The Levels of Granulocyte Colony-Stimulating Factor in the Plasma of the Bone Marrow Aspirate in Various Hematological Disorders

Transcription

1 The Levels of Granulocyte Colony-Stimulating Factor in the Plasma of the Bone Marrow Aspirate in Various Hematological Disorders Kenji Shinohara,a Eiichi Oeda,a Jun Nomiyama,a Hirofumi Inoue; Shinji Kamei,b Mitsuaki Tajirtb Takanori Ichikawa," Tomoki Kuwaki," Katsunori Tachibanac "Division of Hematology, Department of Medicine, Yamaguchi Prefecture Central Hospital, Hofu, Japan; bdepartment of Medicine, Shuto General Hospital, Yanai, Japan; 'Pharmaceutical Development Laboratory and Pharmaceutical Research Laboratory, Kirin Brewery Co., Ltd., Takasaki, Japan Key Words. Granulocyte colony-stimulating factor (G-CSF) Bone marrow Plasma Hematological disorder Abstract. We developed a sensitive method of measurement of granulocyte colony-stimulating factor (G-CSF) by an enzyme-linked immunosorbent assay, which we applied in the plasma of the bone marrow aspirate in 70 patients with various hematological disorders. The lowest limit of detection by this method is 2 pg/ml. G-CSF was detected in all but two of the patients. Compared to the G-CSF level in normal healthy controls, those in non- Hodgkin's malignant lymphoma, aplastic anemia, agranulocytosis and multiple myeloma were significantly higher, while the level in refractory anemia was not different. The G-CSF level in acute myelogenous leukemia patients was either elevated or decreased regardless of the French-American- British subgroup. The level in acute lymphoblastic leukemia was not different from the normal value, as was that in refractory anemia with an excess of blasts, and that in chronic lymphocytic leukemia. A patient with chronic myelomonocytic leukemia showed initial elevation of G-CSF with normalization after entering complete remission. The G-CSF level in chronic myelogenous leukemia was significantly decreased, although one patient in hematological remission who was under a-interferon therapy showed normal levels. The level in polycythemia Vera was not significantly different from the normal value. The G-CSF level for the entire group showed an inverse, although not statistically significant, correlation with the percentages of myeloid cells of the bone marrow (r = , p = 0.170, n = 80). These results are thought to reflect the regulatory mechanism of granulopoiesis in the bone marrow in various hematological disorders, Correspondence: Dr. Kenji Shinohara, Division of Hematology, Department of Medicine, Yamaguchi Prefecture Central Hospital, Hofu 747, Japan. Received January 18, 1995; provisionally accepted February 24, 1995; accepted for publication March 1, OAlphaMed Press /95/$5.OO/O and it is concluded that this method may be of dinical use in the treatment of patients with these disorders and in the selection of candidates likely to benefit from G-CSF administration. Introduction Granulocyte colony-stimulating factor (G-CSF) is secreted by the stromal macrophages, endothelial cells and fibroblasts in the bone marrow (BM) in response to several types of stimuli [l-1. Hematopoietic stem cells proliferate and differentiate in the BM, and mature cells are mobilized into circulation in response to hematopoietic growth factors. G- CSF acts not only as a circulating hormone but also acts locally as a paracrine hormone in areas such as BM. Although several methods of measuring peripheral blood serum G-CSF levels have been reported, the sensitivity of these methods is not high enough to allow detection of the concentration in all healthy controls and in patients with various hematological disorders [4-81. We developed a method of measuring G- CSF concentrations using the 2-step sandwich enzyme-linked immunosorbent assay (ELISA) [9], and measured those concentrations in the plasma of BM aspirate since there are no other reports of a measurement of G-CSF concentration in the BM plasma. This method is sufficiently sensitive to allow detection in the plasma of the BM aspirate in almost all patients with various hematological disorders. In comparison with those measured in the peripheral blood serum, the results obtained using this method are thought to more sensitively and directly STEM CELLS 1995;1:

