Abstract
This meeting was the fourth in this series of Workshops, previous meetings having been held in Versailles, France in 1988, in Elmau, Germany in 1991, and in Bath, UK in 1993. The event has grown in size and significance over the years, to such an extent that it is now the premier European meeting for researchers and practitioners in the internal dosimetry field. Indeed, the description of 'workshop' is now rather a misnomer, the meeting being more accurately described as an international conference.
The Avignon meeting comes at a time of intense activity in internal dosimetry. The International Commission on Radiological Protection (ICRP) has recently completed development and publication of a series of physiologically based age-dependent biokinetic models for a range of important radionuclides. This followed the publication in 1994 of the new Human Respiratory Tract Model (HRTM). These models have now been used to determine the dose coefficients that are given in the EURATOM Directive and in the International Basic Safety Standards. A forthcoming ICRP publication will present predictions of the new models for bioassay quantities of interest, and will give revised guidance on the design and interpretation of internal dosimetry monitoring programmes. Another will give technical guidance on the use of the HRTM in practical situations. Looking further into the future, work on the development of a new age-dependent dosimetric model for the gastro-intestinal tract is now under way. With so many recent and forthcoming developments in modelling, it is not surprising that one of the unifying themes of the meeting was the impending need for internal dosimetry practitioners to be able to use the new models for assessments of doses in the workplace.
Jaak Sinnaeve, Head of the Radiation Protection Research Unit of DG XII in the European Commission, set the scene in his opening address. He remarked that the EC's 4th Framework Programme is currently providing funding of 70 Mecu for radiological protection research, and planning for the 5th Framework Programme, due to start in 1999, is underway. Sinnaeve was confident that all critical aspects of radiological protection would be covered, and that internal dosimetry had a 'good chance' of being one of the funded topics. An important objective of the EC programme will be to ensure that the results of research work are effectively transferred into the regulatory structure, and utilised by operational health physicists. However, there was a concern that too close an interaction between research and regulation could result in a lack of stability, resulting for example in recommended dose coefficients changing too frequently as new models and parameters are introduced.
The meeting's main sessions were organised under five headings: occupational exposure to radionuclides, public exposure to radionuclides, developments in modelling, physical dosimetry and treatment of accidental intakes.
Occupational exposure
Many of the presentations in the occupational exposure sessions addressed the topic of the application of the new models, and the determination of model parameters for use in assessments of dose resulting from occupational exposure. Mike Bailey (NRPB) described the forthcoming ICRP Technical Document on application of the HRTM, which aims to assist and promote the model's use. While default parameter values have been recommended in ICRP publications, the model was designed with the intention that parameter values based on site-specific or material-specific information would be used wherever possible. Sensitivity analyses will be presented in the Document that show which parameters are the most important to specify accurately. These include the set of parameters which describes the time-dependent rate of absorption of material from the respiratory tract to the blood and subsequent uptake to body organs. These parameters depend not only on the chemical compound inhaled, but also on the precise physico-chemical form of the material, which in turn depends on the nature of the industrial process involved. The Technical Document will give examples of the determination of specific parameter values and their use to determine site-specific dose coefficients and bioassay functions. It will also give advice on the design of experiments and the interpretation of data to determine such parameter values. Bailey invited those involved in in vivo and in vitro studies to provide data that could be used as source material for these examples.
Several of the presentations that followed may well help meet this requirement. For example, Eric Ansoborlo (Institut de Protection et de S
reté Nucléaire (IPSN), France) described an investigation of the biokinetics of uranium/uranium oxide aerosols formed at a laser enrichment facility, which involved both in vivo absorption and in vitro dissolution studies. It was found that absorption rates were significantly greater than for uranium oxides from other sources. It may be that this can be attributed to the presence of a significant ultrafine component (in this case, about 20% by mass). There is clearly a need for studies to investigate the extent to which absorption rates depend on particle size (or more fundamentally the ratio of particle surface area to mass).
