Research and analysis

Doses in radiation accidents investigated by chromosomal aberration analysis

Published 8 April 2024

XXVI: Review of cases investigated, 2016 to 2023

Introduction

This report is the twenty-sixth in a series that summarises biological dosimetry investigations undertaken by UK Health Security Agency (UKHSA), and previously by Public Health England (PHE, 2012 to 2021), the Health Protection Agency (HPA, 2005 to 2012) and the National Radiological Protection Board (NRPB, 1970 to 2005). The UKHSA cytogenetics laboratory was established in 1968 in the Health and Safety Branch of the United Kingdom Atomic Energy Authority (UKAEA), which together with the Radiation Protection Division of the Medical Research Council (MRC) were combined to create the NRPB in 1970. Since those very early days, the laboratory has been involved in the development and application of chromosomal aberration analysis as a biological dosemeter for investigating accidental ionising radiation exposure. Reports have been produced at regular intervals, detailing the accident cases investigated by the UKHSA biological dosimetry service (for example, 1, 2).

In common with previous reports in this series, most of the cases are briefly described in an appendix, except for those discussed in detail in the main text. Biological dose estimates are expressed in gray (Gy) and are equivalent whole-body doses unless otherwise stated. The dose estimates are chiefly derived from the frequency of dicentric chromosomal aberrations (DCA) observed in blood lymphocytes, by comparison with an appropriate in vitro dose-response calibration curve and analysis is carried out in accordance with ISO Standard 19238:20141 (3). For suspected exposures dating back more than approximately 3 years from the date of the investigation, fluorescence in situ hybridisation (FISH) analysis has been used to identify levels of stable chromosomal translocations. Where available, physical estimates are also shown in the appendix expressed in sievert (Sv) and are obtained from personal dosemeters. Occasionally these are traditional film badges, but more frequently thermoluminescence based badges (TLD), optically stimulated luminescence (OSL) or personal electronic (PE) dosemeters.

Between 2016 and 2023 UKHSA has taken part in several biological dosimetry inter-laboratory comparison exercises to help maintain our emergency response capability (4, 5, 6). In addition to carrying out biological dosimetry for routine and emergency exposure investigations, UKHSA cytogenetics team members have been instrumental in establishing new techniques within the laboratory. During the period 2016 to 2023, the premature chromosome condensation assay has been introduced into the UKHSA laboratory. This has included the validation, standardisation and simplification of the chemical/drug-induced PCC protocol for suspected high dose exposures (7, 8) and the use of the fusion-induced PCC assay in the recent RENEB (Running the European Network of Biological and Physical dosimetry) intercomparison between biological dosimetry laboratories (6).

Summary of cases investigated

The numbering system for the investigations continues from the 2006 to 2015 report (1). Except for those academically noteworthy cases discussed in the main text, brief details for each investigation are given in the appendix. Table 1 summarises the cases in terms of four categories. Category A scenarios, comprising of 16 of the 19 people investigated during this reporting period (2016 to 2023), are situations where the first indication of a possible overexposure comes from an unexpectedly high reading on a personal physical dosemeter. It is then necessary to determine whether the badge dose truly reflects the dose received by the wearer. 3 individuals fell into Category B, comprised of persons for whom an overdose is suspected but no dosemeter was worn. This situation could arise because a radiation worker omitted wearing their dosemeter or because a non-radiation worker or a member of the public was involved in an accident. Category C covers cases serious enough to merit a full reconstruction of the event, using phantoms incorporating physical measuring devices, for which satisfactory estimates of the whole-body dose can be made from physical measurements. Category D includes individuals for whom internal exposure was suspected.

