In the event of large-scale radiation incidents involving a large number of people, physical methods for dose reconstruction may be difficult or impossible to apply. In incidents involving the public, the lack of dose information complicates accurate assessment of the risks of medical and biological effects of radiation, thus impeding the decision-making process [1][2].

Prompt identification of individuals with a high probability of developing moderate to severe acute radiation sickness (ARS) is essential for their timely referral to specialized facilities. The hematopoietic form of ARS is a treatable form, which occurs at radiation doses ranging 1–10 Gy, with varying degrees of disease severity classified within this range depending on the dose. According to the literature data, clinically recognizable signs of the hematopoietic syndrome of ARS are noted after exposure at doses above 1 Gy. In the exposure dose range of 1–2 Gy, mild ARS is observed, while moderate ARS is observed in the dose range of 2–4 Gy [3][4].

As a rule, medical assessment includes taking a history (including location during the incident), physical examination, documenting clinical symptoms, and, when possible, determining the absolute lymphocyte and/or neutrophil count. Although hematological investigations are helpful in assigning exposed individuals into critical groups, they yield ambiguous results in cases of combined trauma or comorbid conditions [3][5–7].

Assessment of radiation dose is critical for planning medical care. When physical dosimetry is impossible, cytogenetic biodosimetry based on the scoring of unstable chromosomal aberrations in metaphases of peripheral blood T-lymphocytes can be applied [6][8]. Elevated frequencies of dicentric chromosomes and centric rings in the spectrum of observed aberrations indicate exposure to ionizing radiation. The analysis of these types of chromosomal aberrations is referred to as the “gold standard” in biological dosimetry, due to the established dose-effect relationship and the low spontaneous frequency of these aberration types [9]. The cytogenetic method requires specialized laboratory conditions, where cell culture is performed for approximately two days, followed by slide preparation and staining, microscopic analysis with image digitization, and dose calculation using calibration curves [10–13]. Manually, one researcher can analyze no more than ~500 cells per working day, rendering this method labor- and resource-intensive.

In mass casualty incidents, the goal of rapid medical triage requires approaches that significantly reduce resource intensity and increase laboratory throughput, rather than precise estimation of absorbed radiation doses. Therefore, in this study, we set out to create an approach for rapid cytogenetic examination of exposed individuals, who had undergone relatively uniform whole-body gamma-exposure, and identification of those at risk of developing the hematopoietic syndrome of ARS.

Our research is based on the hypothesis about the possibility of classifying in vitro gamma-irradiated blood samples into dose ranges corresponding to the development of mild and moderate severity ARS. This can be carried out by detecting unstable chromosomal aberrations during the analysis of a relatively small number of peripheral blood T-lymphocyte metaphases. It is assumed that an increase in the gamma-irradiation dose is associated with an increase in the probability of detecting dicentric chromosomes in the earliest analyzed T-lymphocyte metaphases. The proposed approach is based on automated slide digitization and data analysis using specialized software, which ensures higher quality of the primary material obtained, reproducibility of results, and allows researchers to verify the detected chromosomal aberrations.

In this work, we aim to develop effective cytogenetic criteria for identifying individuals at increased risk of developing the hematopoietic form of ARS in mass radiation incidents.

The study involved 12 donors:

At the time of blood sample collection, the donors had no history of treatment for diseases using ionizing radiation sources. There were also no severe forms of diabetes mellitus, autoimmune diseases, history of blood transfusion, and radiological examinations in the three months preceding the study.

To identify individuals meeting the above criteria, the potential participants underwent a preliminary interview. In the absence of exclusion criteria, a questionnaire was administered, and consent for the processing of personal data and participation in the study was obtained.

Venous blood, 6 mL in volume, was collected from the cubital vein into sterile heparinized tubes. For each individual, three samples were prepared: two of them were irradiated in vitro on an IGUR-1M unit (Kvant, Russia) at doses of 1 Gy and 2 Gy, while the third sample was not irradiated. The IGUR-1M has the following characteristics: ¹³⁷Cs source, dose rate of 0.014 Gy/s, and irradiation non-uniformity of 5%.

