[Incidence of acute leukemias, lymphomas and thyroid cancers in children under 15 years, living around the Marcoule nuclear site from 1985 to 1995].

Background: The aim of this investigation was to report incidence of childhood leukemia, lymphoma and thyroid neoplasms in children under 15 years of age living in the vicinity of the French Marcoule nuclear reprocessing plant.

Methods: This exhaustive and retrospective survey was carried out between 1985 and 1995 in children aged under 14 at the time of diagnosis and living inside a 35 kilometer zone around the nuclear site. 656 practitioners, 109 medical analysis laboratories and 5 hospitals or cancer institutes were investigated. A panel of experts checked each case.

Results: 48 cases of acute leukemia (39 acute lymphoid leukemia and 9 acute myeloid leukemia), 15 cases of lymphoma (8 Hodgkin lymphomas--53%--and 7 non hodgkinian lymphomas including 5 Burkitt lymphomas), 1 case of chronic myeloid leukemia and 1 case of papillary thyroid cancer, appeared among the 1,116,442 children-years followed. The total incidences of leukemias and lymphomas were respectively 4.12 and 1.29.10(-5). Standardised Incidence Ratios, calculated according to Poisson methods and bayesian inference, with various reference rates did not show any excess of risk: 100.67 (95% confidence interval 72-131) for leukemia. Children under 5 years old and living in non exposed areas to dominant winds or downstream rhodanian water drawing presented a 3 or 4 fold decreased risk of leukemia than others (the latter still having an identical risk to that of the general population). This was not true for lymphomas, nor for the other age groups.

Conclusion: Over the entire zone, children do not have an increased risk of malignant hematology disease but health monitoring by a systematic collection of cases remains useful around Marcoule. The assumption of aquiferous or air contamination thus still remains questionable: further studies investigating models of contamination are needed to take into account all other nonionizing leucemogenic factors (benzene and viral injection in particular) or correlation studies between health indicators and dosimetry.