Table1: Distribution of investigations between the four categories

Category	Description	Previous reports	Present report	Total
A	Possible non-uniform exposure in which the relationship between dose to the physical dosemeter and to the body is uncertain	670	16	686
B	Suspected overexposure of people not wearing a dosemeter	274	3	277
C	Overexposure where satisfactory estimates of the whole-body dose can be made from physical measurements	7	0	7
D	Chronic internal or external exposure	141	0	141
Total		1092	19	1111

Table 2 illustrates the origins of the cases examined during the period 2016 to 2023. The trend has changed from previous years in that while most cases arose from industrial uses of radiation, most involved X-rays rather than gamma sources. For most individuals (~80%) the analysis led to the conclusion that they had received a dose below the minimum detectable level of approximately 100 mGy for the dicentric chromosome assay. This detection limit arises due to a combination of the spontaneous background level of approximately 1 dicentric chromosome in 1000 cells and the statistical uncertainty associated with the scoring of a sample number of cells from the total irradiated population.

Table 2: Origins of the cases and the number of “zero” dose estimates

Case origin	Number of cases (present report)	Number of cases (all reports)		Number of dose estimates < ~100 mGy
Industrial radiography	14	718  (65%)	470
Major nuclear organisations	2	156 (14%)	93
Research, education and health institutions	3	237 (21%)	157
Total	19	1111	720

Of the 19 cases investigated during the last 8 years, there was no evidence of radiation exposure (greater than the 100 mGy detection limit) for 15 people and a dose estimate of less than 200 mGy for 2 cases, both of which showed inconclusive results due to the statistical uncertainty. Positive exposure was confirmed in 2 cases, of which 1 was found to be a partial body exposure, however the exposure was consistent with past radiotherapy treatment. Only 2 cases involved exposure to the individuals from an unknown source. Noteworthy cases are highlighted in the following paragraphs to illustrate the diversity of situations encountered.

Case A540 involved a non-destructive testing technician working with X-rays in real time radiography who experienced a machine malfunction; and may have been exposed for 45 to 60 seconds on the right side of the body. The employer stated that the maximum exposure was likely to be 1 Sv. About a week later, the individual developed a rash down the right side of the body; however, the rash had disappeared before being seen by a medical doctor. No other symptoms were reported. The chromosome test suggested that it was more likely that the actual dose received was 0.05 Gy rather than 0 Gy, with an odds ratio of approximately 4:1. It was also more likely that the body dose received was of the value of 50 mSv (0.05 Gy) than 0.21 mSv as recorded on the badge. The overall probability of a dose greater than 0 Gy was more than 99%.

For penetrating radiation such as X-rays, an approximate whole body absorbed dose of 0.05 Gy, equivalent to 50 mSv of whole body dose, is well below the threshold for deterministic effects and is unlikely to cause skin burns. Based on the International Commission on Radiation Protection expert assessment of the most up-to-date research,  a dose of 1 Sv would represent an increased risk of any type of cancer of approximately 5.5%. The relationship is linear; therefore, 50 mSv would be equivalent to an approximate increase in risk of 0.3%. This should be put into the context of lifetime cancer risk for the general population of approximately 50%.

Despite a high reading on a dosimeter, there was no evidence to suggest that an overexposure actually occurred in case A542. It was presumed that the film badge became dislodged and was inadvertently exposed. Based on comparison with our standard gamma ray dose-response curve, the most likely whole body dose was estimated at 75 mGy. This value is very uncertain with 95% confidence limits of ~0 Gy – 250 mGy. However, the test results suggest that it is much more likely that the actual dose received was 0 mGy than 75 mGy (with an odds ratio of approximately 10:1) or 300 mGy (odds ratio approximately 50:1). These results should be interpreted as not giving clear evidence about whether any actual dose above background was received, but can be interpreted as indicating that it was much more likely that no dose was received than a dose of 300 mGy.

In some cases an investigation can provide calculated doses where no dosimeter dose is  available, as in case A553. An individual working on the quality control of pipe welds could not get the Se-75 source into the transportation container, but nonetheless transported it in their vehicle. The dosimeter was never sent for analysis, but the estimated total body dose was between 20 and 250 mSv. Based on comparison with our standard gamma-ray dose-response curve, the most likely whole body dose was 0.03 Gy, with 95% confidence limits of 0 – 0.15 Gy. The test suggested that the actual dose received was more likely to be 0.03 Gy than 0 Gy (with an odds ratio of approximately 1.2:1), but did not give clear evidence about whether the actual dose received was above background, although the test suggests the dose was not more than 0.15 Gy.

