In a clinical phase II trial GM-CSF was applied to 23 patients with acute leukemias carrying a high risk of early death because of old age or advanced disease. Concomitant laboratory investigations included the analysis of colony assays for normal granulopoietic progenitors (CFU-GM) and leukemic CFU-L with and without the addition of GM-CSF, as well as DNA measurements by flow cytometry (FCM) for the detection of DNA-aneuploidies. The present analysis is focused on one particular patient with acute myeloid leukemia (AML) in whom the detection of a DNA aneuploidy provided the readily accessible means to monitor the response of leukemic blasts to induction chemotherapy and subsequent GM-CSF treatment in vivo complementing the colony assay analyses performed in vitro. Prior to therapy 60% of cells revealed a DNA aneuploidy with a DNA index of 1.26 and in-vitro exposure of leukemic bone marrow blasts to GM-CSF (100 U/ml) enhanced the growth of CFU-L and CFU-GM with a significantly higher stimulatory index for CFU-L (1:19 versus 1:5.5). Three days after the completion of two cycles of induction therapy with thioguanine, cytosine arabinoside and daunorubicin (TAD 9) a morphologic bone marrow evaluation revealed aplasia without leukemic blasts. By FCM DNA analysis, however, 8% residual aneuploid cells were found. No growth of CFU-L was observed at this time point neither spontaneously nor after the addition of GM-CSF which induced the growth of CFU-GM, only. In-vivo application of GM-CSF (250 micrograms/mg2/day by continuous 24-h infusion) led to the recovery of normal granulopoiesis without evidence of a concomitant stimulation of aneuploid leukemic cells. These data indicate a change in the susceptibility of leukemic blasts to GM-CSF before and after chemotherapy. A reduction of the leukemic cell burden below a critical level or a selection of GM-CSF non-responsive early leukemic precursor cells may account for these observations.