Alan Birchall (NRPB) described progress with a joint NRPB/BNFL/Westlakes Research Institute project aimed at aiding implementation of the new biokinetic models. While dose coefficients (ie. dose per unit intake values) determined using these models can be found in ICRP publications, there will still be a need for Dosimetry Services to implement the new models in order to interpret bioassay monitoring data and to calculate doses using site- or material-specific parameter values. The IMBA (Integrated Modules for Bioassay Analysis) suite is a set of software modules, each of which implements a particular aspect of the complete model. This structured approach should ensure that implementation is greatly simplified, and should aid quality assurance.
Alan Britcher (Manager of Environmental and Personnel Protection, BNFL Sellafield) gave the view from a nuclear industry Approved Dosimetry Service of the current system of radiological protection for internal exposures. In a presentation that was clearly intended to question and provoke debate on current trends in model development, Britcher appealed to the research community to ensure that models should be fit for the purpose intended; that is, the assessment of dose to real individuals exposed to radionuclides in an industrial environment, on the basis of monitoring data obtained for that individual. He felt that dosimetric and biokinetic models have become more complicated than is needed by the system of protection. Rather, accessibility, straightforward implementation, reliability and stability over time are the prime requirements.
Public exposure
This session centred on environmental exposures resulting from accidents, notably those at Goiania, Palomares and the Mayak plant. Dunstana Rabelo de Melo (IRD, Brazil) gave an overview of the Goiania accident of 1987, and described the techniques used for whole body measurement of 137Cs, and the use of Prussian Blue to enhance excretion of 137Cs from the body. The large amount of Cs retention data available has enabled a revised model for caesium biokinetics to be developed and the effect of body mass on retention half-times determined.
Antonio Aragon del Valle (CIEMAT, Spain) reviewed a joint CIEMAT/NRPB study of the population exposures that resulted from the actinide releases following the aviation accident over Palomares in 1966. Airborne particle size distributions and lung absorption rates have been assessed, and by combining this data with information on occupancy and working practices, reliable estimates of doses have been derived. Excretion patterns agree well with the predictions of current ICRP models, and the assessed doses were comfortably less than 1 mSv.
Balonov (Institute of Radiation Hygiene, Russia) and Gulko (GSF, Germany) presented papers tackling the problem of dose reconstruction following exposure to 90Sr, 131I and 137Cs after the Chernobyl accident. A combination of in vivo monitoring and contaminated food modelling has been used, with the former given preference. In some areas of the Bryansk region, doses were assessed at between 20 to 40 mSv. Also deriving from the Chernobyl accident was the work of Lennart Johansson (Umeå University Hospital, Sweden), who has studied 137Cs contamination in nursing mothers, breast milk and infants in two regions of Sweden. Concentrations in the infant were about a third of those in the mother, corresponding to an estimated infant dose of 2 µGy per Bq intake by the mother.
No review of public exposure would be complete without mention of radon exposure. Werner Hofmann (University of Salzburg, Austria) reviewed the models available for radon dose assessment and highlighted the different methods for modelling deposition, clearance and dosimetry. He concluded by commenting on the well-known discrepancy between epidemiology-based and dosimetry-based risk estimates. The discussion took this issue further, the consensus being that the difference is not large when compared to overall uncertainties. It was also noted that the radiation weighting factor, wR, for alpha emissions is, by its nature, an average of the RBE values for a wide range of biological effects. This could contribute to the discrepancy when evaluating a risk estimate for the specific effect of lung cancer resulting from radon exposure.
Developments in modelling
The presentations in these sessions could be divided into three broad classifications: reviews of recent developments in biokinetic models, studies of uncertainties in model parameters and the sensitivity of estimated doses to these parameters, and reports on investigations aimed at improving our understanding of radionuclide biokinetics in man.
Bruce Boecker (Lovelace Respiratory Research Institute (LRRI), USA) gave a review of the current status of models describing the biokinetics of incorporated radionuclides. There is a clear trend to greater biological realism in this area, with the new models aiming to describe physiological processes rather than simply giving empirical descriptions of quantities of interest such as organ activities. The recycling of material (for example, from bone back to blood and then to other body organs) is treated explicitly. Short-term as well as long-term behaviour is described, and age dependence is included. Unlike many of the older models, the new models are intended to make realistic predictions of excretion rates, an essential requirement of any model intended for dose assessment purposes.