Cytogenetic preparations from peripheral blood T-lymphocytes were obtained according to a protocol that included four sequential stages: cell culturing to the metaphase stage, hypotonic treatment, fixation of metaphase plates, and preparation of chromosome spreads, as described in [14].

T-cell culturing was conducted in sterile vacuum containers (Medpolymer, Russia). The culture consisted of 2 mL of blood, 5 mL of RPMI-1640 medium (PanEco, Russia), 0.5 mL of qualified fetal bovine serum (PanEco, Russia), and phytohemagglutinin (PHA) at a final concentration of 20 mg/mL (PanEco, Russia). Cells were cultured in a CO2 incubator at 37.5°C for 54 h. Two hours before the end of culturing, a colchicine solution (PanEco, Russia) was added to the culture at a final concentration of 0.03 mg/mL.

For further processing, the cell suspension was transferred to centrifuge tubes and centrifuged at 1500 rpm for 10 min in a centrifuge (ThermoScientific, USA). The supernatant was removed, leaving the pellet. For hypotonic treatment, warm (37°C) 0.55% KCl solution was added, the pellet was resuspended, and the mixture was incubated in a thermostat at 37°C for 30 min. The mixture was then centrifuged for 10 min at 1100 rpm. The supernatant was carefully removed, the pellet was resuspended, and cold (4°C) freshly prepared fixative liquid (ethanol to glacial acetic acid in the ratio of 3:1) was added, bringing the total volume in the tube to 10 mL. The cells were left in the fixative medium for 10 min at 4°C, then centrifuged at 1500 rpm for 10 min. The supernatant was removed, and a fresh portion of fixative agent was added. Thus, the cells were passed through the fixative agent three times.

To obtain chromosome spreads, the total volume of the pellet in the tube was adjusted to 1.5 mL with a fresh fixative liquid, mixed well with a pipette, and the cell suspension was dropped onto a chilled slide (two drops). The slide was dried on a heating plate at 40°C. The quality of the preparation was checked using phase-contrast microscopy. Three slides were prepared per donor. The slides were stained with 2% Giemsa stain (PanEco, Russia) for 5–6 min.

Analysis of chromosome preparations was conducted using light microscopy. To that end, images of high-quality metaphases were obtained using the Metafer scanning and imaging platform (Metasystems, Germany) with a ×63 immersion objective. Image processing was performed using the Ikaros software (Metasystems, Germany).

For cytogenetic analysis, metaphase plates containing 45–47 chromosomes, with good spreading and no overlaps, were selected. During the study, a cytogenetic protocol was completed, recording the following chromosomal aberrations: dicentric chromosomes, ring chromosomes, paired fragments, and acentric rings. For each slide, 500–1000 cells were scored in unirradiated samples and 220–500 cells in irradiated samples. The figure presents images of T-lymphocyte metaphases with chromosomal aberrations.

al aberrations: dic — dicentric chromosome; frag — paired fragment; ring — ring chromosome; Images of T-lymphocyte metaphases were obtained using the Ikaros software (immersion objective, magn. × 63)

The identified chromosomal aberrations were grouped as follows: dicentric chromosomes, unstable chromosomal exchanges (dicentric and ring chromosomes), paired fragments, acentric rings.

The work involved counting T-lymphocyte metaphases with dicentric chromosomes in the order of their occurrence among all metaphases included in the analysis on a digitized slide.

The frequency of chromosomal aberrations was calculated per cell. Standard descriptive statistical methods were used: mean values and the standard error of the mean were calculated for chromosomal aberrations.

