature, Petridou et al. attributed an increased incidence of infant leukaemia in Greece to increased in utero exposure to ionizing radiation arising from the Chernobyl accident, which occurred on 26 April 1986. We see a similar increase in infant leukaemia in western Germany in born after the Chernobyl accident. However, more detailed analyses of different contamination levels and dose rates show no relationship between exposure and incidence. We therefore conclude that the observed effect was not due to ionizing radiation from the Chernobyl accident.
Measurements of radioactive fallout and calculation of effective doses for the population show that large parts of Germany, especially in the south, experienced contamination comparable to that of Greece. The United Nation Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) estimated an effective dose-equivalent of 0.49 mSv attributable to the Chernobyl accident (over the following year) for the population in southern Germany and of 0.33 mSv for most of Greece.
There was a wide variation in levels of exposure in Germany, with UNSCEAR’s first-year dose estimate for northern regions being 0.07 mSv, so it is possible to study potential dose-response patterns. Local fallout pattern and estimates of resulting average external and internal radiation doses are available for the 328 Landkreise of the former Federal Republic of Germany. Despite widespread long-distance transport of even fresh agricultural products in Germany, it has been shown that the variation in total doses closely follows ground-deposition patterns, which thus can be used as a surrogate for dose. We chose levels above 10 kBq/qm Cs137 to represent areas of „high" exposure and below 6 kBq/qm for low exposure. Taking shielding but not iodine contribution (which adds little to blood stem-cell dose) into account, this translates into average in utero doses (total for nine months) of 7 and 55 µSv respectively, for the exposure levels defined above.
Since 1980, a population-based childhood cancer registry has existed, which receives reports of more than 90% of malignancies occurring in children up to15 years of age, for the former Federal Republic of Germany. Therefore, unlike the Greek study, we did not have to ascertain the number of leukaemia have retrospectively. Using the definitions of Petridou et al. for in utero exposure (children born between 1 July 1986 and 31 December 1987) there were 928,649 ‘exposed’ children in the FRG (cohort B). Amongst these, 120,440 lived in the areas of high exposure. Cohorts A and C, defined as ‘unexposed’ groups, consisted of 3,601,176 children born between 1 January 1980 to 31 December 1985 (cohort A) and 2,029,613 born between 1 January 1988 and 31 December 1990 (cohort C). The incidence of infant leukaemia in the ‘exposed’ cohort B is higher than in cohorts A and C (Table 1). The incidence rate did increase from cohort A (23.0 per million) to cohort C (29.6 per million), possibly due some under-reporting in the initial phase of the German childhood cancer registry.
Looking at regions defined by radioactive ground deposition, the increase in leukemia is highest in the regions with the lowest contamination by radioactive fallout. We believe that it is highly unlikely that this observation is due to misclassification, as there was clear geographical distinction between the areas of high and low exposure.
In addition, because there were many short-lived radionuclides, and as surface contaminations in the human environment and on food were quickly washed out,  radiation exposures from Chernobyl showed quite steep gradients in time. The dose rate in the first days after fallout were 23 times higher than in the last month of 1987. Any potential excess of infant leukaemia caused by intra-uterine radiation exposure should thusbe most apparent in the older childrens of cohort B. If we divide B into two subcohorts (birth date July 1986 to March 1987 and April to December 1987) the rate ratio is 1.29 for the first subcohort and 1.67 for the younger subcohort (data for all areas, comparison with the combined cohorts A and C). These results again are not in accordance with the radiation hypothesis.
Analysing the incidence of all leukaemias that occurred up to 5 years old shows that the relative increase of infant leukaemia seems to be compensated for in subsequent years, as the corresponding rate ratio of ‘exposed’ to ‘unexposed’ cohorts is 1.02 for the whole area (95% confidence interval, CI, 0.91-1.15). A decreased incidence in the second year of life was observed for the ‘exposed’ cohort (rate ratio, 0.84;95% CI, 0.61-1.85).
Although our initial analysis is consistent with the observation of Petridou et al., detailed trend analyses for contamination levels and critical time periods fail to correlate exposure levels with increased leukaemia rates. Thus we conclude that the observed increase of infant leukaemia is not caused by an increased in utero exposure to ionizing radiation from the Chernobyl accident.

MICHAELIS, J., KALETSCH, U., BURKART, W., GROSCHE, B.
Infant leukaemia after the Chernobyl accident. Nature 387, 246, 1997.