Early dose stratification 4 h after ionizing radiation exposure using a peripheral blood transcriptional gene panel

Biodosimetry is essential for absorbed-dose estimation following occupational radiation accidents where early clinical decisions determine whether individuals require hospitalization or can be managed on an outpatient basis. Radioinduced transcriptomic alterations are a promising tool for rapid dose estimation. We established a gene expression panel for dose classification at 4 h post-exposure, reducing the time typically required for cytogenetic assays. Peripheral blood leukocytes from healthy individuals were X-ray irradiated and cultured for 4 h. RNA was sequenced and reads were mapped to the human hg38 genome (STAR software). Differentially expressed genes (DEGs) were identified using DESeq2 (false discovery rate, FDR, < 0.05). Radiation specificity was confirmed using eight independent microarray datasets from multiple disease conditions. DEGs were validated by RT-qPCR. A total of 215 DEGs were identified; hierarchical clustering separated samples by absorbed dose. AEN, CD70, CDKN1A, DDB2, FDXR, PCNA, and VPREB3 showed strong dose-dependent transcriptional changes across multiple disease conditions, demonstrating specificity to IR. A LASSO logistic regression model discriminated between low-dose (0–100 cGy) and high-dose (≥ 200 cGy) exposures, achieving AUC-ROC = 0.88 with CDKN1A and CD70, and AUC-ROC = 1.00 with CDKN1A, CD70, and DDB2. This gene panel offers a reliable tool for early classification of radiation exposure, supporting improved treatment of exposed individuals.