Henri Métivier (IPSN, France) outlined plans for a Human Alimentary Tract model, currently being developed by an ICRP Task Group. This revision of the Publication 30 model was motivated by a number of developments. First, the 1990 recommendations of ICRP introduced specific risk estimates for radiation-induced cancer of the oesophagus, stomach and colon. Second, more recent data on gut transit of materials are available, obtained using non-invasive scintigraphic procedures. Third, doses from radionuclides retained in intestinal tissue require consideration. Fourth, uncertainties in doses to sensitive cells, particularly from alpha-emitters require further attention. The new model will be applicable to children and adults under all circumstances of exposure. Reference parameters will be given for a Caucasian population and uncertainties in model parameters will be considered.
Alla Reddy (Defence Laboratory, India) gave a comprehensive introduction to the complexities involved in estimating doses to the fetus following maternal intakes of radionuclides. He outlined the requirements for phantoms for different stages of pregnancy, including organ masses and spatial arrangement in the fetus. For many elements, limited data are available from which to quantify placental transfer and retention by the fetus. John Stather (NRPB) outlined the approach being adopted by ICRP. For the alkaline earth elements and iodine, there are sufficient data to allow the development of biokinetic models for transfer to the fetus. For other elements, fetus: mother concentration ratios are being specified, either for the entire fetal period or for separate trimesters. Doses will be calculated for a range of different intakes: ingestion and inhalation, single and continuous, before and during pregnancy. Doses to the offspring will include contributions received in utero and, where applicable, after birth. An ICRP report on the subject should be completed in 1998.
The results of uncertainty/sensitivity analyses were presented by several authors. Ray Guilmette (LRRI, USA) described an investigation of the effect of variability of deposition in the nasal airways (in other words, the effectiveness of the nose as a filter) on deposition in the deep lung, and hence on effective dose. An investigation of uncertainties in biokinetic parameters and dose coefficients was described by John Harrison (NRPB). This study was part of a joint EC/US Nuclear Regulatory Commission project aimed at assessing uncertainties in the results of assessment codes designed to predict the consequences of accidents at nuclear installations. Panels of experts were asked to provide estimates of median values and confidence intervals for selected parameters, together with supporting documentation. Questions for the internal dosimetry panel concentrated on parameters describing intakes of radionuclides by inhalation and ingestion, and distribution and retention of radionuclides after systemic uptake. The data obtained will provide a valuable resource for the assessment of uncertainties in doses from intakes of radionuclides for other applications.
With the trend to increased biological realism in models, the need for experimental investigations of the behaviour of radionuclides in the human body is becoming ever more important. One of the areas of respiratory tract modelling where experimental data is sparse is clearance from the extra-thoracic airways. George Etherington (NRPB) reported on a human volunteer study to determine more realistic nasal clearance parameters, and discussed the implications of these results for assessment of inhalation doses. The contribution to clearance by nose blowing, and the relationship between nose-blow activities and inhaled activity, is not well understood. Preliminary results indicate that, contrary to the established view, it may be possible to make quantitative assessments of inhalation dose from nose-blow data if a suitable sampling programme is instituted.
A fundamental assumption underlying the Human Respiratory Tract Model is that the rates of absorption of material from respiratory tract to blood can be extrapolated from animal data to man. Wolfgang Kreyling (GSF Institute for Inhalation Biology, Germany) reported on progress with an interspecies comparison (humans, rhesus monkeys, beagle dogs and rats) using inhaled terbium oxide particles, which is aimed at testing this assumption. The results of the study will be used to determine the kinetics of the two competing clearance mechanisms of absorption and particle transport, and absorption parameters for each species will then be compared.
Little information is available on the dynamics of the absorption of ingested radionuclides and the rates of transfer to blood. Paul Roth (GSF, Germany) described experiments to provide such information in which human subjects were given oral and intravenous administrations of mainly stable isotopes of elements. Measurements of concentrations in blood showed rapid transfer after ingestion of elements in solution after an overnight fast. Conversely, transfer to blood was delayed when the element was ingested together with food. In addition, evidence for intestinal retention of iron was presented, suggesting that doses to gut regions from 59Fe and radioisotopes of related elements may be dominated by the contribution from the retained component.
