Colony-stimulating Factor on Primary Human Tumors and Assessment of the Effects of Granulocyte-Macrophage In Vitro

One hundred eighty-nine human tumor specimens were tested in a human tumor cloning assay to determine their growth response to human recombinant granulocyte-macrophage colony-stimulating factor. Of these samples 48 were évaluable for response. Growth stimulation to >150% of controls was noted in 1 of 12 lung cancers (8%) and 1 of 14 breast cancers (7%) but in no other instances for an overall rate of 2 of 48 (4.2%). A dose-response effect was not seen with each of the two stimulated samples responding only at the two lowest concentrations tested. In addition, 7 cell lines derived from human tumors were tested using a metabolic CO«production assay without evidence of growth stimulation. Samples of normal bone marrow displayed the usual dosedependent stimulation whether grown in agar or assayed metabolically. We conclude that human recombinant granulocyte-macrophage colonystimulating factor has minimal effect on the growth of the solid tumors tested and that clinical trials to reduce chemotherapy-associated myelosuppression may proceed without undue concern for enhancement of tumor growth. INTRODUCTION GM-CSF3 is a cytokine produced by a wide array of body tissues which stimulates the proliferation and differentiation of myeloid progenitors (1). In addition to endothelial cells and fibroblasts in the bone marrow stroma, activated T-cells have been shown to secrete GM-CSF (2). Now available in a human recombinant form this agent has been used in disorders such as aplastic anemia and acquired immunodeficiency syndrome with dramatic increases in numbers of circulating neutrophils, eosinophils, and monocytes (3, 4). Several studies have also confirmed the utility of rhGM-CSF in association with intensive chemotherapy in reducing both the severity and duration of induced neutropenia (5-7) and have laid the groundwork for further clinical trials. The optimal dosage is not universally agreed upon but appears to be in the 3to 15-jig/kg/day range with corresponding peak blood levels as high as 20 ng/ml and sustained levels in the 1to 10-ng/ml range (5-9). As our use of rhGM-CSF expands, a significant concern must be the potential stimulatory effect on malignancies we seek to expunge. Disorders whose clonal participants are primary tar gets for this protein, such as acute myelogenous leukemia and myelodysplasia states, seem at greatest hazard. Although in vitro data suggest that a prominent differentiation effect occurs (2), in one clinical trial in patients with myelodysplasia, GMCSF may have induced frank leukemia in 2 of 13 patients (10). Received 3/20/90; accepted 6/29/90. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ' Support provided by the Anne Guttentag Fund. : To whom requests for reprints should be addressed, at UTHSC-SA, Depart ment of Medicine, Division of Oncology, 7703 Floyd Curl Drive, San Antonio, TX 78284-7884. 3The abbreviations used are: GM-CSF, granulocyte-macrophage colony-stim ulating factor; rhGM-CSF, recombinant human GM-CSF; G-CSF, granulocyte colony-stimulating factor; HTCA, human tumor-cloning assays. In contrast, nonhematopoietic normal body tissues have not been shown to be responsive to GM-CSF or to have surface receptors. However, several reports suggest that a small per centage of human tumors may indeed respond, some with accentuation and others with inhibition of growth (11-19). Baldwin et al. (11) reported the presence of G-CSF and GMCSF receptors on the surface of small cell lung cancer cells. Ruff et al. (12) and Yamashita et al. (13) demonstrated the inhibition of such cells at high concentrations of GM-CSF. A wide range of human tumor-derived cell lines have been tested with moderate stimulation noted in some instances (14-15), while others have been unable to document increased prolifer ation (16). Most recently two reports have described the re sponse of primary human tumors to rhGM-CSF in in vitro HTCA (17, 18). Jorashkewitz et al. (17) noted growth stimu lation in 3 of 7 samples in contrast to 4 of 33 specimens reported by Salmon and Liu (18). The latter authors also noted growth inhibition in 10 of their studies with marked inhibition occurring in 3 cases. The data presented here extend these observations but suggest less growth modulatory effects of rhGM-CSF than have been reported previously. MATERIALS AND METHODS Primary Tumor Samples. One hundred eighty-nine consecutive hu man tumor specimens submitted to the Cancer Treatment Research Center, San Antonio, TX, for the assessment of in vitro response to chemotherapeutic agents were also tested for their response to rhGMCSF. The specimens were collected between September 1987 and December 1990 and were submitted either as fresh solid tumors, pleural effusions, ascites, or bone marrow aspirates. Solid tumors were minced and forced through no. 50 steel mesh and needles of successively increasing gauge to generate single cell suspensions prior to assessment on the cloning assay. All samples were granted exempt research status by the Institutional Review Board of the University of Texas Health Science Center at San Antonio. Human Tumor-Cloning Assay. As previously described (19)5 X 10* cells were suspended in 1 ml of RPMI 1640 culture medium containing 10% fetal calf serum and 0.3% agar and plated over an equivalent volume of feeder layer in 0.5% agar. The concentration of rhGM-CSF was adjusted to 0.1, 1.0, or 10 ng/ml in the plated cell suspension to allow for continuous exposure. Control and treatment samples were plated in triplicate and assessed 12-14 days after plating for colony formation. Colonies were defined as cell clusters containing >50 cells. To assure the presence of a single cell suspension, positive controls were prepared using 200 fig/ml of vanadium (Na3VO4) which com pletely inhibits colony growth. In order for an experiment to be consid ered évaluable, untreated control plates must yield >20 colonies/plate and vanadium-treated plates must have <30% of the number of colonies on untreated controls. Inhibition was defined as a reduction in growth to <50% of control. No effect was reported when growth was 50-150% of control and stimulation was noted when 150% of control colony formation was exceeded. GM-CSF. Human recombinant GM-CSF (Escherichia coli, nonglycosylated, Schering-Plough/Sandoz) (Leucomax) was obtained from Schering-Plough Research (Bloomfield, NJ), lot 20605-75. It exhibited with a specific activity of 2 x 10s units/mg protein.