The effects of human recombinant colony-stimulating factors (r-CSFs), interleukin 3 (IL-3), granulocyte-macrophage colony-stimulating factor (GM-CSF) and granulocyte colony-stimulating factor (G-CSF) on inducing the growth of colonies derived from patients with acute myeloid leukemia (AML) (CFU-L) were investigated and compared to the proliferative response of CFU-GM derived from highly enriched normal blast cell populations. The effects of GM-CSF and IL-3 alone were similar. Both only minimally stimulated normal colonies derived from CFU-GM when compared to stimulation with MoCM (a mean of 28% of the total colonies and 17% of the colonies greater than 100 cells obtained with MoCM). Similarly, the number of leukemic colonies was substantially less than with MoCM (less than 30% of MoCM) in all but 3/10 AML patients and both were only able to significantly stimulate CFU-L derived colonies greater than 50 cells from 2/10 patients. G-CSF alone stimulated some CFU-L derived colony growth in 9/10 patients but the number stimulated was minimal relative to MoCM in five of the patients and significant stimulation of colonies greater than 50 cells occurred in only one patient. The mean number of normal CFU-GM derived colonies stimulated by G-CSF was 41% of the total colonies and 34% of the colonies greater than 100 cells generated by MoCM. The combination of G-CSF with GM-CSF and G-CSF with IL-3 resulted in a synergistic or additive increase in the number of CFU-L in 5/10 and 7/10 patients, respectively, and a synergistic increase in the size of CFU-L in 5/10. The same combinations resulted in a significant synergistic effect on size of normal CFU-GM derived colonies. There was no evidence of a synergistic increase in the number or size of CFU-L and CFU-GM derived colonies stimulated with GM-CSF in combination with IL-3. In addition, a combination of all three (G-CSF + GM-CSF + IL-3) did not enhance the effect of G-CSF + GM-CSF or G-CSF + IL-3. These results suggest that there is significant heterogeneity among AML patients in the pattern of responsiveness of the leukemic cells to the recombinant growth factors. In addition, their responsiveness does not significantly differ from that of normal progenitors. In view of the current clinical trials with r-CSFs and cytotoxic drugs in AML patients, this issue is important and worthy of further investigation.(ABSTRACT TRUNCATED AT 400 WORDS)