There was one case, B157, where two former employees at a nuclear facility suggested their medical conditions were radiation-induced. The allegations were investigated by the country’s Nuclear Regulator and although the review was inconclusive, it was recommended the individuals undergo cytogenetic biodosimetry testing. As the potential for radiation exposure occurred more than 20 years previously the FISH translocation assay was employed to estimate the individuals lifetime dose.

Table 3. Scoring data for case B157

		Person (i)	Person (ii)
	Number of stable cells scored	2996	3003
Stable aberrations
	Translocations	14	9
	Inversions	1	1

Translocations are stable, that is, the number increases with age, and are representative of biological damage observed after exposure to ionising radiation or a variety of other agents including chemicals such as alcohol and tobacco, thus the observed number of aberrations must be adjusted for the expected background level. Based on age, the expected incidence of translocations was approximately 12 (range 9 to 14) and 15 (range 12 to 18) in 1000 full genome equivalent cells, for persons (i) and (ii) respectively (9).

The potential source of exposure was unknown in this case, however, based on comparison with UKHSAs standard 250 kVp and Co-60 gamma ray dose-response curves, the most likely estimate of  body dose was less than the minimum detectable dose for the assay with these assumed exposure conditions. In these circumstances, dose estimation is highly uncertain, and the 14 and 9 translocations observed are well below ISO standard detection limit for the assay in these circumstances of approximately 26 and 30 translocations in 1000 cells respectively (10).

The overall conclusion, in this case, was that as the observed number of translocations was significantly less than the detection limit, it is unlikely that persons (i) and (ii) had received considerable acute whole body doses of the type expected to lead to acute health effects; however, given the uncertainties associated with the radiation source, time and length of exposure, other exposures could not be ruled out.

Case B158 was interesting as a concerned individual suspected an overexposure occurred that was related to an adverse incident during their radiotherapy treatment, undertaken approximately 2 years prior to biodosimetry being requested. Dicentric chromosome aberrations  persist in blood with a half-life of approximately 3 years, hence, in this case, comparison with the expected background rate of dicentrics was not possible, and dose estimation only gave an indication of the combined dose to blood received from radiotherapy.

Table 4a. Scoring data and statistical information on the dicentric distribution (DD)

Number of cells scored	Number of dicentrics	Number of centric rings	Number of excess acentrics	DD=0	DD=1	DD=2	Var:mean	± SE	U
1000	23	3	27	981	15	4	1.326	0.044	7.453

Notes: Var:mean = variance to mean ratio, an indication of departure from a Poisson distribution of dicentrics and thus a partial body exposure (var:mean for Poisson = 1); SE = standard error of the measurement; U = the u test statistic (values > 1.96 indicate statistically significant over dispersion).

Table 4b. Whole and partial body dose estimates

Whole body dose estimate (Gy)	LCL (Gy)	UCL (Gy)	Partial body dose estimate (Gy)	LCL (Gy)	UCL (Gy)	Percent of body exposed
0.333	0.219	0.446	2.1	1.8	2.4	~10

Notes: LCL= lower confidence limit; UCL = upper confidence limit

The scoring data and dose estimates are shown in Table 4a and 4b. Based on comparison with our standard X-ray dose-response curve, the most likely whole-body equivalent dose was 0.333 Gy. This value is very uncertain with 95% confidence limits of 0.219 Gy to 0.446 Gy. However, the distribution of the dicentrics was significantly over-dispersed indicating a partial body exposure to approximately 10% of the body, as might be expected from radiotherapy. The estimated partial body dose was on the order of 2.1 Gy, with 95% confidence limits of 1.8 to 2.4 Gy. Previous work has shown that radiotherapy can lead to blood doses assessed by the dicentric chromosome assay of approximately 2 Gy (11). Overall, the results showed that the estimated dose was above background, but there was no way to separate the suspected overexposure from the likely radiotherapy dose. However, there was no evidence from comparisons of the treatment plans and delivery records or between the different delivery records, that indicated that a higher dose than expected was delivered during one of the fractions; nor was the difference between the dose to blood estimated by biodosimetry and the estimated delivered dose sufficient to cause concern.

