key: cord-006444-eq56zhtd authors: nan title: Abstracts of oral presentations and posters date: 1993 journal: Ann Hematol DOI: 10.1007/bf01695978 sha: doc_id: 6444 cord_uid: eq56zhtd nan Since the blood cells are primarily concerned with host defense, the introduction of the CSFs as therapeutic agents offers the opportunity to develop unique therapeutic strategies designed to enhance overall host defense, particularly with relevance to cancer and AIDS. Administration of CSFs is associated with profound changes in cellular function, and treatment strategies will need to consider the potential deleterious effects of heightened host cell activity and potential effects on nonhematopoietic cells. Memorial Sloan-Kettering Cancer Center, New York, NY 10028, USA A75 3 CYTOKINE The majority of hemopoietic stem cells in the steady state marrow are dormant in the cell cycle. Using serial observation (mapping) of blast cell colony formation from bone marrow cells of mice that have been treated with a high dose 5-fluorouracil (5-FU), we have identified a number of cytokines that appear to control the cell cycling of the primitive hemopoietic progenitors. The early-acting cytokines may be divided into three groups. The first group would consist of IL-3, GM-CSF, and IL-4. The second group consists of IL-6, G-CSF, IL-11, IL-12, and LIF/DIA. The third group consists of steel factor (SF). According to our studies in culture, cytokines in each group can interact with those in other groups to initiate cell division in the cell cycle dormant primitive progenitors. While studies using retrovirally-labeled murine stem cells demonstrated unequivocally the presence of lymphohemopoietic progenitors that are capable of producing both lymphoid and myeloid progenies, it has not been possible to identify and quantitate these progenitors in culture. Recently, we have developed a two-step methylcellulose culture method to quantitate murine lymphohemopoietic progenitors that are capable of producing myeloid cells and and pre-B-cells. After establishing the primary culture system initially with medium conditioned by pokeweed mitogen stimulated spleen cells, we characterized combinations of cytokines that are able to maintain the B-lymphoid potentials of the primary colonies. We observed that two-factor combinations including SF such as SF plus IL-6, SF plus G-CSF, SF plus IL-11, SF plus IL-12 were effective in maintaining the proliferation of B-cell progenitors. Somewhat less effectively IL-4-based combinations such as IL-4 plus IL-6 and IL-4 plus IL-11 also supported the B-cell potentials of the the primary colonies. Interestingly, IL-3-based combinations were unable to maintain the B-cell potentials of the primary colonies even though the cells in myeloid lineages proliferated strongly. We also found that addition of IL-3 to an effective two-factor combination such as SF plus IL-11 inhibit the B-cell potentials of the primary colonies. Our cell culture for the murine lymphohemopoietic progenitors may provide an important tool for studying the mechanisms regulating the early process of lymphopoiesis. The survival and proliferation of haemopoietic stem cells, induced by growth factors, occurs concomitantly with differentiation and developmeat of the stem cells and their progeny, into mature blood cells. Are the growth factors simply permissive for these processes? Or do they have an inductive role to play in lineage commitment of stem cells and subsequent maturation into phenotypically mature cells? From various experiments, it is clear that the outcome of the response (that is the types of mature cells produced) is a reflection of the range of growth factors to which the cells are exposed-suggesting that combinations of growth factors may well influence the choice of lineage options taken by multipotent cells. The problem, of course, is that in all systems studied to date, no detailed examination has been made of cell death, and that even in growth factor combinations where, for example, no erythropoiesis is found, it is possible that erythroid progenitors are being produced as a consequence of differentiation of the multipoteni cells but that these cells are then dying due to lack of availability of the appropriate survival (growth factor) stimulus. We have recently been able to circumvent some of these problems and have shown that, orovided the cells receive a survival stimulus, differentiation can occur hi the absence of added growth factors and also that proliferation it not a prerequisite for acquisition of a mature cell phenotype. In other words, the growth factors may act primarily as survival and mitogenic stimuli and not as "inducers" of differentiation. Expression of p210BCR/ABL by transfection converts interleukin-3 (IL-3)-dependent cell lines to factor independence and transforms immature hematopoietic cells in vitro. We tested the hypothesis that p210 BCR/ABL may induce factor-independence by constitutively activating signal transduction pathways which are normally regulated by IL-3/GM-CSF. In both the IL-3-dependent murine myeloid cell line, 32Dcl3, and the IL-3/GM-CSF-dependent human line, MO7e, p210 BCR/ABL induces rapid factor-independence despite continuous growth in IL-3-containing medium. One-and two-dimensional antiphosphotyrosine immunoblotting showed that most proteins tyrosine phosphorylated by p210BCR/ABL are different that those phosphorylated in response to IL-3. Several signaling molecules have been found to be activated or phosphorylated by both IL-3/GM-CSF and p210BCR/ABL, including Raf-1, MAP kinase, SHC, vav, and probably PI3K. Other signal transducing proteins were found to be phosphorylated only by pZ10BCR/ABL (p120rasGAP, two rasGAP associated proteins, and c-fes), or only by . In order to better define the biochemical activities of p210BCR/ABL which lead to mitogenesis, a series of cell lines were constructed in which the functional expression of pZ10BCR/ABL was inducible. The uninduced cell lines had a wild-type phenotype while the induced cell lines displayed markedly reduced apoptosis in the absence of growth factor, and some were hyper responsive to growth factors. The phenotypes of these cell lines have been stable in culture, and the lines should be useful to define biochemical activities of p21 Obcr/abl which are important for mitogenesis. R.E. Donahue, M.R. Kirby, P.D. Lawman, S.E. Sellers, S.W. Kessler, and M.J.P. Lawman A novel factor has been purified to homogeneity from a cell line derived from a human mixed germ cell tumor. By Western blot analysis, using a polyclonal rabbit antibody raised to the purified native protein, SCPF was found to be expressed both as a 32kD secreted and as a 37kD membrane-bound protein. To further evaluate SCPFs' ability to support CD34+ cell growth in culture, SCPF was used in short-and long-term cultures using immunoselected CD34 + cells. For short-term culture studies, CD34+ cel~s were evaluated prior to and subsequent to a six day exposure to either media alone or media supplemented with IL-3, SCPF, or IL-3+SCPF. The greatest expansion of CD34+ cells was in those expressing OD38. Compared to pre-culture, cultures maintained in SCPF, IL-3, or IL-3+SCPF had, respectively, a 0.7, 2.9, or 0.7-fold increase in CD34+CD38+ numbers. There was also a consistent increase in the ratio of large CD34+38+ cells to small CD34+CD38+ cells. Presumptively, this change represents an increase in the number of cells in either the G2/M or S phase of the cell cycle. Of the cytokine combinations evaluated, only the combination of IL-3 + SCPF led to a t .8-fold expansion of CD34 + CD38-Cells above baseline values. By themselves, SCPF and IL-3 led to a 0.4 and 0.7-fold reduction in CD34+CD38-numbers. When compared, however, to the number of CD34 + CD38-cells present in media alone after the 6 days in culture, SCPF, IL-3, and IL-3 + SCPF hat respectively a 3.0, 5.7, and 12.3-fold greater number of CD34 +CD38-cells. In long-term culture assays in the absence of SCPF, the cultures deteriorated rapidly and were test by day 18. in the presence of SCPF, cell numbers were maintained over the initial 8-14 days in culture, with proliferation becoming evident 16 days post-culture. At day 21, some of the cells were removed from culture media containing SCPF and replatad in culture media alone. After an additional 30 days in culture the cells that were no longer exposed to the SCPF had differentiated. Interestingly, the cells that were cultured in SCPF continued to proliferate. After 50 days in culture, these cells were predominantly CD34 +, CD33 +, CD38 +, CD45 +, CD71 +, and HI.A-Dr +, and failed to express Thy t, CD4, CD8, CD14, OD20, or CD56. Karyotypic analysis demonstrated that these cells hat multiple chromosomal aberrations, Civin, A. Gewirtz, P. Rockwell, L. Witte We have cloned the cDNA for STK-1 (Stem cell Tyrosine Kinase 1), a human growth factor receptor tyrosine kinase, and investigated its expression in bone marrow, leukemias, and leukemic derived cell lines. STK-1 expression is restricted to the CD34+ fraction of normal human bone marrow, the fraction containing all of the hematopoietic stem cell activity of marrow. Experiments in which CD34+ cells grown on irradiated bone marrow stromal feeder layers were exposed to STK-1 antisense oligonucleotides resulted in inhibition of colony formation. STK-1 is also expressed in most cases of AML, B lineage ALL, and T cell ALL. A number of hematopoietic tissue culture cell lines which express STK-1 have been identified, including KMT2, KGla, KG1, ML-1, HL-60, Nalm-16, and REH. ML-1 cells stop growing and differentiate after exposure to phorbol esters and other agents. STK-1 expression is completely shut off by this treatment. Antipeptide antibody generated against several regions of STK-1 identifies a doublet of proteins of 155kD and 130kD, probably corresponding to different degrees of glycosylation, in several of the cell lines. These data imply a possible role for the STK-1 receptor in the normal proliferation of hematopoietic stem cells and the abnormal proliferation of leukemic cells. Further antisense experiments and the isolation of the growth factor for this receptor will be necessary to fully understand its role in hematopoiesis and leukemia. Murine long term repopulation and double transplantation assays have clearly demonstrated irreversible damage to early stem cells by repeated doses of alkylating agents or anti-metabolites. The ability to protect from the short term effects of chemotherapy-induced myelosuppression by administration of hematopoietic growth factors has obscured the potential problem of long term stem cell insufficiency. Indeed in murine models involving repeated cyclophosphamide administration, CSF administration has been reported to protentiate stem cell damage. The development of techniques for isolation of human hematopoietic stem cells in a CD34 + Lin -re fraction of marrow and blood, together with long term culture-initiating assays on marrow stroma or long term ex vivo expansion with cytokine combinations, permits quantitative analysis of human stem cell proliferation potential. It is becoming apparent that extensive chemotherapy treatment gravely compromises the population of primitive hematopoietic stem cells as reflected in their impaired capacity to peripheralize and to be represented in the blood CD34 + population following CSF treatment with or without cytoxan. Five strategies are currently under evaluation: 1) Upfront harvesting of marrow and/or elicited peripheral blood prior to onset of chemotherapy with subsequent "rescue" following chemotherapy; 2)Fine tuning of cytokine and chemotherapy administration to take advantage of "rebound quiescence" of stem cells; 3)Administration of negative regulators to suppress stem cell proliferation. Transforming growth factor 6, macrophage inflammatory protein e and tumor necrosis factor have all proved protective in preclinical models. 4) Utilize cytokines, eg IL-1, that protect stem cells by increasing drug enzymatic inactivation, decreasing drug influx and/or increasing drug efflux, and inducing DNA repair or decreasing DNA damage. 5) Ut ze gene therapy to introduce into hematopoietic stem cells drug resistance genes such as mutated dihidrofolate reductase that confers methotrexate resistance or enhance the expression of the multi-drug resisting gene (MDR) expression. Interleukin-11: A novel hematopoietic cytokine possessing multiple biological activities in vitro and in vivo. J.P. Leonard and S.J. Goldman, Genetics Institute, Cambridge, MA. Interleukin-1 1 is a multifunctional hematopoietic cytokine which was originally identified in the conditioned medium from an IL-1 stimulated primate stromal cell line. The human cDNA was subsequently cloned from a fetal fibroblast cell line enabling the expression and purification of the human protein. The in vitro biological activities of rhlL-11 result predominantly from synergistic interactions with other growth factors. In combination with other cytokines, rht[,-1 1 has been shown to support the formation of primitive human blast ceil coIonies from bone marrow, promote erythroid burst formation and stimulate both early and late stages of megakaryocyte proliferation and differentiation. In addition, rhlL-11 alone directly increased the size and ploidy of enriched megakaryocytes. Although rhlL-1 1 has no inherent 13 cell growth factor activity, rhlL-1 1 has been shown to stimulate immunoglobulin producing B cells both in vitro and in vivo. rhlL-I 1 is biologically active in mice, rats, dogs and primates when administered as a single agent in vivo. The predominant effect of rhlL-1 1 in naive animals was on cells of the megakaryocyte lineage, increasing the number of bone marrow megakaryocyte progenitors, stimulating megakaryocyte endoreplication and increasing peripheral platelet counts in a dose dependent fashion. In a variety of murine rtiodels of myelosuppression, the effects of rhtL-1 1 were multitineage, stimulating the recovery of megakaryocyte, erythroid, and granulocyte and macrophage progenitors in the bone marrow, rhlL-I 1 administration reduced the platelet and hematocrit nadirs and the overall duration and severity of thrombocytopenia and anemia in these models. In a murine bone marrow transplant model, rhlL-1 1 also accelerated neutrophil recovery. The results from ongoing preclinical studies continue to confirm the broad spectrum of biological activities possessed by rhiL-1 1 in vitro and suggest this cytokine may be an effective agent in the treatment of myelosuppression associated with cancer chemotherapy and bone marrow transplantation. PRECLINICAL BIOLOGY OF HUMAN IL-10 T.L. Nagabhushan, S.K. Narula and M.I. Siegel Human IL-10 (hulL-10) is a 160 amino acid polypeptide synthesized by a number of different cell types. It is a pleiotropic factor with both immunosuppressive and immunostimulatory activity. Recombinant hulL-10 expressed in CHO cells is not glycosylated, and when expressed in E. coil the protein retains the biological activity of the CHO-derived product. In the presence of antigen presenting cells, hulL-10 inhibits cytokine synthesis in T cells. HulL-10 downregulates ,f-IFN and IL-4 induced MHC Class II antigen expression on monocytesmacrophages. In contrast, it has no effect on Class II expression on purified tonsillar and peripheral B cells. HulL-10 also inhibits the synthesis of IL-1% IL-1/L IL-6, IL-8, TNF-m GM-CSF and G-CSF at the protein and RNA levels in monocytes activated by LPS or LPS and .r The y-IFN or GM-CSF induced phagocytosis of opsonized yeast particles by human peripheral blood-derived macrophages and granulocytes is downregulated by hulL-10. IL-4 induced IgE synthesis by PBMNC is inhibited by hulL-10. The protein also potentiates a strong LAK activity in human PBMNC. These results will be discussed along with some early in vivo biology in rodents. A monomeric pentapeptide (pEEDCK) inhibits murine hematopoiesis in vitro and in vivo while its dimer counterpart (pEEDCK): formed through a disulphide bridge has a stimulatory effect in the same systems. Stable peptides of the rimer have been made by substitution of the disulphide bridge with a dimethylene bridge. In vitro the dimer seems to have no direct effect on GM-CFC or on purified Lin-Scal + cells in the HPP-CFC assay (400 cells per culture, 15-20% PE). The monomer does not affect GM-CFCs, but inhibits approximately 50% of the purified HPP-CFCs. The dimer stimulates human or mouse stromal cells to produce M-CSF and possibly other cytokines which augment colony formation in vitro and also activate human PBL as measured by an increased expression of CD1 lb. In ex vivo experiments it was found that the dimer increases and the monomer decreases cell cycle rate of CFU-GM and CFU-S in the bone marrow. Intraperitoneal injection of the dimer (1 to 100 ng/kg) into mice, led to increased progenitor cell number in the bone marrow and also increased survival of mice given a lethal dose of cyclophosphamide (550 mg/kg). At present receptors of the peptides have not been identified. These studies have shown that the dimer has an indirect effect on hematopoiesis as a stimulator of cytokine production, while the monomer seems to act both as an antagonist of dimer action and is also able to directly inhibit early myeloid progenitor cells. The possibility that these two compounds have therapeutic efficacy in diseases involving a myelosuppressed bone marrow is indicated. There is evidence that basic fibroblast growth factor (bFGF) plays a role in the regulation of normal blood cell proliferation and differentiation. Basic FGF is produced by and is a potent mitogen for human slromal cells. It is found in megakaryocytes and cells of the granulocyte lineage, in vivo, and it enhances megakaryopoiesis and myelopoiesis in human long term bone marrow cultures. It stimulates progenitor cell growth and augments the proliferation of progenitor cells when added in conjunction with other hematopoietic growth factors. In addition, it counteracts the suppressive effects of transforming growth factor beta. Basic FGF also synergizes with stem cell growth factor to augment granulocyte macrophage colony stimulating factor stimulated progenitor cell growth. Based on the observations in normal hematopoiesis, the role of bFGF in malignant bematopoiesis is presently an area of active investigation. SCF is one of the earliest-acting hematopoietic growth factors. Pre-clinical studies have demonstrated its multi-lineage hematopoietic effects. We have conducted a phase I trial of SCF and report now on the first 21 patients (pts) with stage IIIB (n=7) or IV (n=14) breast cancer. The study was designed to evaluate the toierabllity and biologic effects of SCF. Pts received SCF prechemo (cycle O) and following cycles 2-6 of C/A chemotherapy. Cohorts of 5 lOtS were randomized in a 4:1 ratio to receive either SCF (n=17) by subcutaneous injection at dosages of I 0, 25, and SO/~g/kg/day for 14 days or no SCF as a parallel C/A control group (n=4). Various physiologic, biochemical, pharmacokinetic, and hematologic parameters were studied. Bone marrow (BM) and peripheral blood (PB) progenitors were assayed. SCF administration was associated with PB progenitor mobilization in cycle 0 at all dose levels. Absolute neutrophil counts demonstrated modest doserelated increases over baseline ranging from median values of 20% (10 pg/kg/d) to 120% (SO/Jg/kg/d). No reproducible effects were seen on red blood cells or platelets. The principal adverse events were derrnatologic in nature. They included mild to moderate reactions at the injection site in all pts, and moderate to severe reactions distant to the site in I0 pts (primarily at higher doses). At 50 pg/kg/d, 4/10 pts experienced dose-limiting respiratory symptoms including cough, hoarseness, and laryngospasrn. Because of SCFs known mast cell effects, prophylaxis with H1/H2 blockers and bronchodilators is being evaluated. No pts developed antibodies to SCF. Evaluation of effects following chemotherapy is ongoing. We conclude that SCF is an active hematopoietin capable of stimulating production of BM and PB progenitor cells as well as peripheral neutrophils. G. Trinchieri NKSF/IL-12 is a heterodimeric cytokine produced by monocytemacrophages, B cells and other accessory cells in response to various stimuli including bacteria and bacterial products. NKSF/IL-12 acts on T cells and NK cells inducing cytokine production, proliferation, and enhancement of cytotoxic activity. NKSF/IL-12 is a particularly efficient inducer of IFN-7 production, acting alone or in synergy with other IFN-7 inducers such as IL-2, antigens, anti-CD3 or anti-CD16 antibodies, mitogens, and phorbol diesters. NKSF/IL-12appears to play a major role in regulation of natural resistance: when produced by monocyte-macrophages, it directly activates the cytotoxic activity of NK cells and induces both NK and T cells to produce IFN-7 and other cytokines with important effects on activation of phagocytic cells. The important role of NKSF/IL-12 in response to bacterial products is clearly demonstrated by the ability of anti-NKSF/IL-12 antibodies to inhibit in vivo in mice the iFN-y production induced by LPS ii~ a toxic shock model. NKSF/IL-12 is also an important factor in the regulation of adaptive immune response, by inducing the differentiation and growth of T helper cells type 1 (Th-1) and by preventing the differentiation of IL-4 producing Th-2 cells. A possibly obligatory role of NKSF/IL-12 for Th-1 cell differentiation can be demonstrated in vitro in the human lymphocyte response to allergens or bacteria-derived antigens and both in vitro and in vivo in the murine system, e.g. in the immune response to Leishmania infection. Production of NKSF/IL-12 by accessory ceils is stimulated by IFN-7, a product of Th-1 cells and suppressed by IL-10 or IL-4, products of Th-2 cells. These results suggest that NKSF/IL-12 represents an important link between natural resistance and adaptive immunity and is at the center of a cytokine network that regulates the equilibrium between cellular and humoral immunity. Work with alpha-interferon in CML started at MDACC in 1981 and has evolved over the years to allow us interpretation of response rate in a large number of patients with relatively long follow up. Newly diagnosed patients ( 3 pts per dose step. Before CT rhlL-6 is given dl iv and d2-7 sc. CT (mitoxantrone 10 mg/m 2 and thiotepa 40 mg/m 2, iv dl) starts d15, q 3 wks. After CT rhlL-6 is given d4-15 se in all cycles (c). Before CT 16 pts were treated. RML~ related side effects were: fever WHO grI-II (100%), headache (56%), myalgia (25%), local erythema (100%) and starting at 2.5 #g nausea (56%), increase of liver enzymes WHO grI-II (44%) and anemia (33%). Symptoms were controllable with acetaminophen (n=16) and antiemetics (n=2). Upto 10 #g them was a dose response related platelet increase compared to dl (r=0.75, p<0.002), with highest count after rML-6 cessation. At 0.5 /~g max. increase was 25.7+5.6 (SE), at 10 ~g 114.0+38.0%. On d3 a leukocyte increase compared to dl was observed starting at 2.5 #g upto 10 /~g (dl 5.80+0.51, d3 8.76+0.45, p<0.001) mainly due to neutrophil increase (p<0.001). At d8 increase of lymphocytes occurred compared to dl (dl 1.02+0.19, d8 1.24+0.19, p<0.05). There was a reversible hemoglobin reduction especially at 5 and i0 ,gg, starting d3 with a max. decrease at dg (5/zg 24.0+7.4 g/l, 10 #g 23.3+5.5 g/l), at d15 initial values were almost reached without intervention. A dose related acute phase response was observed, with max. value of CRP (mg/l) 67• at 0.5 /zg and 602• at 10 /tg (r=0.87, p<0.002), and max. value of Serum Amyloid A (rag/l) at 0.5 /zg of 258• and at 10 #g 838+143 (r=0.80, p<0.002). After CT 2 c (no c=22) were evaluable upto dose 5#g, in these e rhlL-6 related side effects were: fever (82%), headache (73%), myalgia (27%), local erythema (68%), nausea (27%) and increase in liver enzymes (14%). Uptu c2 there was no difference in platelet