Simultaneous administration of granulocyte-macrophage colony-stimulating factor and cytosine arabinoside for the treatment of relapsed acute myeloid leukemia

The treatment of patients with relapsed or refractory acute myeloid leukemia (AML) with high dose cytosine arabinoside (ara-C) results in short-lived complete response rates of 30-50%. We have previously shown that entry of myeloid leukemic cells into S phase can be accelerated in vitro through the use of recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF), resulting in enhancement of ara-C-mediated cytotoxicity. In order to evaluate the in vivo biological and clinical effects of this strategy in patients with high risk AML, we treated three patients with either refractory or relapsed disease with a continuous infusion of rhGM-CSF (0.45 micrograms/kg/h aglycoprotein) for 18 h, followed by the institution of high dose ara-C and continuation of rhGM-CSF throughout the 4 day duration of ara-C treatment. Prior to therapy, no patient had detectable levels of circulating rhGM-CSF, and there was no evidence of GM-CSF receptor occupancy in leukemic myeloblasts. After 18 h of rhGM-CSF therapy, all patients had biologically active levels of circulating rhGM-CSF (7.9-12.0 ng/ml), and two patients showed a significant degree of leukemic GM-CSF receptor occupancy without evidence of GM-CSF receptor down-regulation. A significant rise in the S phase fraction of leukemic myeloblasts was observed at 18 h of rhGM-CSF treatment in all three patients (29-56% increment). The toxicity of combined rhGM-CSF/ara-C therapy included pericarditis and cerebellar degeneration in one patient, fever and mild renal dysfunction in two patients, and mild hepatic dysfunction in all three patients. Each patient showed a transient rise in the absolute neutrophil and blast count during rhGM-CSF/ara-C administration, followed by profound, but clinically tolerable, myelosuppression. No patient developed clinical evidence of leukostasis. There was one death related to pericardial tamponade, one death related to refractory disease, and one clinical and cytogenetic remission. These results suggest that exogenously administered rhGM-CSF is capable of rapidly mobilizing leukemic cells into S phase in vivo and theoretically should be useful in overcoming kinetic resistance to ara-C. Clinical trials of this regimen in patients with high risk AML who are not already pharmacologically resistant to ara-C are warranted.

Differentiation-associated expression of two functionally distinct classes of granulocyte-macrophage colony-stimulating factor receptors by human myeloid cells

Granulocyte-macrophage colony-stimulating factor (GM-CSF) activates a broad range of myeloid cells through binding to high affinity surface membrane receptors. The effects of this hematopoietin are dependent upon the differentiation status of the myeloid cell and range from proliferation of early myeloid progenitor cells to activation of neutrophil and monocyte function. In addition, many of the biological effects of GM-CSF are shared with interleukin-3 (IL-3), a distantly related lymphokine. In this study, we have characterized the GM-CSF receptor of myeloid cells at various stages of differentiation by comparing the binding characteristics and surface regulation of this receptor in early versus late myeloid cells. Scatchard analysis revealed a single class of high affinity receptors on normal neutrophils, monocytes, and myeloblasts from patients with acute myeloid leukemia. Neutrophils expressed significantly higher numbers of receptors, with an approximately 2-fold lower affinity, when compared with other myeloid cells. Two different patterns of GM-CSF receptor regulation and binding were observed. In the first pattern, the GM-CSF receptor of neutrophils was rapidly down-regulated by GM-CSF itself, by phorbol myristate acetate (PMA), and by the calcium ionophore A23187, and it was not competed for by IL-3 (class I receptor). In contrast to the neutrophil receptor, the GM-CSF receptor of the myeloblast demonstrated resistance to the down-regulatory effects of GM-CSF itself, PMA, and A23187, and it was completely competed for by IL-3 (class II receptor). In some cases of acute myeloid leukemia and monocytes, a mixed pattern of partial PMA responsiveness and partial competition by unlabeled IL-3 was observed, suggesting the coexpression of both class I and II receptors in these cells. In these cells, after down-regulation of the class I receptor by PMA, the remaining receptors were shown to be completely cross-competed for by IL-3, further supporting the hypothesis that these cells have a mixture of class I and II receptors. Chemical cross-linking of radiolabeled GM-CSF to myeloid cells revealed the labeling of three proteins (156, 126, and 82 kDa) which were identical in cells expressing either class I or II binding sites. These data show that there are differentiation-associated differences in the regulation of the GM-CSF receptor which may have important physiological consequences.

Regulation of surface expression of the granulocyte/macrophage colony-stimulating factor receptor in normal human myeloid cells

Recombinant human granulocyte/macrophage colony-stimulating factor (GM-CSF) exerts stimulatory effects on hematopoietic cells through binding to specific, high-affinity receptors (Kd = 30-100 pM). By using radiolabeled GM-CSF with high specific activity, we have investigated the factors and mechanisms that regulate GM-CSF receptor expression in normal human neutrophils, monocytes, and partially purified bone marrow myeloid progenitor cells. The neutrophil GM-CSF receptor was found to rapidly internalize in the presence of ligand through a mechanism that required endocytosis. Out of a large panel of naturally occurring humoral factors tested, only GM-CSF itself, tumor necrosis factor, and formyl-Met-Leu-Phe were found to down-regulate neutrophil GM-CSF receptor expression after a 2-hr exposure at biologically active concentrations (95% +/- 1%, 34% +/- 5%, 48% +/- 8% receptor down-regulation, respectively). GM-CSF also down-regulated its own receptor on monocytes and myeloid progenitor cells. Since formyl-Met-Leu-Phe is known to stimulate neutrophil protein kinase C activity, we also tested the ability of protein kinase C agonists to modulate GM-CSF receptor expression. Phorbol 12-myristate 13-acetate, bryostatin-1, and 1,2-dioctanoylglycerol were found to induce rapid down-regulation of the GM-CSF receptor in neutrophils, monocytes, and partially purified myeloid progenitor cells, suggesting that this effect may be at least partially mediated by protein kinase C. These data suggest that certain activators of neutrophil function may negatively regulate their biological effects by inducing down-regulation of the GM-CSF receptor.

