Skip to main content
EnglishEN
European Commission logo print header
Content archived on 2024-04-15

PROCEDURES TO ASSESS INTAKES OF RADIONUCLIDES FROM SAMPLES OF AIRBORNE RADIOACTIVITY AND STATISTICAL STUDIES OF RADIATION RISK

Objective

THE OBJECTIVE OF THIS PROJECT IS TO DEVELOP THE TECHNIQUE OF ALPHA-TRACK REGISTRATION ON CR-39 PLASTIC TO THE STAGE WHERE IT IS SUITABLE FOR ROUTINE ASSAY OF LONG-LIVED ALPHA-ACTIVITY COLLECTED ON PAS FILTERS.THIS TECHNIQUE HAS THE POTENTIAL FOR MEASURING SIGNIFICANTLY LOWER LEVELS OF ACTIVITY THAN THE COUNTING METHODS USED AT PRESENT.

INITIALLY,THE EFFECTS OF CLUMPING OF ACTIVITY IN DISCRETE PARTICLES ON THE EFFICIENCY OF TRACK REGISTRATION USING CR-39 WAS INVESTIGATED.THIS POTENTIAL PROBLEM WAS SOLVED BY DETERMINING THE OPTIMUM SPACING BETWEEN THE ACTIVE FILTER AND THE CR-39 DETECTOR TO AVOID OVERLAP OF BTRACKS AND YET PRESERVE INFORMATION ON PARTICLE AGREGATION.AUTORADIOGRAPHS OF PAS FILTERS HAVE BEEN ANALYZED USING A CYTOSCAN IMAGE ANALYZER.SOFTWARE WAS WRITTEN TO ALLOW THE CYTOSCAN TO RECOGNISE THE TRACKS,TO DISCRIMINATE BETWEEN GENUINE TRACKS AND BACKROUND,AND TO PLOT THE TRACK DENSITY ACCROSS THE AUTORADIOGRAPH.THE POSITION ON THE FILTER PAPER AT WHICH AEROSOLS ARE COLLECTED DEPENDS TO SOME EXTEND ON THEIR AERODYNAMIC DIAMETER,SO POSITIONAL INFORMATION ON TRACK DENSITY ALLOWS SOME SIZE DISCRIMINATION.INFORMATION ON ACTIVITY PER PARTICLE CAN ALSO BE DERIVED FROM THE CLUMPING TRACKS.THE ABILITY OF THE CYTOSCAN TO IDENTIFY TRACKS AND PLOT THEIR LOCATIONS HAS BEEN DEMONSTRATED AND THE POTENTIAL OF THE TECHNIQUE FOR USE IN ROUTINE MONITORING WILL BE EVALUATED.

In projecting cancer risks observed in high dose studies over time, 2 approaches have generally been used. The relative (or multiplicative) risk projection is based on the assumption that the relative cancer risk observed in the period of follow up for the study will remain constant into the future. The absolute (or additive) risk projection assumes that the annual absolute excess cancer risk observed in the study will remain constant over time. Since for most solid cancers the underlying risk increases sharply with age, the annual absolute excess risk also increases with attained age under the relative risk model, whereas under the absolute risk model it remains constant. Consequently these 2 methods can yield quite different estimates of lifetime risk, particularly if the relative or absolute risks also depend on age at exposure. The United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) (1988) have presented results based on both projection models, using data from the Japanese atomic bomb survivors; the lifetime radiation induced risk for fatal cancers of all types at high doses and high dose rates varied from 4% Gy{-1} (low linear energy transfer (LET)) based on an age dependent absolute risk model to 11% Gy{-1} (low LET) based on an age dependent relative risk model.

A formal statistical approach to distinguishing how well the relative and absolute risk models fit to data from a follow up study has been developed. This involves the use of a wider class of risk models, of which the relative and absolute models are special cases. The goodness of fit of the the best fitting model within this wider class can then be compared with that of each of the relative model and absolute model separately, in order to test whether either the relative risk model or the absolute risk model (or both) is consistent with the data. These methods were initially applied to data from earlier follow ups of the United Kingdom ankylosing spondylitis patients g iven X-ray therapy and the Japanese atomic bomb survivors in Hiroshima and Nagasaki; generally the data on solid cancers are consistent with models under which the relative risk either remains constant over time or ultimately tails off, rather than with the modelunder which the absolute risk is constant over time.