nadir, moment and duration of platelets <100 x109/l between 0.5-1 /~g and 2.5-5 #g doses. The same was observed for leukocytes. In conclusion, rML-6 upto 10 #g/kg/d has a acceptable toxicity profile and results, before CT, in dose dependent platelet and acute phase proteins increase. After CT, evaluable now for the first cycle, there is a faster platelet recovery at higher dose steps. University Hospital Groningen, Division of Medical Oncology, Department of Internal Medicine, Oostersingel 59, 9713 EZ Gronlngen, THE NETHERLANDS. G-CSF IN ACUTE LYMPHOBLASTIC LEUKEMIA. D.Hoelzer, O.Ottmann rhG-CSF given in acute lymphoblastic leukemia (ALL) after induction/consolidation therapy or allogeneic/autologous BMT can shorten the recovery time of neutrophils by 4-9 days, with reductions of the infection rate in some studies. Granulocytopenia is the major dose limiting toxicity not only after but also during chemotherapy. For example, severe neutropenia occurs in 66~ of patients during the intensive 8 wk induction regimen of the German ALL trials. Therefore, the effect of parallel administration of G-CSF (Filgrastim) and chemotherapy was explored, rhG-CSF (5~g/kg) was given concurrently with cyclophosphamide, araC, 6-mercaptopurine, prednisone and prophylactic cranial irradiation, rhG-CSF was continued until neutrophil counts exceeded 1000/~I, but for at least seven days following the last dose of chemotherapy. In a pilot study of 15 pts, the median duration of granulocytopenia < 500/~I was 6 days compared to 12 days in a historical control group. When rhG-CSF was applied throughout the entire 8 wk induction treatment (Scherrer et al) in 14 pts, the CR rate was 93%, the tim~ of nm,trophil reeoverv tO > ~O~l~l after the end of chemotherapy was reduced from 29,5 days to 14,5 days, and the infection rate was lower in the G-CSF treated patients. From these studies it was concluded that the concomitant administration of rhG-CSF and intensive chemotherapy in adult ALL is leasable, well tolerated and without major side effects; in particular, there was no evidence of prolonged neutropenia due to stimulation of hemopoietic stem cells by G-CSF and their subsequent elimination by chemotherapy. The clinical effects of rhG-CSF given concomitantly to chemotherapy require confirmation in a randomized trial. In an ongoing randomized multicenter trial~ an interim analysis of 40 pts revealed a highly significant reduction of the total duration of severe neutropenia (< 500/~l) in the patients receiving G-CSF. Further analysis of the potential clinical benefits of this treatment modality is in progress. We performed a phase H trial to assess the ability of G-CSF -mobilized PBPC to rapidly and completely restore hemopeiesis after high dose chemotherapy in the absence of bone marrow infusions, with selection for PBPC-only infusions based on yield of granulocyte -macrophage colony -forming cells (GM-CFC) after G-CSF treatment. Twenty-nine adults with acute lymphoblastic leukemia, non-Hodgkin's lymphoma, or Hodgkin's disease who were eligible for autologons bone marrow transplantation were treated. G-CSF was given at 12 or 24 Ixg/kg/d for 6-7 d and mononuclear cells collected by apheresis on days 5,6,7 or 4,6,8. Yield of PBPC was assessed by assay for granulocyte-macrophage colony forming cells (GM-CFC). High yield was defined as total GM-CFC collected >30 x 104/kg. High dose bnsulfan (4mg/kg/d x 4d) an& cyclophosphamide (60mg/kgtd x 2(:1) were administered and hemopoiedc cells infused• and recovery of hemopoiesis mouitored. Progenitor cell yield was high in 11 of 29 patients. Patients given infusions of G-CSF-mobilized PBPC, but without bone marrow infusion, experienced recovery of hemopoiesis in all cases. No patient given PBPC alone required bone marrow infusion in up to 24 months of follow up. Kinetics of recovery of both the platelet and neutrophil counts were more rapid in the high yield (PBPC-alone) group than in the low yield group (PBPC pins bone marrow). Pintelels recovered to >50 x 109/1 at a median of 11 days and neutrophils to >0.5 x 109/1 at 9 days in the high yield group, compared with 37 days and 10 days in the low yield group (p=0.0016 and 0.0308 respectively). Fewer platelet transfusions were required in the high yield group (median 11 packs per patient vs. 39, p---0.0138). We conclude that in patients with a high yield of PBPC after G-CSF therapy, infusion of G-CSF-mobilized PBPC results in rapid and sustained restoration of hemopoiesis. Use of G-CSF mobilised PBPC to support multiple cycles of HDC for treatment of high-risk stage II and III breast carcinoma is now being examined in a phase 1I study pursued by our group. Our data on PBPC moblisation without chemotherapy allows consideration of G-CSF-mobilised PBPC for haemopoietie ceil allotmnsplantation and for gene therapy trials. Our approach for high-dose (HD) chemotherapy is to first treat patients eligible for dose intensification with a standard dose chemotherapy (VIP: VP26 = etoposide: 500 mg/m 2, ifosfamide: 4 g/m 2, cis-platinum: 50 mg/m 2) followed by the application of colony stimulating factors (G-CSF, GM-CSF or IL-3 + GM-CSF) in order to combine a regimen with broad anti-tumor activity with the recruitment of peripheral blood progenitor cells (PBPCs). PBPCs are reinfused into the patients after high dose intensification chemotherapy (according to underlying disorder: HD-VIP, VIC-E or BEAM). Highest numbers of PBPC were recruited by the sequential administration of 11_-3 + GM-CSF (median of 420 CD34 + cells/iJI and 24.000 total progenitors/ml, including multilineage as well as megakaryocytic progenitors). CSF-mobilized PBPCs include lineage negative CD34 + cells, 4-HC resistant precursors as well as long term culture initiating cells (LTC-IC). So far, a total of 47 patients were supported with PBPOs and CSFs following HD-chemotherapy. The period of severe neutropenia (<100//JI) as well as thrombocytopenia (<20.00O//JI) was reduced to a median of 4 or 5days respectively. In order to decrease the number of potentially contaminating tumor cells in the leukapheresis preparations, we have started to positively select CD34 + cells by an avidin-biotin immunoadsorption column (provided by R. Berenson, CellPro, USA). In all these patients, recovery data were comparable to the patients who received unseparated cells. To provide sufficient numbers of PBPCs for repetitive use or in patients with low progenitor cell yield as well as to possibly avoid leukapheresis, we investigated the ability of hematopoietic growth factor combinations to expand PBPCs ex vivo. Chemotherapy + G-CSF recruited CD34 + cells from 18patients were enriched by immunoadsorption (purity: 89.2 -+ 4.8 ) and cultured in suspension. A combination of stem cell factor, erythropoietin, interleukin-16, IL-3, and IL-6 was identified as the optimal combination for the expansion of clonogenic progenitors. Proliferation peaked at day 14 with a mean 190-fold increase (range 46-930) of clonogenic cells. Interferon-gamma synergized with the 5-factor combination, whereas the addition of GM-CSF or G-CSF decreased the number of clonogenic cells. Large-scale expansion of CD34 + cells in auto-Iogous plasma supplemented with the 5-factor combination resulted in an equivalent expansion. Our data indicate the feasibility of large-scale pro+genitor cell expapsion in cancer patients, starting from small numbers of CD34 ceils. Given the pressures on health care resources new health technologies need to be assessed to establish whether their benefits justify their costs. G-CSF is an important new technology that has profound impacts on cancer chemotherapy and patients outcome. As clinical studies show, the main economic benefits of G-CSF come from * a significant reduction of hospitalisation during chemotherapy, * reduced expenditure on inpatient and outpatient i.v. and oral antibiotic treatment, * and reduced platelet and RBC transfusions. In addition, it is expected that a therapy with G-CSF during cancer chemotherapy shows intangible and non-monetary benefits: * increase in the rate of treatment as intended (time, number of cycles) * increase of quality of life of patients * reduction of drop-outs * ability to intensify dosage Recently coducted economic cost-cost studies show that the cost of managing neutropenia and infection and the cost of chemotherapy are lower for patients receiving G-CSF than those who do not. Future studies have to focus mainly on the quality-of-life cost-effectiveness of G-CSF. The aim of this study is to evaluate the effects of recombinant human Erythropoietin (rhu EPO) on hematopoietic regeneration after allogeneic or autologous bone marrow transplantation. Patients were randomized to receive either 100 U rhu EPO/kg body weight or placebo as continuous intravenous infusion from day 1 after BMT until independence from erythrocyte transfusions for 7 days or until day 42. The randomization was performed per each center and stratified according allogeneic or autologous BMT and major ABO-blood group incompatibility. At the time of the planned interim analysis with 205 patients treated, the time to erythrocyte transfusion independence after allogeneic BMT was shorter in group A (n = 52) than in group B (n = 55). After autologous BMT no difference between group A (n = 49) and B (n = 49) could be detected so far concerning time to transfusion independence or the number of transfusions applied, considering either erythrocyte or thrombocyte transfusions. There were no major differences in side effects between groups A and B. As of October 1992, the study was finished with a total of 329 patients. Since 1978 164 RCC patients have been screened for entry into clinical trials and entered onto surveillance as first management policy until symptoms or serial radiology indicated tumour progression. 3+4(4%) have demonstrated unexplained "spontaneous" CR+PR(median duration 17 mths), and 7(4%) "stable" disease(median 28mths). None of 19 not nephrectomised regressed, while CR+PR was 11% of 40 nephrectomised at diagnosis of metastasis, 1% of 76 developing metastases 1-17 mths after nephrectomy and 10% of 29 developing metastases more than 18mths after nephrectomy. It was 12% of 47 with lung only verses 1% of 117 with other sites. Possible evidence for relief of potentially immunosupressive influences have been demonstrated in 4 of 7 patients demonstrating unexplained "spontanous regression. Study of HLA class I and II antigen abnormally on these tumours and their influence on tumour infiltrating lymphocytes will be reviewed. 80 patients seen to progress on surveillance were entered into treatment trials, the remainder being too old, sick or having bone or brain metastases needing radiotherapy. 2+6(10%) achieved CR+PR. 2+5 of 14(50%) of patients with lung only verses 0+1 of 66(1.5%) of those with other sites of metastasis. Although these results suggest that no harm has come from a period of preliminary surveillance, the fact that the therapeutic benefits including durable complete remission from therapy are confined almost entirely to the good risk small volume asymptomatic patients makes it difficult to justify a policy of surveillance for such patients. Auto BMT, as alternative of allo BMT, lack{ immunotherapical effect delivered by allo graft. However immune reconstiturion after auto BMT shows features recalling the post allo setting. Futhermore in vitro and in vivo in rlL2 is able to stimulate the proliferation and the lyric functions of NK and T cells, major effectors of this GVL effect. For these reasons from 1990, we have so far treated 41 patients with acute leukemia (AML-17;ALL:24) in CR1 with auto BMT followed with rlL2. All conditioning regimen included TBI. Median rime between diagnosis and BMT was r~.2 mths (4-9 mths).arlL2 was started as soon as platelets reach 50x10~/1 and ANL 0.5x10~/1. rlL2 was given in 5 cycles of 5 days (C1) and 2 days (C2 to C5), starting on D1,D15,D29,D43,D5~. rlL2 was given as a continuous infusion at a median of 16M IU/m~/D (12M-24M). No patient died of rlL2 related toxicity. Platelet toxicity was obvious during the starting cycle but did not impair the long term hematological recovery. Immune stimulation was major and intense for both NK and T cells and both LAK and NK activity (all p < 0.05). Long term infusion conducts to privilegiate NK stimulation. With a median follow-up of 24 mths, relapse and survival probabilities are respectively 45% and 65% fo ALL and 27% and 89% for AML comparying favorably with historical control ; these data invited us to set up a randomised study in CR1 AML and ALL after auto BMT. Started in 1991 in France, Italy and England, 135 pts have been so far included. We have shown that local, inhalative IL-2 application in combination with low dose systemic IL-2 and IFNa is highly effective and has low toxicity (J Urol 147,344, 1992) . Here we report about longterm follow-up from 4 to 28 month of an outpatient schedule in 17 patients with pulmonary metastasis of renal cell carcinoma. Treatment protocol based on 5 time daily inhalation of IL-2 (5 x 200.000 I J) combined with low dose i.v. IL-2 (4x106 U/4 days every 2 weeks) in 5 patients or low dose IL-2 s.c. injection of 100.000 U per day in 12 patients. All patients received 5 x 106 s.c. IFNa three times weekly. Median treatment duration was 17 month (range 1-24 month). Toxicity of IL-2 inhalation was low, no fever, no vasculary leakage. Even after up to 24 month of continuous inhalative treatment no evidence for irreversible pulmonary damage due to IL-2 induced immunomodulation occured in 17 patients. In pulmonary metastases 1 complete response (6 month), 9 partial response (4,5,8,13, + 18, 18, 24,24,24) and 7 stable disease (1,6,7,10 + 16, 20, 23, +) were achieved. 4 of 17 patients responded with nonpulmonary metastasis. Overall tumor response considering both, pulmonary and nonpulmonary metastases was 1 complete response, 7 partial responses 6 stable diseases, 2 mixed responses and 1 progressive disease. While IL-2 per inhalation had to be stopped because of grade II toxicity in 1 patient only, IFNa systemically had to be stopped in 10 of 17 and IL-2 i.v. cycles in 5 of 5 patients. 13 patients are alive with median actual survival of all patients of 16.9 month which seems to be prolonged compared to risk factors. Inhalafive treatment with IL-2 combined with low dose systemic cytokines represents a highly valuable model for the low toxicity of lccal IL-2 application, resulting in an effective long term treatment schedule with long term responses. Recent clinical trials for the biological therapy of solid tumors have used recombinant human cytokines in combination with conventional chemotherapy. In patients with metastatic renal cell carcinoma, we established a 3-drug combination of subcutaneous recombinant human interferon-u (IFN-u), interleukin-2 (IL-2) and 5-fluorouracil (5-FU) as outpatient regimen. Treatment consisted of eight weeks each of IFN-~ (10 million U/m' x3 per week SQ) combined sequentially with IL-2 (5-20 million IU/m' x3 per week SQ for four weeks) and 5-FU (750 mg/m 2 IV weekly for four weeks). Among 39 consecutive patients treated, there were 6 complete (15.4%) and 12 (30.8%) partial responders, with an overall objective response rate of 46.2% (95% confidence interval, 30-63%). Median response duration was calculated at 10+ months, and no relapse has occurred among complete responders. Systemic toxicity was mild to moderate, with no severe 5-FU related mucositis or diarrhea. There were no dose limiting adverse effects of SQ IL-2 and no toxic deaths. In summary, this outpatient biochemotherapy was as effective as the most aggressive inpatient IV IL-2 regimen; it appeared to significantly improve the therapeutic index in patients with metastatic renal cell carcinoma. Medizinische Hochschule, D-3000 Hannover, Germany Metastatic melanoma: immunotherapy with IFNa and IL-2, DTIC/IFNa as second line treatment, IFNe(/IL-2 + DTIC or CDDP. U Keilholz, C Scheibenbogen, W Tilgen, D Maclachlan, P Brossart, Th M0hler, W Hunstein. This abstract summarises our 3-year experience in the treatment of metastatic melanoma with sequential or combined chemo-immunotherapy. Patients with progressing metastatic melanoma have been treated with IFNa and IL-2. 10 mill U/sqm IFN(~ were given s.c. on days 1-5, and a new decrescendo regimen of IL-2 was used: lmg/sqm/6hours, followed by 1mg/sqm/12 hours, and 1mg/sqm/24 hours, and 0.25mg/sqrrV24 hours x 3. The current response rate is 33% (3 CR, 12 PR, 16 MR/SD, 14 PD). Patients not responding to IFNedlL-2 (SD and PD) were eligible for subsequent chemotherapy with DTIC, 850 mg/sqm day 1, followed by IFNa, 3 Mill U/day 2-6. The response rate for this second line regimen is 18% (1 CR, 3 PR, 5 SD, 13 PD, n=22). Using this sequential approach, the overall response rate in this cohort is 51%, and the median survival is 17 months. In preparation of a randomized trial comparing chemo-immunotherapy and immunotherapy a pilot study was performed. Patients not responding to the standard IFNedlL-2 regimen received a single dose of DTIC, 850 mg/sqm (n=6) or CDDP, 100 mg/sqm (n=7) on day one, followed by IFN~IL-2 according to the identical protocol as previously without chemotherapy. In the case of CDDP, grade 3 nephrotoxicity was observed in 2/7 patients. Pharmacokinetics of IL-2 were not influenced by previous chemotherapy, except in the patients with CDDP-associated nephrotoxicity. Induction of secondary mediators (TNFa, IFN'~, neopterin, sCD25) by IL-2 was not diminished by previous chemotherapy. 3 patients unresponsive to immunotherapy alone showed tumor shrinkage upon chemo-immunotherapy. Conclusion~: With initial immunotherapy followed in nonresponders by chemotherapy a response rate above 50% is achieved. Combined chemoimmunotherapy is feasible, and the immunologic response to IL-2 is not diminished by previous chemotherapy. A randomised trial is necessary to determine, whether combined chemo-immunotherapy is superior to immunotherapy alone. Inflammation is characterized by an accumulation of leucocytes at the site of injury. It hab been well established that the disease associated lesions are caused by a plethora of soluble mediators released by both the infiltrated leukocytes as well by tissue cells, including many degrading enzymes, lipid mediators, reactive oxygen species, or cytokines. Among those, cytokines play a crucial role in the inflammatory process as their release appears to represent the initial response which mediates the secretion of subsequent effector molecules responsible for the pathophysiological effects (e.g. changes of vascular resistance) in an autocrine or paracrine fashion. This holds true for the acute inflammation during which circulating granulocytes or monocytes are activated directly by a noxious agent; an example being the pivotal function of tumor necrosis factor (TNF) in septic shock. Chronic inflammatory diseases are initiated and perpetuated by immune reactions. Cytokines represent an important link between immune reactions and the recruitment and activation of blood borne infiltrating inflammatory ceiLs. Thus interferon r and tumor necrosis factor 8, secreted by activated T lymphocytes, are potent stimulators of mononuclear phagocytes. The activated macrophages themselves secrete cytokines such as Interleukin-t (IL-t), TNF a or Interleukin-8 (IL-8) which have been termed collectively as inflammatory cytokines, iL-1 and TNF again in an autocrine or paracrine way upreguiate the synthesis of the ultimate pathogenic mediators in macrophages-e.g. the secretion of degrading enzymes as in rheumatoid arthritis-. Equally important they induce functional changes in vascular and autochthonous tissue cells which are thus recruited to contribute to the inflammatory lesions. IL-8, and similarly the lymphocyte derived RANTES, chemotactically attracts to the site of lesion and activates granulocytes, which then participate in the inflammatory reaction. While this rapidly recruits circulating phagocytic cells, the secretion of colony stimulating factors by activated T lymphocytes and macrophages also increases the pool of granulocytes and monocytes by stimulating their synthesis in the bone marrow. In turn IL-1 an TNF a synthesized by monocytes (and also some tissue cells) are coactivators of T lymphocytes, and thus contribute to al local immune reaction. Besides this general role of cytokines in inflammation their involvement in specific situations becomes increasingly apparent, an important example being the allergic inflammation. Cytokines synthesized by T lymphocytes such as Interleukin-4 or Interleukin-5 not only regulate immunogtobulin synthesis of B lymphocytes, including IgE, but equally important modulate the functions of mast cells or eosinophilic leukocytes to become effective effector cells of allergic reactions. The detailed functions of cytokines in different acute or chronic inflammatory diseases will be discussed in the subsequent contributions. CLINICAL RESULTS E Holier, R Hintermeier-Knabe, B Ertl, HJ Kolb, M Kaul, U Behrends, S Thierfelder, W Wilmanns Experimental as well as clinical studies suggest dual involvement of proinflammatory cytokines as TNFalpha and ILl in complications and GvHD following allogeneic BMT: Nonspecific activation of recpient tissues during pretransp~ant conditioning results in early release of TNFa which strongly enhances donor T cell activation; in the effector phase, T cell stimulation is amplified by IFNgamma and LPS mediated release of TNF and ILl. The role of these mediators as critical effectors of GvHD related tissue damage could be confirmed by application of cytokine antagonists early after murine BMT: Both, anti-TNFa and ILl-receptor-antagonist (ILlra, as shown by J Ferrara, Boston) resulted in a >50% reduction of GvHD related mortality in mice, while pentoxifylline proved to be ineffective. In human BMT, phase l/ll studies analyzing anti-TNFa and ILlra in refractory GvHD report substantial improvement in 50 -70% of patients, though reoccurrence of symptoms after cessation of treatment in most patients indicates a late effector function of cytokines at least in advanced GvHD. Contribution of TNFa to recipient related induction of GvHD is further suggested by a recent phase 1/11 study performed in our institution as prophylactic application of anti-TNFa during pretransplant conditioning suppressed occurrence of early GvHD in high risk patients as compared to historical controls. Though these data indicate a future role of a variety of cytokine antagonists in management of clinical complications of BMT their impact on long term survival has to be evaluated and compared to classical strategies (e.g.the use of corticosteroids and T-cell-manipulation) in randomized studies. The cytokines IL-1 and TNF play central roles in inflammatory responses which can lead to tissue injury, Naturally occurring antagonists such as soluble cytokine receptors or receptor antagonists are also produced during inflammatory responses. These soluble receptors and antagonists may act as a buffer system to reduce and limit tissue injury induced by cytokines. However, in some disease states this naturally occurring buffer system may not be sufficient to inhibit the detrimental actions of an inflammatory response. Responses to IL-1 are mediated via the type 1 receptor (IL-1R, type I). The type II receptor (IL-1R, type II) has never been shown to signal, but instead appears to be shed as an IL-1 antagonist. Both recombinant soluble IL-1R, type I an IL-1R, type II are capable of binding IL-1 and inhibiting responses in vitro. In a Phase I clinical study evaluating soluble 1L-1R (type 1) in modifying cutaneous allergic responses, IL-1R was a potent inhibitor of allergen induced latephase inflammation in the skin, with a high safety profile. Responses to TNF are mediated via the pS0 or pt0 TNF receptor. Soluble forms of both the p80 and 1960 TNFR occur naturally and are increased in many inflammatory states. Dimeric constructs of the p80 TNF have been engineered to form a soluble Fc fusion protein with two monomeric TNFR extraceUular portions contained on a truncated Ig heavy chain (TNFR:Fc). The TNFR:Fc possesses higher affinity than a monomeric soluble receptor and the Ig-like Fc structure imparts a longer half life. Animal studies have shown that soluble IL-1R, type I and TNFR:Fc are effective antagonists of inflammation. In animal models for arthritis and pulmonary inflammation, IL-IR and TNFR:Fc reduced inflammatory celt infiltration and tissue damage. Both molecules are currently in clinical trials and hold promise for treatment of autoimmune and allergic diseases such as rheumatoid arthritis and asthma. Recently, evidence was raised that pentoxifylline (POF} is able to suppress the synthesis of tumor necrosis factor-alpha [TNF} in cell cultures, in vivo, and to protect experimental animals against endotoxin shock: It was found that POF selectively inhibits the formation of TNF but not interleukin-6 (IL-6]. We could confirm these pharmacological effects of POF in humans under controlled conditions of endotoxlnemia. Ten healthy volunteers recieved a bolus injection of endotoxin [100 ng of LPS of S.a.e.], which caused a transient increase of circulating TNF and IL-6. 