Machine learning approach for quantitative biodosimetry of partial-body or total-body radiation exposures by combining radiation-responsive biomarkers

Validation of a blood biomarker panel for machine learning-based radiation biodosimetry in juvenile and adult C57BL/6 mice

Following a large-scale radiological event, timely collection of samples from all potentially exposed individuals may be precluded, and high-throughput bioassays capable of rapid and individualized dose assessment several days post-exposure will be essential for population triage and efficient implementation of medical treatment. The objective of this work was to validate the performance of a biomarker panel of radiosensitive intracellular leukocyte proteins (ACTN1, DDB2, and FDXR) and blood cell counts (CD19+ B-cells and CD3+ T-cells) for retrospective classification of exposure and dose estimation up to 7 days post-exposure in an in-vivo C57BL/6 mouse model. Juvenile and adult C57BL/6 mice of both sexes were total body irradiated with 0, 1, 2, 3, or 4 Gy, peripheral blood was collected 1, 4, and 7-days post-exposure, and individual blood biomarkers were quantified by imaging flow cytometry. An ensemble machine learning platform was used to identify the strongest predictor variables and combine them for biodosimetry outputs. This approach generated successful exposure classification (ROC AUC = 0.94, 95% CI: 0.90-0.97) and quantitative dose reconstruction (R2 = 0.79, RMSE = 0.68 Gy, MAE = 0.53 Gy), supporting the potential utility of the proposed biomarker assay for determining exposure and received dose in an individual.

Biomarkers for Radiation Biodosimetry and Correlation with Hematopoietic Injury in a Humanized Mouse Model

After a large-scale radiological or nuclear event, hundreds of thousands of people may be exposed to ionizing radiation and require subsequent medical management. Acute exposure to moderate doses (2-6 Gy) of radiation can lead to the hematopoietic acute radiation syndrome, in which the bone marrow (BM) is severely compromised, and severe hemorrhage and infection are common. Previously, we have developed a panel of intracellular protein markers (FDXR, ACTN1, DDB2, BAX, p53 and TSPYL2), designed to reconstruct absorbed radiation dose from human peripheral blood (PB) leukocyte samples in humanized mice up to 3 days after exposure. The objective of this work was to continue to use the humanized mouse model to evaluate biomarker dose-/time- kinetics in human PB leukocytes in vivo, at an earlier (day 2) and later (day 7) time point, after exposure to total-body irradiation (TBI) doses of 0 to 2 Gy of X rays. In addition, to assess hematological sensitivity and radiation-induced injury, PB leukocyte cell counts, human BM hematopoietic stem cell (HSC) and progenitor cell [multipotent progenitor (MPP), common myeloid progenitor (CMP), granulocyte myeloid progenitor (GMP), megakaryocyte/erythrocyte progenitor (MEP) and multi-lymphoid progenitor (MLP)] levels were measured, and their correlation was also examined as the BM damages are difficult to assess by routine tests. Peripheral blood B-cells were significantly lower after TBI doses of 0.5 Gy on day 2 and 2 Gy on days 2 and 7; T-cells were significantly reduced only on day 2 after 2 Gy TBI. Bone marrow HSCs and MPP cells showed a dose-dependent depletion after irradiation with 0.5 Gy and 2 Gy on day 2, and after 1 Gy and 2 Gy on day 7. Circulating B cells correlated with HSCs, MPP and MLP cells on day 2, whereas T cells correlated with MPP, and myeloid cells correlated with MLP cells. On day 7, B cells correlated with MPP, CMP, GMP and MEP, while myeloid cells correlated with CMP, GMP and MEP. The intracellular leukocyte biomarkers were able to discriminate unirradiated and irradiated samples at different time points calculated by receiver operating characteristic (ROC) curve. Using machine learning algorithm methods, combining ACTN1, p53, TSPYL2 and PB-T cell and PB-B cell counts served as a strong predictor (area under the ROC >0.8) to distinguish unirradiated and irradiated samples independent of the days after TBI. The results further validated our biomarker-based triage assay and additionally evaluated the radiation sensitivity of the hematopoietic system after TBI exposures.

