Adaption of the Cytokinesis-Block Micronucleus Cytome Assay for Improved Triage Biodosimetry

The purpose of this work was to adapt a more advanced form of the cytokinesis-block micronucleus (CBMN) cytome assay for triage biodosimetry in the event of a mass casualty radiation incident. We modified scoring procedures for the CBMN cytome assay to optimize field deployability, dose range, accuracy, speed, economy, simplicity and stability. Peripheral blood of 20 donors was irradiated in vitro (0–6 Gy X ray, maximum photon energy 240 keV) and processed for CBMN. Initially, we assessed two manual scoring strategies for accuracy: 1. Conventional scoring, comprised of micronucleus (MN) frequency per 1,000 binucleated (BN) cells (MN/1,000 BN cells); and 2. Evaluation of 1,000, 2,000 and 3,000 cells in total and different cellular subsets based on MN formation and proliferation (e.g., BN cells with and without MN, mononucleated cells). We used linear and logistic regression models to identify the cellular subsets related closest to dose with the best discrimination ability among different doses/dose categories. We validated the most promising subsets and their combinations with 16 blind samples covering a dose range of 0–8.3 Gy. Linear dose-response relationships comparable to the conventional CBMN assay (r² = 0.86) were found for BN cells with MN (r² = 0.84) and BN cells without MN (r² = 0.84). Models of combined cell counts (CCC) of BN cells with and without MN (BN^+MN and BN^–MN) with mononucleated cells (Mono) improved this relationship (r² = 0.92). Conventional CBMN discriminated dose categories up to 3 Gy with a concordance between 0.96–1.0 upon scoring 1,000 total cells. In 1,000 BN cells, concordances were observed for conventional CBMN up to 4 Gy as well as BN^+MN or BN^–MN (about 0.85). At doses of 4–6 Gy, the concordance of conventional CBMN, BN^+MN and BN^–MN declined (about 0.55). We found about 20% higher concordance and more precise dose estimates of irradiated and blinded samples for CCC (Mono + BN^+MN) after scoring 1,000 total cells. Blinded sample analysis revealed that the mean absolute difference (MAD) of dose estimates and the number of dose estimates outside the ±0.5 Gy interval based on CCC (Mono + BN^+MN) was comparable to conventional CBMN for doses ≤4 Gy when scoring 3,000 total cells or more. At doses >4–8.3 Gy, the MAD of CCC (Mono + BN^+MN) declined to half of the MADs observed for conventional CBMN, suggesting that the combined cell counts approach improved the discrimination ability. Conventional CBMN at 1,000 total-cell counts performed as efficiently as counting 1,000 BN cells. Discriminating and counting only BN cells with and without MN after 1,000 BN cells at ≤4 Gy revealed performances similar to conventional CBMN. After 3,000 total cells were scored, CCC (Mono + BN^+MN) was superior to conventional CBMN at >4 Gy up to about 8 Gy. Our modification of CBMN evaluations saved time compared to the widely established semiautomated MN scoring and extended the dose range up to approximately 6 Gy for triage biodosimetry.