It should be emphasised that even in those cases where the relative risk appears to be constant over the period of follow up, there can still be great uncertainty in the estimation of lifetime risks. As mentioned later in relation to the development of risk estimates from the 1988 UNSCEAR report, lifetime risks based either on a relative risk projection or simply on the follow up to date of the Japanese atomic bomb survivors can differ by more than a factor of 3.

Ankylosing spondylitis patients:
Analysis of age and time trends in radiation induced cancer risks has been undertaken using data from the latest follow up of United Kingdom ankylosing spondylitis patients who received X-ray therapy. For the grouping of all cancers other than leukaemia and colon cancer, the relative risk at a given time since exposure was found to decrease with increasing age at exposure, in line with results from the Japanese atomic bomb survivors. However, the trends in risk with time since exposure were such that both the relative and absolute excess risk began to tail off about 20 years following exposure. Thus different methods for projecting risks beyond the current period of follow up yielded similar lifetime risks on the basis of these data. The above result contrasts with the most recent follow up of the Japanese atomic bomb survivors, where the absolute excess risk of solid cancers is continuing to increase with increasing time since exposure and the relative risk is approximately constant up to 40 years post exposure. However, a tailing off in the radiation induced risk as in the spondylitics has also been seen, for example, in lung cancer among uranium miners a nd bone cancer in United States radium luminisers.

Data on the Japanese Atomic Bomb Survivors:
The full data set on cancer mortality up to 1985 among the Japanese atomic bomb survivors, as based on the new DS86 dosimetry, is awaited from the Radiation Effects Research Foundation (RERF) in Japan. However, some analyses have been carried out using data in RERF Technical Report 9 to 87. These data were used to examine a claim that the cancer risk per unit dose in the Japanese survivors increases to a statistically significant extent with decreasing dose. It was found when the mortality rates were calculated correctly and adjusted for sex, age and time since exposure that the data for both leukaemia and the grouping of all other cancers were consistent with linear dose response relationships.

Risk estimates developed from the 1988 UNSCEAR report:
In the wake of recent publications by the RERF in Japan on cancer risks among the A-bomb survivors based on the new DS86 dosimetryy system and of the 1988 UNSCEAR review of risk estimates, health effects models applicable to a United Kingdom population have been developed. For cancer, risk estimates were derived mainly on the basis of the Japanese data, but using other studies where these were more informative for particular cancers (eg using data on radium luminisers in the case of bone cancer, and on patients given X-ray treatment for enlarged thymus in the case of thyroid cancer). Owing tothe uncertainty in the prediction of lifetime risks, 2 sets of risk estimates were calculated for solid tumours; 1 set based on the follow up of the A-bomb survivors to date (40 years postexposure) and the other set based on the assumption that relative cancer risks will remain constant until the end of life. Based on an analysis of radiation induced cancer rates and baseline rates for individual cancer types in the Japanese survivors and United Kingdom ankylosing spondylitis patients, it was decided for solid tumours to transfer ri sk coefficients derived from the former population to a United Kingdom population on the basis of relative risk, since this parameter was more stable across populations than absolute risk. In order to extrapolate to low doses and low dose rates, a dose rate effectiveness factor (DREF) of 3 was used in most cases, as suggested by a review of animal studies; however, the DREF was taken equal to 2 for breast cancer.

The computer program ARFAR (At Risk For Any Reason) has been developed at The National Radiological Protection Board (NRPB) to assist in the analysis of epidemiological data for those exposed occupationally to ionising radiation. The program allows person years at risk to be stratified by cumulative radiation dose, as well as by standard variables such as age, calendar period, etc. On the basis of the output from the program, the internal test for trend in cancer rates with dose can be performed. It is intended that the program be used in the analysis of the United Kingdom's National Registry for Radiation Workers (NRRW).

