Assessment of retrospective dose estimation, with fluorescence in situ hybridization (FISH), of six victims previously exposed to accidental ionizing radiation.

The present study aims to evaluate the use of the fluorescence in situ hybridization (FISH) translocation assay for retrospective dose estimation of acute accidental exposure to radiation in the past. Reciprocal translocation analysis by FISH with three whole-chromosome probes was performed on normal peripheral blood samples. Samples were irradiated with 0-5Gy (60)Co γ-rays in vitro, and dose-effect curves were established. FISH-based translocation analyses for six accident victims were then performed, and biological doses were estimated retrospectively by comparison with the dose-effect curves. Reconstructed doses by FISH were compared with estimated doses obtained by analysis of di-centrics performed soon after exposure, or with dose estimates from tooth-enamel electron paramagnetic resonance (EPR) data obtained at the same time as the FISH analysis. Follow-up FISH analyses for an adolescent victim were performed. Results showed that dose-effect curves established in the present study follow a linear-quadratic model, regardless of the background translocation frequency. Estimated doses according to two dose-effect curves for all six victims were similar. FISH dose estimations of three adult victims exposed to accidental radiation less than a decade prior to analysis (3, 6, or 7 years ago) were consistent with those estimated with tooth-enamel EPR measurements or analyses of di-centrics. Estimated doses of two other adult victims exposed to radiation over a decade prior to analysis (16 or 33 years ago) were underestimated and two to three times lower than the values obtained from analysis of di-centrics or tooth-enamel EPR. Follow-up analyses of the adolescent victim showed that doses estimated by FISH analysis decrease rapidly over time. Therefore, the accuracy of dose estimates by FISH is acceptable only when analysis is performed less than 7 years after exposure. Measurements carried out more than a decade after exposure through FISH analysis resulted in underestimation of the biological doses compared with values obtained through analysis of di-centrics and tooth-enamel EPR.