Response of the blast stem cells of acute myeloblastic leukemia to G-CSF, GM-CSF, or the ligand for C-KIT, alone or in combination.

Growth factors are commonly included in protocols for the treatment of acute myeloblastic leukemia (AML). Because the response of blast stem cells in culture to growth factors might influence the contribution of factor to clinical outcome, we studied 42 patients with AML or severe myelodysplasia. Peripheral blood blast cells were cultured in a clonogenic assay at three cell concentrations and with the following combinations of growth factors: no added growth factor (NF), G-CSF, GM-CSF, Kit ligand (KL), G-CSF + KL, GM-CSF + KL, and G-CSF + GM-CSF + KL. The slope of the line relating cell number plated to colony formation was calculated by least squares. The slopes were used to form three equally sized groups of patients. Marked heterogeneity was found in response of the blast populations to factor. A few general conclusions emerged: (1) autonomous blast populations are very rare; (2) although usually a population responds better to one of the growth factors than to others, seldom is the response exclusively to one factor; (3) when more than one factor is included in the cultures, synergism is usually seen. Significant associations were seen between successful remission induction for low slope values in cultures with NF or KL alone. For remission, but not survival, associations were found with intermediate values of slope in cultures with G-CSF + KL and GM-CSF + KL. We conclude that measurements of growth factor response are feasible and yield clinically useful data.