When calculating the number of T-lymphocyte metaphases in which dicentric chromosomes were identified in the order of their occurrence during slide analysis, the data are presented as the median (Me) and the 95% confidence interval (CI). The CI for the median estimate was calculated using the bootstrap procedure with 1000 iterations. The 10th and 90th percentiles (%), as well as the minimum and maximum individual values for the number of T-lymphocytes with dicentric chromosomes in the order of their occurrence during slide analysis, were also assessed.

The expected number of cells in which dicentric chromosomes can potentially be identified (Dn) was calculated using formula (1):

(1)

wherein X — the expected number of cells in which dicentric chromosomes can be identified in the order of their occurrence on the slide; Dn — sequence number of dicentric chromosomes during slide analysis (Dn = 1, 2, 3, 4, 5, 10, 15, 20, 25, 30); F — frequency of dicentric chromosomes, calculated per cell.

To calculate the probability of encountering dicentric chromosomes in unirradiated samples for each donor, the average number of dicentric chromosomes was analyzed using the moving average method with a window size of 10, 20, 30, 50, and 100 cells. Subsequently, based on these values, the probability of the frequency of dicentric chromosomes was calculated using formula (2):

(2)

wherein P — probability of dicentric chromosome occurrence; m — number of identified dicentric chromosomes; k — number of analyzed cells (window size).

For statistical analysis, the Past 4.01 and SPSS Statistics 21 statistical software were used.

During the study, dicentric chromosomes were detected in 33% of the examined individuals (4 out of 12) in unirradiated samples. The frequency of dicentric chromosomes ranged 0–1 abs. unit per 500 cells and 0–2 abs. units per 1000 cells. The mean frequency of dicentric chromosomes and unstable chromosomal exchanges was 0.0007 ± 0.0003 per cell, of paired fragments — 0.016 ± 0.003 per cell, and of acentric rings — 0.0003 ± 0.0002 per cell.

After in vitro irradiation of blood samples at a dose of 1 Gy, the mean frequency of chromosomal aberrations increased statistically significantly compared to unirradiated samples (p < 0.0001). The mean frequency of dicentric chromosomes was 0.073 ± 0.008 per cell, unstable chromosomal exchanges — 0.081 ± 0.009 per cell, paired fragments — 0.17 ± 0.01 per cell, and acentric rings — 0.009 ± 0.002 per cell.

For in vitro irradiation at a dose of 2 Gy, the mean frequency of dicentric chromosomes was 0.28 ± 0.02 per cell, unstable chromosomal exchanges — 0.3 ± 0.02 per cell, paired fragments — 0.64 ± 0.03 per cell, and acentric rings — 0.03 ± 0.005 per cell (Table 1).

In unirradiated samples (among those individuals in whom dicentric chromosomes were detected), the expected number of T-lymphocytes with one dicentric chromosome ranged 412–500 cells. The first dicentric chromosome was identified in T-lymphocytes with sequence numbers 77, 169, 341, and 314. The second dicentric chromosome was identified in two donors: in a T-lymphocyte with sequence number 341 (one cell contained two dicentric chromosomes) and in cell number 492.

Table 2 presents the expected number of cells in which dicentric chromosomes can be identified in the order of cell scoring following in vitro irradiation at doses of 1 Gy and 2 Gy. According to the presented data, after irradiation at a dose of 1 Gy, two dicentric chromosomes were detected in T-lymphocytes between the 17th and 69th cells scored, Me = 27 [ 95% CI: 21; 41]. To detect 5 dicentric chromosomes, it was necessary to analyze 67 T-lymphocytes [ 95% CI: 41; 83]; for 10 dicentric chromosomes — 135 T-lymphocytes [ 95% CI: 103; 206]; for 20 dicentric chromosomes — 270 T-lymphocytes [ 95% CI: 206; 414]; and to identify 30 dicentric chromosomes — 405 T-lymphocytes [ 95% CI: 309; 620].