While the stratification of person years by variables such as age that vary linearly with time is relatively straightforward, the additional stratification by a variable such as cumulative radiation dose that generally changes in a nonlinear manner over time is more complicated. It is in its ability to stratify person years by dose that ARFAR differs from most other programs that perform cohort analyses. The calculations are based on histories of annual doses received by the workers. By assuming that dose is received at a constant rate over any year, a cumulative dose history can be constructed for each worker that is a piecewise linear (rather than linear) function of time. This allows person years to be split according to various levels of cumulative dose.

Certain enhancements have been made to ARFAR and the ancillary programs. 1 refinement allows more general latency distributions for risk following exposure to be accommodated. In particular, rather than assuming that a radiation induced cancer risk suddenly commences at a specific time following exposure, ARFAR can now stratify person years on the basis of a model for which the risk increases linearly with time until it reaches a plateau. While the former model is routinely used in analyses of occupational studies and so will be used in the first analysis of the NRRW, the latter model is more flexible and allows the sensitivity of the latency assumptions to be assessed.

Another modification to ARFAR has been made that will reduce considerably the computing time arising from multiple runs of the program.
A program has also been written to calculate median doses within dose groups, such that trends in mortality can be examined in relation to these median doses.
Among more recent enhancements, ARFAR can now perform analyses on subgroups of workers, such as all those monitored for internal emitters. ARFAR has also been modified to allow workers to enter the study some time after commencing radiation work.

The National Radiological Protection Board (NRPB) multichannel diffusion battery was modified to allow measurements of the activity size distribution down to 0.5 nm. The larger wire diameter for channel 1, with correspondingly larger gaps between wires, allows a larger fraction of the aerosols in the range 0.5 to 5 um to penetrate. The computer routine used to unfold the size distribution was modified accordingly.

The instrument was used in the living room of a representative house. Activity median diameters (AMD) in the range 150 to 200 nm were found for the attached component. In the case of the unattached component, AMDs in the range 0.8 to 2.9 nm were found with the unattached fractions, from 37 to 8 per cent, varying inversely with condensation nucleus concentration. Although activity measurements were not made during the night, the condensation nucleus concentration was found to fall to a minimum of 200 particles cm{-3} around 0200 hours, about an order of magnitude lower than the daytime value, so unattached fractions are likely to be even higher at night. Trimodal size distributions were observed when cooking was taking place in the adjacent kitchen. The additional mode, at about 6 nm, is assumed to be due to the growth of aerosols from vapours generated during cooking. These distributions were similar to those observed in earlier measurements in kitchens.

The instrument was also used, during the period, in an intercomparison held in a realistic mining environment at Limoges, France, which was sponsored by the Commission of European Communities (CEC). The equilibrium factor found in the mine was around 0.4 similar to that in homes, but the unattached fraction or radon daughters was lower, at 1 to 3.5%. This implies that for this mine, at least, lung doses for a given potential alpha energy concentration will be lower than in homes. Further measurements in homes and mines will be carried out to determine how typical are these conditions.

During the last few years there has been considerable interest in reports claiming to show excesses of childhood cancer, and in particular, leukeamis, in the vicinity of British nuclear installations. The interpretation of such epidemiological studies is generally not straightforward, particularly if the number of cancer cases is small. The choice of boundaries for the relevant study area, time period, ages and disease grouping, as well as the choice of control group, can often affect the statistical significance of the results. The sensitivity of the result with respect to these choices has therefore been assessed when examining reports concerning various installations. A review has been conducted of the different methodologies, particularly concerning the choice of geographical units, employed in several recent studies. Also, a computer program has been written for demonstration and teaching purposes which allows a distribution of cases of a rare disease to be generated across a geographical grid. In this way it is possible to show how apparent 'clusters' can occur by chance.