3 weeks later the voluntecrs were again injected with endotoxin and POF was also infused. Due to POF administration, there was no rise in circulating TNF, whereas IL-6 levels rose in paraUel with body temperature, comparable to those seen in the first part of the study. Treatment of allograft transplant recipients with the murine anti-CD3 monoclonal antibody OKT3 leads to an acute cytokine release syndrome. Especially TNF seems to play a pivotal role in the pathophysiology of the OKT3 first-dose reaction. Pretreatment of 8 patients with POF prior to OKT3 administration was able to reduce the endogetlous TNF formation significantly as compared to controls and, thus, prevents severe clinical side effects, whereas IL-6 formation and febrile response were not affected. Severe pulmonary tuberculosis is associated with a chronic cytokine release syndrome [elevated levels of circulating TNF and IL-6]. In 8 patients POF treatment inhibited chronic TNF release selectively, and, thus, reduced TNF-dependent caehexia without affecting chronic IL-6 formation and related symptoms, such as fever and night sweat. In conclusion, we suggest that POF may improve therapeutic strategies in cases of acute and chronic cytoklne release syndromes. Tumor Necrosis Factor (TNF) plays a central role in the maintenance of the inflammatory events in rheumatoid arthritis. We evaluated the expression of p75 and p55 TNF receptors (TNFR) and of TNF-a and TNF-I~ on the surface of synovial fluid mononuclear cells in patients with rheumatoid arthritis (RA) (n = 9), spondylarthropathy (SPA) (n = 11 ), and traumatic effusions (n=3). Synovial T-lymphocytes from RA patients express in all cases the p75 TNFR on the cell-membrane, in 4/9 cases also a weak expression of the p55 TNFR is detectable, both mRNAs can be detected by polymerase chain reaction (PCR). Synovial macrophages also express the p75 TNFR and low amounts of the p55 TNFR. Patients with active RA also have circulating p75 TNFR positive T-lymphocytes in their blood. High concentrations of soluble TNFR (sTNFR) are found in the joint effusions of RA patients: up to 40 ng/ml of p75 sTNFR and ~Jp to 54 ng/ml p55 sTNFR. Significantly lower sTNFR levels are found in SpA effusions. Both receptors are also more elevated in the serum of RA patients (2.59 _+ 0.28 ng/ml p75 sTNFR and 4.49 + 0.55 ng/ml p55 sTNFR) as compared to SpA patients (1.41 _+ 0.09 ng/ml p75 sTNFR and 1.78 + 0.08 ng/ml p55 sTNFR, p < 0.001). TNF-a could be detected in the synovial fluid of RA patients (up to 140 pg/ml), but not in the serum. The soluble TNFR are biologically active and neutralize the effects of TNF-a in a cytotoxicity assay. The high levels of soluble TNFR in the inflammatory effusions may reflect TNF neutralizing activity in an environment where enhanced TNF synthesis has occured. We have generated several anti-sense-TNF-a oligonucleotides (as-TNF), in order to down-regulate TNF biosynthesis at the mRNA level. With as-TNF-3 we could achieve more than 90% inhibition of TNF secretion in PHA-stimulated peripheral blood or synovial fluid lymphocytes. The effects of as-TNF-3 on the expression of TNFR and on the synthesis of other cytokines are currently being investigated. To investigate the effects of G-CSF (r-metHuG-CSF, Aragon, Thousand Oaks, CA) on neutrophil production, blood distribution, survival and function, daily subcutaneous injections of G-CSF were administered to healthy, young (Y) (20-30 years) and healthy elderly (O) (70-80 years) subjects for 14 days at three dose levels, 0 mcg (n=7), 30 mcg (n=10), and 300 mcg (n = 10). Daily CBCs and serial measurements of neutrophil oxidative activity by chemiluminescence were made. In addition, blood neutrophil kinetics were determined (day 6), transit time of the marrow neutrophil post mitotic pool (NTT) (days 7 to 14), neutrophil migration to skin chambers (days 0 and 5), and blood colony forming cells (CFU-GM) (day 0 and 5), as well as routine marrow morphology studies (days 0, 5 and 15) were performed. Baseline neutrophil counts were similar in the Y and O subjects and counts increased similarly for the two age groups, with peak counts of 51.6 --+-6.7 for the 300 mcg dose. G-CSF significantly shortened the NTT from 5.7 4-1.6 days (control) to 4.2 4-0.3 days (30 mcg) and 2.9 4-0.4 days (300 mcg) (p<0.05). A concomitant of the shortened NTT was a dose dependent increase in the chemiluminescence responses, reflecting higher myeloperoxidase activity. The shortened NTT was also reflected by a decreased proportion of marrow cells in the post mitotic pool (metas, bands and PMNs) and apparently lengthened blood half life of the circulating PMNs. Neutrophil migration to cutaneous inflammatory sites was also decreased in a dose dependent fashion. Comparison of the age groups showed the only significant difference to be in the mobilization of blood colony forming cells with blood CFU-GM increasing 80 fold in the young versus 40 fold in the elderly over 5 days (300 mcg). No significant toxicities were observed in these normal subjects. These studies demonstrate the dose dependent stimulation of neutrophil production with G-CSF administration, which is not affected by the age of the subjects. The neutrophils released into the peripheral blood have enhanced oxidative responses but somewhat reduced migratory capacities, probably reflecting the accelerated production and early release of the developing neutrophils. These changes are remarkably similar to changes in neutrophil production and function observed with bacterial infections in hematological normal individuals. The systemic regulation of the host response during acute inflammatory states remains poorly understood. Among the regulatory systems that are likely to play a role in controlling host responses to bacterial infection is the neuroendocrine axis. The pituitary for example, is ideally situated to integrate central and peripheral stimuli and among other effects initiates the systemic increase in glucocorticoid production that accompanies host stress responses. We studied the secretory response of the murine pituitary cell line AtT-20 in vitro and whole pituitaries in vivo after endotoxin (LPS) stimulation. We identified macrophage migration inhibitory factor (MIF), a previously described T-cell cytokine, as a major secreted protein of AtT-20 cells that were stimulated by sub-nanogram amounts of LPS. To study the expression of pituitary-derived MIF in vivo, BALB/c mice were injected ip with sub-lethal amounts of LPS. Pituitaries and serum were collected at intervals and pituitary MIF mRNA and pituitary and serum MIF protein were measured. Pituitary MIF mRNA specifically increased with time and reached a plateau 16-24 hr after LPS challenge. MIF protein, which is present constitutively in the pituitary, decreased within 20 hr. Determination of serum MIF in normal, athymic and hypophysectomized BALB/c mice indicated that the pituitary is an important source of serum MIF that appears in the post-acute phase (6-20 hr), whereas T cell MIF contributes primarily to the hyper-acute rise in serum MIF (2 hr). These data suggest that MIF plays a central role in the systemic response to endotoxin and that pituitary MIF is likely to reflect the interplay of diverse neurohumoral stimuli that regulate acute and chronic inflammation. Pretreatment of mice or rats with granulocyte colony-stimulating factor (G-CSF) protected against endotoxin-induced lethal shock or against endotoxin-induced liver injury in galactosamine-sensitized animals. In the animals protected by such pretreatment, the systemic tumor necrosis factor 0r (TNF) bioactivity was significantly suppressed. Various macrophage populations taken ex vivo from G-CSF-pretreated animals showed an attenuated TNF release following LPS stimulation. However, G-CSF had no such effects on macrophages when directly added in vitro to the cells. These findings show that G-CSF protects against septic shock via TNF suppression in a way requiring the participation of additional circulating cells or factors. Pretreatment of rodents with granulocyte macrophage colony stimulating factor (GM-CSF) greatly enhanced the susceptibility of the animals to endotoxin and led to a tremendous increase of the systemic TNF found following an LPS challenge. An anti-GM-CSF antibody significantly protected animals against septic organ failure. Considerable amounts of endogenous GM-CSF are released following endotoxin chaitenge with a maximum at 2 h. The enhancement of LPS-inducible TNF release from rnacrophages takes also place in vitro. It is concluded that GM-CSF is a pivotal mediator of sepsis as well as a directly acting enhancer of LPSinducible TNF release. We have previously demonstrated that protein production and mRNA expression of GM-CSF, G-CSF, and IL-3 were decreased in activated MNC from umbilical cord (C) compared with adult (A) peripheral blood (Cairo, et al, Pediatr Res 30:362, 1991, and 31:574, 1992) . RhG-CSF + recombinant rat Stem cell factor (rrSCF) administration in newborn rats has, however, resulted in a significant induction of neutrophilia, an increase in bone marrow post-mitotic pool, and is synergistic with antibiotics during experimental group B streptococcal sepsis (Cairo, et al, Blood 76:1788 , 1990 , and 80:96, 1992 . To assess the toxicity and efficacy of rhG-CSF in newborns with presumed sepsis, 42 NB <3 days (26-40 wks) with a diagnosis of presumed sepsis were randomized to either placebo (P), 1.0, 5.0, or 10.0 pg/kg q 24 hrs, 5.0 or 10.0/lg/kg q 12 hrs of IV rhG-CSF x 3 days. CBC, differential, and platelet counts were obtained at time 0, 2, 24, 48, 72, and 96 hrs. Tibial bone marrow aspirates were performed at 72 hrs and bone marrow NSP, NPP, CFU-GM, and CFU-GEMM were determined. Serum for G-CSF levels was obtained before and 2, 6, 12, 14, 16, 18, and 36 hrs after rhG-CSF dosing and measured by a sandwich ELISA assay. RhG-CSF induced significant neutrophilia at 24 hrs vs. placebo (85 + 19%) following both 5.0 (203 + 40%, p<0.04) and 10.0 IJg/kg/d (351 -+ 89%, p<0.02). Significant neutrophilia continued at 48 and 72 hrs at both 5.0 (p<0.03 and <0.013) and 10 h'g/kg/d (p<0.0S and <0.014) respectively. RhG-CSF also significantly induced a dosedependent increase in the BM post-mitotic pool (P vs. 10 IJg/kg/d) (18 +-3.7 vs. 63 + 7.5%) (p<0.005). T1/z of G-CSF in the NB was 4.42 _+ 0.44 hrs (r= -0.9) with peak G-CSF levels occurring within 2 hrs. To date there has been no evidence of acute or chronic toxicity secondary to rhG-CSF. Future studies are required to determine the clinical implications of the biological efficacy of rhG-CSF in newborns with presumed sepsis. Children's Hospital of Orange County, Orange, California USA Depenbrock, T. Block, H. Vogelsang, Ch. Fellbaum, J. Rastetter, A.-R. Hanauske TGF-Po is known to function both as inhibitory and stimulatory factor depending on the type of cell line investigated. The purpose of our study was to determine the effects of TGF-~ 1 and TGF-62 (concentrations: 0.1 -1 -10 ng/ml) on soft agar colony formation of freshly explanted I~uman tumors in vitro. Of 89 specimens, 6 had to be excluded from further analyses (4 confirmed benign, 2 bacterial/fungal contamination). 50 of the remaining 83 tumors showed evaluable growth in control capillaries (60%). At 10 ng/ml, TGF-Po 1 inhibited colony formation (_< 0.5 x control) in 12 specimens (24%): 5/19 renal, 4/6 Non Hodgkin's lymphoma, 1/7 breast, 1/2 ovarian, 1/2 melanoma (median: 0.37 x control, range: 0.22 -0.49). At 10 ng/ml, TGF-E,= showed inhibition of colony formation in 17 specimens (34%) with a similar spectrum of activity (median: 0.36 x control, range 0.06 -0.48). Stimulation (colony formation _> 1.5 x control) was observed in only 1/50 specimens. Combination of TGF-Po 1 with Epidermal Growth Factor (EGF) reversed the inhibitory activity in 4/5 specimens (80%). Combination of TGF-P~ with Platelet Derived Growth Factor (PDGF) reversed the inhibitory activity in 2/5 specimens (40%). Similar results were observed for TGF z. We conclude that TGF-I~ 1 and TGF-I~ 2 inhibit a subgroup of freshly explanted clonogenic tumor cells. Their activity, however, appears to be modulated by other growth factors. A Korfel. Z yon Marschall. D Ol~erbera. E Thiel. and WE Berdel MIP-I~ is a member of a family of proinflammatory cytokines produced by activated macrophages which has been shown to be a negative regulator of early hematopoietic stem cell progenitors. Our group investigates the interactions between hematopoietic cytokines and non-hematopoietic malignant cells. Here, we describe results testing rhMIP-lc~ (rh stem cell inhibitor, SCI; kindly provided by Dr. Wolpe, Genetics Institute, Cambridge, MA, USA) on the clonal growth of different human non-hematopoietic tumor cell lines in vitro. Cell lines tested included the following histologies: gtioblastoma 6x, neuroblastoma lx, head and neck carcinoma 2x, lung carcinoma lx, colorectal carcinoma 2x, gastric carcinoma lx, pancreatic carcinoma l x, breast carcinoma l x, bladder carcinoma lx, prostate carcinoma lx, choriocarcinoma lx, ovary carcinoma lx, osteosarcoma lx, melanoma 3x. MIP-1 (0, 2, 20, 200 ng/ml) was tested in human tumor cloning assays (HTCA) in mixtures of methylcellulose and agar. HTCA has previously been shown to detect positive and negative growth control by cytokines. Plating efficacy of the cells in the controls was > 2 % (median 8 %, range 2.2 -46.2 %) in this series of experiments: Tumor cells were continuously exposed to the cytokine for the complete assay period. Clonal growth of none of the celt tines was significantly and reproducibly stimulated or inhibited by MIP-lm Since MIP-lc~ may enter clinical trials for indications such as protecting hematopoietic stem cells from damage caused by cytotoxic chemotherapy in tumor patients, further experiments should be performed in vitro and in vivo. DFG Be 822/4-2 Klinikum Steglitz, Freie Universitaet Berlin, Berlin, Germany Clinical studies have demonstrated the activity of single hematopoietic growth factors (HGF) in restoring bone marrow function after chemotherapy. These studies have prompted clinical trials using multiple growth factors to promote the maturation of precursor cells at various stages of differentiation. However, HGF also have the capability to stimulate growth of non-hematopoietic tumor ceils at least in long-term cell cultures. We have assessed the growth-modulating activity of combinations of various HGFs on freshly explanted human tumor colony forming units in vitro. A total of 86 tumor specimens were exposed for 21 -28 days to IL-3, GM-CSF, G-CSF, SCF (a~l at final concentrations of 100 ng/ml) and IL-113 (final concentration: 10 ng/ml) or combinations of these HGFs using a capillary soft agar cloning system. 54 specimens (63%) showed evaluable growth in control capillaries. Stimulation of colony growth was observed in 17/1026 tests (1.7%) with 6/54 (11%) specimens expressing sensitivity to individual HGFs, 3/54 (6%) to combinations of two HGF, and 6/54 (11%) to combinations of more than two HGFs. Inhibition of colony growth was observed in a total of 24/1026 tests (2.3%) with 1/54 (2%) specimens expressing marginal sensitivity to individual HGFs, 6/54 (11%) to combinations of two HGF, and 7/54 (13%) to combinations of more than two HGFs. For inhibitory effects, median colony survival for combinations > 2 HGFs was 0.46 x control (range 0.31-0.49). For stimulatory effects, median colony survival for combinations > 2 HGFs was 2.0 x control (range: 1.6 -2.5). Our data indicate that combinations of HGF will not substantially alter the pattern of clonal proliferation of the majority of freshly explanted tumor cells in vitro. However, growth modulation may occur in a minority of tumors. Klinikum rechts der Isar der Technischen Universit&t M0nchen, Ismaninger Str. 22, 8000 M0nchen 80 Supported by grants W41/88/Ha-1-2 from the Deutsche Krebshilfe and 90.055.1 from the Withe~m-Sander Stiftung IS GRANULOCYTE COLONY-STIMULATING FACTOR AN ANGIOGENIC FACTOR IN HUMAN GLIOBLASTOMA? S. Corbacioglu, K. Welte and T. Pietsch Granulocyte-colony stimulating factor (G-CSF) belongs to a family of glycoproteins winch regulate growth, differentiation and function of hematopoietic ceils of the myelomonocytic lineage. In addition, G-CSF induces migration and proliferation of endothelial cells in vitro. To investigate the effects of G-CSF on vascularization of glioblastoma in vivo we transplanted the human glioblastoma cell line U-87 MG which is capable of producing G-CSF in high amounts. After s.c. inocculation of these cells into the backs of athymic mice, they developped highly vascularized solid tumors. In order to block the effect of G-CSF directly or to inhibit the production of G-CSF by tumor cells neutrolizing monoclonal antibody (mAb) 75A against G-CSF or dexamethasone were injected intravenously. At 2-day intervals the tumor volumes were measured. After seven days the mice were sacrificed and the tumors were explanted. Blood was collected for differential blood counts and serum was tested for G-CSF. Fresh frozen sections of the tumors were specifically stained for capillaries using Bandeiraea (Griffoina) Simplicifolia Lectin I Isolectin B 4 (BSL B4). Morphometric studies of the stained sections were performed in order to quantitate the vascularization of the tumors. The differential blood counts showed significantly increased neutropinls due to the effect of human G-CSF produced by the glioblastoma cells. Tins effect was inhibited by anti-G-CSF mAbs or dexamethasone. However neither G-CSF mAbs nor dexamethasone could inhibit tumm growth compared untreated tumor bearing mice. Dexamethasone significantly decreased the tumor vascularization whereas anfi-G-CSF mAbs did not have any effects on tumor vascularization. These findings suggest that G-CSF is not an essential angiogenic factor in vivo. Pediatric Hematology/Oncology, Medical School Hannover, Konstanty-Gutschow-Str. 8, 3000 Harmover 61, Germany Michael Martin, Torsten Spencker, Karen Welch*, and Andreas Strasser* The SPGM1 cell line was established from a transplantable mouse progranulocytic/promacrophage tumor (D0hrsen U et al. 1989, Leukemia 3: 796-803) . Extensive phenotyping of this mouse progenitor line revealed the properties of a typical CD5 positive pre-B cell, SPGM1 being positive for PB76, B220 (CD45R), and the pre-B cell immunoglobulin receptor complex, comprising p-heavy chains, XS-and Vpree-surrogate light chains and the IgMa and Igl~ co-receptor molecules. Southern Blot analysis revealed clonal rearrangement of the p-heavy chain locus and germline light chain loci. Interestingly, SPGM1 readily formed blast cell, macrophage and occasional granulocytic colonies in soft agar in the presence of Interleukin 3 (IL-3). In suspension cultures IL-3 also induced macrophage differentiation, the cells becoming larger, adherent and independent of 2-mercaptoethanol in the culture medium. IL-3 induced an initial burst of proliferation during differentiation, accompanied by loss of self renewal capacity, subsequently followed by a decrease and cessation of proliferation. The earliest changes were detectable at 24 hours by Northern Blot analysis. IL-3 treatment increased Mac1 mRNA, induced cfms mRNA (M-CSF receptor) and down regulated mRNA for p, X 5, vpree, and mbl. After 2 to 4 days the cells morphologically, phenotypically and functionally resembled macrophages, expressing strongly Mac1 and F4/80, and phagocytosing latex beads (Martin Met al. 1993 J.Immunol. in press). Thus, IL-3 induced the CD5 positive pre-8 cell line SPGM1 to switch its differentiation program in a coordinate fashion from a pre-B cell to a macrophage. iGF is known to be mitogenic for a variety of cell lines in vitro. We have studied the effects of Insulin-like-Growth-Factor I [IGF-I] and Insulin-like-Growth-Factor II [IGF-II] on 92 freshly explanted human tumors using a capillary soft agar cloning system. 6 specimens had to be excluded from further analyses (4 confirmed benign, 2 bacterial/fungal contamination). 51/86 specimens (59%) showed adequate growth in controls (17 renal, 7 breast, 7 lymphoma, 6 colorectal, 14 miscellaneous) with a median colony formation of 18.7 colonies/capillary (range: 4.2 -120.0). At final concentrations between 10 lr M and 10 .8 M, IGF-I significantly stimulated colony formation (_> 1.5 x control) in 10/51 evaluable specimens (20%) with a median of 2.05 x control (range: 1.57 -3.22) and inhibited colony formation ( _< 0.5 x control) in 4/51 specimens (8%, median: 0.33 x control, range: 0.08 -0.49). IGF-II stimulated 8/51 specimens (16%, median: 2.66 x control, range: 1.50 -3.71 ) and inhibited 8/51 specimens (16%, median: 0.32 x control, range: 0.05 -0.47). The optimal concentration for stimulation was found to be 10 -~ M for IGF-I and IGF-II. Of 39 specimens not significantly stimulated by either IGF-I (10 g M) or Epidermal Growth Factor (EGF, 10 1~ M), 4 (10%) were significantly stimulated by the combination of the two factors. 