Polyamine synthesis inhibitors, such as a-difluoromethylornithine (DFMO), inhibit tumor cell growth in vitro and in vivo. However, upon cessation of treatment, tumor growth resumes. We hypothesized that incorporation of radioactive polyamines might kill the growth-arrested cells. This hypothesis was previously tested in rat 9L brain tumor cells in which DFMO increased both the uptake and the retention of [3H] putrescine. In these rat cells, DFMO-induced retention of high-specific-activity [3H] putrescine for 20 days resulted in several logs killing. In the present studies all of the 5 different human glioma cell lines tested with DFMO treatment also showed enhanced uptake of exogenous [3H] putrescine, reduced cell counts and enhanced killing of colony forming cells (CSF). Extending the time of DFMO treatment of cells that had taken up high-specific-activity (80 Ci/mmol) [3H] putrescine further increased the killing. A 10-day extension resulted in a 10,000-fold reduction in cumulative cell growth. A 5-day extension resulted in a 2-3 log decrease in numbers of surviving CFC. These data further support the hypothesis and suggest that DFMO-induced cell cycle arrest enhances cellular retention of [3H] putrescine, increasing the effective internal radiation dose enough to cause proliferative death. In a clinical setting, the short (approximately 1 microm) path-length of the tritium beta particle should limit effects to the tumor cells and spare adjacent normal cells. These results support the concept that treatment with the combination of polyamine inhibitors and radioactive polyamines might be a useful adjunct to current therapies for glioblastoma multiforme.