BAX and DDB2 as biomarkers for acute radiation exposure in the human blood ex vivo and non-human primate models

There are currently no available FDA-cleared biodosimetry tools for rapid and accurate assessment of absorbed radiation dose following a radiation/nuclear incident. Previously we developed a protein biomarker-based FAST-DOSE bioassay system for biodosimetry. The aim of this study was to integrate an ELISA platform with two high-performing FAST-DOSE biomarkers, BAX and DDB2, and to construct machine learning models that employ a multiparametric biomarker strategy for enhancing the accuracy of exposure classification and radiation dose prediction. The bioassay showed 97.92% and 96% accuracy in classifying samples in human and non-human primate (NHP) blood samples exposed ex vivo to 0-5 Gy X-rays, respectively up to 48 h after exposure, and an adequate correlation between reconstructed and actual dose in the human samples (R2 = 0.79, RMSE = 0.80 Gy, and MAE = 0.63 Gy) and NHP (R2 = 0.80, RMSE = 0.78 Gy, and MAE = 0.61 Gy). Biomarker measurements in vivo from four NHPs exposed to a single 2.5 Gy total body dose showed a persistent upregulation in blood samples collected on days 2 and 5 after irradiation. The data indicates that using a combined approach of targeted proteins can increase bioassay sensitivity and provide a more accurate dose prediction.

Comparison of Novel Proteomic Expression Profiles for Radiation Exposure in Male and Female C57BL6 Mice

There is a need for point-of-care diagnostics for future mass casualty events involving radiation exposure. The development of radiation exposure and dose prediction algorithms for biodosimetry is needed for screening of large populations during these scenarios, and exploration of the potential effects which sex, age, genetic heterogeneity, and physiological comorbidities may have on the utility of biodosimetry diagnostics is needed. In the current study, proteomic profiling was used to examine sex-specific differences in age-matched C57BL6 mice on the blood proteome after radiation exposure, and the usefulness of development and application of biodosimetry algorithms using both male and female samples. Male and female mice between 9-11 weeks of age received a dose of total-body irradiation (TBI) of either 2, 4 or 8 Gy and plasma was collected at days 1, 3 and 7 postirradiation. Plasma was then screened using the SomaScan v4.1 assay for ∼7,000 protein analytes. A subset panel of protein biomarkers demonstrated significant (FDR < 0.05 and |logFC| > 0.2) changes in expression after radiation exposure. All proteins were used for feature selection to build predictive models of radiation exposure using different sample and sex-specific cohorts. Both binary (prediction of any radiation exposure) and multidose (prediction of specific radiation dose) model series were developed using either female and male samples combined or only female or only male samples. The binary series (models 1, 2 and 3) and multidose series (models 4, 5 and 6) included female/male combined, female only and male only respectively. Detectable values were obtained for all ∼7,000 proteins included in the SomaScan assay for all samples. Each model algorithm built using a unique sample cohort was validated with a training set of samples and tested with a separate new sample series. Overall predictive accuracies in the binary model series was ∼100% at the model training level, and when tested with fresh samples, 97.9% for model 1 (female and male) and 100% for model 2 (female only) and model 3 (male only). When sex-specific models 2 and 3 were tested with the opposite sex, the overall predictive accuracy rate dropped to 62.5% for model 2 and remained 100% for model 3. The overall predictive accuracy rate in the multidose model series was 100% for all models at the model training level and, when tested with fresh samples, 83.3%, 75% and 83.3% for Multidose models 4-6, respectively. When sex-specific model 5 (female only) and model 6 (male only) were tested with the opposite sex, the overall predictive accuracy rate dropped to 52.1% and 68.8%, respectively. These models represent novel predictive panels of radiation-responsive proteomic biomarkers and illustrate the utility and necessity of considering sex-specific differences in development of radiation biodosimetry prediction algorithms. As sex-specific differences were observed in this study, and as use of point-of-care radiation diagnostics in future mass casualty settings will necessarily include persons of both sexes, consideration of sex-specific variation is essential to ensure these diagnostic tools have practical utility in the field.