1139 (3%) specimen was stimulated by a combination of IGF-II (10 -~ M) and EGF. We conclude that IGF modulates the clonogenic growth of a subgroup of freshly explanted human cancer cells in vitro. The myeloid growth factors G-CSF and GM-CSF are increasingly introduced in therapy trials in neutropenic disorders and in MDS. In a series of 4 therapy trials 12.5% up to 36.4% of patients treated with GM-CSF displayed a stimulation of blast cells (Antin 1988 , Estey 1991 , Ganser 1989 , Vadhan-Raj 1987 . Recently, we could demonstrate a growth advantage for monosomy 7-cells in GM-CSF containing bone marrow cultures (Haase, 1990) which is supported by results from others (Andreeff 1991). Now, further data are available corroborating an association of monosomy 7 and leukemic blast stimulation by myeloid growth factors. In a cytogenetic follow-up study of 13 patients with MDS under GM-CSF (in preparation) we could observe a cytogenetic, clinical, and cytologic progression in a patient with monosomy 7 within 34 days. In a patient with Kostman's syndrome, receiving G-CSF, we performed sequential cytogenetic analyses. The patient's disease progressed to MDS and finally to AML. He initially had a mosaic karyotypeof normal and monosomy 7 cells but displayed a rapid karyotype evolution with supersession of normal cells and gain of additional abnormalities. A recent publication from a Japanese group adds further information to a possible association of monosomy 7 with stimulation of leukemia in 14% of neutropenic children treated with G-CSF (Kojima, 1992) . Besides the need for further cytogenetic follow-up studies in growth factor therapy trials, data are accumulating that monosomy 7 is a risk factor in GM-CSF and G-CSF therapy. Ii-3 is a cytokine with multilineage activity and stimulates proliferation of immature hemopoietic progenitors (Yang, 1989) . Recently, IL-3 has been introduced in clinical trials in MDS (Dunbar, 1990; Ganser, 1990) . As in GM-CSF therapy one major concern attributes to an IL-3 mediated stimulation of leukemic cells, since a rise in blast counts has been observed in two of 32 patients reported (Ganser, 1992) . However, as yet it is not known whether the blasts stimulated belong to the normal or leukemic population. We performed cytogenetIc analyses of bone marrow cultures with and without addition of i00 ng/ml IL-3 (Behringwerke, Marburg). The influence of this cytokine on the clonal compostion in 7 patients (3 ANLL, 4 MDS) with mosaic karyotypes was examined. In all 7 patients independent from diagnosis and chromosome abnormality the normal cell population was promoted by IL-3 with different intensity. Individual data are outlined on the The effect of Stem Cell Factor (SCF) on peripheral blood B-CLL-CelIs were studied in vitro by bromodeoxyuridine/propidiumiodide (BrdU/PI) cell-cycle-analysis. Peripheral blood cells from 41 patients with B-CLL were examined with CD-markcrs and prepared as follows: after Ficoll centrifugation and lysis of monocytes by leucine-methyl-ester (LME) T-Lymphocytes were depleted with a CD3 monoclonal antibody and magnetic cell sorting. The cells were grown in serum free culture medium (CG-medium) containing 10 #mol/l BrdU, and 0.1 ng/ml up to 100 ng/ml SCF. Controls were grown without cytokines. Samples were drawn repeatedly at 20, 68, 92, 116, and 140 hours. Cell-cycle-analysis was performed after double DNA staining with propidiumiodide and anti-BrdU-antibodies and determinated by flow cytometry. Minimal alterations were observed with T-cell depletion, the B-CLL cells from 31 patients were stimulated by SCF during the first 92 hours reaching a maximum of -2.77 % compared with the controls after 68 hours in cultures containing 100 ng/ml SCF. On the other hand, cultures without T-cell depletion showed no effect for SCF. We conclude that SCF has only a minimal stimulatory, early effect in inducing the proliferation of B-CLL cells. Stem cell factor (SCF) is known to promote proliferation of hematopoieticprogenitors and mast cells.However, the spectrum ot its biological effects is not completely understood. Since SCF may be able to accelerate hemopoietic recovery after chemotherapy or in other situations where severe immunosuppression is present, we were interested in its effects on T cells. Thus, we have studied the influence of human recombinant SCF on T cell proliferation invitro. When cultured for 6 days in serum-f~ee medium, 3H-thymidine incorporation of unstimulated peripheral blood mononuclear ee ells s (PMNC) resulted in 1480-+689 epm without growth factors, 1189+_119 clam with 11.-3 (50ng/mL), 2386+_1149 cpm with SCF (lOn~/mL), and 3828_+1660 c.m with both 11.3 and SCF (n=fi~ Addition of SCF to one way" mixed lymphocyte cultures (MLC) increased thymidine incorporation by 27% (0-182%); addition of SCF plus ID3 increased thymidine incorporation by 324% (48-734%; ID3 alone 0%, n=5). After depletion of monocytes and the majority of B cells from PMNC, the proliferation of the remaining eeU fraction, which consisted mainly of T cells, was not enhanced by ID3, SCF or IL-3 + SCF. We conclude that SCF can promote proliferation of unstimtdated or allostimulated T cells in the presence of ID3. Since this effect requires monoeytes or other accessory ceils, a direct influence of SCF on T cells does not become evident from our data. It is, however, still unclear whether LTB 4 acts in this regard directly or indirectly by stimulating the release of chemotactic and inflammatory cytokines. Here we report that LTB4 induces synthesis of interleukin (IL)-6 by human blood monocytee through transcriptional activation of the IL-6 gene. We furthermore demonstrate that this process involves activation of the transcription factor NF-rB and, to a lesser extent, of NF-IL6, white the activity of the transcriptior~ factor AP-1, shown to otherwise confer IL-6 inducibility, appeared to be unaffected by LTB4. Involvement of NF-~:B and NF-IL6 in induction of IL-6 transcripts by monocytes was demonstrated using deleted forms of the IL-6 promoter. Activation of the IL-6 promoter by LTB4 was not only associated with accumulation of the respective transcripts but resulted in synthesis of functional IL-6 protein as well. In addition, LTB4 mediated transcactivation of a heterologous promoter construct containing the NF-~:B or the NF-IL6 enhancer, but not the AP-1 enhancer. The signalling events mediating this effect appeared to involve the release of H202, since LTB4 failed to induce NF-rB or NF-IL6 in the presence of the scavenger of H202, N-acetyI-L-cysteine. Both Interleukin-3 (IL-3) and granulocyte-macrophage colony-stimulating factor (GM-CSF) have been previously identified to induce rapid phosphorylation of the MAP-kinase (Blood 79:2880 (Blood 79: ,1992 . However, little is known about signalling events initiated by IL-3/GM-CSF which occur downstream of the MAP-kinase. MAP-kinase has been shown to phosphorylate the AP-1 transcription factor and to activate two kinases designated insulin-stimulated protein kinase-1 (ISPK-1) and MAPKAPkinase 2. We here report that IL-3 and GM-CSF induce MAPKAP-kinase 2 activity in the human megakaryoblastic leukemia cell line MO7E and phos-phorylate the human small heat shock protein hsp27 on serine residues. In contrast, neutrophils failed to phosphorylate hsp27 upon IL-3, while GM-CSF induced hsp27 phosphorylation in a similar range as observed in MO7E cells suggesting that MAPKAP-kinase 2 mediated hsp27 activation occurs independently of proliferation. Hsp27 phosphorylation is dose-dependent and occurs as early as 5 minutes following exposure to IL-3 or GM-CSF. Moreover, hsp27 phosphorylation is inhibited by tyrosine kinase inhibitors such as genistein or herbimycin A. In addition, we show that protein tyrosine phosphatase and protein phosphatase 2A (PPA2) interfere with the ability of IL-3 or GM-CSF to induce serine phosphorylation of hsp27. Taken together, our findings indicate that tyrosine phosphorylation of MAP-kinase is a prerequisite for serine phosphorylation of hsp27 mediated by MAPKAP-kinase 2. Hsp27 is localized in the nucleus and has been linked to the cellular stress response. Its precise function, however, is largely unknown. Our data identify hsp27 as a target of IL-3/GM-CSF stimulation via MAP-kinase and MAPKAP-kInase 2. Further-more, our results indicate that hsp27 may also exert phosphorylation-activation functions involved in growth signalling pathways in unstressed cells. In recent years it has been shown that the mechanism of betalactam antibiotic-induced agranulocytosis involves a direct inhibition of the replicative DNA polymerases alpha, delta, and epsilon. In this report we examine, as a representative of these antibiotics, the effect of freshly prepared ceftazidime (CEF) and degradation products of CEF on myelopoiesis. We investigated freshly dissolved CEF and CEF incubated for 96 hours at 37 + (CEFd) on the production of myeloid cells in the supernatant (SN cells) and colony stimulating activity (CSA) by the stroma. One week after drug treatment of the mLTBMC a significan~ dosedependent inhibition of the myeloid cell production (xl0~ and CSA (assessed by CFU-GM assay) occured, as summarized in the following We here report that human lung fibroblasts respond to X-ray treatment (XRT) with release of Interleukin (IL)-6. Synthesis of IL-6 upon ionizing radiation is preceded by an increase of IL-6 transcript levels resulting from transcriptional activation of the IL-6 gene. Analysis of deleted fragments of the IL-6 promoter indicated that transcriptional activation of the IL-6 promoter was due to enhanced binding activity of the transcription factor nuclear factor (NF)-~B. Although activation protein (AP)-I did not participate in the rapid induction of the IL-6 promoter, its binding activity was also enhanced by XRT. In contrast to binding kinetics observed with NF-~:B, AP-1 binding following XRT was biphasic with the second peak being dependent on de novo protein synthesis. In contrast, however, NF-IL-6 activity was not enhanced by XRT in fibroblasts. The introduction of both the NF-~B-and the AP-1 recognition sequence, conferad inducibility by XRT to a heterolgous promoter, with reporter gene activity being maximal 24 hours or 48 hours following XRT, respectively. Sequential acitvation of two distinct transcription factors might thus contribute to synchronize transcriptional activation of different genes participating in the X-ray (XR) response. On the basis of study of functional and morphological characteristic of bone marrow stromal tissue of human fetuses, 8-27 week gestation, and of adults aged 16-60 years, and in experiments on animals, the role of the sinusoidal endothelium, reticular, fat and endosteal cells in hemopoiesis regulation has been concretely defined. The endothelium of sinuses forms the histo-hematic barrier "bone marrow-blood', ensures the Wanscellular migration of stem cells and mature blood cells, releases hemopoiesis-regulating factors and is involved in the erythroid cell maturation. Bone marrow reticular cells participate in the formation of intramedallar supporting frame-work, regulate the transvasal and intramedullar cell migration, form the extracellular matrix, produce humoral regulators of hemopoiesis, effect the cell differentiation by way ot their intercellular contacts with hemopoietic precursors and give origin to adipocytes, lntramedullar adipocytes present an energetic depot of hemopoietic and stromal tissues and in the stage of preadipocytes they effect, by means of contacts, the granuloeyte development. The endosteal cells of bone marrow are the source of intramedullar stromal tissue, they participate in the anchorage of the hemopoietic stem cells and form the microenvironment of the latter, regulate the endosteal Ca ion levels and might be a possible source of hemopoietic tissue (population of cells of the residual embryonal mesenchyma). The thesis on the mechanism causative of the impaired regulation of precursor proliferation and differentiation in hematologic diseases is proposed. The self-renewal capacity of CFU-S (spleen colony-forming unit) following the treatment of (CBAxC57B1)F1 female mice with recombinant human granutocyte colony-sffmulating factor (rhG-CSF) was investigated. The possibility of synergism between erythropoietin (EPO) and rhG-CSF in blood CFU-S mobilization was also studied. During the investigation the following observations were made: 25 and 250 mkg/kg/day injection of rhG-CSF expended the number of CFU-S-11 in circulation 4-and 32-fold, accordingly. The self-maintenance potential of peripheral blood CFU-S-I1 did not change significantly. The treatment of mice with 250 mkg/kg/day of rhG=CSF resulted in two fold increase of spleen cellularity and 15 fold augmentation of CFU-S-11 number, without noticeable changes in their self-renewal capacity. Moderate changes in CFU-S-11 number were observed after EPO administration in spleen and peripheral blood, however no significant synergistic effect of EPO with both doses of rhG-CSF was detected. The multifold increase of CFU-S-11 number in peripheral blood following rhG-CSF administration with no reduction in their self-maintenance potential suggests that mobilized with rhG-CSF blood stem cells provide an appropriate source for reconstitution of the hematopoietic system. We used the BrdU-Incorporation method to show the effects of 1L-2 (100, 1000, 3000 U/ml),IL-4 (0.1, 1, 10 ng/ml) and IL-2 (3000 U/nil) plus IL-4 (10 ng/ml) on B-CLL-ceIIs. After Ficollseparation, lysis of the erythrocytes (NH4CI) and lysis of monocytes (1-1eucin-methyl-ester), cells were divided into two groups. Group 1 was cultured in a serum free medium (+BrdU +cytokine) without any T-cell depletion. Group 2 was cultured in a serum free medium (+BrdU +cytokine) after T-cell (CD3 +) elimination by the MACS (Magnetic Activated Cell Sorting). Samples were taken after 20h, 68h, 92h, l16h and 140h. After staining with anti-BrdU FITC and propidiumiodide (PI) proliferation was measured by flow cytomemy (FACS scan). .00 P=proliferation N=no effect I=inhibition Conclusions: I[,-2 shows a proliferative effect on B-CLL-cells independent of T-cells. IL-4 shows heterogeneous effects. By itself it has most often no effect on proliferation, but sometimes it inhibits or increases the prolifoation. This effect does not seem to depend on T-cells. It could depend on the dosage or some unknown patients' characteristics. Further on IL-4 inhibits 11,-2 induced proliferation in nearly all cases independent of T-cells. Ii-i and TNF are inflammatory cytokines with overlapping biological activities. Human vascular endothelial cells express Ii-I and TNF receptors and respond to Ii-i and TNF stimulation by elaboration of colony stimulating factors such as G-CSF and GM-CSF.However quantitative data are required in order to evaluate the contribution of Ii-i and TNF to the activation of inflammatory hemopoietic cells such as granulocytes and macrophages by CSFs. Therefore we quantified the production of G-CSF and GM-CSF in endothelial cell cultures subsequent to treatment with IL-l~ (0.l-200U/ml) and TNF(10-2000U/ml)or ll-1 in combination with TNF.A dosedependent stimulation of G-CSF and GM-CSF secretion was found following ll-i and TNF treatment of endothelial cells.ll-i was a more potent inducer of G-CSF secretion than was TNF using approximately equipotent doses of IL-I (100U/ml) and TNF (1000 U/ml) regarding the induction of GM-CSF.In addition significant snperadditive stimulation of G-CSF and GM-CSF production was found with a low dose of Il-1 ( IU/ml) and a saturating dose of TNF(10OOU/ml) in combination.However the costimulatory effect of II-l(IU/ml) was significantly more pronounced with G-CSF than with respect to GM-CSF.In summary the differential modulation of G-CSF and GM-CSF production ~n endothelial cells by Ii-1 and TNF may indicate 'a disparate role of II-i and TNF in vascular inflammatory processes and atherosclerosis regarding recruitment and activation of inflammatory leukocytes. Cytokines are known to be involved in the pathophysiology of graft versus host disease (GHVD) and to effect lymphohematopoetic progenitor cell growth after bone marrow transplantation. The use of T-cell depleted marrow in human transplantation is associated with suppression of GVHD but decreased rates of engraftment. We have shown in rodent models that UVB irradiation (UVB) of donor inoculum inhibits GVHD while preserving engraftment. To determine the effects of UVB on eytokine production by cells associated with GVHD, human marrow mononuclear cells isolated by ficoll density gradient were UVB-irradiated at doses of 10-200 J/m2 and then stimulated with PHA/LPS or allogeneic cells. ELISA assays were used to measure the production of GM-CSF, IL-3, LIF, IL-lbeta, IL-2, IL-6, TNF-alpha, and lFN-gamma by stimulated cells . Two week methylcellulose cultures were used to determine viability of CFU-GM, BFU-E and CFU-Gemm progenitor cells after UVB. All results were compared to non-UVB-irradiated marrow and to marrow depleted by soybean agglutination and sheep erythrocyte rosetting. Progressively increasing doses of UVB produced progressively decreasing levels of all cytokines except IL-6, which remained unchanged. T-cell depleted marrow produced decreased levels of all cytokines except IL-6. UVB at 75 J/m2 resulted in higher levels of GM-CSF and IL-3 as compared to T-cell depleted marrow. This same dose of UVB essentially preserved CFU-GM, BFU-E, and CFU-Gemm colonies. We conclude that UVB may inhibit the cytokine cascade active in GVHD while preserving progenitor cell growth at UVB 75 Jim2. UVB may serve as GVHD prophylaxis in clinical marrow transplantation and in vivo studies on non-human primates are in preparation. The ability to migrate is fundamental for the acquisition of invasive properties by tumor cells. A tumor derived cytokine was identified by its ability to induce direct and random migration via a receptor mediating signal pathway, i.e. autocrine motility factor (AMF). We identified the receptor for AMF (HAMF-R) and found ~at hAMF-R plays a role in invasiveness and metastasis in human bladder carcinomas. We investigated the expression pattern of hAMF-R in 83 fresh frozen bladder cancer specimens by immunofluorescence technique on the translation level and found a strong correlation (p<0.01) to tumor stage and grade. Furthermore we have shown that patients who were found to be hAMF-R positive have an increased risk for early tumor progression (p<0.05). A large number of substances from the working place and in the general environment such as quartz and coal mine dusts are causing silicosis and leading to lung fibrosis. Alveolar macrophages are the primary target for the noxious effect of quartz dust. Quartz dust incubated human macrophages release in vitro a cytokine, which stimulates cell proliferation of human lung fibroblasts (Wistar 38). In recent studies we found that beside fibroblasts also epithelial cells of the alveolar unit, such as pneumoeyte type II cells (A-549) respond to the cytokine with strong proliferation. Supernatants of quartz dust exposed macrophages were concentrated by ultrafiltration and thereafter fractionated by gelfiltration (Sephadex G150, Pharmacia). Biological activity of the factor eluted within a molecular range of 75-102 kD. Furthermore, the factor was purified by anion exchange chromatography (Q-Sepharose). Fractions revealing biological activity were further analysed by SDS-gel eleetrophoresis (PAGE) and showed two protein bands with apparently molecular masses of 76 and 79 kD,respectively. After addition of the supernatant initially spindle shaped pneumocytes were detected, followed by epithelial-like cells when cell proliferation progressed. Induction of spindle-shaped pneumocyte type II cells could also be seen after addition of pure Tumor Necrosis Factora. However, in this case no cell proliferation was observed. We assume that beside the cytokine, which is responsible for the induction of cell proliferation TNF~ is present in supernatant. differences in a variety of immunologic parameters. In peripheral blood however differences to healthy persons have hardly been described. In this investigation we compared serum levels and concentrations of Interferon-7 (fiN-y) and IL-l--ct in stimulated samples of whole blood. 50 Ixl whole blood of 38 healthy controls (REF) and 39 patients with sareoidosis, 27 without treatment (POT) and 12 under corticoid medication (PUT), was incubated in medium at 37~ and 5% CO 2. Con A was used for stimulation of IFN-y and LPS for IL-1--m Concentrations were determined with an ELISA. Results: Without stimulation no measatalg, e amounts of IFN--y could be found. After stimulation REF showed median coneentrations of 127 ng/ml, POB 10 ng/ml and PUT 2 ng/ml. The difference between REF and PUT respectively POT was statistically significant. IL-l-m without LPS the differences were not significant. Under stimulation POT had si~ificantly higher values (220 pg/ml) compared to REF (170 pg/ml) and also to PUT (155 pg/ml). In conclusion we were able to demonstrate that in conlrast to serum levels stimulation of peripheral whole blood reveals significant differences in concentrations of IFN--y and IL-1--ct between patients with sarcoidosis and healthy references. We established a modified polymerase chain reaction protocol for the detection and semiquantitative assessment of mRNA-transcdpt levels for IL-2, IL-4, IL-5, IL-6, IFN-7, TNF-cL, GM-CSF, TGF-13 and IL-2-receptor-c{ (IL-2R). The method was shown to distinguish 10-fold differences in template concentration after 4 rounds each of amplification in the range from 20 to 40 cycles; the lower threshold of sensitivity was at 10 copies per PCR-reaction. Reproducibility was >95% for a positive result after 32 rounds of amplification; it decreased to 80% after 36 rounds. This corresponded to the detection of 1,000 and 100 template copies, respectively, per PCR-reaction. Human peripheral blood mononuclear cells (PBMC) and tumor cell lines were evaluated for mRNA-expression with or without stimulation and these results were compared to secretion of the corresponding cytokine. For PBMC, constitutive mRNA-expression was found positive for TNF-o~, IFN-7, IL-4, IL-6, TGF-13 and IL-2R, whereas detectable expression of IL-2, IL-5 and GM-CSF was induced only after stimulation. Using ~-actin as an intemal standard, the PCR could demonstrate relative differences in cytokine transcripts after stimulation with (A) 100 ]U/ml IL-2, (B) 10% lymphocyte-culture conditioned media (CCM) and (C) PMA (10ng/ml) plus A23187 (100ng/ml). For IL-2 transcripts no detectable expression was found without stimulus or after addition of IL-2, whereas CCM resulted in a 1,000-and PMNA23187 in a 10,000-fold increase. Other mRNAtranscripts increased t0-fold (TNF-(~) up to 10,000-fold (GM-CSF) with or without differences between the stimulating agents. The cell lines CAKI-1 (renal cell carcinoma) and Daudi (Burkitt lymphoma) also expressed comparable levels for cytokine transcripts, with a strong induction after stimulation with PMNA23187. The relative IFN~ mRNAcontent in CAKI-1 increased from 0 to 1,000, for GM-CSF from 0 to 10,000. The influence of different cytostatics on IL-6 production by peripheral blood mononuclear cells (PBMC) was studied. PBMC were pre-incubated with or without phytohemagglutinin P (PHA-P), then pulse exposed during 1 hr to different cytostatics in their therapeutical concentrations, washed three times and incubated additionally during 24-72 hrs. Then the supernatants were collected and IL-6 biological activity was measured in MTTmodified B9-cells biological assay. Though significant individual variations of the IL-6 production were found, all the studied drugs can be separated on severa) groups: 1) adriamycine and methotrexate induced late increase of IL-6 production ( 48-72 hrs ); 2) cytarabine strongly increased the early as well as the late IL-6 production ; 3) the pretreatment with etoposide and rubomycine suppressed subsequent production of IL-6 by PBMC during 24-48 hrs, the late production was increased; 4) in contrast, the cyclophosphamide pretreatment stimulated the early production and strongly suppressed the late one. The changes of IL-6 production by PBMC was not due to the cellular death. The PHA-P stimulated PBMC produced usually more IL-6 that unstimulated cells did. These data suggest that the cytostatics possess the different effects on IL-6 production by PBMC that could be important in the therapy of malignancies. Recent studies implicating a deficiency of interleukins in several diseases have underlined the importance of measuring in vitro the DNA synthesis and the cytokine production (IL-1, IL-2, IL-6, TNFalpha) in the same cell system. Previously had found that normal peripheral blood mononuclear cells (MNC) of patients suffering from high-malignant Non-Hodgkin lymphomas showed a decreased capability to proliferate after mitogenic stimulation (PHA, Con A, PWM). Here we have studied the DNA synthesis and the production of different cytokines (IL-1, IL-2, IL-6, TGF-i~ and TNF-alpha) by pokeweed mitogen (PWM) stimulated MNC from 15 healthy control subjects and from 14 patients with NHL. The IL-2 production of PWM-stimulated MNC of patients with NHL was found to be significantly decreased, wheras the IL-1, IL-6 and TNF-alpha release were not changed significantly. These data showed a good correlation with the reduced capability of MNC from patients with NHL to proliferate after mitogenic stimulation. The multifunctional cytokine Transforming Growth Factor-II (TGF-i3) is known to inhibit the DNA synthesis, as well as the IL-2 production of mitogenstimulated MNC. However, TGF-I~ release was not significantly changed in call culture