At an in vitro irradiation dose of 2 Gy, the expected number of T-lymphocytes that need to be analyzed to detect a comparable number of dicentric chromosomes was statistically significantly lower than that at an in vitro irradiation dose of 1 Gy (p < 0.001): 2 dicentric chromosomes were detected between the 5th and 12th cells; the median value for T-lymphocytes was 8 [ 95% CI: 6; 9]; 5 dicentric chromosomes were detected upon analysis of 19 T-lymphocytes [ 95% CI: 16; 22]; 10 dicentric chromosomes — upon analysis of 38 T-lymphocytes [ 95% CI: 31; 45]; 20 dicentric chromosomes — upon analysis of 77 T-lymphocytes [ 95% CI: 63; 90]; and 30 dicentric chromosomes — upon analysis of 115 T-lymphocytes [ 95% CI: 94; 134].

Table 3 presents data on the number of T-lymphocyte metaphases with detected dicentric chromosomes in the order of their analysis on digitized cytogenetic slides. Upon in vitro irradiation at a dose of 1 Gy, the median number of cells in which dicentric chromosomes were detected differed statistically significantly from the expected values only for the 1st, 25th, and 30th dicentric chromosomes in the order of their identification during analysis. For instance, the expected median value for T-lymphocytes was statistically significantly (p = 0.002) higher than the experimentally obtained value, i.e., 0.032 and 0.019 cells, respectively.

Upon in vitro irradiation at a dose of 2 Gy, the median expected number of T-lymphocytes with one dicentric chromosome was statistically significantly higher than that obtained from direct slide analysis (4 vs. 2, p = 0.033). In all other cases, no statistically significant differences were found.

The analysis showed that identifying the first dicentric chromosomes in in vitro gamma-irradiated T-lymphocytes at doses of 1 and 2 Gy requires an analysis of a relatively small number of cells. According to our and literature data, the frequency of dicentric chromosomes in unirradiated samples ranged 0–2 per 1000 analyzed cells [15].

Table 4 shows the probability of detecting the first dicentric chromosome in unirradiated samples. Thus, when analyzing 50 cells, it was 4.9%; when analyzing 100 cells, it was 9.4%. The probability of detecting two dicentric chromosomes in 100 T-lymphocyte metaphases was 2%.

Thus, given the probability of detecting two dicentric chromosomes upon analyzing the first 100 T-lymphocyte metaphases in unirradiated samples, it is necessary to increase the number of identified dicentric chromosomes to indicate radiation exposure. Based on the data presented in Table 3, to identify radiation exposure and classify a sample into a respective dose range, it is proposed to evaluate the number of T-lymphocyte metaphases containing the fifth dicentric chromosome in the order of cytogenetic slide analysis.

For instance, the median number of T-lymphocyte metaphases with the fifth dicentric chromosome was 56 [ 95% CI: 27; 85] for in vitro gamma-irradiation at a dose of 1 Gy and 16 [ 95% CI: 12; 25] for in vitro gamma-irradiation at a dose of 2 Gy (p < 0.001). The choice of the median and its corresponding confidence interval as the criterion for dividing exposed individuals into dose ranges is based on its robustness against the influence of outliers, which are characteristic of distributions containing inter-individual differences in radiosensitivity. Although the accuracy of the median estimate is clearly linked to sample size, increasing the latter will only partially reduce the uncertainty in estimating the central tendency. We believe that the reduction in the confidence interval will be insignificant due to individual responses to radiation. Therefore, the use of median values and the confidence interval appear to be an optimal solution at this stage of the research.

To make a decision regarding the classification of a potentially irradiated sample into one of the dose ranges, it is necessary to analyze the number of cells in which five dicentric chromosomes are identified. Based on the preliminary results, the following algorithm for analyzing T-lymphocyte metaphases is proposed:

Table 5 presents the proportion of examined individuals for whom the sequence number of the T-lymphocyte containing the fifth dicentric chromosome falls within the range calculated according to the algorithm described above.