2 Shinohara et al. 422 reflect the regulatory mechanism modulating the relationship between the production of G-CSF and the response of hematopoietic stem cells in the various pathological granulopoietic conditions, and to provide more information regarding the clinical significance of G-CSF in hematological disorders. Materials and Methods Patients The subjects included 70 patients with various hematological disorders, as follows: 12 with aplastic anemia (AA), each with non- Hodgkin s malignant lymphoma (ML) and acute myelogenous leukemia (AML), 9 with multiple myeloma (MM), 7 with refractory anemia (RA), 6 with chronic myelogenous leukemia (CML), each with agranulocytosis (Agra), acute lymphoblastic leukemia (ALL), refractory anemia with excess of blasts (RAEB), chronic lymphocytic leukemia (CLL) and polycythemia Vera (PV) and one with chronic myelomonocytic leukemia (CMML). In addition, plasma samples from the normal control were also evaluated. The AML patients consisted of 2 with MO, 1 each with MI and M2,4 with M4, 1 with M5 and 1 with hybrid leukemia according to the French-American-British (FAB) classification. The patients with ML were at the clinical stages I and I1 according to the Ann Arbor classification; some patients had received previous chemotherapy, and all of the patients were hematologically normal at the time of this study. The patients with AA and RA showed moderate to severe grades of severity; some of these received anabolic steroids and showed improvement in the Hb level, but not in leukocytes, granulocytes or platelets. None of the patients with MM, AML, ALL, RAEB, CMML, CLL, CML or PV had received any previous chemotherapy. Some of the patients with CML had received a-interferon, and one was in hematological remission. None of the patients had infection at the time of this study except two patients with Agra who had fever due to infection. Methods The BM was aspirated from the patient after informed consent was obtained. Since the BM aspirate plasma may become diluted with peripheral blood plasma, we attempted to aspirate only about 2 to ml of BM aspirate. It was difficult to obtain plasma in less than this volume of BM aspirate, because of the sticky nature of plasma, due to the presence of fat and fibrous tissue. We did not measure the G-CSF levels simultaneously in the peripheral blood in all of the patients, which had already been measured in several reported studies [4-81. In our preliminary study of simultaneous measurements in the peripheral blood plasma and BM plasma in a limited number of patients, G-CSF levels were well-correlated (r = 0.954, n = 12, data not shown). We did not quantify the number of cells in the BM aspirate since the method is not always reliable in view of the variable cellularity of BM [lo], but rather measured the percentage of myeloid cells in the BM smear to approximately measure the conditions of granulopoiesis. The heparinized BM plasma was obtained by centrifugation of the aspirate. The concentration of G-CSF was measured using a modified 2-step sandwich ELISA method [9]. Briefly, 96-well microliter plates were coated with anti-g-csf monoclonal antibody immunoglobulin G (IgG) and left overnight at 4 C. The plates were then washed and blocked by incubation with skim milk solution. After washing, BM plasma samples and G-CSF standard solutions were then added to the plates and incubated for 5 h at room temperature. The plates were then washed, and the solution containing anti-g-csf polyclonal antibody (FAB ) conjugated with P-galactosidase was added to the plates and further incubated overnight at 4 C. After washing, fluorescence substrate solution (4-methyllumbelliferyl-~-D-galactoside) was added to the plates and reacted for 6 h at 7 C. The reaction was stopped by addition of glycine- NAOH buffer (ph.), and fluorescence intensity was measured at the excitation wavelength of 50 nm and absorbance emission wavelength of 460 nm. The G-CSF level was determined with reference to the standard graph using known concentrations of recombinant G-CSF (Fig. 1). The variability of the assay was of minimum range. The data were analyzed according to the Wilcoxon 2-sample test. Results The lowest concentration of G-CSF in the BM plasma detectable by our method was found