supernatants from patients with NHL in comparison to healthy controls. We conclude that the suppressed DNA synthesis and 11_-2 production of MNC from patients with NHL is not the consequence of a deceased TGF-5 level secreted by these cells. is a co~on problem after chemotherapy and requires supportive care until normal hemopoiesis has recovered. To study the importance of endogenously produced cytokines for regeneration of bone marrow progenitors we measured serum levels of G-CSF and Ii-6. Blood samples were obtained before and 24 hours after chemotherapy, during the stage of leukopenia (<1000/ul) and recovery (>1500/ul)of leucocyte counts. In 6/7 patients we found a more than 5-fold increase in serum G-CSF levels at the stage of leukopenia. Highest amounts were observed in two patients with lethal outcome. There was no correlation between thrombocytopenia and levels of G-CSF or IL-6. serum G-CSF (pg/ml, mean, range) before (leuko>1500) after (<1000/ul) chemotherapy AML DAV 25 (5-i00,n=14) 2000 (50-8000,n=7) ALL 'H~izer" 100 (10-300,n=4) 1000 ( We measured different hematopoietic cytokines as G-CSF, GM-CSF and IL-6 in amniotic fluid and cord blood to ctearify their physiological and palhophysiological role in fetal and neonatal development Amniotic fluid was available from the 15 th to the 21 st week of gestation (n=53), cord blood from the 26 th to the 42 nd week of gestation (n-60). Activity levels of cytokines were determined by stimulation of the ceII lines NFS-60, 77D1, and TF-I. which are responding specifically to G-CSF, IL-6, and GM-CSF. Specificity has been proved by neutralizing antibodies. Calculation of cytokine levels was done by standards of recombinant growth factors. Sensivity for G-CSF and IL-6:5 pg/mL for GM-CSF: 1OO pg/ml. In amniotic fluid G-CSF tanged from 40 to I23.OOO pg/m[ and IL-6 from 20 to 12.000 pg/ml, whereas GM-CSF was not detectable. In cord blood G-CSF was between 5 and 23.000 pg/ml and [L-6 between 5 and 30 pg/mh In most of the samples (QO%~ GM-CSF was beyond the sensivity limit. In 92% (11 of 12 cases) G-CSF levels were elevated over 200 pg/ml and associated with amnion infection syndrome, while green amnioflc fluid alone during delivery did not stimulate the production of G-CSF, The levels of the hematopoietic cytokines showed no influence by the gastational age. Identical twins without maternal infection showed the same G-CSF levels. Inflammation of the amniotic membrane and maternal sepsis is associated with elevated G-CSF levels in cord blood. GM-CSF can normally not be detected in cord btcod and amniotic fluid. Detectable levels of GM-CSF in cord blood and amniotio fluid maybe give a hint for a pathological situation during pregnancy. Total number of 110 children with ALL, 68 boys and 52 girls, aged from 5 to 15 years were included to the study. TNF production was studied acc. the method based on growth inhibition of sensitive to TNF L 929 mice fibroblasts, ILL-1 production ace. method based on inhibition of autologous rosette formation by thymoeyteg of CBA mouse and 1L-6 production ace. to conventional EL1SA Genzyme-test. Twenty five healthy children served as the control group. It was found that in children with ALL during the whole period of therapy the IL-1 and IL-6 production, was significantly lower than that observed in the control group of healthy children (p 0.005). The TNF production in ALL children before therapy was lower in comparison with the control group values. During cytostafic therapy was higher and grew up above the normal limits after cessation of the therapy. IL-1 production grew up after the end of the therapy but never reached the value of the control group. EFS at 54 ruth in ALL children with IL-1 production < 10/~ before therapy was higher than that in children with IL-1 < 10/~ (EFS 90 % v 70 %). The IL-1 production seems to be a good prognostic marker. (pts) with active autoimmune disorders as well as with malignant lymphomas. In addition, CD23 and its soluble form (sCD23) is thought to be involved in the regulation of B-cell proliferation. Therefore, we examined sCD8, sCD25, and sCD23 in pts with B-cell chronic lymphocytic leukemia (B-CLL) in order to assess their role as indicators of disease activity. Fifty-five pts with B-CLL were studied so far. Staging was performed according to the classification systems of Rat and Binet, respectively. Serum samples were freshly stored in liquid nitrogen until further processing. Levels of sCD8, sCD25, and sCD23 were measured by a sandwich ELISA technique using commercially available assays (Biermann, Germany). Advancing Rat stages were associated with a progressive increase of all the three serum factors (sCDS: Rat 0 556+165 U/mt vs Rat IV 1310+150 U/ml, sCD25: Rat 0 1326-1-190 U/ml vs Rat IV 8850__.1330 U/ml; sCD23: Rat 0 1265+455 U/ml vs Rat IV 6500+_1800 U/ml). This progression was also evident when Binet's classification was applied. Occurrence of B-symptoms was associated with high levels of sCD25 (p<0.001), whereas sCD8 and sCD23 were found also to be increased but without statistical significance. High levels of all the three factors strongly correlated with a lymphocyte doubling time < 12 months, a lymphocyte count >50.000/tzl, and with the presence of hepato-and splenomegaly. Interestingly, occurrence of bulky lymph nodes (i.e. at least one nodule of > 8 cm in diameter) was linked with high levels of sCD23 only (p<0.002). In summary, (1) progressive serum levels of sCD8, sCD25, and sCD23 correlate with advancing stages of disease in B-CLL. (2) B-symptoms were associated with high levels of sCD25. (3) We found sCD23 to be the more sensitive marker of the total tumor load than sCD8 and sCD25. Thus, sCD23 may be useful in monitoring pts with B-CLL. Cytosine arabinoside (Ara-C) is one of the most active single agenls in the treatment of acute myeloid leukemia (AML). Its cytotoxic activity mainly depends on its phosphorylation to Ara-CTP and on its incorporation into the DNA. Based on recent in vitro studies showing that hematopoietic growth factors like GM-CSF and IL-3 enhance the cytotoxicity of Ara-C on clonogenic leukemic cells, the GM-CSF priming concept is currently explored in clinical phase II and III studies. In an ongoing study at the Universities of Muenster and Goettingen GM-CSF (250/~g/m2/d) is started 24 hrs before induction chemotherapy (TAD9/HAM) until recovery of blood ceil counts. This study provided a means to assess the effect of GM-CSF on the intraceUnlar Ara-C metabolism in vivo in 23 pts with AML. Enzyme activities of deoxycytidine kinase (dCK), thymidine kinase (TK), dcoxycytidine deaminase (DCD), DNA polymerase (Pol) and DNA polymerase alpha (Pola) were determined before therapy, 24 hrs after the administration of GM-CSF and 48 hrs after the administration of Ara-C. In addition, Ara-C incorporation into the DNA was measured after 48 hrs Ara-C administration. Enhancement of enzyme activity was observed in 11/15, 9/17, 10/16, 6/16 and 7/16 cases for Pola, Pol, TK, dCK and DCD, respectively. Increases ranged from 13-630% for Pola (median 127%), 13-188% for Pol (median 77%), 14-820% for TK (median 113%), 10-670% for dCK (median 43%) and 29-1350% (84%) for DCD. Inadequate blast cell reduction after TAD9 (>5% blast cells on day 10 or 21) was associated with significantly higher DCD blast cell activities compared to the DCD activity values obtained for pts with adequate blast cell clearance (median values: 5.12 vs 0.6 nmol/min x mg, p<0.01). Cases with DCD activities <5 nmol/min x mg showed significantly higher Ara-C incorporation into the DNA compared to pts with DCD activities >5 nmol/min x mg (5.2 vs 0.65 ng/107 cells). Furthermore, inadequate blast cell clearance was associated with lower Ara-C incorporation into the DNA (median 0.8 vs 5.3 ng/107 cells) and lower Pola activities (median 1.3 vs 3.1 pmol/min x mg). In 11 pts we investigated simultaneously the effect of GM-CSF pretreatment on Ara-C metabolism in vitro. Enzyme activities of Pola, TK and Pol correlated significantly in vivo and in vitro (rP~ rTK=o.64, rP~ p<0.05). These data demonstrate that GM-CSF enl/ances DNA synthesizing enzyme activities in vivo and in vitro. Furthermore, these data suggest that GM-CSF might improve the therapeutic response to induction chemotherapy by increasing DNA polymerase alpha activity and thereby increasing the Ara-C incorporation into the DNA. The effects of Interferon-alpha (IFN-alpha), Interferon-beta2/ lnterleukin-6 (IL-6) and Interferon-gamma (IFN-gamma) in inducing megakaryocytic differentiation of blast cells from acute megakaryoblastic leukaemia (AMegL) patient determined by the increase in CD41 and CIM2b expressions using monoelonal antibodies in APAAP technique were investigated in liquid suspension culture. After six days of culture, the percentage of CD41 and CD42b positive cells increased in control cultures from 15,2 % and 10,6 % on day 0 to 32,0 i 4,3 % and 22,1 • 2,5 %, respectively. The addition of IFN-alpha significantly increased the number of CIMI and CIM2b positive cells by about two to three fold compared to control cultures, p <0,01 and by about four to six fold compared to day0, p <0,001. Similarly, IFN-beta2/IL-6 induced a significant increase in CD41 and DC42b positive cells. On the other hand, IFN-gamma failed to increase the number of CIM1 and CD42b positive cells in comparison to control cultures on day 6 and instead induced a significant increase in the number of monocytes/macrophages from only 3,7 _+ 1,9 % in control cultures to 47,9 + 2,5 %, 52,3 + 1,4 %, 54,5 • 2,9 % and 57,5 • 2,3 % in 1, 10, 100 and 1000 Units/nil IFN-gamma-treated cultures, respectively, p <0,001. The present results suggest that megakaryocytic differentiation of blast cells in AMegL could be induced by IFN-alpha and IL-6 and support a clinical role for IL-6 in the treatment of AMegL patients. Also, the present results showed that monocytic differentiation of blast cells in AMegL could be induced by IFN-gamma~ supporting the multipotent stem or progenitor cell origin of the AMegL subtype of acute myeloid leukaemia. A monoclonal antibody-based ELISA and bioassay were used to measure leukemia inhibitory factor (LIF) protein levels, activity and the functional role of LIF in superuatants of cultured stromal cells derived from patients with acute and chronic myelogenous leukemia, myelodysplastic syndrome, and hairy cell leukemia and from normal controls. We demonstrate that biologically active LIF protein is constitutively produced and secreted by coltured bone marrow stromal cells from all subjects studied. In addition, adherent-layer conditioned-media LIF protein levels were significantly elevated in samples from patients with all hematologic malignancies studied as compared to samples from normal controls. Confluent adherent layers exposed for 24 hours to interleukin (IL) 113 or tumor necrosis factor-9 (TNF-a) showed a significant increase in LIF protein levels, whereas exposure to IL-4 (Sterling Drug Inc., Great Valley, PA) resulted in a dose-dependent decrease in LIF levels. 3.0 (i.6-8.0) 2.5 (0.9-5.8) 6 0.036 Interestingly, neutralizing antibody against LIF caused a 25% reduction in normal progenitor proliferation derived from the superuatant but not from the adherent layers, and this effect was reversible by the addition of recombinant LIF protein. We conclude that (i) biologically active LIF protein is constitutively produced by adherent layers from normal donors, (ii) TNF-ct and IL-113 increase and IL-4 decreases adherent layers LIF protein levels, (iii) the steady state levels of LIF protein produced by adherent layers from leukemic patients is significantly elevated, and (iv) LIF may participate in the interaction between adherent layers and hematopoietic progenitors to maintain normal hematopoietic colony growth. It is well known that bone marrow stromal cells have crucial impact on haemopoietic cell proliferation. Little is known about stromal humoral factors leukemic cell proliferation. The aim of this study is to evaluate the effect of stromal cell conditioned media (SCCM) on the 3H-thymidine uptake by normal and leukemic target cells. 21 patients with AML were studied treated with "7+3" based regimens. Long term bone marrow cultures were established in non-leukemic and leukemic patients (before and during treatment). Target cells for SCCM were normal haemopoietic cells and leukemic blasts. The results are the comparison of the effects of leukemic and non-leukemic stromal cells. A part of the patients revealed high stimulative activity upon non-leukemic cells (+125 + 2 %, p<0.001) and inhibited proliferation of leukemic cells (-70 • 3 %, p<0.001) This group entered complete remission. SCCM of another group of patients inhibited proliferation of nonleukemic cells (-84 _+ 3 %, p<0.001) and stimulated blast cell proliferation (+ 146 + 5 %, p<0.01). The magnitude of the figures was even more profound later: during treatment, bone marrow aplasia, recovery. This group of patients failed to achieve remission. It seems that stromal cells has significant on impact on restoration of normal or leukemic hemopoiesis after chemotherapy. Peripheral B-CLL-cells from 51 patients were investigated upon the proliferating effect of IFN~, TNF~ or combination of both in serum free culture medium (CG-medium). Blood cells were drawn from patients and lymphocytes separated by Ficoll Hypaque and monocyte-lysis (Leucinemethyl-ester incubation). T-cells were depleted using a, CD3 mAB and MACS (Magnetic Activated Cell Separat9 r, Miitenyi Biotec). At each step heterogeniety of the population was controlled by FACS analysis with 7 different mAB (CD 3, 4, 8, 1O, 14, 19, 20, 56) . The homogeneous population (contamination less than 1%) was co-incubated with both cytokines (0.1 -10 ng/ml) and Bromodeoxyuridine (BrdU) in CGmedium. After 20, 60, 92, 112 and t40 hours cells were harvested an analysed for BrdU-incorporation into the genome. IFNu and TNFc~ measurings (51 in total) were almost similar: 24 patients were non-responder and showed no stimulatory effect on cells; 5 patients showed an inhibitory effect; cells from 21 patients were responding upon cytokine cultivation. The combination of both IFN~ and TNF~ produced in these cells an additive effect (14 out of 21). Best results could be observed when the control population (without cytokine) was minimal proliferating compared to no proliferation. A high correlation was observed between cytokine response and pre-treatment: without glucocorticoid treatment of patients prior to measurements the influence of cytokines on resting B-CLL-cells was significantly higher (with methyl-prednisolone 15%, without we-treatment 61% were responders). Functional defects of the monocyte/macrophage system probably contribute to the increased rate of severe infections in patients with myelodysplastic syndromes (MDS). Therapeutic trials with hematopoietic growth factors (HPGF) have resulted in substantial improvement of cytopenia, especially neutropenia. However, little is known about the alterations of the monocyte/macrophage system during these therapeutic interventions. It therefore was the aim of the present study to analyze the capacity of monocytes/macrophages from MDS patients prior to and after HPGF therapy to secrete IL-11~, TNFe, IL-6, and IL-8 upon in vitro stimulation with lipopolysacharide (LPS). Sixteen patients were studied: 1t had a refractory anemia, 5 had a refractory anemia with excess of blast cells. Prior to therapy, the capacity of adherent monocytes/macrophages to secrete IL-11~, TNFc~, IL-6, and IL8 was significantly reduced by 50-70 percent as compared to normal controls. On the other hand, oxygen radical release was normal in 11 MDS patients tested. Treatment with GM-CSF (15-250/~g/m2/d SQ x 7-14; n=7), IL-3 (60-500/~g/m2/d SQ x 15-18; n=3), and G-CSF (1-3 Fg/kg/d SQ x 84 in combination with all-trans retinoic acid; n=6) normalized the potenital of monocytes/macrophages to secrete IL-tI~, TNF~, and IL-6. IL-8 secretion was only improved by IL-3 or GM-CSF dosages _> 250 /~g/mZ. Oxygen radical release was significantly stimulated by both GM-CSF and IL-3. These results indicate that treatment of MDS patients with GM-CSF, IL-3, and G-CSF (the latter in combination with all-trans retinoic acid) can restore deficient monocyte/macrophage secretory function to normal. Depletion of CD6 positive T ceils has been used in human patients for prevention of GVHD. We studied depletion of CD6 + cells from canine marrow for induction of GVH-tolerance across a complete DLA-haplotype difference. Prompt engraftment and fatal GVHD occurs in a littermate combination of DLA-homozygous donors and DLA-heterozygous recipients when undepleted marrow is given. AIIogeneic marrow depleted with a crossreactive antibody to human CD6 and immunomagnetic beads was given to 6 dogs. One dog died with haemorrhage on day 48 due to thrombocytopenia, 5 dogs showed complete hematopoietic recovery. 4 dogs became tolerant chimeras and one dog died with GVHD due incomplete depletion. Chimerism was mixed early after transplantation, became complete later and is still complete in 2/4 dogs after 1-2 years. 4 dogs received CD6 depleted marrow grafts and lOIJg/kg/d s.c. r-canine G-CSF starting on day 2 after transplantation. Although all dogs had fast recovery of granulocytes, 2 dogs receiving 0.5 x 108 MNC/kg died of marrow aplasia on days 47 and 62 without recovery of thrombopoiesis. Two dogs receiving 1 x 108 MNC/kg had sustained engraftment with delayed recovery of thrombocytes compared to dogs without G-CSF. FACS analysis of depleted marrow showed complete depletion of CD4 + cells but about 2% of CD8 + cells; CFU-C growth and NK-activity was retained after depletion. CD6 depleted marrow inhibited the generation of cytotoxic cells. These experiments indicate that qualitative T-cell depletion is effective, since CD6 recognizes only subpopulations of T-cells. The application of r-canine G-CSF enhanced the recovery of granulocytes but led to graft failure in dogs receiving a low number of marrow cells. GSF-Inst. fDr Immunologic, Marchioninistr. 25, 8000 MQnchen 70 Supported by the Wilhelm-Sander Foundation We wished to analyse the factors which may affect the yield of PBPC (CFU-GM) to be collected by leukapheresis following high-dose Cyclophosphamide (HD-CYC: 7g/m2). We retropectively studied the following criteria in 31 patients with high-risk MM of which 10 received GM-CSF (Sandoz SA/ Schering-Plough) after HD-CYC: time from diagnosis to HD-CYC, number of chemotherapy cycles (CTC) (Fermand, 1992) , B2 microglobulin, bone marrow plasma cell count before HD-CYC, administration of GM-CSF after HD-CYC, "slow" or "fast" rate of platelet and WBC recovery (Jagarmath, Schwartzenberg, 1992), "poor" or "good" mobilization ofPBPC (Jagannath, 1992), differential WBC count between "day x ~ and "day x-l" during haematopoietic recovery. Each variable was studied as continuous (regression analysis) and discontinuous (t-or chisquare tests). When the differential WBC count was < 1300 WBC/IaL, 37% of the leukapheresis procedures performed on day x yielded more than 30x 104 CFU-GM vs 90% when it was _> 1300/tJL (p<0.05). The infusion of GM-CSF was asaocAated with a higher yield of CFU-GM (BIdT, 1993). The patients with "good ~ PBPC mobilization (> 50 x 104 CFU-GM in >_ 2 leukapheresis) could all be transplanted with PBPC alone (vs 47% of those with ~poor" mobilization). They had a shorter duration of aplasia after transplantation than the other patients (p<0.03). The "fast" WBC recoverers had a higher yield of CFU-GM than the other patients (p<0.05). When only the patients who did not receive GM-CSF after HD-CYC were considered, a higher yield of CFU-GM was achieved in patients who underwent < 6 crc as compared to those who underwent > 6 CTC before HD-CYC (p lxlOg/l. A positive correlation of the rising and decreasing subpopulation counts within the MNC's were noticed (r=.65-.85), however the CD3* were in inversed ratio to the CD14 § cells (r =-.66). The percentages of CD3 * and especially of the CD25 + cells showed an increment immediately after Ctx, where the proportions of CD14" and CD34 § cells tended to fall. There was also a correlation between CD34" and CD14 § cells (r=.76, p<.001) and between CD34 § ceils and CFU-GM growth (r= .82, p<.001 ). An increased clonogenity was associated with low numbers of CD34 * cells: CD34~: CFU-GM 38:1 before Ctx; 18:1 immediately after Ctx; 38:1 during the regeneration phase. The same phenomenon could be seen by intensively pretreated compared to less intensively pretreated patients. The hematopoietic reconstitution parameters (median) were as follows in the PBSC-rescued vs non-rescued pts: platelet transfusions -18 vs 28 (p<.02), RBC transfusions -6 vs 8, days with platelets <50000//11 -6 vs 13 (p<.01), duration of neutropenia with WBC< 1000//JI -5 vs 9, days with fever -6 vs 8.5. The augmentation of CD34 § cells correlated with rising numbers of MNC's and especially of CD14 § but not with CD3*. The correlation between CD34" cells and CFU-GM in peripheral blood was convincing. Dose-escalated cytotoxic therapy with stem cell support may be considered for patients with stage II1 multiple myeloma, because of the poor median survival of only 3.5 years with conventional treatment. A threshold quantity of 5x106 CD 34+ eells/kg bw is necessary for a rapid and sustained engr~ent following myeloablative conditioning therapy. Since june 1992 six patients (median age: 46 years, range 31-53) with MM received 5pg G-CSF/kg bw (Neupogan R, Amgen) so. daily at the time of best response with conventional treatment. The content of CD 34+ cells in the peripheral blood was monitored by FACS each day. Leukapheresis were started when a detectable population of CD 34+ cells appeared. In 2 of 9 steady-state leukaphereses, more than 1,0xl0 ~ CD 34+cells were harvested. After the therapy with high dose eyclophosphamide ( 5 pts 4g/m 2, 1 pt 7g/m 2) plus G-CSF more than 1,0xl06 CD 34+ cells were collected in 11 of 19 leukapheresis. To date two pts have undergone myeloablative conditoning therapy with hyperfractioned total body irradiation (14,4 Gy) and melphalan (140 mg/mZ). One patient received 200 mg/sqm melphalan as eonditoning therapy. After the reinfusion of the G-CSF-mobilized PBSC, a rapid engraftment was achieved with median time of 13 days ( range 9-16) to reach 0,5 x 109/1 neutrophils and 9 days (range 5-14) for 20.0x109/1 platelets. No hematopoietic growth factors were given post-transplantation. In this pilot study, high dose eyclophosphamide and G-CSF is an effective method for harvesting PBSC. Evaluation of the mobilization of hematopoietic stem cells during steady-state hematopoiesis using higher doses of G-CSF is planned. were treated with Mtx, Ifo, Ara-C, prednisolone and escalating doses of VP16. Pts. (n = 12) with relapsed or advanced non-seminomatous germ cell tumors were treated with cisplatinum, escalating doses of VP16 and Ifo. The protocol design was similar: G-CSF before Ctx (2x12 pg/kg/die s.c.) with PBSC-aphereses at days 5 to 7 (NHL) or days 5, 6 (NSGCT} followed by 2 to 4 Ctx-courses. Cytaphereses were also performed after Ctx when the total WBC's recovered above lxl0~ The Ctx-cycles were followed by reinfusion of the previously collected PBSC (n = 44) and application of G-CSF (5 pg/kg/d.; n = 57) up to the last day of the subsequent stem cell collection. The CD34 § cells, the clonogenic peripheral blood progenitor cells (CFU-GM & BFU-E) and light density cell (LDC) counts were determined in 150 cytapheresis samples. The Ctx/G-CSF courses contributed to substantially higher progenitor cell amounts than G-CSF alone (n=75/21; p<0,01), without a difference in the collected LDC (1.3• vs 1.7• LDC; 3.9• vs 0.8• ~ CD34 § cells; 10.9+2.1 vs 9.0• 1.5x104 CFU-GM; 11.6 • 4.3 vs 6.9 • 0.9x104 BFU-E/kg/apheresis; n = 102/48; mean + SEM), but with approximately two times lower clonogenity. Two or three leukaphereses were enough to rescue 3 Ctx-courses with a minimal dose of 2x106 CD34 + cells/kg/patient. The optimal time to initiate PBSC-collection after Ctx in the studied patient group was proved to be at the l't to 4 ~h day after reaching leukocyte > 1000 pL. The original total leukocyte, light density cell (LDC) and platelet counts in the peripheral blood at start of leukapheresis played an essential role for the efficiency of the procedure (EFF), as shown by regression analysis: EFF to total leukocyte count correlation was r2=-0.48 (p<0.Ot); EFF to LDC-count r 2 =-0.55 (p1000 pL and 12 days -platelets>50OOO pL; duration of neutropenia was 5 days; and 8 days to become platelet transfusion-independent (median). Total number of 25 children, 17 girls and 8 boys, with neoplasia aged from 1-18 years were treated with GM-CSF-Leucomax Sandoz and G-CSF Filgrastim Hoffmann-La Roche, during severe myelosupression occuring after intensive polychemotherapy. In 4 children GM-CSF was applied twice after 2 consequent courses of chemotherapy. One child received GM-CSF four times after 4 chemotherapy courses. Twenty children with malignancies served as historical control group. GM-CSF was given at dose 5 #g/kg, G-CSF 5-10 #g/kg daily s.c. Duration of therapy ranged from 2-28 days with median 10 days. After cytokines therapy increase of mean and median numbers of total WBS, neutrophils, monocytes and eosinophils was observed. The median time to hematopoietic recovery was shorter in the group of children treated with cytokines when compared with the control group. (9 v 16 days). In 19 of 25 children signs of infection disappeared even before granulocyte count increase. Also shorter median time of febrile days, 4 v 12 days, in comparison with the control group was observed. No serious side effects during cytokines therapy were noticed. Only in one patient local erythema in injection place was observed. In two children transient retrosternal pain was seen. Our results showed that GM-CSF and G-CSF administered in children with neoplasia after chemotherapy shortens the period of neutropenia and infection duration. For both Hodgkin's disease and non-Hodgkin's lymphoma the outcome of chemotherapy has been shown to correlate closely with the dose intensity of treatment. However, dose intensification is limited most often by severe myelosuppression with the subsequent risk of fever and infections. We performed a clinical trial in 17 patients with Hodgkin's disease or non-Hodgkin's lymphoma to evaluate whether r-metHuG-CSF could facilitate the safe and timely administration of an intensive chemotherapy regimen. Patients who developed neutropenia _< 0.5 x 109/L for more than two days and / or fever _> 38.2~ and / or signs of infection after a cycle of chemotherapy (CEBOPPNIM protocol administered at intervals of 21 days), as well as patients in which chemotherapy had to be delayed due to an ANC _< 1.5 x 109/L on day 1, were eligible for treatment with r-metHuG-CSF. In the subsequent cycles r-metHuG-CSF was given subcutaneously at a dose of 5 #g/kg/d from day 11 to 20. 