The c-myb protooncogene encodes proteins that are critical for hematopoietic cell proliferation and development. Disrupting c-myb function might, therefore, prove an effective therapeutic strategy for controlling leukemic cell growth. Antisense oligodeoxynucleotides have been utilized for this purpose in vitro, but their in vivo efficacy has not been reported. We therefore established human leukemia-scid mouse chimeras with K562 cells and treated diseased animals with phosphorothioate-modified antisense oligodeoxynucleotides. K562 cells express the c-myb protooncogene, which served as the target for the antisense DNA. They also express the tumor-specific bcr-abl oncogene that was utilized to track the human cells in the mouse host. Once circulating leukemic blast cells had been detected, the survival of untreated control mice was 6 +/- 3 days (mean +/- SD). The survival of animals treated for 7 or 14 days with either sense or scrambled-sequence c-myb oligodeoxynucleotides was not statistically different from the control animals. In distinct contrast, animals treated for similar lengths of time with antisense c-myb oligodeoxynucleotides survived at least 3.5 times longer than the various control animals. In addition, animals receiving antisense c-myb DNA had significantly less disease at the two sites most frequently manifesting leukemic cell infiltration, the central nervous system and the ovary. These results suggest that phosphorothioate-modified antisense DNA may be efficacious for the treatment of human leukemia in vivo, and by analogy, for the treatment of other human neoplasias.