3 42 The Levels of G-CSF in the Plasma of the Bone Marrow Aspirate in Various Hematological Disorders AML patients, there was no correlation 0.01 I Concentration of G-CSF (pglml) Fig. 1. The standard graph for determination of G-CSF concentration. Recombinant G-CSF of known concentration was used as the standard solution. to be 2 pg/ml. The G-CSF levels, mean and standard error of mean (SE) in normal controls and in each group of patients are shown in Table I. The level of BM plasma G-CSF in normal controls was pg/ml, and was higher than that from our results (.5 _+ 0.6 pg/ml) for the peripheral blood serum of normal controls (p < 0.01) [9]. The G-CSF level reported hereafter for each group is that compared with the level for normal controls. The G-CSF values in ML were significantly higher (p < 0.01). The G-CSF values in the patients with AA and MM showed a wide range, but the values were significantly higher (both, p < 0.01). Values in RA, ALL, RAEB, CLL and PV were not significantly different. The G-CSF value was extremely high in two of the three Agra < 0.05). Of the patients with AML, 2 with MO and 1 each with M4 and M5 showed an elevated G-CSF level, while 1 each with MI, M2 and hybrid leukemia and with M4 showed reduced G-CSF values. In the between the percentage of leukemic blasts and G-CSF level (Fig. 2). The G-CSF level in the CMML patient was high, but returned to normal (. pg/ml) after complete remission was entered. The G-CSF level for the CML patients was significantly reduced (p < 0.05), although one patient in hematological remission who was under a-interferon therapy had normal levels. In general, the G-CSF levels of the BM plasma in the patients with hematological disorders and normal controls were inversely correlated with the percentages of BM myeloid cells in the smears although not statistically significant (r = , p = 0.170, n = 80) (Fig. ). An inverse, although statistically not significant, correlation was also observed in AA patients (r = , p = , n = 12) (Fig. 4). Table I. The levels of G-CSF in the plasma of the BM aspirates in patients with various hematological disorders Diagnosis Number SE p value* Normal Control ML AA RA Agra MM AML ALL RAEB CMML CLL CML PV < 0.01 < 0.01 < 0.05 < 0.01 < 0.05 *compared to normal controls : not significant Data are expressed as mean * SE. The p values are those for comparison with the normal controls

4 Shinohara et al. 424 G-CSF(p(yml) r= p= 0.8 n= G-CSF(pg/ml) 1 ooooj 1 00 A 1 I= p= n= Bone Marrow Myeloid Cells (Yo) I, I I I,,,,, Bone Marrow Mveloblasts (%I Fig. 4. The G-CSF levels and the percentage of myeloid cells in the BM in the patients with AA. An inverse correlation was observed. Discussion Fig. 2. The G-CSF levels and the percentage of leukemic myeloblasts in the BM in patients with AML. MO, M1, M2, M4 and M5 denote the subgroups of AML according to the FAB classification. Hy denotes hybrid leukemia. No correlation was observed. G-CSF(W/ml) r= x Normal 0 ML AAP A 'mo~o p= ) I.t( *MI 0 am.i mu1 0 20,, ,, I, Bone Marrow Myeloid Cells (%) Fig.. The G-CSF levels and the percentage of myeloid cells in the BM in normal control and various hematological disorders. An inverse correlation was observed. The detection limit of previously reported methods for measuring peripheral blood serum ranges from pg/ml to 0-50 pg/ml [4-81, and with these methods the G-CSF level could be measured in only a minority of normal controls and in some, but not all patients, remaining undetectable in the rest. Because of the increased sensitivity of measurement of our method, the BM plasma G-CSF levels could be measured in all except two of the patients. The patients showed a wide range in the G-CSF levels in BM plasma. The levels in the patients with non-hodgkin's lymphoma were higher than the normal control levels. In most of the patients with ML, the peripheral blood serum G-CSF level was not high and was below the detection limit [6]. There may have been an effect of the past chemotherapies on granulopoiesis in some patients, since peripheral blood serum level has been observed to increase during chemotherapy or preconditioning-induced neutropenia and to decrease following the granulocyte recovery phase after chemotherapy [5, 61 and BM transplantation [7, 8, 1 I]. The G-CSF level of AA patients is reported to be elevated in the peripheral blood serum; of 6 patients had elevated levels [5], and the mean level was 260 pg/ml [6]. We also observed elevated values, but the G-CSF level in the BM