16 of 17 patients were evaluable, one patient had received only 1 day of r-metHuG-CSF treatment. 15 of the 16 evaluable patients experienced neutropenia with an ANC of less than 0.5 x 109/L during the chemotherapy course preceeding r-metHuG-CSF treatment, whereas only 5 patients had ANCs _< 0.5 x 109, L after the subsequent therapy with r-metHuG-CSF (p<0.01).Overall analysis showed that the duration of ANC nadirs < 0.5 x 109/L was on average 3.27 days in 41 cycles without r-metHuG-CSF compared to 1.78 days in 52 cycles with r-metHuG-CSF treatment. The administration of chemotherapy had to be delayed only for 2.47 days (mean value) during cycles with r-metHuG-CSF. Side effects probably related to r-metHuG-CSF, were moderate muscle and joint pain in 3 patients and chills in one patient. In general, r-metHuG-CSF was well tolerated. Under this treatment regimen 11 patients reached complete remission, 4 patients reached partial remission and one patient had stable disease. One patient was treated adjuvantly after gastrectomy. In conclusion, r-metHuG-CSF allowed the safe and timely administration of this intensive chemotherapy regimen and reduced myelosuppression for patients with Hodgkin's disease and non-Hodgkin's lymphoma. Based on previous studies (Cancer Res. 1991 :51, 116, Blood, 1992 :80, 1141 we know that many patients (pts) cannot receive CT consisting of carboplatin (CBDCA) 300 mg/m 2 and cyclophosphamide (Cyclo) 750 mg/m 2 for OC every 4 wks without hematopoietic growth factor support. The desirable dose of rhIL-3 based on a phase I/II study in this setting was 5 or 10 gg/kg/day. A study was designed to determine whether rhIL-3 would allow CT administration every 3 wks with 17 pts treated to date. Cyclo was administered 750 mg/m 2 and CBDCA was dose adjusted to creatinine clearance: 60-80 mL/min: 257 mg/m 2, 80-120 : 300 mg/m 2, 120-140:340 mg/m 2, > 140:385 mg/m 2. A total of 6 cycles (c) was administered. RhIL-3 (5 or 10/zg/kg/d) was given sc d2-11 in each c. At 5 and 10 #g doses are 10 (46c) and 7 (33c) pts evaluable for toxicity and I0 (43c) and 7 (27c) pts for efficacy. Side effects were fever and headache controllable with acetaminophen. At 5 #g rhlL-3 in three c (2 pts) and at 10 #g in six c (5 pts) urticaria occurred. In 4 episodes dyspnea and/or oedema was observed. This reaction only occurred during c34~ and was controlled with antihistamine and prednisolone. CT could be administered every 3 wks in 43/70 c (25/43 c at 5/~g, and in 18/27 c at 10 /zg (NS)), every 4 wks in 13/70 c and > 4 wks in 14/70 c. No platelet transfusions were required. Thus, in 61% of c it was possible to give a CT dose intensification of 33%. If full dose CT were to be given every 4 wks it would have been possible to administer in 80% of c in time. Conclusion: with rhIL-3 CT dose intensification of 33% is possible by reducing CT intervals, while no platelet transfusions were required with rhIL-3. Dept. of Medical Oncology. University Hospital Groningeu. Oostersingel 59, 9713 EZ Groningen. The Netherlands. K. Mempel, A. Reiter, E. Yakisan, E. Odenwald, M. Pfetsch, G. Schwab, H. Riehm, K. Welte. In the multicenter trial ALL-BFM 90, we have initiated a phase III study of rhG-CSF in children with High Risk ALL. High risk (HR) patients are characterized by at least one of three criteria: 1. Prednisone poor response (2 1000/mm 3 absolute blasts number in the blood at day 8 after 7 days' exposure to prednisone), 2. failure to achieve complete remission at day 33 of induction therapy, and 3. t(9;22). The primary objective was to test whether rhG-CSF reduces the incidence of febrile neutropenic episodes. The second objective was to examine whether rhG-CSF administration allows closer adherence to planned dosing schedule and to determine the overall response to chemotherapy. HR-ALL pts are randomized to receive either 9 cycles of chemotherapy (HRG I) or 9 cycles of chemotherapy (day 1-6) followed by rhG-CSF (day 7-20; HRG li). Up to date, 20 pts have been treated according to this protocol (HRG I: 10 pts, HRG I1:10 pts). In HRG II, rhG-CSF is well tolerated without G-CSF related adverse events. In each arm, one pt relapsed. The incidence of neutropenia was 37% in HRG I and 10% in HRG I1. More importantly the incidence of febrile neutropenia was 14% in HRG I and 5% in HRG II. These data demonstrate that rhG-CSF allows for reduction of the incidence of febrile neutropenia in HR-ALL-patients. The patient has experienced complete resolution of stomatitis, fever and malaise. The administration of G-CSF in patient with idiopathic neutropenia significantly increased the absolute ueutrophil counts (p < 0.001). G-CSF was effective in reducing the severity of neutropenia and infectious complications in our patient. Hansen F., Stenbygaard L. and Skovsgaard T. Twenty patients with recurrent metastatic breast cancer treated with high-dose myelosuppressive antJr3eoplastic drugs (Cyclophosphamide 2,5 g/m 2 or Epirubicin 130 mg/m 2 both q 3. weeks) as first or second line chemotherapy were randomized in a prostective study to GM-CSF (n=11) 5 microg/ kg/dag for ten days after cessation of chemotherapy or control (n=9). Compared to the control-group highly significant reduction in granulocyte nadir duration (two days (0-5) with GM-CSF vs. seven (2-11) days) and severity (WBC 0.4 x 109/I with GM-CSF vs. 0.2 x I09/I) was found. No difference in frequency of neutropen fever or antibiotic use could be observed. Even though the patients treated with GM-CSF at random were more heavily pretreated with chemotherapy, there was a surprisingly higher responserate in these patients as compared to the control-group, namely 64% vs. 22%, resp. No severe side effects were seen, but presumably due to GM-CSF one patient developed an allergic type I reaction and one patient developed a possible pericardia[ exudation. Both were fully reversible after cessation of GM-CSF treatment. Keywords: GM-SCF, chemotherapy, breast cancer. Twelve adult patients with chronic neutropenia, including 7 patients with idiopathic sporadic neutropenia, 2 with idiopathic familial neutropenia and 3 with cyclic neatropenia have been treated with rhu-met-G-CSF (Amgen, Thousand Oaks, USA). Treatment has been started in all patients with 3 #g/kg/d sc once daily. Doses have been modified according to WBC. All patients had a rapid increase of absolute neutrophil counts. Data are shown for idiopathic neutropenia (base line, after 1, 2 and 4 weeks). Doses required ranged from 0.1 to 9 #g/kg/d. Treatment has been continuously given up to three years in patients with severe preceeding infections. The clinical efficacy of the treatment was excellent with abrogation of significant infections. One patient with idiopathic i sporadic neutropenia recovered after 151 days of treatment with an ANC of "~ ~o0o >2000/#1 after stop of G-CSF. In a 1 patient with familial cyclic neutropenia cycle length shortened from 21 to 14 days. In another patient with acquired idiopathic cyclic neutropenia the cycle length of 120 days remained constant but re t w 2 w 4 w the nadir of ANC rose from 0 to 400/#1. This patient was taken off therapy because of urticaria related to G-CSF on day 850. There were no further significant adverse events. No loss of effectivity was observed during long-term treatment. We conclude that G-CSF is safe for long-term treatment of idiopathic neutropenie with severe preceeding infections. As response to treatment is quic, it may also be an effective interventional treatment in acute infections in these patients. Cytokines and growth factors are widely used to promote growth and proliferation ofbematologic cell populations. Improvement or wonnu healing by stimulation with G-CSF has been relxn'ted in patients suffering from Kostmann syndrome, Felty syndrome or from neutro-~ nia due to chemotherapy. e report on two patients with MDS/RA (HA, female, age 76; SH, male age 82); duration of disease was 3 months and 5 years, respectively. HA was admitted for neutropenia (neuU'ophils: 0.45-0.9 G/l), epistaxis and a growing ulcerous wound in the pubic area (diameter 60 mm) already pretreated with antibiotics for 10 days. Surgery was not possible due to poor heart condition and thrombocytopenia refractory to donor platelets. 30 MU G-CSF were administered subcutaneously daily for 5 days resulting in neutrophil counts of 3.15 G/1 and effective wound granulation and epithelialisation. The patient died of cardiac failure on day 10. SH was adnutted for infected hematoma of the left thigh. Subcutaneous infection progressed due to severe neutropenia (neula-ophils < 0.5 G/l). Incision and resection (ulcus diameter 30 ram with deep invasion into the fascia) was performed. 20 days later the defect measured 120 x 50 mm, reaching the knee joint, and the patient underwent a second surgical interventaon. Enterococcus, staphylococcus epidermidis and bacterium xerosis could be cultured from direct swabs. Therapy with G-CSF at a dose of 30 MU s.c. was started on day 61. Neutrophils reached 19.5 G/I and G-CSF was reduced to every other day. Complete wound healing without any further surgical intervention was achieved by day 88 and SH was dismissed. After discontinuation of G-CSF the patient is well and has normal differential blood counts. We conclude that G-CSF is successful in promotion of wound healing in MDS patients due to enhancement of neutrophil production. Leukemic conversion of MDS during G-CSF was not observed. We report on a 44 year old male patient presenting in 5/88 with moderate thrombocytopenia~ transfusion dependent macrocytic anemia and normal., WBC. Trephine biopsy showed hyperceUular marrow without fibrosis wzm trilineage dysplasia (MDS/RA). Cytogenetic analysis was 46,xy. In 10/88 vasculilas ~as diagnosed. From 10/88 to 12/88 three cycles of GM-CSF (250 #g/m s.c. -14 days) were administered, resulting in both transient leukocytosis and increased platelet counts_ Bone marrow aspirations showed dysplasia but no blast proliferation. In 5/89 vasculitis progressed, splenomegaly and hemolytic anemia developed requiring prednisolon. In 3/90 high dose erythropoietin was started (400 IU/kg i.v. twice weekly) and continued till 6/90. There was no change clinically, bone marrow smears and cytogenetics. In 10/90 the patient complained of pain in the lumbosacral region and neuralgia in both legs developed; a CT scan was negative. Both pain and neurological symptoms (paraplegia and sensibility disorders) progressed. ACT scan and an MRT showed an intraspinal tumor (D2 -D11). "Although severe thrombocytopenia refractory to high dose i.v. immnnoglobulin and platelet support (HLA class I & lymphocytotoxic antibodies) developed, therapeutic laminectomy was performed in 6/91, but only a part of the tumor could be resected. Histologically the tumor consisted only of erythropoiesis with dysplasia without excess of blasts. Wound healing was without complicatmns. After surgery gamma irradiation and therapy with IFNc~ (6 #g s.c. 3 times/week) were performed. The patient recovered totally from neurological disorders and is still alive but Iransfusion dependent because of severe cytopenia. We conclude: intraspinal extramedullary hematopeiesis is a rare symptom in MDS. Althoughthis infiltration was diagnosed months after GM-CSF a_ad HD-EPO therapy, it Could be induced by cytokine therapy. ( monoclonal antlbodles elther' with or without GaM crosslinking. In addition, we added the cytokines. The phenotypic change of expression of Fcy receptors was measured. ~O 2 production and calcium flux using the dihydrorhodamine (DHR) and Fluo-3 AM methods, resepectively. There were no changes in expression of Fcy receptors, but a significant enhancement of PMN activation via Fc receptors by all three cytokines. We observed an increase of H202 production 4.5 fold by G-CSF, 2 fold by GM-CSF and 3.5 fold by IL-8. A FcyRIII-B specific monoclonal antibody fgr PMN (ID3), which was alone unable to mobilize Ca I+, together with all three cytokines was a potent stimulator. The effects of GM-CSF and G-CSF were calcium independent, in contrast, IL-8 also enhanced calcium mobilization significantly. In summary, all three cytokines potentiate the Fcy receptor activation of PMNs and therefore play a significant role in inflammatory granulocyte activation as in leukocytoclastic vasculitis. G-CSF is a hematopoetic growth factor required for proliferation and differentiation of hematopoeitic progenitor cells. It is now being successfully used to overcome neutropenias of various etiologies. Recently, we demonstrated that rhG-CSF induced neutrophils from patients with severe congenital neutropenia showed altered surface marker expression (upregulation of FcyRI (CD64} and CD14 and downregulation of FcvRIIlICD16)) as well as decreased chemotaxis towards a variety of chemoattractants including fMLP. To separate the effects of the underlying disease from those of the rhG-CSF therapy, we investigated neutrophils from patients receiving cytotoxic chemotherapy (n = 6) and healthy adults (n = 5) after application of rhG-CSF. Results: Neutrophils from 6 patients receiving daily application of rhG-CSF (Neupogen ~ 300#g sc.) were studied ex vivo one day before, three times during and 3-5 days after cessation of rhG-CSF treatment. Expression of FcvRI, CD14 and CD54 (measured by flow cytometry) increased during therapy reaching a maximum at 3-5 days after initiation of rhG-CSF therapy, whereas expression of FcvRll! decreased to a minimum after 6-9 days. Chemotaxis of neutrophils under agarose towards fMLP was also reduced during therapy. Investigation of surface marker expression and chemotaxis 3-5 days after cessation of rhG-CSF revealed return to levels before therapy. To exclude the possibility that the observed alterations were caused by the underlying disease or chemotherapy, five healthy adults were treated with a single dose of rhG-CSF (Neupogen', 300pg, sc.). A continuous upregulation of FcvRI and CD 14 starting 3h after application with a maximum after 48 hours (Fc~l) and 18 hours (CD14) and a downmodulation of FcrRIII reaching a minimum at 48 hours was observed. Chemotaxis towards fMLP decreased 3 to 4 h after application and returned to normal after 6 h, whereas expression of FcrRI, CD14 and F%RIII showed baseline values after 96 hours. Conclusions: The results obtained from the healthy test subjects clearly demonstrate that neither the malignant disease nor chemotherapy, but rhG-CSF induced the profound alterations of Fcr receptor and CD14 expression and chemotaxis in neutrophils in vivo. The continuous character of the surface marker alterations without appearance of subpopulations and the increase in CD14 expression suggests that preactivation rather than immaturity of rhG-CSF induced neutrophils alone might be responsible for the observed phenomena. Fraunhofer Institute ITA, Nikolai-Fuchs-Sral3e 1, W-3000 Hannover 61 Severe congenital neutropenia (SCN) is a disorder of myelopoiesis characterized by a maturation arrest on the level of promyelocytes with absolute neutrophil counts below 200/pt in the peril~heral blood. In this study we investigated the expression of receptors for the granulocyte colony-stimulating factor (G-CSF) on neutrophils from patients with SCN during G-CSF therapy. The normal G-CSF receptor expression on neutrophils is in the range of 480-1200 receptors per cell. Neutrophils from SCN patients express increased numbers of receptors in the range of 2100-3900 receptors per cell. The dissociation constant of the binding of G-CSF to the G-CSF receptor is not altered as compared to healthy donors. In contrast neutrophils from patients suffering from cyclic neutropenia express normal G-CSF receptor numbers (400-900 receptors per ceil). In addition, we have compared the G-CSF receptor cDNA of neutrophils from healthy donors and SCN patients using the polymerase-chain-reaction technique. We could not detect any major alterations in the G-CSF receptor cDNA in SCN patients. Preliminary cDNA sequencing data also did not reveal any point mutation. From this data we conclude that there is no defect in G-CSF receptor expression and no alteration in the sequence of the G-CSF receptor mRNA in SCN. Pediatric Hematology and Konstanty-Gutschow-Str. 8, D-3000 Hannover 61 Oncology, Medical School Hannover, Severe congenital neutropenia is a disorder of myelopoiesis characterized by severe neutropenia secondary to either a maturational arrest of myelopoiesis at the level of promyelocytes (Kostmenn's-Syndrome; SCN) or regular cyclic fluctuations in the number of blood neutrophils with a median ANC below 500/1~1 (cyclic neutropenia). We have treated 32 patients with SCN and 4 patients with cyclic neutropenia. Thirty of 32 patients with SCN and all 4 patients with cyclic neutropenia responded to rhG-CSF treatment with an increase of the median ANC to above 1000/ixl. The doses needed to achieve and maintain the response varied between 0.8 and 120 i~g/kg/d. Long-term treatment did not exhaust the myelopoiesis: The mean ANC remained stable up to 5 years of treatment. The increase in ANC was associated with dramatic clinical responses: significant reduction of severe bacterial infections, reduction of intravenous antibiotic treatment episodes, and reduction of hospitalizations. No severe bacterial infections occured in any of the rhG-CSF responders during long-term treatment. Severe adverse events, most likely associated with the underlying disease, included the development of MDS/Leukemia in two patients, and osteopenia/osteoporosis in 12 patients. These results demonstrate the benefical effects of rhG-CSF treatment in severe congenital neutropenia patients. Fifty-two patients (pts) (median age -51 years) with Philadelphia chromosome positive (Ph 1+) chronic myeloid leukemia (CML) have been treated with IFN (5 x 106 units/m 2) within six months of diagnosis (median 1.5 months (mths), range 1 -6 mths). We divided the pts into three groups according to Sokars classification: low risk group (n = 24), intermediate risk group (n = 19) and high risk group (n = 9). Forty-three pts acheived a complete hematological response (CHR) as defined by the Houston criteria. The cytological response was evaluable in 40 pts: 23 pts (57.5 %) demonstrated a partial or major eytogenetical response (more than 65 % Ph I negative metaphases). The hematological and cytogenetical responses were influenced bsr the risk factors, os the percentage of CHR and cytogenetical responses was higher for the pts from the low or intermediate risk groups (88 % and 41% respectively) than for the high risk group (55% and 22 %). Transformation occurred in four pts who did not demonstrate a cytogenetical response. The estimated chance of surviving at three years was 83% for the overall population. Toxicity was mild but IFN had to be internpted in four pts for cardiac (n = 2), liver (n = t) or neurological (n = 1) tocxic effects. These results confirm that IFN is a very effective treatment for CML. The effect of rhG-CSF on platelets was studied in 20 healthy volunteers with the thrombometer, a specially developed device which is described in detail. Additionally, conventional aggregation tests were performed Low doses of rhG-CSF enhance functional platelet activity, as shown by significant acceleration of the occlusion of the thrombometer channel. Similar results were found in conventional aggregation tests using collagen for induction. With G-CSF concentrations of 0,1 and 1,0 ng/ml the time of response was significantly accelerated and the maximum response was observed in a higher proportion ofplatelets. However, the second phase of aggregation induced by epinephrine was significantly inhibited by 1,0 ng/ml G-CSF. The expression of CD41, CD42 and CD 62 on platelets' surface was determined in ten patients before and niter administration of G-CSF (FACscan flow cytometer).Quality controls were done by calculating the events positive for CD41 and CD42, which were expressed in nearly t00 % of the platelets without being changed by the cytokine. The expression of CD62 in the platelets' surface however was significantly enhanced indicating a depletion of the c~-granules. No platelet aggregation was observed. CD62 expressed on thrombocytes' or damaged endothel cells' membrane is a receptor for macrophages. This property facilitates rapid adhesion of leucocytes to endothelium at regions of tissue injury as well as platetetleucocyte interactions at areas of inflammation and hemorrhage.-In contrast CD62 can also have an antiinflammatory function because exposure of TNFa-activated neutrophils to plasmatic CD62 inhibits their CDl8-dependent adhesion to resting endothelium and superoxide production. IFN a has a unique activity in CML leading to complete and partial remissions in 15-30% of the patients. To improve these results, we are currently treating patients with Ph'+ CML with a combination of cytosine-arabiuoside at a maximum dose of 20 mg/m z SC on 5 days per week and IFN c~-2b. IFN u-2b is started at a dose of 3 MUIm 2 SC dally and escalated to the maximum tolerated dose. 48 patients (25 male, 23 female, median age 44 years) have been entered into the trial. 