Following contamination from the Chernobyl accident in April 1986 excess infant leukemia (0-1 y) was reported from five different countries, Scotland, Greece, Germany, Belarus and Wales and Scotland combined. The cumulative absorbed doses to the fetus, as conventionally assessed, varied from 0.02 mSv in the UK through 0.06 mSv in Germany, 0.2 mSv in Greece and 2 mSv in Belarus, where it was highest. Nevertheless, the effect was real and given the specificity of the cohort raised questions about the safety of applying the current radiation risk model of the International Commission on Radiological Protection (ICRP) to these internal exposures, a matter which was discussed in 2000 by Busby and Cato and also in the reports of the UK Committee examining Radiation Risk from Internal Emitters. Data on infant leukemia in the United Kingdom, chosen on the basis of the cohorts defined by the study of Greece were supplied by the UK Childhood Cancer Research Group. This has enabled a study of leukemia in the combined infant population of 15,466,845 born in the UK, Greece, and Germany between 1980 and 1990. Results show a statistically significant excess risk RR = 1.43 (95% CI 1.13 < RR < 1.80 (2-tailed); p = 0.0025) in those born during the defined peak exposure period of 01/07/86 to 31/12/87 compared with those born between 01/01/80 and 31/12/85 and 01/01/88 and 31/12/90. The excess risks in individual countries do not increase monotonically with the conventionally calculated doses, the relation being biphasic, increasing sharply at low doses and falling at high doses. This result is discussed in relation to fetal/cell death at higher doses and also to induction of DNA repair. Since the cohort is chosen specifically on the basis of exposure to internal radionuclides, the result can be expressed as evidence for a significant error in the conventional modeling for such internal fetal exposures.
Keywords: Chernobyl; child leukemia; infant leukemia; ionising radiation.