5 425 The Levels of G-CSF in the Plasma of the Bone Marrow Aspirate in Various Hematological Disorders plasma was not so high as compared with that in the peripheral blood serum. Subtle changes of the G-CSF concentration in the BM plasma may more directly reflect the effects of the process of granulopoiesis in the BM. The peripheral blood serum G-CSF level in the AA patients was inversely correlated with the absolute neutrophil count [6], while we also observed a similar tendency between the BM plasma G-CSF level and the percentages of myeloid cells in the BM aspirate smears. The clinical use of erythropoietin (Epo) in patients with AA has not always had successful results [ 121, since the endogenous production of Epo is extremely high in many of these patients; the mean value was one hundred times higher than the normal value 111. Our results in turn suggest that G-CSF may be clinically more useful than Epo, and that exogenously administered G-CSF may increase the blood G-CSF levels and stimulate granulopoiesis if responding hematopojetic stem cells remain. However, granulocyte-macrophage colony-stimulating factor (GM-CSF) has not always been effective when administered to patients with severe BM failure, and in most patients only a transient increase in leukocyte and granulocyte counts has been observed [ 14-17]. Prolonged administration of G-CSF alone [ 181 or in combination with either Epo [ 191 or other hematopoietic growth factors, such as stem cell factor, interleukin 1 (IL-1) and IL , or with immunosuppressive agents, such as cyclosporine or antithymocyte globulin [2-251, may hold promise in restoring hematopoiesis in the BM. Although 6 of 12 reported patients with myelodysplastic syndrome (MDS) had detectable or elevated levels in the peripheral blood serum [6], the subtype of MDS was not delineated. Our study revealed insignificant elevations of BM plasma G-CSF level in RA patients, but lower elevations than those found in AA patients, although another study revealed no elevation in the peripheral blood serum [26]. These results suggest the possibility that G-CSF is clinically useful in RA patients. Indeed, the administration of GM-CSF or G-CSF in patients with RA has been found to substantially increase the numbers of leukocytes and granulocytes during administration [27-11, although they decreased to the pretreatment level after discontinuation, indicating that maintenance treatment is necessary [0]. The G-CSF level in the MM patients in our study was moderately elevated while it was found to be elevated in another study [6]. The anemia in MM is well controlled by administration of Epo [2], as neutropenia may be by G-CSF. The G-CSF level in 2 of the Agra patients was extremely high. There is a report of increased G-CSF levels in patients with acute infection without hematological disorders []. Our data in Agra patients indicate that the nadir phase of granulopoiesis is associated with infection. The reason for the variation in G-CSF levels among AML FAB subgroups and the absence of any relationship between these levels and the percentages of myeloblasts in the BM remains to be solved. Leukemic myeloblasts have the receptor for G-CSF [4]. High peripheral blood G-CSF levels have been reported for the patients with M4 and MS subgroups [S]. There is a wide variation in the frequency of spontaneous colony formation in vitro, depending on the patient and FAB subgroup [5]. Leukemic myeloblasts secrete GM-CSF as well as G-CSF [6], and show autocrine growth [5-71 as well as paracrine growth 181 in response to the secreted cytokines. Our results may represent the diversity of response to G-CSF by each type of leukemic myeloblast in vitro. The G-CSF levels in ALL and CLL patients were normal, and the growth of these cells may not be affected by G-CSF. In our patient with CMML, the G-CSF level was high before chemotherapy but returned to normal after the patient entered complete remission. The conditioned medium of the CMML cells contained elevated