Granulocyte-macrophage colony-stimulating factor enhances the cytotoxic effects of cytosine arabinoside in acute myeloblastic leukemia and in the myeloid blast crisis phase of chronic myeloid leukemia

A strategy designed to stimulate myeloid leukemic blasts into active cell cycle may increase the effectiveness of S phase-specific agents such as cytosine arabinoside (ARA-C). Since recombinant human granulocyte-macrophage colony stimulating factor (GM-CSF) is known to stimulate the growth of myeloid leukemic cells in vitro, we have evaluated the ability of this growth factor to enhance leukemic clonogenic cell kill in the presence of ARA-C. In seven patients studied, GM-CSF increased the fraction of myeloid leukemic blasts in S phase as measured by propidium iodide DNA staining, bromodeoxyuridine incorporation, or ARA-C suicide techniques. Six of these seven patients demonstrated clonogenic cell growth in agar in response to GM-CSF. In five of these six patients, the combination of GM-CSF and ARA-C treatment in vitro resulted in a significant increase in leukemic clonogenic cell kill when compared to treatment with ARA-C in the absence of GM-CSF. Similar results were observed with the combination of GM-CSF and hydroxyurea, another S phase specific agent, further suggesting that the observed enhancement of cytotoxicity was due to the ability of GM-CSF to increase the number of leukemic cells in S phase. These data provide a rationale for investigating the toxicity and efficacy of combined GM-CSF and ARA-C therapy in patients with high-risk myeloid leukemia.

Monocytes enhance gamma-interferon-induced inhibition of myeloid progenitor cell growth through secretion of tumor necrosis factor

In this study, we have examined the effects of autologous monocytes and T-lymphocytes on gamma-interferon (gamma-IFN)-induced inhibition of granulocyte-monocyte progenitor cells (CFU-GM) in vitro. Depletion of adherent cells from the mononuclear fraction of normal bone marrow (NBM) resulted in a significant reduction in the inhibitory effects of gamma-IFN on CFU-GM growth, whereas T-lymphocyte depletion had no effect. Adding back autologous monocytes to the underlayer fraction of agar culture resulted in a concentration-dependent enhancement of gamma-IFN-induced CFU-GM inhibition that did not require cell-cell contact. Adding back autologous T-lymphocytes had no effect and did not synergize with monocytes in enhancing gamma-IFN-induced inhibition. Based on the use of indomethacin and the pattern of CFU-GM subset growth, it was determined that prostaglandin E was unlikely to be the humoral inhibitory factor involved in this process. However, the effects of monocytes were completely reversed in the presence of a neutralizing monoclonal antibody to tumor necrosis factor (TNF), suggesting that monocyte-derived TNF was responsible for the enhancement of gamma-IFN-induced CFU-GM inhibition. This observation was further supported by the ability of gamma-IFN to induce an eightfold increase of baseline monocyte TNF secretion in agar culture. These data suggest that gamma-IFN may inhibit progenitor cell growth in vitro through indirect humoral mechanisms involving monocyte-derived TNF, as well as through direct inhibitory effects on CFU-GM proliferation. Because monocytes are a component of the bone marrow microenvironment, the ability of gamma-IFN to induce biologically relevant levels of monocyte-derived TNF may play an important role in the negative regulation of hematopoiesis.

Human granulocyte-monocyte colony-stimulating factor and interleukin 3 stimulate monocyte cytotoxicity through a tumor necrosis factor-dependent mechanism

Human colony-stimulating factors (CSF) exert multiple effects on the proliferation, differentiation, and function of myeloid lineage cells. In this study, the effects of three recombinant human CSFs (granulocyte-monocyte CSF [GM-CSF], interleukin 3 [IL-3], and granulocyte CSF [G-CSF]) on antibody-independent monocyte tumoricidal activity were investigated by using WEHI 164 fibrosarcoma cells as monocyte-sensitive targets. None of the CSFs directly induced monocyte cytotoxicity, although both GM-CSF and IL-3 were found to significantly enhance monocyte killing in response to a second stimulatory event (endotoxin). No effect was seen with G-CSF. Antitumor necrosis factor antibody completely abolished CSF-enhanced monocyte cytotoxicity, which suggests that this effect was mediated through increased release of tumor necrosis factor (TNF). As previously shown for GM-CSF, IL-3 was found to induce cytoplasmic accumulation of TNF messenger RNA (mRNA) after 18 hours of exposure. These results suggest that GM-CSF and IL-3 may stimulate monocyte killing indirectly by enhancing expression of TNF mRNA, thereby leading to augmented TNF protein secretion in response to a second activation signal.