Biokinetic and dosimetric modelling in the estimation of radiation risks from internal emitters

The International Commission on Radiological Protection (ICRP) has developed biokinetic and dosimetric models that enable the calculation of organ and tissue doses for a wide range of radionuclides. These are used to calculate equivalent and effective dose coefficients (dose in SvBq^-1 intake), considering occupational and environmental exposures. Dose coefficients have also been given for a range of radiopharmaceuticals used in diagnostic medicine. Using equivalent and effective dose, exposures from external sources and from different radionuclides can be summed for comparison with dose limits, constraints and reference levels that relate to risks from whole-body radiation exposure. Risk estimates are derived largely from follow-up studies of the survivors of the atomic bombings at Hiroshima and Nagasaki in 1945. New dose coefficients will be required following the publication in 2007 of new ICRP recommendations. ICRP biokinetic and dosimetric models are subject to continuing review and improvement, although it is arguable that the degree of sophistication of some of the most recent models is greater than required for the calculation of effective dose to a reference person for the purposes of regulatory control. However, the models are also used in the calculation of best estimates of doses and risks to individuals, in epidemiological studies and to determine probability of cancer causation. Models are then adjusted to best fit the characteristics of the individuals and population under consideration. For example, doses resulting from massive discharges of strontium-90 and other radionuclides to the Techa River from the Russian Mayak plutonium plant in the early years of its operation are being estimated using models adapted to take account of measurements on local residents and other population-specific data. Best estimates of doses to haemopoietic bone marrow, in utero and postnatally, are being used in epidemiological studies of radiation-induced leukaemia. Radon-222 is the one internal emitter for which control of exposure is based on direct information on cancer risks, with extensive information available on lung cancer induction by radon progeny in mines and consistent data on risks in homes. The dose per unit ²²²Rn exposure can be calculated by comparing lung cancer risk estimates derived for ²²²Rn exposure and for external exposure of the Japanese survivors. Remarkably similar values are obtained by this method and by calculations using the ICRP model of the respiratory tract, providing good support for model assumptions. Other informative comparisons with risks from external exposure can be made for Thorotrast-induced liver cancer and leukaemia, and radium-induced bone cancer. The bone-seeking alpha emitters, plutonium-239 and radium isotopes, are poorer leukaemogens than predicted by models. ICRP dose coefficients are published as single values without consideration of uncertainties. However, it is clear that full consideration of uncertainties is appropriate when considering best estimates of doses and risks to individuals or specific population groups. An understanding of the component uncertainties in the calculation of dose coefficients can be seen as an important goal and should help inform judgements on the control of exposures. The routine consideration of uncertainties in dose assessments, if achievable, would be of questionable value when doses are generally maintained at small fractions of limits.