amounts of IL-6 and GM-CSF, suggesting that these act in vitro as autocrine growth factors for CMML cells [9, 401. Our results suggest that G-CSF may also be involved in this process. Most of our patients with CML showed significantly decreased BM plasma G-CSF levels, and one patient who had a normal level was under a-interferon therapy and in hematological remission, as in another report [6]. The peripheral blood serum G-CSF level has been reported to be undetectable in almost all CML patients, except a few with elevated G-CSF levels [S, 61. The reason for this difference is unknown. The significantly decreased G-CSF levels in these disorders indicate the presence of a negative feedback mechanism in the control of granulopoiesis. The G-CSF level in the PV patients was not significantly decreased. In general, the BM plasma G-CSF level is inversely correlated with the percentages of

6 Shinohara et al. 426 myeloid cells in the BM smear, as peripheral blood serum G-CSF is inversely correlated with the absolute neutrophil count in patients with AA [6] and as we found in the BM plasma. These results for the BM plasma G-CSF level may more directly reflect the pathophysiological mechanisms of disordered granulopoiesis than do those for the G-CSF level of the peripheral blood plasma. We conclude that our method of G-CSF measurement may facilitate the understanding of hematological disorders from the standpoint of granulopoiesis and the selection of candidates who are likely to benefit from the clinical use of G-CSF. References 1 Zsebo KM, Yuschenkoff VN, Schiffer S et al. Vascular endothelial cells and granulopoiesis: interleukin 1 stimulates release of G-CSF and GM-CSF. Blood 1988;71: Fiebbe WE, van Damme J, Billaian A et al. Interleukin I induces human marrow stromal cells in long-term culture to produce granulocyte colony-stimulating factor and macrophage colonystimulating factor. Blood 1988;71: Yang Y-C, Tsai S, Wong GG et al. Interleukin 1 regulation of hematopoietic growth factor production by human stromal fibroblasts. J Cell Physiol 1988;14: Sallerfors B, Olofesson T. Granulocytemacrophage colony-stimulating factor (GM-CSF) and granulocyte colony-stimulating factor (G-CSF) in serum during induction treatment of acute leukemia. Br J Haematol 1991;78: Omori F, Okamura S, Shimoda K et al. Levels of human serum granulocyte colony-stimulating factor and granulocyte-macrophage colony-stimdating factor under pathological conditions. Biotherapy 1992;4: Watari K, Asano S, Shirafuji N et al. Serum granulocyte colony-stimulating factor levels in healthy volunteers and patients with various disorders as estimated by enzyme immunoassay. Blood 1989;7: Haas R, Gericke G, Witt B et al. Increased serum levels of granulocyte colony-stimulating factor after autologous bone marrow or blood stem cell transplantation. Exp Hematol 199;21: Miksitis K, Beyer J, Siegert W. Serum concentrations of G-CSF during high-dose chemotherapy with autologous stem cell rescue. Bone Marrow Transplant 199;11: Ichikawa T, Kuwaki T, Tachibana K et al. Establishment of a highly sensitive enzyme immunoassay for granulocyte colony-stimulating factor in human plasma. Exp Hematol (in press). Dacie JV, Lewis SM, eds. Practical Haematology. Edinburgh: Churchill Livingstone, Cairo MS, Suen Y, Sender L et al. Circulating granulocyte colony-stimulating factor (G-CSF) levels after allogeneic and autologous bone marrow transplantation: endogenous G-CSF production correlates with myeloid engraftment. Blood 1992;79: Bessho M, Jinnai I, Matsuda A et al. Improvement of anemia by recombinant erythropoietin in patients with myelodysplastic syndromes and aplastic anemia. Int J Cell Cloning 1990;8: Urabe A, Saito T, Fukamachi H et al. Serum erythropoietin titers in the anemia of chronic renal failure and other hematological states. Int J Cell Cloning 1987;5: Vadhan-Raj S, Buescher S, Broxmeyer HE et al. Stimulation of myelopoiesis in patients with aplastic anemia by recombinant human granulocyte-macrophage colony-stimulating factor. N Engl J Med 1988;19: Nissen C, Tichelli A, Gratwohl A et al. Failure of recombinant human granulocyte-macrophage colony-stimulating factor therapy in aplastic anemia patients with very severe neutropenia. Blood 1988;72: Champlin