All non-irradiated samples were classified into the “< 1 Gy” group; 100% of them were assigned to the group where the radiation dose is close to zero. At the same time, according to the algorithm, two (16.7%) samples irradiated at a dose of 1 Gy were classified into the “< 1 Gy” group, and two (16.7%) samples irradiated at a dose of 2 Gy were classified into the “≥ 1 Gy or higher” group.

The preliminary results of our study revealed a potential issue of false-negative identification, associated with a possible underestimation of the calculated dose range for a certain proportion of irradiated samples. For medical triage, however, a conservative approach is preferable, where an exposed individual is assigned to a dose group with higher exposure values. A conservative approach reduces the risk of underestimating ionizing radiation exposure and ensures that all potentially exposed individuals receive the necessary medical care. At the next stage of the research, it will be necessary to refine the criteria for assigning an irradiated sample to a particular dose group, taking into account the requirements of conservatism.

According to the literature, the lower limit of detection for the method of scoring unstable chromosomal exchanges depends on the number of analyzed cells and averages 100–250 mGy for sparsely ionizing radiation. At higher radiation doses (5–6 Gy and above), significant impairment of cellular proliferation occurs, resulting in fewer cells reaching metaphase and being included in the cytogenetic analysis [16–18]. Medical triage of exposed individuals based on the frequency of dicentric chromosomes or unstable chromosomal exchanges in T-lymphocytes should be conducted by reducing the number of analyzed T-lymphocyte metaphases to 50 per person, which will significantly increase laboratory throughput. The use of 20–50 cells for analysis increases the confidence intervals for estimating radiation doses assessed using a calibration curve. However, the dose intervals of 0–0.75 Gy and 2–3 Gy are clearly distinguishable when analyzing 50 cells [19][20], which is consistent with the results obtained in this study.

It is important to note that the present study did not aim to accurately determine accidental exposure doses based on the dose-effect relationship. Under conditions of limited equipment and human resources, it is necessary to develop criteria that, by reducing the number of analyzed cells, will enable timely assignment of the exposed population into groups at risk of developing the hematopoietic form of ARS. This will facilitate the planning and adjustment of further medical follow-up. Following cytogenetic triage, it is possible to increase the number of analyzed cells for biodosimetry purposes.

In future studies, we aim to validate the developed method on other sample sets, conduct inter-laboratory comparisons, and, based on the data obtained, optimize the algorithm for assessing dose intervals.

One of the tasks in providing emergency medical care to exposed individuals during mass radiation incidents is to conduct medical triage aimed at identifying persons at high risk of developing ARS. In this work, we have developed a preliminary cytogenetic triage algorithm, based on determining the number of analyzed T-lymphocyte metaphases containing the fifth dicentric chromosome in the order of their analysis.

It was established that when the fifth dicentric chromosome is identified within the first 26 analyzed T-lymphocyte metaphases, the radiation dose falls within the range of 2 Gy or higher. When the fifth dicentric chromosome is identified between the 27th and 85th analyzed T-lymphocyte metaphase, the radiation dose falls within the range of 1 Gy or higher. In the case of analyzing 85 T-lymphocyte metaphases with fewer than 5 dicentric chromosomes identified, the radiation dose falls within the range below 1 Gy. In cases where no dicentric chromosomes are detected upon analysis of 50 T-lymphocyte metaphases, the radiation dose falls within the range below 1 Gy and/or is close to zero values. This algorithm is preliminary and requires further refinement, taking into account a conservative approach to reduce the risk of underestimating ionizing radiation exposure. Continuation of the research is planned to refine the algorithm and validate the results obtained.

Authors’ contributions. All the authors confirm that they meet the ICMJE criteria for authorship. The most significant contributions were as follows: Yulia R. Akhmadullina — research planning, statistical analysis of results, results interpretation, draft writing, editing and final manuscript preparation; Yana V. Krivoshchapova — conducting laboratory research, obtaining primary data, statistical processing of primary data, draft writing; Anastasiya V. Kupriyanova — conducting laboratory research, obtaining primary data, draft editing.