Assessment of Micronuclei Frequency in the Peripheral Blood of Adult and Pediatric Patients Receiving Fractionated Total Body Irradiation

The cytokinesis-block micronucleus (CBMN) assay is an established method for assessing chromosome damage in human peripheral blood lymphocytes resulting from exposure to genotoxic agents such as ionizing radiation. The objective of this study was to measure cytogenetic DNA damage and hematology parameters in vivo based on MN frequency in peripheral blood lymphocytes (PBLs) from adult and pediatric leukemia patients undergoing hematopoietic stem cell transplantation preceded by total body irradiation (TBI) as part of the conditioning regimen. CBMN assay cultures were prepared from fresh blood samples collected before and at 4 and 24 h after the start of TBI, corresponding to doses of 1.25 Gy and 3.75 Gy, respectively. For both age groups, there was a significant increase in MN yields with increasing dose (p < 0.05) and dose-dependent decrease in the nuclear division index (NDI; p < 0.0001). In the pre-radiotherapy samples, there was a significantly higher NDI measured in the pediatric cohort compared to the adult due to an increase in the percentage of tri- and quadri-nucleated cells scored. Complete blood counts with differential recorded before and after TBI at the 24-h time point showed a rapid increase in neutrophil (p = 0.0001) and decrease in lymphocyte (p = 0.0006) counts, resulting in a highly elevated neutrophil-to-lymphocyte ratio (NLR) of 14.45 ± 1.85 after 3.75 Gy TBI (pre-exposure = 4.62 ± 0.49), indicating a strong systemic inflammatory response. Correlation of the hematological cell subset counts with cytogenetic damage, indicated that only the lymphocyte subset survival fraction (after TBI compared with before TBI) showed a negative correlation with increasing MN frequency from 0 to 1.25 Gy (r = -0.931; p = 0.007). Further, the data presented here indicate that the combination of CBMN assay endpoints (MN frequency and NDI values) and hematology parameters could be used to assess cytogenetic damage and early hematopoietic injury in the peripheral blood of leukemia patients, 24 h after TBI exposure.

Biomarker integration for improved biodosimetry of mixed neutron + photon exposures

There is a persistent risk of a large-scale malicious or accidental exposure to ionizing radiation that may affect a large number of people. Exposure will consist of both a photon and neutron component, which will vary in magnitude between individuals and is likely to have profound impacts on radiation-induced diseases. To mitigate these potential disasters, there exists a need for novel biodosimetry approaches that can estimate the radiation dose absorbed by each person based on biofluid samples, and predict delayed effects. Integration of several radiation-responsive biomarker types (transcripts, metabolites, blood cell counts) by machine learning (ML) can improve biodosimetry. Here we integrated data from mice exposed to various neutron + photon mixtures, total 3 Gy dose, using multiple ML algorithms to select the strongest biomarker combinations and reconstruct radiation exposure magnitude and composition. We obtained promising results, such as receiver operating characteristic curve area of 0.904 (95% CI: 0.821, 0.969) for classifying samples exposed to ≥ 10% neutrons vs. < 10% neutrons, and R2 of 0.964 for reconstructing photon-equivalent dose (weighted by neutron relative biological effectiveness) for neutron + photon mixtures. These findings demonstrate the potential of combining various -omic biomarkers for novel biodosimetry.

Performance Evaluation of Radiation-Shielding Materials and Process Technology for Manufacturing Skin Protection Cream

Personnel using X-ray devices, the main source of radiation in medical institutions, are primarily affected by scattered rays. When interventionists use radiation for examinations/treatments, their hands may enter the radiation-generating area. The shielding gloves used for protection against these rays restrict movement and cause discomfort. Here, a shielding cream that directly adheres to the skin was developed and examined as a personal protective device; further, its shielding performance was verified. Bismuth oxide and barium sulfate were selected as shielding materials and comparatively evaluated in terms of thickness, concentration, and energy. With increasing wt% of the shielding material, the protective cream became thicker, resulting in improved protection. Furthermore, the shielding performance improved with increasing mixing temperature. Because the shielding cream is applied to the skin and has a protective effect, it must be stable on the skin and easy to remove. During manufacturing, the bubbles were removed, and the dispersion improved by 5% with increasing stirring speed. During mixing, the temperature increased as the shielding performance increased by 5% in the low-energy region. In terms of the shielding performance, bismuth oxide was superior to barium sulfate by approximately 10%. This study is expected to facilitate the mass production of cream in the future.