Examination of data on the geographical variation in cancer rates throughout the country would be of great help in trying to interpret the results around nuclear installations. An important issue is whether the number of disease cases in any specified area follows a Poisson distribution (as is usually assumed in studies around nuclear installations) or whether it follows a distribution with a larger variance. A search of the statistical literature indicated that a test derived by Potthoff and Whittinghill should have good properties at detecting such overdispersion in the numbers of cases. However, this test does not appear to have been used in an epidemiological setting until now. Some computer simulations were performed that confirmed the good statistical power of the technique; these results were presented at a Royal Statistical Meeting on Cancer near Nuclear Installations, held in Lond on in May 1989. The above test will be applied to a database on the geographical distribution of diagnoses of childhood leukaemia and lymphoma throughout Great Britain.

An examination was made of a paper written by members of the Department of Social Medicine at the University of Birmingham (Knox et al, J Soc Radiol Prot, 7, 177 to 189 (1987)), which contained a reanalysis of data on prenatal X-rays and childhood cancer obtained from the Oxford Survey of Childhood Cancers (OSCC). In particular, it was suggested in the paper that adjusting in the analysis for the effects of both drugs administered to the mother and maternal illnesses during pregnancy increased the excess relative cancer risk associated with prenatal X-rays by more than a factor of 2.
As a result of discussions with the authors and further analyses of the data it was found that an error had been made in their calculations; in particular, the cohort specific cancer risks and the relevant years of birth had been mismatched. When these were aligned correctly it was found that the relative risks were similar to those obtained in earlier analyses of this study (ie a relative risk of about 1.4).

A 6-channel, parallel, multiscreen diffusion battery, utilising silicon diffused junction detectors to detect alpha particle emissions from filters downstream from the scrrens, has been used to determine the activity size distribution of the aerosol in the atmosphere of mines and homes to which radon-222 daughters are attached. The battery was modified to allow resolution of the unattached fraction of radon daughters, having diameters in the range 0.5-3 nm. Previously reported measurements of unattached fraction of potential alpha energy concentration (PAEC), fp, in homes have given values of 0.04 to 0.34. Measurements in a kitchen while cooking was in progress yielded a mode with activity median diameter of 11 nm. In living rooms close to kitchens, a mode between 6 and 14 nm occurs during and following cooking. The accumulation mode in the activity size distributions measured had a modal diameter between 110-130 nm. Aerosol conditions in above ground workplaces, (eg schools) were similar to those in homes.

Limited data has been obtained on activity size distributions of radon daughters in mines. In homes, the average radon gas concentration is a better estimate of dose to the lungs than is measurement of average total PAEC, but this is not so in mines. In homes, factors affecting the equilibrium factor, F, and unattached fraction, fp, act in opposite senses, so that increasing fp, which would increase the dose to the lungs, is offset by a falling F, tending to reduce the dose. In mines, because of high aerosol concentrations, independent of varying ventilation conditions, fp is always low (less than 0.02). In homes, an annual average radon concentration of 20 Bqm{-3} results in an effective dose equivalent of 1 mSv.

The efficiency of the human nose as a filter of small aerosol particles has been investigated using hollow anatomical casts. Penetration through the nasal airways has been measured using monodisperse sodium chloride particles and silver particles in the range 5-150 nm. The average fractional penetrations of particles in this range were measured at flow rates of 5, 11.5 and 18 litres per minutes and found to be 0.4 0.46 and 0.51 respectively. With this model, the equivalent diffusion diameter of unattached polonium-218 was found to be approximately 2 nm. It was concluded that the fraction of unattached radon daughters that deposits in the human nose is approximately 0.5 with little dependence on flow rate over the normal respiratory range, and that radon daughter activity carried by particles larger than 10 nm diameter penetrates the nasal passages without loss.

The deposition of unattached radon daughters in nasal and oral airways were studied further using 5 plastic casts of human upper airways of varying ages. There was substantial deposition of unattached radon daughters in both nasal and oral airways, which need to be taken into account when estimating doses to the lower regions of the respiratory tract. The penetration through the oral cavity of the sectioned cast was similar to that through the nasal casts, implying that the lung dose from unattached radon daughters is not strongly dependent on the manner of breathing, by nose or mouth.

