Tumor radiosensitivity prediction by the cytokinesis-block micronucleus assay.

An in vivo to in vitro cytokinesis-block micronucleus assay technique using cytochalasin B (Cyt-B) was established in xenografted human and murine tumors, and the correlation between radiosensitivity measured by this assay and that measured by a colony-forming assay was investigated. Tumors were irradiated in situ, excised immediately, and disaggregated to single cells that were plated for the micronucleus and colony-forming assays. Some of the tumor cells were irradiated in vitro rather than in vivo. For the micronucleus assay, Cyt-B (0.5-3 micrograms/ml) was added to dishes soon after plating or in vitro irradiation and the cells were subsequently fixed and stained at intervals (12-144 h). The micronucleus frequency in binucleate cells was evaluated under conditions of maximum yield of the binucleate cells. The micronucleus frequency after irradiation was quite variable depending on the tumor type and the average number of micronuclei per single binucleate cell after 4 Gy ranged from 0.2 to 1.4. The results of in vitro irradiation were not significantly different from those of in vivo irradiation for all tumors. A good correlation was found between the radiosensitivity determined by the micronucleus assay and that found with the colony-forming assay in six human tumors (r = 0.94 approximately 0.98) but not in four murine tumors because of one exceptional tumor. When this tumor was excluded, a correlation was also found for the remaining nine tumors (r = 0.62 approximately 0.96). These results indicated that the cytokinesis-block micronucleus assay has some promise as a rapid predictive assay of radiosensitivity.