RE, Nimer SD, Ireland D et al. Treatment of refractory aplastic anemia with recombinant human granulocyte-macrophagecolony-stimulating factor. Blood 1989;7: Guinan EC, Sieff CA, Oette DH et al. A phase YII trial of recombinant granulocyte-macrophage colony-stimulating factor for children with aplastic anemia. Blood 1990;76: Kojima S, Fukuda M, Miyajima Y et a]. Treatment of aplastic anemia in children with recombinant human granulocyte colony-stimulating factor. Blood 1991;77: Bessho M, Toyoda A, Ito Y et al. Trilineage recovery by combination therapy with recombinant human granulocyte colony-stimulating factor (rhg-csf) and erythropoietin (rhepo) in severe aplastic anemia. Br J Haematol 1992;80: Williams DA. Ex vivo expansion of hematopoietic stem and progenitor cells. Blood 199;81: Srour EF, Brandt JE, Briddell RA et al. Long-term generation and expansion of human primitive

7 427 The Levels of G-CSF in the Plasma of the Bone Marrow Aspirate in Various Hematological Disorders hematopoietic progenitor cells in vitro. Blood 199;81 : Muench MO, Firpo MT, Moore MAS. Bone marrow transplantation with interleukin l plus kitligand ex vivo expanded bone marrow accelerates hematopoietic reconstitution in mice without the loss of stem cell lineage and proliferative potential. Blood 199;81: Kojima S, Fukuda M, Miyajima Y et al. Cyclosporine and recombinant granulocyte colony-stimulating factor in severe aplastic anemia. N Engl J Med 1990;2: Bertrand Y, Amri F, Capdeville R et al. The successful treatment of two cases of severe aplastic anemia with granulocyte-colony stimulating factor and cyclosporine A. Br J Haematol 1991 ;79: Doney K, Storb R, Appelbaum FR et al. Recombinant granulocyte-macrophage colony stimulating factor followed by immunosuppressive therapy for aplastic anemia. Br J Haematol 199;85: Verhoef GEG, Schouwer PD, Ceuppens JL et al. Measurement of serum cytokine levels in patients with myelodysplastic syndromes. Leukemia 1992;6: Vadhan-Raj S, Keating M, LeMaistre A et al. Effects of recombinant human granulocytemacrophage colony-stimulating factor in patients with myelodysplastic syndromes. N Engl J Med 1987;17: Ganser A, Volkers B, Greher J et al. Recombinant human granulocyte-macrophage colony-stimulating factor in patients with myelodysplastic syndromes-a phase YII trial. Blood 1989;7: Kobayashi Y, Okabe T, Ozawa K et al. Treatment of myelodysplastic syndromes with recombinant human granulocyte colony-stimulating factor: a preliminary report. Am J Med 1989;86: Negrin RS, Haeuber DH, Negler A et al. Maintenance treatment of patients with myelodysplastic syndromes using recombinant human granulocyte colony-stimulating factor. Blood 1990:76: Yoshida Y, Hirashima K, Asano S et al. A phase I1 trial of recombinant human granulocyte colonystimulating factor in the myelodysplastic syndromes. Br J Haematol 1991;78: Ludwig H, Fritz E, Kotzrnann H et al. Erythropoietin treatment of anemia associated with multiple myeloma. N Engl J Med 1990;22: Kawakami M, Tsutsumi H, Kumakawa T et al. Levels of serum granulocyte colony-stimulating factor in patients with infections. Blood 1990;76: Piao YF, Okabe T. Receptor binding of human granulocyte-colony stimulating factor to the blast cells of myeloid leukemia. Cancer Res 1990;50: Young DC, Wagner K, Griffin JD. Constitutive expression of the granulocyte-macrophage colonystimulating factor gene in acute rnyeloblastic leukemia. J Clin Invest 1987;79: Young DC, Griffin JD. Autocrine secretion of GM-CSF in acute myeloblastic leukemia. Blood 1986;68: Murohashi I, Tohda S, Suzuki T et al. Autocrine growth mechanisms of the progenitors of blast cells in acute myeloblastic leukemia. Blood 1989;74: Oster W, Cicco NA, Klein H et al. Participation of the cytokines interleukin 6, tumor necrosis factora, and interleukin I-p secreted by acute myelogenous leukemia blasts in autocrine and paracrine leukemia growth control. J Clin Invest 1989;84: Everson MP, Brown CB, Lilly MB. Interleukin 6 and granulocyte-macrophage colony-stimulating factor are candidate growth factors for chronic myelomonocytic leukemia cells. Blood 1989;74: Akashi K, Shibuya T, Harada M et al. Interleukin 4 suppresses the spontaneous growth of chronic myelomonocytic leukemia cells. J Clin Invest 1991 ;88:22-20.