Intercomparison exercises

UKHSA is a member of the (RENEB network)[https://www.reneb.net/]. The laboratories that make up the network can provide mutual assistance in individual dose estimation in the event of a large scale radiological or nuclear emergency. RENEB holds regular inter-laboratory comparison (ILC) exercises to maintain the networks readiness to respond to an emergency, to ensure harmonisation across the laboratories and for training purposes. Participants in an ILC are sent coded samples, which they process in their laboratory and produce estimates of dose. During 2016 to 2023 several ILC have been held and published in peer reviewed literature (4, 5, 6) and these are briefly described below.

In the ILC 2017 (4), 3 coded samples were sent to 30 laboratories of which 20 belonged to the RENEB network. After scoring 500 cells per sample, dicentric dose estimates and associated uncertainties were sent to the organising RENEB partner. The trained and harmonised RENEB partners obtained better results than the non RENEB laboratories, with 85% satisfactory results compared to 61%, respectively. The ILC in 2019 (5) was a joint venture between RENEB and EURADOS (The European Radiation Dosimetry Group) Working Group 10 (Retrospective Dosimetry) and involved a field exercise simulating real-life exposure scenarios either close, distant, lateral or partially shielded from an Ir-192 source. The dicentric assay was used to produce dose estimates and 95% confidence intervals, after scoring 500 cells per sample. All 17 RENEB laboratories that participated in the dicentric assay ILC were able to successfully estimate doses and provide information on the exposure scenarios. A large ILC was held in 2021 by RENEB [6], which aimed to provide rapid, but approximate, dose estimates and to place 3 coded samples into clinically relevant groups of low (< 1 Gy), medium (1 – 2 Gy) and high (> 2 Gy) exposure. Several biological dosimetry assays were employed and 33 laboratories from around the world took part in the dicentric assay assessment using both manual and semi-automated scoring methods, analysing 50 and 150 cells per sample respectively. All the laboratories correctly placed the 0 and 3.5 Gy samples into the correct group. The 1.2 Gy sample was assigned to the correct group by 74% (manual scoring) and 80% (semi-automated scoring) of the laboratories (12). In addition, 6 RENEB laboratories took part in the ILC using the gamma-H2AX radiation-induced foci assay for rapid dose assessment, which can be used to distinguish the exposed from the worried well prior to a more in-depth assessment. All the laboratories correctly identified the unirradiated and irradiated samples (13). Overall, an ILC exercise is a useful tool to identify areas for improvement.

Premature chromosome condensation assay capability at UKHSA

The cytogenetics team at UKHSA continually seek to improve the existing biological dosimetry tools or to establish new techniques within the laboratory. A drawback of using the dicentric assay for biological dosimetry is the potential to underestimate the radiation dose, especially at high doses (> 5 Gy), as a result of radiation induced mitotic delay and cell death during the two-day cell culture process needed to observe chromosomes at their first mitosis. The cell-fusion induced premature chromosome condensation (PCC) assay can overcome this problem by causing chromosomes to condense prematurely before reaching mitosis, so reducing the chance of mitotic delay and cell death (14). Without the need to culture cells, this method allows rapid dose estimates to be provided. Chromosomes can also be induced to condense prematurely at any stage of the cell cycle by a chemical/drug-induced method, using inhibitors of DNA phosphorylation such as Calyculin A (15). This method requires cells to be cultured for two days, however, owing to the high induction efficiency it can be used when a limited volume of blood is available and especially when the suspected overexposure is greater than 5 Gy.

Both cell fusion-induced and chemical/drug-induced PCC techniques have now been introduced into the biological dosimetry assay ‘toolkit’ used by the UKHSA cytogenetics laboratory. In particular, the cell fusion-induced PCC method has been used in the RENEB ILC held in 2021. The PCC assay results from the UKHSA laboratory correctly placed the three coded samples into the appropriate exposure group (6). Steps have also been taken to further standardise and simplify the chemical/drug-induced PCC protocol and both PCC methods have been used to assess homogeneous ex vivo high dose exposures to gamma rays by scoring ring chromosomes (7, 8). Additionally, scoring excess objects in Calyculin A induced PCC is being investigated as both low dose and suspected high dose exposures can be scored using the same set of microscope slides or digitalised images (16, 17). Further work is ongoing to investigate inter-individual variation in the spontaneous background frequency and the dispersion of aberrations, that could impact the minimal detectable dose limit and the applicability to partial body dose estimation respectively.

