What Dose-Response Modeling Can, and Cannot, Tell Us About the Biological Mechanisms of Radiation Health Effects

Speaker
Dale L. Preston

Description
Efforts to develop quantitative models for radiation effects on disease rates are largely based on descriptive models with no explicit consideration of biological mechanisms. Most commonly radiation  effects are described in terms of the relative rate (ratios of the rates in an exposed group to those in a comparable unexposed population) or excess relative rate (the proportional increase in rates in an exposed group relative to those in an appropriate unexposed population) with some allowance for effect modification and confounding. These models are typically formulated and interpreted without any explicit consideration of biological mechanisms. We will discuss how the multistage model of carcinogenesis suggested by Armitage and Doll in the 1950s has motivated the basic form for the cancer baseline rate models used as a starting point in radiation effect modeling. We will also touch on how Kellerer and Rossi’s ideas about dual radiation action influenced the form of most widely used radiation dose response models. Other, rather more sophisticated, biologically motivated models, such as Moolgavkar’s two stage model, have also been used to describe radiation effects on cancer rates. While results from these models have occasionally been used to motivate thinking about the implications of the excess (relative) rate patterns that one sees, in our view, they have had little impact on thoughts about biological mechanisms of radiation carcinogenesis. We will talk about the challenges in using the radiation effect models to provide insight into the underlying mechanisms of radiation carcinogenesis, including the limited amount of information on radiation effects in virtually all epidemiological studies; the importance of choosing the right scale (excess rates or excess relative rates) on which to describe effects; and the need to look carefully at the excess risk patterns in providing ideas about mechanisms. We will also touch upon how use of richer classes of models might provide insights into the biological mechanisms of radiation effects. Despite the rapid advances in radiation-biology and, more generally, the understanding of the biology of carcinogenesis and the sophistication and usefulness of many of the descriptive models used in radiation epidemiology, more should and can be done to link those involved in these areas together.

Education Discount
Discounts are available for educational or classroom use. Contact Laura Atwell at [email protected]