Urine metabolomics based prediction model approach for radiation exposure

Area of Science:

• Biomarkers
• Metabolomics
• Radiation Detection

Background:

• Radiological incidents necessitate rapid, non-invasive methods for radiation dose assessment.
• Existing radiation-induced metabolic markers require further validation for reproducibility.

Purpose of the Study:

• To assess the reliability and reproducibility of metabolic markers for radiation dose detection using NMR spectroscopy.
• To develop a logistic regression model for radiation exposure diagnosis based on identified metabolites.

Main Methods:

• Whole-body gamma irradiation of C57BL/6 mice, followed by urine sample collection at 24 hours post-exposure.
• High-resolution NMR spectroscopy coupled with multivariate analysis to identify metabolic changes.
• Receiver Operating Characteristic (ROC) curve and logistic regression for diagnostic model development.

Main Results:

• Fifteen distinct metabolites and three metabolic pathways (TCA cycle, taurine/hypotaurine metabolism, bile acid biosynthesis) were significantly altered post-irradiation.
• A diagnostic model was established using tau, citrate, alpha-ketoglutarate (α-KG), and fumarate, achieving 1.00 sensitivity and 0.964 specificity.
• The model demonstrated high accuracy in distinguishing irradiated from sham-irradiated controls.

Conclusions:

• NMR-based metabolomics provides a reliable and reproducible method for detecting radiation exposure.
• The developed logistic regression model shows promise as a mass screening tool for triage following radiological events.
• This approach validates the potential of metabolomics in biodosimetry and emergency preparedness.