Related Experiment Video
Updated: Sep 2, 2026
Expedited Radiation Biodosimetry by Automated Dicentric Chromosome Identification (ADCI) and Dose Estimation
Published on: September 4, 2017
Accurate estimation of clinically significant radiation doses up to 5 Gy using qRT-PCR based gene expression analysis
Ivonne Romero Aguilera1, Omar García Lima2, Pia Loren Reyes3
1Depto de Ciencias Básicas, Facultad de Medicina, Universidad de La Frontera, Temuco, Chile.
Purpose:
The main aim of this study is to evaluate the capability of the qRT-PCR based on FDXR and DDB2 expression to accurately dose reconstruction over 2 Gy within clinically relevant exposure intervals.
Materials And Methods:
To obtain calibration data, whole blood samples were irradiated in vitro (0-25 Gy) and incubated at 37 °C for 24 h. Peripheral Blood Mononuclear Cells were isolated, from which total RNA was extracted, cDNA was synthetized and gene expression (FDXR, DDB2) relative to 18S was quantified by qRT-PCR (ΔCt). Saturation dose was determined using a combined approach including effect-threshold analysis, local slope reduction, and four-parameter logistic (4PL) modeling. Three calibration models (asymptotic exponential, four parameters logistic Hill-type (4PL-H), and log-linear) were fitted and compared using residual standard error (RSE), AIC and BIC. Dose estimations were performed from data obtained from an independent set of blood samples irradiated at 1, 2.5, 5 and 7.5 Gy. Dose estimation accuracy was assessed by 95% CI, ±20% deviation from physical dose, and correct allocation into clinically relevant categories.
Results:
Both genes exhibited progressive ΔCt reduction with increasing dose followed by plateau formation. The median saturation dose was 6.9 Gy for FDXR and 7.5 Gy for DDB2, indicating diminishing sensitivity above approximately 7 Gy. Although the log-linear model provided the lowest RSE and most favorable AIC/BIC, it failed to reconstruct the 0 Gy control within its 95% CI. The 4PL-H model correctly identified unexposed samples and was therefore selected for calibration model. For dose estimation, FDXR achieved accurate reconstruction up to 5 Gy and correctly classified doses into relevant clinical intervals, demonstrating large uncertainty and overestimation at 7.5 Gy, consistent with saturation.
Conclusions:
The 4PL-H model provides superior reconstruction validity incorporating biological asymptotes. Within 0-5 Gy, FDXR demonstrated clinically robust interval classification performance.