15 patients have been pretreated with other regimen for a median time of 32 months. 33 patients are without pretreatment. The treatment has been well tolerated. Besides the IFN a related side-effects some patients experienced gastrointestinal toxicity with nausea and vomiting after prolonged Ara-C application. The median observation time in the study is now 8 months and patients are still entered. Up to now complete hematologic remissions have been achieved in 24 patients, and partial ones in 15 patients. The rate of complete hematologic remissions was higher in patients without pretreatment (55 %) compared to patients who have been pretreated (40%). Five partial eytogenetic remissions have been observed and 3 minor reductions in the Ph'+ cell done. All cytogenetic responses have been found in patients without pretreatment. We conclude that a combination of cytosine-arabinoside and IFN u-2b is well tolerated in patients with CML. Early results are encouraging. Longer follow up times are necessary to evaluate whether combination therpy will give superior results compared to a txeatment with IFN a alone. During rlFN-~2 therapy a minority of patients develops high-titered antibodies neutralizing the injected rlFN-a2. The rlFN-c~-andbodies from six of such patients, who lost their clinical response to rlFN-c~2 and showed a relapse of their leukemia (3 CML, 3 HCL) despite continuous rlFN-c~2a-therapy, as well as IFN-c~-specific antibodies from two patients with systemic lupus erythematusus (SLE) were characterized. The anti-IFN activity was purified by sequential Protein G -and rIFN-c~2 affinity chromatography and was found to consist only of IgG-antibodies. These antibodies were further tested for their capacity to neutralize the antiviral and antiproliferative activity of various rIFNs-c~-subtypes. All six sera tested showed a common pattern of neutralization (MDBK-VSV bioassay) distinct from the SLEantibodies. All six neutralized rIFN-c~A and rIFN-aK consistently with a higher titer against aA. Three of the six sera neutralized aA, aK, ~C, ~C/J1 and M, but not M, M1 and some other subtypes. Therefore, from the structure of the C/Jl-hybrid, it seems that one epitope recognized by these three sera is at the NH2-terminal half of the molecule. In contrast, the SLE associated antibed]es neutralized the antiproliferative and antiviral activity of every subtype tested. These data indicate that the therapy-induced antibodies against rIFN-c~2 recognize very selected epitopes on the rIFN-cd-molecule suggesting that only a part of the rIFN-~2 molecule is immunogenic. In vitro experiments indicate higly synergistic effects of combining IFN with cytostatic drugs such as anthracylines (A). In a phase I/II-study 31 patients (pts) with pro$ressive inoperable HCC were treated with E 20mg/m z weekly x 4 and IFN 3 Mio IU/m 2 s.c. 5 x weekly for 4 weeks, followed by one week off treatment. In case of at least no change (NC) and tolerable toxicity the therapy was continued. Escalation of E in steps of 5mg/m z per cycle was attempted. PTS characteristics: Median age 56 years (21 -69); male 20, female 11 PTS. Pretreated with A 3 pts. Toxicity and treatment: Total number of cycles 92; median 3 (1-9) per PT. Worst toxicity per PT (WHO): WBC ~ 38%, ~ 6%; platelets ~ 9%, ~ 0%; diarrhoea ~ 5%, ~ ~%; n~Isea/vumitinq ~ 5%; ~ 0%; IFN related fever (maximal ~ in 53%. IFN-related wasting syndrome 2 PTS, no severe organ toxicity. Divisione di Ematologia -Ospedale San Camillo -Roma From November 1987 to October 1989 six patients with acute leukaemia, who achieved their first complete remission with standard chemotherapy followed by autologous bone marrow transplantation (ABMT), were consecutively treated with r.interferon alfa-2a (rA-IFN). Patients (4 ALL and 2 ANLL) were from ii to 44 years old, 4 of them (2 ANLL and 2 ALL) were reinfused with autologous bone marrow purged with ASTA-Z i00 mcg/ml/ 2 x 107 cells and in the remaining 2 ALL patients immunomagnetic purging was employed. Conditioning regimens were BUCY in 4 patients and CY-TBI in the others, rA-IFN started at median time of 6 months (4-11) after ABMT when complete consolitated hemopoietic recovery occurred. The rA-IFN dose was 500.000 IU/sqm 3 times a week for 2 years. None of the 6 patients presented significant toxicity and only 3 short suspensions occurred for fever or ALT level increase. The incidence of infectious complications were particularly low compared with other autotransplanted groups of patients who received similar antinfectious prophilaxis. One case six months after ABMT and 20 days after acyclovir prophilaxis interruption presented a mild herpes-zoster complication which required new acyclovir therapy and resolved i0 days later. The amount of these patients is extremely low because the study was early interrupted to start a new protocol including IL-2; but the long duration of the good continous complete remission (3/5 years after ABMT) in all these unselected and consecutively treated patients is very interesting. Surgical procedures may be associated with an inceased risk of tumor spreading due to surgical mobilisation of the tumor and transient postoperative immunosuppression. Recurrences may result either from early growth of micrometastases already present at the time of surgery or from the seeding of malignant cells shed during operative manipulation of the tumor. Imrnunomodulators have been proposed to correct the immunological impairement induced by surgical procedures. From 1/92 to 7/92, 23 patients with advanced stage cancer underwent surgical resection with peri-operative interferon-alpha administration. Patients received interferon alpha-2a (Roferon-A), by daily subcutaneous injection for fourteen days, starting on three days before surgery. Incremental doses were 2,3,5,7,9 and 12 x 10 6 IU for 4,4,3,3,3 and 6 patients respectively. Peripheral blood lymphocyte (Pbl) subset numbers were assessed using flow cytometric analysis the day before injection (D-3), before surgery (D-I), at the day 4 and 12. Absolute numbers of total T cell (CD 3+) and NK cell (CD 56+, CD 3-) were determined, as well as auxiliary T cell (CD 4+), activated T cell (CD 3+, Dr+), and B cell (CD 19+) counts. Short-term cytotoxicity of PBMC against K 562 and DAUDI target celts in a 4-hour standard chromium release assay were determined. No W.H.O. grade IV toxicity were observed. A significant post-operative fall of the total Pbl count, of CD3+; CD4+; CD19+; CD3-56+ occured from D-3 to D4. The decrease were not significant for CD3+Dr+. Values were not significantly different, between D-3 and D12, only for CD3Dr+ and CD3-56+. Cytotoxicity against K 562 and Daudi target ceils increased significantly from D-3 to D-l, and from D-I to D1, with a significant fall of cytotoxicity against Daudi from D-3 to D4. Peri-operative interferon alpha administration is well tolerated even at the 12.10 6 IU doses. In spite of treatment and increasing cytotoxicity activity, we observed a post-operative fail for the majority of the imunological parameters. Further studies are necessary to compared with a control group, with more patients treated with 12.10 6TIIU daily. Patients: Six patients have been treated (med. age 29 yrs.; 24-53 yrs.) Four patients had acute myeloid leukemia (AML M2: 2, M4: 1, M4Eo: 1). Two patients had acute lymphocytic leukemia (both T-ALL). All patients had manifest disease with more than 50% blasts in the bone marrow. Patients were selected not to have rapidly progressive diesease allowing the application of the cytokine. Treatment and toxicities: IFN a was given for a median of 31 (14-114) days at a dose of 5-10 MU SC daily. The following toxicities _>grade 3 WHO were observed: fever 5, GPT 2, pulmonal 1, infection 4 (1 pneumonia), pains 1. The patients with AML ware transsusion dependent for platelets and erythrocytes. No significant additional bleeding was observed. Results: One AML patients had stable disease and three had disease progression. Of the ALL patients one had disease stabilization and one progression. Conclusions: IFN a is well tolerated in patients with refractory AML and ALL if they ere in a relatively stable condition. The effectivity of IFN a-2b as a single agent is poor in patients with refractory AML or ALL. Freund et al. Eds, Springer-Verlag, 1992) have confn'med the wide range of clinical usefulness of IFN alphabased therapy in cancer patients (pts). We present in this paper a retrospective analyses of 102 cases with advanced neoplasias treated between 1986-1992 by a sequential administration of IFN alpha and standard chemotherapy. There were 71 W, 31 M, aged 18-76 y, with solid cancers 99 pats and lymphomas 3 pts. The treatment consisted of a sequential association of IFN alpha (also with RA all-trans) and CHT plus Tamoxifen (for appropriate cancers). The IFN schedule was of monthly series, one series consisting of 3 million IU/d for 3 consecutive days. CHT, appropriate for each primary cancer also administered in monthly series. The results are grouped according to the status of the disease (subsets of pts) at the onset of IFN-based therapy, and refer especially to long term survival (1-5 y +). For minimal residual disease (MRD) from 39 cases there were 37 CR (24/24 breast ca, 6/6 CR melanomas, each STS and RCC 2/2 RC, 2/4 gastric and head and neck cancers together, 1/t lymphoma. For progressive disease, pre-lFN-based therapy there were 63 pts, and post IFN-CHT treatment there were 58 SD and 5 PD. In 3 of failure pts association of IFN alpha and bropirimine (IFN-inducer) appeared an unusual good response. Conclusions: 1) IFN alpha-based therapy is a very useful one especially in MRD. 2) The therapy must be individualized for selected subsets of pts and for each patient day to day. Authors have used INF alpha 2b in cases of haematological malignancies for three years. The number of cases not too high, and non fo CML cases was so called "early" CML. Our cases are: 5CML, 3Non-Hodgkin Malignant Lymphoma, 4 Myeloma multiplex, 3 Essential thrombocythemia. The TNF alpha 2b was used as monotherapy in the cases of low grade Non-Hodgkin Malignant Lymphoma, and Essential thrombocythemia, and was combined with chemotherapy in the other cases. On the basis of our initial results, we recommend the INF alpha 2b in the treatment of haematological MALIGNANCIES in suitably selected cases. Department of Haematology and Oncology, Hospital of Ministry of Interior, Vftrosligeti fasor 9-11 The influence of low oral doses of human leukocyte derived interferon alpha on the immune system of chronically HBV infected patients with depressed immunological response. Low oral doses &the interferon were given to a group &seven children with limphoblastic acute leucaemia in the state of remission, chronically HBV infected. Interferon alpha was produced by Hayashibara Biochemical Laboratories Inc. Okayama, Japan, in tablets of 50 IU and 100 IU respectively. Immunological response was checked by measuring population and subpopulations of limphocytes, level ofimmunoglobulin and complement C 3 fraction. A distinct stimulation of cellular immune response was observed: the fraction of activated limphocytes T increased significantly, index CD4/CD 8 became normal and population of limphocytes B increased gradually. There was no influence of interferon treatment on immunoglobulin and complement C 3 fraction serum level. The interferon treatment improved the patients' general condition and shortened the period of intoxication after cytostatic treatment. No side effects were observed. None of the children eliminated the HB virus during the six months treatment. Cytokines play an important role in activating the immune system against malignant cells. One of these cytokines, IL-4 has entered clinical phase I trials because of its immunoregulatory potency. In the present study we report that rhlL-4 has direct antiproliferative effects on some human lung cancer cell lines in vitro as measured by a human tumor cloning assay (HTCA). This activity could be abolished by neutralizing antibody against rhlL-4. The biological response of the tumor cells to the cytokine is correlated with expression of receptors for blL-4 on both the mRNA level and the protein level. The most responsive cell line CCL 185 secretes IL-6 after being incubated with rhlL-4. On the other hand, neutralizing antibodies against IL-6 showed no influence on the growth modulatory efficacy of rhIL-4 in this cell line. Furthermore, CCL 185 does not show detectable production ot IL-1, TNF-~ or IFN-y alter incubation with rhlL-4. Thus, the response to rhlL-4 is not mediated through autoeriee production of these cytokines triggered by rhlL-4. In a next series of experiments the ceillines were xenotransplanted to BALB/c nu/nu mice. Subsequently, the mice were treated for >12 days with twice 0,5 mg/m2 rhlL-4 (rhlL-4 was a kind gift from Dr. Urdal, Immunex, Seattle, USA) or control vehicle subcutaneously per day. Treatment with rhlL-4 yielded a significant inhibition of tumor growth versus control in the responsive cell lines CCL 185 and HTB 56, but no therapeutic effect in the non-responsive cell lines. Plasma levels of rhtL-4 were sufficient for in vitro growth-inhibition in the responsive lines. Histology of the tumors in both groups showed no marked infiltration of the tumors with murine hematopoietic and lymphocytic cells consistent with the species specificity of IL-4. We conclude that rhlL-4 has direct antiproliferative effects on the growth of some human lung tumor cell lines in vitro and in vivo which together with its regulatory effects on various effector cell populations makes this cytokine an interesting candidate for further investigation in experimental cancer treatment. The combination of systemic chemotherapy and immunotherapy comprising interleukin-2 and alpha-interferon leads to significant tumor regressions in patients with advanced malignant melanoma. In contrast to chemotherapy by itself, the combination produces a significantly extended remission duration in the majority of treatment responders. We conducted 2 phase II studies.to assess the potentially additive or synergistic effects of chemotherapy and immunotherapy in metastatic malignant melanoma patients: The first study comprised two cycles of carboplatin (400mg/m 2) and dacarbazine (750mg/m~); the sesond study included up to four cycles of cisplatin (25mg/m 2 x3 days), dacarbazine (220mg/m 2 x3 days), BCNU (150mg/m 2, cycle I+3) and tamoxifen (20mg daily). Chemotherapy was followed by up to 2 cycles of a 6-week immunotherapy comprising interleukin-2 (5-20 million IU/m ~ sc 3x weekly) and alpha-interferon (3-6 million U/m 2 sc 3x weekly). Among 25 evaluable patients in study I, there were 9 (12%CR, 24%PR) objective responders; median remission duration was 18+ months for complete, and 10+ months for partial responders. Chemotherapy intensification in study II lead to an increased response rate of 50% (9 out of 18 patients). In both studies, the progression free interval was significantly extended when compared to patients who received chemotherapy, only (historic controls). The role of immunotherapy as maintenance in patients with advanced metastatic malignant melanoma is currently being evaluated in a prospective randomized trial. Integrin receptors play a crucial role in cell-cell and cell-matrix adhesive function, and thus are supposed to influence invasion and metastasis. Very little is known about the impact of interleukins on integrin regulation in tumor cell lines. Therefore, we investigated the expression of 7of and 41I integrin subunits on well (HT29) and poorly differentiated (SW620) human colon cancer cell lines using a panel of specific monoclonal antibodies and cDNA probes. HT29 and SW620 expressed similarly high levels of ~1, a2, ~ gt, and f~4 subunits on the cell surface. No a4, ~2, and 1~ 3 was detected on either cell line. While a 5 was not expressed on HT29, SW620 showed higher levels of the laminin receptor ~694. The poorly differentiated cell line SW620 was resistant to IL-4, whereas HT29 was sensitive. Treatment with IL-4 induced a decrease in ~2, a3, '~6, ~v, l~l, and 134 integrin expression. However, ~1 subunit was markedly upregulated. In contrast to IL-4, there was no evidence that IL-lfI could modulate integrin expression on these celt lines. The function of integrin receptors was assessed by measuring adhesion to collagen, laminin, vitronectin, and fibronectin. IL-4 significantly increased the adhesion of HT29 to fibronectin, while attachment to collagen, laminin, and vitronectin remained unchanged. These results suggest differential integrin expression pattern on well and poorly differentiated tumor cell lines. We provide evidence that integrin expression may be selectively regulated by IL-4, but not by IL-IB. Furthermore, IL-4 can alter adhesive behavior of tumor cells. Since IL-4 is currently studied in clinical trials, the metastatic potential of malignant tumors should be monitored thoroughly. Immunotherapy with IFNcz and IL-2 is an active regimen in malignant melanoma and has shown response rates of 20 -30%. In previous studies no prognostic parameters for response could be identified. 72 patients with progressive metastatic melanoma have been enrolled in various immunotherapy trials including IFNo~ and high dose IL-2 since 1987 with an overall response rate of 30%. 59 patients with MM treated in our phase II trials could be analysed to identify possible prognostic parameters for response. Patients were divided into three groups: responder (3 CR/14 PR), stable disease (16 SD/2MR), and nonresponder (24 PD). All patients had measurable tumor, a Karnofsky index of > 70%, no CNS metastasis, and no severe cardiorespiratory or renal disease. We examined the following pretreatment parameters for prognostic relevance of response: age, sex, performance status, time from diagnosis to onset of first metastases/ to begin of immunotherapy, tumor toad, number of metastatic sites, organ sites of metastases, LDH, AP, ESR, and HLA-type. Of these several variables were found to significantly correlate with response: tumor load (p=0.023), number of metastatic sites (p=0.045), serum LDH (p=0.005) and AP (p=0.013). Tumor load, LDH and AP are no independent parameters. While time from diagnosis to onset of first metastasis is of no prognostic significance for response, the time between first diagnosis and begin of immunotherapy, usually reflecting metastatic disease necessetating systemic treatment, significantly correlates with probability of response (p=0.018). Since several HLA class I alleles have been shown to function as restriction elements for recognition of melanoma cells by specififc T cells in vitro, namely A1, A2, B44, and Cw7, we compared the frequency of these HLA antigens between responder and non-responder. We found A1, B44 and Cw7 to be increased in responder vs. non-responder. Our results indicate that in patients with MM tumor load, number of metastatic sites, LDH, and time from diagnosis to begin of immunotherapy are prognostic parameters for response to immunotherapy. These parameters may be useful to determine patients with good and poor risk for response to immunotherapy and are of relevance for stratification in randomized clinical trials. Dept of Medicine, University of Heidelberg, Hospitalstr. 3, 6900 Heidelberg, Germany Surgery of metastatic melanoma following successful IL-2 based immunotherapy. U Keilholz 1 , E Stoelben 2, C Scheibenbogen 1 , HD Saeger 2, K Neumann 3, W Hunstein 1 Surgery of advanced metastatic melanoma is of limited value and usually not recommended. Immunotherapy using high dose IL-2 is effective in a substantial proportion of patients, however, the duration of responses is limited, and benefits in survival are not yet proven. This evaluation was done to determine the value ot resection of residual tumor lesions following successful immunotherapy. 72 patients with progressive metastatic melanoma have been enrolled in various immunotherapy trials including IFNa and high dose IL-2 since 1987. 37 patients showed evidence of antitumor response (4 CR, 15 PR, 18 MR/SD). In patients responding to immunotherapy, residual lesions were resected, whenever technically possible and patients agreed to surgery (13 patients). 22 of the 24 responding patients without surgery relapsed, the median time to progression was 5 months (range 2-14), almost all initial relapses occured locally, 8 patients died so far. 11 of 13 patients who underwent surgery (9 PR, 4 MR/SD) were converted into CR by surgery. 4 of the 11 patients disease-free after surgery relapsed, 3 locally (2, 5, and 9 months after surgery), and one CNS 9 months after surgery. 7 patients are still free of recurrence (3+,3+,6+,12+,16+,18+,28+ months after surgery) and 9 of 11 are still alive. In the patient with CNS relapse complete resection of this lesion was again possible, and there was no evidence of recurrence for 9 months after this second surgery. Histology revealed vital tumor cells in almost all resected specimens, however in 11 of 13 patients profound necrosis of the tumor tissue was observed. Of special importance is the observation that 3 patients with minor response or SD according to imaging procedures were found to have an almost complete response histologically. Interestingly, almost all metastases resected after immunotherapy had developed a fibrous capsule. Surgical reevaluation and resection of residual lesions should be considered in patients with partial response after immunotherapy, and in selected cases also with stable disease. This approach offers the chance for extended disease free survival, and may be curative in certain patients. T-cell-receptor (TCR) Vcz~ usage of tumorinfiitrating T-cells in primary, regressing and progressing melanoma metastases following [mmunotherapy with IFN~ and IL-2: evidence for a specific T-cell response. MShler, T., Willhauck, M., Scheibenbogen, C., Pawlita, M.#, Bludau, H.#, Brossart. P.. Keilholz. U. The identification and characterization of immunological effector cells mediating tumor regression in immunotherapy with IL2 is of great interest for understanding and further development of this therapeutic approach. Tumor infiltrating lymphocytes specific for autologous tumor cells can be expanded from certain melanoma tissues. T cells recognizing the same antigen use a limited TCR repertoire with a certain Vc~ and [3 variable region, determining the specifity of their receptor. We therefore analyzed T-cell receptor V-Regio,q distribution in tumor tissue from melanoma patients prior to and following immunotherapy with IL-2. We used a highly sensitive RNA-PCR method. After RNA-extraction from tissue and subsequent cDNA-synthesis semiquantitative PCR with different primers for all known Vc~-and Vii-T-cell receptor gene families (18 Vc~ and 20 V~) was pedormed. 12 tumor tissue samples were analysed including 6 samples of primary malignant melanoma and tumor samples of three patients after immunotherapy. The results were compared to control tissues (peripheral blood, unatfected skin, and liver tissue). The analysis of primary malignant melanoma tissue showed a weak overexpression of different V[3-families. Preferential usage of different TCR-V~,genes was more obvious in tumor tissue of patients alter immunotherapy. Of special interest is a patient with a mixed response to immunotherapy with progressing and regressing skin metastases, in the regressing lesion we could demonstrate a predominant usage of TCR-V[~11-Gene almost lacking in the progressing lesion. This suggests a role of V[311-expressing T-cells in mediating tumor regression in this patient. Cloning and sequenzing analysis are currently performed to assess wether this represents a true clonal T-cell proliferation. Recently a highly sensitive assay combining reverse transcription and polymerase chain reaction (RT/PCR) to assess for melanoma cells in peripheral blood has been developed. The detection of tyrosinase mRNA, a tissue spezific enzyme in melanocytes and melanoma cells in peripheral blood indicates the presence of melanoma cells, We used RT/PCR assay to determine malignant melanoma ceils in peripheral blood of 43 patients with malignant melanoma in different stages of disease. in none of 8 patients with stage I (localised tumor) but in 5 of 14 patients in stage II (regional lymph node metastases) tyrosinase transcripts were detected. Tyrosinase mRNA was found in all 21 patients with distant metastases (stage III). This method may be helpful to define a group of patients at high risk for development of hematogenous metastases, that would be a possible target group to explore adjuvant treatment strategies. We then examined blood samples and bone marrow aspirates of 28 patients with metastatic malignant melanoma for presence of melanoma ceils prior to and after therapy with IFN-a and IL-2. 