Appendix

See the appendix for a summary of individual cases investigated in 2012 to 2023

References

1. Sun M and others. ‘Doses in radiation accidents investigated by chromosome aberration analysis XXV. Review of cases investigated, 2006 to 2015’. Chilton, PHE-CRCE-025 2016

2. Lloyd DC and others. ‘Doses in radiation accidents investigated by chromosome aberration analysis XXIV. Review of cases investigated, 2003 to 2005’. Chilton, HPA-RPD-012 2006

3. International Standards Organisation. ‘Radiological protection – performance criteria for service laboratories performing biological dosimetry by cytogenetics’. ISO 19238:2014

4. Gregoire E and others. ‘RENEB inter-laboratory comparison 2017: limits and pitfalls of ILCs’. International Journal of Radiation Research 2021: volume 97, issue 7, pages 888-905

5. Endesfelder D and others. ‘RENEB/EURADOS field exercise 2019: Robust dose estimation under outdoor conditions based on the dicentric chromosome assay’. International Journal of Radiation Research 2021: volume 97, issue 9, pages 1181-1198

6. Port M and others. ‘RENEB inter-laboratory comparison 2021: Inter-assay comparison of eight dosimetry assays’. Radiation Research 2023: volume 199, issue 6, pages 535-555

7. Sun M and others. ‘Scoring rings in the cell fusion-induced premature chromosome condensation (PCC) assay for high dose radiation exposure estimation after gamma-ray exposure’. International Journal of Radiation Biololgy 2019: volume 95, issue 9, pages 1259-1267

8. Sun M and others. ‘A simplified Calyculin A-induced premature chromosome condensation (PCC) protocol for the biodosimetric analysis of high-dose exposure to gamma radiation’. Radiation Research 2020: volume 193, issue 6, pages 560-568

9. Sigurdson AJ and others. ‘International study of factors affecting human chromosome translocations’. Mutation Research 2008: volume 652, issue 2, pages 112-121

10. ‘Radiological protection – Performance criteria for laboratories using Fluoescence In Situ Hybridization (FISH) translocation assay for assessment of exposure to ionizing radiation’. ISO 20046:2019

11. Moquet J and others. ‘Dicentric dose estimates for patients undergoing radiotherapy in the RTGene study to assess blood dosimetric models and the new Bayesian method for gradient exposure’. Radiation Resesearch 2018: volume 190, issue 6, pages 596-604

12. Endesfelder D and others. ‘RENEB inter-laboratory comparison 2021: The dicentric chromosome assay’. Radiation Research 2023: volume 199, issue 6, pages 556-570

13. Moquet J and others. ‘RENEB inter-comparison 2021: The gamm-H2AX foci assay’. Radiation Research 2023: volume 199, issue 6, pages 591-597

14. Pantelias GE and Maillie HD. ‘The use of peripheral blood mononuclear cell prematurely condensed chromosomes for biological dosimetry’. Radiation Research 1984: volume 99, issue 1, pages 140-150

15. Kanda R, and others. ‘Easy biodosimetry for high-dose radiation exposure using drug-induced prematurely condensed chromosomes’. International Journal of Radiation Biology 1999: volume 75, issue 4, pages 441-446

16. Sun M and others. ‘A faster and easier biodosimetry method based on calyculin A-induced premature chromosome condensation (PCC) by scoring excess objects’. Journal of Radiological Protection 2020: volume 40, issue 3, pages 892-905

17. Sun M and others. ‘The applicability of scoring calyculin A-induced premature chromosome condensation (PCC) objects for dose assessment including for radiotherapy patients’. Cytogenetic Genome Reseach 2023: doi:10.1159/000534656. Epub ahead of print. PMID: 378793