Personalized late health risks of radiation exposure

Currently, the risk models for exposure to radiation with doses below 4 Gy are mainly based on analyses of the atomic bomb survivors of Hiroshima and Nagasaki with empirical models of the excess relative and absolute risk. However, further cohorts provide additional information regarding the risk and individual risk factors for specific organs. Up to now this data is not considered in a systematic way in the risk models. Mechanistic models of the pathogenesis provide another important component for risk assessment and new developments in this direction will be considered. Baseline risks for particular organs clearly differ between the Japanese and the German population and transfer of risk between the populations must be investigated separately for each organ. It is necessary to assess model uncertainties in order to exclude artifacts which could occur if using only a single model.

Radiation exposures with doses above 4 Gy occur virtually only in radiation therapy. Long-term risks related to such radiation exposures are not well quantified. Increased risk has been observed for a number of cancer sites and for cardiovascular diseases in various (mostly case-control) studies, however, a systematic description of related risks does not exist. Risk models will therefore be developed from published as well as individual data, which include individual risk factors (as for instance a possibly performed chemotherapy). Since significant dose gradients can occur within organs, it will be investigated which dose parameters are most informative for assessment of late health risks. The objective of WP4 is the development of risk models that can be used to define the excess relative and absolute risk at given times after radiation exposure as well as lifetime risks. Risks will be estimated for the most important (second primary) tumors as well as for cardiovascular diseases. In these models, individual risk factors such as family predisposition, lifestyle habits, hypertension or excess weight will be considered insofar as data with sufficient quality is available.

Involved partners:
Institut für Strahlenschutz, Helmholtz Zentrum München
Institut für Medizinische Biometrie, Epidemiologie und Informatik, Johannes-Gutenberg Universität Mainz