The technique of alpha particle registration on CR-39 plastic was developed with a view to its use for the routine assay of long lived alpha activity collected on personal air samplers (PAS), which are used for routine individual monitoring of workers at risk of exceeding 30% of the annual limit on intake (ALI) of long lived alpha emitters. The activity on the filters is usually measured using alpha drawer counters, but, in monitoring chromic exposures to plutonium, the sensitivity is too low, the minimum detectable activity being equivalent to about0.1 ALI. The results of the study were as follows:
autoradiography of PAS filters could provide a sensitivity at least a factor of 5 greater than present counting techniques;
activity attached to aerosols of greater than 400 nm could be assessed separately;
the dynamic range of the technique was limited by tracks overlapping in the centre of the autoradiograph on high activity samples; these could be remeasured by exposing them to CR-39 for a shorter period;
even with automatic image analysis, the technique was more labour intensive than present techniques;
clusters of tracks were difficult to identify among evenly spread tracks unless there was a large difference in track density;
quality control of etched-track counting was difficult;
the technique entailed a delay of several weeks before obtaining a result;
although antoradiography of PAS filters provided useful extra information, it was deemed unsuitable for routine use.
Project 1 : Plate out of radon daughter aerosols in domestic and mine environments

The risk of lung cancer following the exposure to radon daughters is determined by the dose absorbed by bronchial tissue. This depends on two factors : intake of potential alpha-energy and the distribution of potential alpha-energy with aerosol particle size. The objective of this project is to assess radon gas concentrations measured in large scale surveys of dwellings in terms of effective dose equivalent. This requires an understanding of the physical parameters determining aerosol characteristics in room air in order to model the range of dose conversion coefficients applicable to domestic environments.

In order to study the radioactive aerosols in dwelling and mines, it was first necessary to develop appropriate instruments. Under this contract, a five channel parallel diffusion battery was developed to allow the size distributions of the radioactive aerosols to be studied. An aerosol generator capable of producing monodisperse aerosols of 10 to 125 nm diameter was constructed and used to calibrate the parallel diffusion battery.

Project 2 : Deposition of aerosols in the upper respiratory tract

In order to calculate lung doses from inhalation of radioactive aerosols, it is necessary to know the proportion of the aerosols removed by the nasal and oral cavities as a function of size. This is particularly important for radon daughter aerosols, where a large fraction of unattached daughters may be removed.

In this study, the penetration of a wide range of aerosol sizes through nasal casts is being examined. Three nasal casts are being used : two obtained from cadavers and one made by carrying out nuclear magnetic resonance imaging of a live subject. Sheets of polymethyl methacrylate were cut according to the slice images from NMR and assembled to make an airway model which included an oral cavity.

Project 3 : The application of CR-39 etched track detectors to low background counting and particle sizing of air samples

Personal Air Samples (PAS) are currently used for routine individual monitoring for workers who are at risk of exceeding 30% of the annual limit on intake (ALI) of long lived alpha-emitters. In monitoring chronic exposures to plutonium, however, the sensitivity of the PAS is unsatisfactorily low, owing to uncertainty in correcting the observed alpha counts for counter background (ie it is equivalent to about O.1 ALI). The objective of this project is to develop the technique of alpha track registration on CR-39 plastic to the stage where it is suitable for routine assay of long lived alpha-activity collected on PAS filters. This technique has the potential for measuring significantly lower levels of activity than the counting methods used at present.

Fields of science

Programme(s)

Topic(s)

Data not available

Call for proposal

Data not available

Funding Scheme

CSC - Cost-sharing contracts

Coordinator

NATIONAL RADIOLOGICAL PROTECTION BOARD (NRPB)
EU contribution
No data
Address
CHILTON
OX11 0RQ DIDCOT
United Kingdom

See on map
Total cost
No data

Participants (2)

Share this page

Download

Last update: 23 January 1994

My Booklet

Permalink: https://cordis.europa.eu/project/id/BI6*0116

European Union, 2024