Report on a workshop to examine methods to arrive at risk estimates for radiation-induced cancer in the human based on laboratory data. Jointly sponsored by the Office of Health and Energy Research, Department of Energy, and Columbia University.

This summary is a record of discussions and the general consensus reached by the participants. The views expressed are those of the attending scientists and do not necessarily represent the policy or opinion of the Department of Energy. An urgent need is recognized for better estimates of the risk of cancer from low levels of radiation. This need arises because of the ongoing nuclear energy option, the hazard of naturally occurring radon, and the possibility of an increasing number of lawsuits by individuals exposed to radiation in the past. The most recent estimates of risk evaluated by UN-SCEAR and BEIR V rely heavily on epidemiological studies of the A-bomb survivors which have recognized limitations and cannot provide direct risk information for chronic low-dose-rate exposure such as that experienced in occupational and medical settings. It was the consensus view that an effort to obtain information relative to risk estimates for radiation-induced cancer in humans, based on laboratory data, would be both important and timely: important, because of the opportunity to extend existing epidemiological data and overcome existing limitations; and timely, because of advances in cellular and molecular biology. In the short term, such an effort could supplement epidemiological data by providing information on the variation of cancer risk estimates with radiation dose rate and radiation quality and by providing guidance on the extrapolation of data measured at high doses to low dose regions where direct measurements are not feasible. In the long term, it may be possible to use new information about the genome from cellular and molecular studies to refine epidemiological data, i.e., to integrate classical epidemiological approaches with cell and animal biology as well as molecular genetics. Laboratory-based studies may be able to supplement epidemiological studies by: (1) identifying the molecular lesions involved in radiation-induced cancer and resolving dose, dose-rate, and radiation quality effects on their repair; (2) suggesting more biologically realistic models to describe cancer induction; and (3) addressing the problem of individual susceptibility to radiation effects and identifying radiosensitive and/or radioresistant subpopulations. It is appropriate and timely that the DOE develop and fund a research program to target areas where laboratory data are required to supplement the epidemiological cancer risk estimates. Broad-based support will be necessary for research in many fields where efforts are already ongoing, with special focus on two newer areas: attempts to understand individual susceptibility to radiation effects, and the development of model systems to investigate the mechanisms of radiation carcinogenesis.