Fetal models for iodine (Vladimir Berkovski, Radiation Protection Institute, Ukraine) and for alkaline earth elements (Tim Fell, NRPB) were described. Both involve bidirectional transfer across the placenta and are based on human and animal data. Validation of the alkaline earth model included comparison with data presented by Eugena Tolstyck (Urals Research Centre for Radiation Medicine, Russia) for skeletal concentrations of 90Sr in stillborn human fetuses and their mothers, following consumption of Techa river water.
Physical dosimetry
In this session, Keith Eckerman (ORNL, USA) reviewed developments in methods for the estimation of tissue dose. He described the series of mathematical phantoms that have been developed and how these are being further developed to enable the calculation of doses to the fetus. A recent development is the use of body imaging techniques such as magnetic resonance imaging (MRI) to generate mathematical phantoms. Eckerman considered that such techniques could be used to improve the realism of phantoms, but felt that this would not necessarily be by the direct route of generating voxel (volume pixel) phantoms. Rather, this information could be used in conjunction with new methods for handling geometric structures to build better mathematical phantoms. Nevertheless it might prove advantageous to include some voxel features where the phantom was too complex to model mathematically.
David Jones (NRPB) and Maria Zankl (GSF, Germany) both discussed voxel phantoms and their use for the calculation of Specific Absorbed Fractions (SAFs). Significant differences with values calculated using MIRD-type phantoms have been found. Jones showed that the differences arose as the result of differences in inter-organ distances between the two phantoms. Zankl went on to apply her values to calculations with radionuclides and showed that in the case of absorption in the source organ, the effects of phantom differences may be reduced where a dominant part of the absorbed dose is due to beta emission.
In a presentation describing the determination of SAFs for radio-iodine in the developing thyroid, Ulanovsky (Institute of Radiation Medicine, Belarus) demonstrated the importance of including electron transport. This is a consequence of the small organ size, which gives rise to a significant difference between the dose determined using the kerma approximation and the true absorbed dose. In this study the children varied in age from fetus to five years old, and the doses were determined using the MCNP4A code, applied to Ulanovsky's mathematical model of the thyroid and his modified version of the MIRD5 phantom.
Treatment of accidental intakes
The session opened with an expert review by Pat Durbin (Lawrence Berkeley National Laboratory, USA) on developments in the decorporation of actinides from the 1950s to the present day. Substantial progress has been made in recent years, mainly because of the advent of analogues of siderophores (ie. sequestering agents produced by micro-organisms in order to obtain Fe(III) from their environments) synthesised at the University of California. The rationale for this approach is that in mammals the biokinetics of the actinides are usually associated with Fe(III) transport and storage mechanisms. Hence the formation constants of the actinide-ligand complexes are likely to be much higher than with the current agent of choice, DTPA.
The achievements of European laboratories collaborating under the EC 3rd and 4th Framework Programmes were described in oral presentations by Neil Stradling (NRPB) and Marie-Helene Hengé-Napoli (IPSN, France), with further information given in poster presentations by Beatrice Ramounet (CEA, France), François Paquet (IPSN, France) and Inka Rencova (National Institute of Public Health, Czech Republic). Animal studies have shown that some of the siderophore analogues, particularly the hydroxypyridonate code-named 3,4,3-LI(1,2-HOPO), are considerably more effective decorporation agents than DTPA for plutonium, americium and thorium after inhalation or simulated wound contamination. On the other hand, the decorporation of uranium and neptunium by these and other substances has been much less successful, and this issue remains one of the outstanding practical problems in radiological protection. A report on the current status of chelation therapy and medical treatment is currently being prepared by a European Task Group under the 4th Framework Programme.
Conclusion
This was a lively and successful meeting, and IPSN are to be congratulated for the local organisation. The conference proceedings are to be published shortly in Radiation Protection Dosimetry. For this author (GE), the most important theme to arise during the week was the question of whether the needs of the health physics practitioner are being met adequately by the internal dosimetry research community. Speaking as someone who has a foot in both camps, I believe this will be an important issue to address in the run up to the new Ionising Radiations Regulations. The question was not fully answered at the meeting, but it is important that it has been raised.