The excessive proliferation of the myeloid marrow compartment in Philadelphia chromosome (Ph)-positive acute and chronic leukemias has been largely attributed to a hyperactive and autonomously acting hybrid tyrosine kinase BCR-ABL, a product of the fusion between the second exon of the c-ABL proto-oncogene and 5' portions of the BCR gene on chromosome 22. This specific molecular event, amenable to attack with specifically designed inhibitors, has recently been successfully influenced by the drug CGP-57148 in mammalian cells transfected with full-length BCR-ABL gene and expressing full-length p210Bcr-Abl protein, as well as in primary human leukemic cells expressing p210Bcr-Abl fusion protein. In view of the heterogeneity of BCR-ABL transcripts associated with various phenotypes, we investigated the effect of CGP-57148 on p190Bcr-Abl- and p210Bcr-Abl-expressing, patient-derived cell lines and primary intact blast cells. In particular, we were interested in whether the variations in molecular events and/or the phenotype of Ph-positive cells would affect their susceptibility to the specific tyrosine kinase inhibitor CGP-57148. We have demonstrated that the sensitivity of human cells with lymphoblastic immunophenotype expressing p190Bcr-Abl protein is comparable to that for leukemic myeloid cells expressing p210Bcr-Abl protein. After documenting profound and phenotype-independent suppression of both autophosphorylation and cell growth, we explored the importance of time and dose of exposure on the manifestation and stability of the induced events. Although there were variations between target cells, in vitro exposure for 24-48 h induced extensive and apparently irreversible apoptosis in BCR-ABL-expressing but not other normal or BCR-ABL-negative leukemic cells. These findings support the potential use of CGP-57148 to purge Ph-positive cells from autologous bone marrow in vitro. Another important finding was the comparable suppressive effect of temporary CGP-57148 exposure on both clonogenic KBM-5 cells and the whole cell population. Exposure time and dose appeared to be important variables among various cell types. Moreover, effective doses appeared uniformly harmless to cells lacking BCR-ABL protein functioning as tyrosine kinase. Thus, the continuous exposure of target cells, at least during the initial period of 24-48 h, may prove to be an important variable in the design of in vitro and in vivo therapy using tyrosine kinase inhibitors.