10 patients showed antitumor response to immunotherapy: 3 complete remissions (CR) and ? partial remissions (PR R. Hilse, M.Meffert, J.Grosse, H.Kirchner, H.Poliwoda, and J.Atzpodien We investigated the use of PCR for a semiquantitative estimation of cytokine expression patterns in PBMC before and after administration of IL-2 to patients with advanced renal cell carcinoma or malignant melanoma, mRNA of 9 cytokines was measured using a modified polymerase chain reaction protocol, which could detect 10-fold differences in mRNA-contents of stimulated PBMC in vitro. Weekly RNA-samples of 7 patients receiving a total of 11 treatment cycles were examined for long term changes, in 2 patients frequent samples were taken immediately after tL-2-administration for transcript-kinetics, mRNAexpression for IL-4, IL-5, IL-6, IFN-)', TNF-c~, GM-CSF, TGF-~ and IL-2receptor-c~ was clearly detectable in most of the samples, including four healthy donors. However, our method could not detect significant changes in transcript-levels of PBMC during 3 days following injection of (A) 36 Mio.lU or (B) 2x18 MioJU dL-2 daily. This was in marked contrast to cytokine secretion assayed by ELISA. Thus, serum IL-2 peaked 2-4 hours after administration followed by secondary cytokines with a peak 2-16 hours later. Increases for TNF-m IFN-7, IL-6 and IL-2R serum levels were significant (p<0,05) with the highest response found for IL-6, increasing 35-(A) and 32-fold (B) at day 1, or 8-/ 14-fold at day 2. Comparing normal individuals to patients, only small differences in constitutive cytokine expression were seen (<10-fold) with no distinct pattern. During therapy, changes could be seen for all cytokines except for IL-2 and TGF-I 3. In one patient, a 100-fold increase for IL-6, TNF-c~ and IFN-7 transcripts was observed during week 4 of the second treatmentcycle, other changes were approximately 10-fold. Abt. H&matologie und Onkologie, Medizinische Hochschule Hannover, D-3000 Hannover 61, Germany REGIONAL IMMUNOTHERAPY: PERFUSION OF LIVER METASTASES WITH LAK CELLS U. Keilholz, C. Scheibenbogen, M. Brado, W. Tilgen, and W. Hunstein A regional approach of adoptive immunotberapy with interleukin-2 and lymphokine activated killer cells for the treatment of liver metastases is reported. The treatment consists of continuous infusion of interleukin-2 i.v. or into the splenic artery, and transfer of ex vivo generated lymplaokine activated killer cells into the portal vein or the hepatic artery. 15 patients with malign ant melanoma, 2 with renal cell carcinoma, and 1 with thyroid carcinoma have been treated. All had progressive liver metastases. Trafficking studies using indium-oxine labelled cells revealed that > 80 % of the LAK cells remained in the liver after regional adoptive transfer. In 9 patients with liver metastases of cutaneous melanoma, 2 CR (24 and 23+ months), 1 PR (9+ months, converted to CR by surgery), 2 SD (10 and 11 months), ad 4 PD were observed. The lesion in the patient with PR was resectable after two cycles of treatment, and histology revealed almost completely necrotic tumor tissue surrounded by a dense fibrous capsule. No responses were observed in 6 patients with liver metastases of ocular melanoma, suggesting an immunologic difference between these two melanoma subtypes. 1 PR (6 months) and 1 SD (10 months) were achieved in 2 patients with renal cell carcinoma, and 1 SD (6 months) in the patient with thyroid carcinoma. Evidence for the crucial role regional cell transfer is provided by the observation in a patient with an anatomic variation of hepatic blood supply in whom we achieved complete and durable tumor regression. In this case anti-tumor responses were only observed in anatomic areas of the liver which were perfused with LAK cells. Depts. of Internal Medicine, Diagnostic Radiology, and Dermatology, Hospitalstrage 3, 6900 Heidelberg, Germany In a randomized phase II study we evaluated the response and side effects of a combined administration of Interleukin-2 (IL-2) and Interferon-alpha 2b (IFN-alpha 2b) versus Interferon-gamma (IFN-gamma) in patients with metastatic renal cell cancer. Patients in group A received subcutaneous (sc.) 200 meg IFN-gamma once a week. In group B patients were treated with a (sc.) combination therapy of IlL-2 (36 x 106 IU/m 2 in week 1 and 4, 18 x' 106 IU/m 2 in week 2,3, 5 and 6, twice a day for 5 days) and IFN-alpha 2b (5 x 106 1U/m 2 ) over 6 weeks once a day 3 times a week. Up to now 60 patients were treated, 30 patients in each group. Toxicity of IFN-gamma treatment was absent. The therapy with IL-2 and IFN-alpha 2b led to sideeffects grade 2 (WHO): Fever, chivering, fatigue and weight-loss. Treatment were withheld in 20 %, Follow up after 11 months (3-22 months) showed stable disease in 12 patients and progression inl 8 patients in group A In group B there were 4 complete remissions, 2 partial remission and 24 patients with progressive disease. Although the combination therapy showed 20 % objective response (p<0.05, Fisher-Test) no significant improvement on survival was seen (p = 0.342 logrank test). Patients with locally advanced renal carcinoma are at high risk of relapse after initial radical surgery. We initiated a clinical phase II trial using autologous tumor vaccines for the surgical adjuvant therapy of renal cancer patients. Seventy-two patients (pts) (25 female, 47 male; median age, 56 yrs; range, 28-77 yrs) with locally advanced renal carcinoma (pT3b-4pN0 or pTxNI-2M0) received autologous Newcastle Disease Virus modified and lethally irradiated tumor vaccines in combination with 1.8 million IU of IL-2 and 1.0 million U of IFN-~2, once weekly over 10 consecutive weeks. Toxicity was very mild with transient flu-like symptoms. Among 55 evaluable patients, there were 5 relapses (2pts, pT3aNI-2; 3pts, pT3bN0); the median relapse-free survival was 22+ months with a range from 6 to 41+ months; survival probability in this vaccine treated cohort was significantly better than in all historic controls. Using Western blot analyses, we could demonstrate a vaccine specific in-vivo B-cell response in all patients receiving NDV tumor vaccine. A subset of peripheral blood natural killer (NK) cells has been found to exhibit high density surface expression of the NK associated CD56 antigen; it has been suggested that these NK cells respond to lower concentrations of IL-2 when compared to the majority of NK cells expressing cell surface CD56 at low density. We evaluated density of the CD56 antigen on circulating NK cells of 47 patients with advanced renal cell carcinoma by flow cytometry. Patients received a combination of low-dose subcutaneous recombinant interleukin-2 (rIL-2) at 18 million IU/m2/day on days 1 and 2, followed by 3.6 million IU/m2/day, 5 days per week, over 6 consecutive weeks, in combination with recombinant ~-interferon (rIFN-~) at 5 million IU/m 2, three times weekly. Antigen density of CD56 before therapy was found 2.2-fold higher (p<0.005) in patients who subsequently achieved a complete or partial tumor remission (n=10) when compared with patients who presented with progressive disease on therapy (n=ll). After a 6-week treatment cycle, NK cells of treatment responders expressed significantly (2.l-fold; p<0.005) more CD56 antigens than NK cells in nonresponding patients. These results suggested a potential role of both pre-and posttreatment NK antigen density levels as a biologic correlate to treatment response in tumor patients receiving low-dose rIL-2 and rIFN-a. Intravesical immunotherapy against superficial bladder tumor recurrences and carcinoma in situ is a recognized and highly effective regimen in urology. To further clarify the mode of action of this approach, the local immune response of patients was investigated: The cytokines IL-1, IL-2, and TNF were determined in the urine before and after intravesical instillation by ELISAs and biological assays. Furthermore, bladder biopsies taken before and after the treatment course were analysed by means of immunohistology for the presence of mononuclear cell subsets. The results show a significant increase of urinary cytokines with a maximum 4-8 hours after the instillation of BCG which returned to baseline values within 24 hours. This intense locai immune activation was further reflected by the accumulation of activated mononuclear cells, predominated by T cells as demonstrated with bladder biopsies. The local T-helper/T-suppressor cell ratio shifted towards the T-helper subset. These changes persisted for more than 1 year after the initial treatment course. In conclusion, this local immune response may be associated with the therapeutic success of BCG. Further analyses will dissect the role of each factor with regard to antitumor cytotoxicity against bladder carcinoma. Enhancement of therapeutic effect of Intarleukin-2 (IL-2) by association with Cyclephcsphamide (Cy) was studied on EL-/+ lymphoma maintained in ascitic form in syngeneic C57BL/6 (H-2 b) mice and lymphoreticulosarcoma (spontaneous origin) maintained in solid form in syngeneic CBA (H-2 k) mice. Immunotherapy with IL-2 (obtained by in vitro stimulation of EL-4 lymphoma cells with phorbol myristate-acetate) was applied 24 hours after transplantation of EL-4 lymphcma and I0 days after transplantation of lym phoreticulosarcoma bE administration i.p. (intratumorel) in EL-4 bearing mice and s.c.(peritumoral) in lymphoreticulosarcoma bearing mice for three consecutive days. Cyclophosphamide was administered at a dose of 180 mg/kg i.p. six hours before the immune treatment with IL-2. The results obtained demonstrated that the prolongation of the survival rate expressed by the median survival time (MST) and the percentage of increasing llfe span (ILS) of the groups treated with IL-2 associated with Cy was significantly higher than that of the groups which receveid a single treatment with IL-2 or Cy. Enhancement of therapeutic effects of IL-2 in association with Cy on lymphoreticulosarcema was revealed by inhibition of tumor growth with a marked regression in the vol~tme of the established tumors and even resorption in some cases. We conclude that the antitumoral effect of IL-2 treatment was enhancemented by association with C2 a well known cytoreductive drug which selectively removed T-suppressor l~mphocytes from the tumor bearers. This could be conside red as an alternative to immune-or chemotherapy in cancer. To investigate the toxicity and clinical efficacy of aerosolized niL-2 (Biotest) 15 patients presenting with advanced malignancy were entered into a phase I trial. 13 patients suffered from metastasizing renal cell carcinoma, 2 patients from advanced bronchial carcinoma. At start the patients received either 50000, 150000 or 300000 U niL2 applied as a single dose. If no adverse events were observed, treatment was continued with the same dose 5 times daily for six weeks. In addition to standard invetigations detailed evaluation of the respiratory function was performed once weekly. Soluble interleukin 2 receptor serum levels and the effect on numbers and/or phenotype of lymphocytes in the bronchoalveolar lavage fluid were measured for assessment of biological response to inhalative niL-2 treatment. Treatment with aerosolized niL-2 was well tolerated. Most prominent toxicity appeared to be resistant cough in all patients treated with 5x300000 U/d. No febrile reactions or other constitutional side effects were observed. A dose-dependent increase of the numbers of memory T lymphocytes, macrophages and eosinophil granulocytes could be demonstrated in BAL fluid. In addition, the treatment resulted an increased expression of adhesion molecules on lymphocytes. 1 patient suffering from renal cell carcinoma achieved a partial remission after 6 weeks of treatment with 5x50000 U/d. We conclude that treatment with aerosolized niL-2 is biologically active and well tolerated and should be further tested in clinical phase II trials. Divisions of Hematology and Pulmology of the IIIrd Department of Internal Medicine, Medical Center of the Johannes Gutenberg University, W-6500 Mainz, Department of Urology, Univ. Hospital Eppendorf, Hamburg. Soluble interleukin-2 receptors (sIL-2R) exert a potential role in immunoregulation. We investigated the ex vivo effects of sIL-2R on several interleukin-2 (IL-2)-dependent activation events. Proliferation of the IL-2-dependent mouse cell line CTLL-2 and isolated human PBMC stimulated with recombinant IL-2 (rIL-2) was suppressed by sIL-2R added to the culture medium in a dose-dependent way. Preincubation of sIL-2R with rIL-2 did not enhance this suppression. Cytotoxicity of rIL-2-stimulated human PBMC against the human cell lines K562 and Daudi was correlated inversely to the concentration of sIL-2R in the culture medium during rIL-2 stimulation. sIL-2R concentrations higher than 4.0 pM produced a significant decrease in cytotoxicity (p<0.01). Light microscopy of IL-2-stimulated PBMC revealed no signs of cellular activation when high dosages of sIL-2R had been added. The effect of different sIL-2R concentrations added to cultured human PBMC on secondary IL-2 and sIL-2R production was tested by ELISA. Initial supply with high sIL-2R dosages yielded weak increase and subsequent slow reduction of IL-2 levels. In contrast, strong secondary IL-2 production followed by rapid clearance was observed when low sIL-2R concentrations had been added. Endogenous shedding of sIL-2R in response to rIL-2 was abrogated by the initial exogenous addition of high amounts of sIL-2R whereas low exogenous addition of sIL-2R was followed by a continuing endogenous production of sIL-2R after five days of culture. Our studies may lead to a better understanding of IL-2-related immunoregulation in the preclinical and clinical settings. We investigated the effect of interferons (IFN) on expression of IL-8 and other cytokines regulating inflammatory responses in various cellular models in vitro and in vivo. In peripheral blood mononuclear cells (PBMNC) of healthy individuals IL-8 gene expression which was upregulated in vitro was significantly reduced in presence of IFN-o~. In dose titration experiments a reduction of the IL-8 protein was detected at IFN concentrations as low as 30U/ml. In CML patients with constitutive expression of IL-8 a reduction of IL-8 mRNA expression was seen after therapeutic administration of IFN-a. By contrast, in LPS stimulated granulocytes IFN failed to inhibit IL-8 expression in vitro. We investigated the mechanism of IL-8 inhibition in the THP-1 cell line more in detail. Nuclear run on assays and RNA decay analysis in presence of acinomycin D suggested that the effect was regulated predominant/y at a posttranscriptional level. De novo protein synthesis was not required since the inhibitory effect was also detected in presence of cycioheximide. In addition to IL-8 expression we studied the effect of IFN-o~ on the synthesis of IL-1, TNF, IL-6 and IL-1RA in PBMNC and bone marrow stromal cell cultures. These experiments revealed an antagonistic effect of IL-1 action by IFN4x at two levels which was most striking in bone marrow stromal cells. Expression of IL-1 mRNA was downregulated whereas the production of IL-1RA was enhanced by IFN-c~. In contrast expression of IL-6 and TNF was enhanced by IFN. We conclude that IFN-o~ differentially regulates proinflammatory cytokines. The inhibition of IL-8 and IL-1 action suggest an antiinflammatory role of type I IFNs. In a phase I clinical trial of recombinant human Interleukin-6 (IL-6), 21 patients were entered to receive daily subcutaneous injections of IL-6 over 7 days followed by a two week observation period and another 4 weeks of daily IL-6 injections. Doses varied between 0.5 and 20 pg/kg body weight. 18 patients were evaluable for studying immune functions. At all dose levels IL-6 administration led to a marked increase in serum levels of C reactive protein and complement factor C3. Natural killer (NK) cell activity was reduced at doses exceeding 5 pg/kg. Similarly lymphokine activated killer (LAK) cell activity induced by in vitro culture over 4 days in the presence of 200 U/ml Interleukin-2 (IL-2) was suppressed at 10 and 20 p.g/kg, as was the proliferative response to IL-2 in vitro. However no changes were observed in the proliferation induced by phytohaemagglutinin, pokeweed mitogen or fixed staphylococcus aureus. There were no changes in peripheral blood lymphocyte subpopulations as measured by CD4 and CD8, nor in the expression of HLA DR. Serum levels of immunoglobulins IgA, IgM and IgG remained unaffected by IL-6 treatment. In contrast we found consistent elevations in levels of IgE all over the dose range. We conclude that IL-6 inhibits NK and LAK activity in vivo which may be of interest in future studies with cytokine combinations and that the role of IL-6 in IgE related diseases might be more important than previously thought. One of the most potent stimulatory agents for the induction of cytokines in myeloid cells is the bacterial ceUwallproduct LPS (Liopopolysaccharide or Endotoxin). In the bloodstream it forms a complex with LBP (LPS Binding Protein) and is recognized by effector cells via the CDI4 receptor. Here we report on studies performed with human peritoneal macrophages that were stimulated in vitro with LPS and a synthetic LPS homologue in the presence and absence of serum. As revealed by ELISA-based analysis of the cellsupernatants, strong, serum-dependent resonses were seen for TNF-, IL-6, IL-8 and G-CSF production, while unstimulated cells produced basically only IL-8. Repeated stimulation of the cells with LPS resulted in adaptation that was different for certain groups of cytokines. The "adapted" ceils produced much less TNF and IL-6 while IL-I and G-CSF was superinduced. Stimulation of the cells with the Lipid A anolog MRL 593 showed a similar picture, given that MRL had to be used in higher concentration. "Adapatation" of the cells with MRL also resulted in an "adapted" response to a challenge with a high dose of LPS so that it might be useful as a therapeutic agent for preventing the septic shock syndrome, e.g. Northern blot analysis showed that the differentiated response of the cells towards a low dose LPS stimulation regarding cytokine production after an LPS challenge occured on transcriptional level, as mRNA levels were regulated accordingly. These results give evidence that two different pathways for LPS dependent stimulation of myeloid cells for cytokineproduction exist and experiments to further elucidate this phenomenon and possibly discover CDI4 independent and dependent pathways are underway. The neutrophil-activating peptide 2 (NAP-2), a member of the "intercrine"-family of chemotactic and reparative host defense cytokines, represents one of several N-terminally truncated cleavage products that originate from platelet-derived 6-thromboglobulin through proteolytic processing. Here we present evidence that there exists also a naturally occurring C-terminally truncated form of NAP-2 that is about four times more potent in eliciting neutrophil degranulation than the original cytokine. The novel molecule was detected in concentrates of culture supernatants from peripheral blood mononuclear ceils and could be separated from authentic NAP-2 by several steps of column chromatography. According to amino acid sequence analysis it had a N-terminus identical to NAP-2, whereas electrophoretic analyses indicated a lower molecular weight as well as a higher isoelectric point. Immunochemical analyses performed with epitope-characterized antibodies raised against NAP-2 C-terminal synthetic peptides identified limited truncation at the C-terminus of the variant molecule. Comparison of reactivity patterns of these antibodies in Western blots as well as in a NAP-2 biologic assay (PMN degranulation assay) confirmed that the variant NAP-2 was truncated by at least one and by maximally three residues. Thus, there is for the first time evidence that proteolytic processing at the N-terminus is not necessarily the only mechanism regulating the formation of neutrophilactivating peptides, but that modification at the C-terminus may assist in the fine-adjustment of biological activity. Cytokines like Interleukin-1 -beta (ILl), Interleukin-6 (IL6), Interleukin-8 (IL8) and Tumor-Necrosis-Factor-alpha (TNF) are involved in the pathogenesis of fever and infection. However, intra-and inter-individual values differ considerably and there is only limited data on early.cytokine serum levels and their evolution in febrile neutropenic patients. Therefore. we measured cytokine levels in 14 adults with chemotherapy-induced aplasia and fever. Concentrations of ILl. IL6 and TNF were determined by IRMA. IL8 by ELISA in 11 specimens per patient. ILl and TNF were elevated in 5 of 10 patients with peak values of 113 pg/ml and 147 pg/ml, respectively. IL6 and IL8 were elevated in all patients with a maximum of 41.300 pg/ml (median 416 pg/ml, range 104 -41 316 pg/ml) for IL6 and 25.6oo pg/ml (median 1100 pg/ml, range 150 -26.000) for IL8. Both cytokines showed a high correlation (r=0.89). The individual IL6concentration-time curve closely paralleled the temperature curve. In 10 of 14 patients IL6 was elevated before onset of fever and peaked at or one hour before the temperature maximum in seven cases. So also in the cytopenie patient lacking a main source of cytokine producing cells and cytokine target cells consistently high IL6 and IL8 serum levels can be detected very early in the course of fever and infection. The biological and clinical significance of this cytokine response and its regulation mechanisms remain to be determined. Interleukin 8 (IL-8) and the neutrophil-activating peptide 2 (NAP-2) are two closely related members of the "intercrine" family of host defense cytokines. The expression of at least two different receptor classes for IL-8 on human neutrophils (PMN) exhibiting similar affinities has been demonstrated recently. Using iodinated ligands we could directly demonstrate that 12SI-NAP-2 specifically bound to PMN with two different affinities, characterized by Kd-values of about 1.4 nM and 9.6 nM, respectively. Cold 72-residue IL-8 competed with iodinated NAP-2 for binding to the high affinity site(s) with practically equivalent efficacy, while it was significantly more effective in displacing 12%NAP-2 from its low affinity site(s) than was cold NAP-2 itself. As new findings, unlabeled IL-8 could completely displace ~251-NAP-2 and vice versa, indicating that there are no distinct binding sites for either cytokine on PMN. in contrast to iL-8, NAP-2 did not induce PMN degranulation at concentrations (2 nM), engaging solely its high affinity site. However, short-term priming of PMN with the same amount of NAP-2 dramatically down-regulated degranulation inducible by higher concentrations of NAP-2 as a secondary stimulus. The IL-8-induced secondary response was also diminished, but to lower extents. These phenomena correlated with the rapid downregulation and internalization of NAP-2 high affinity binding sites from the cell surface. Thus, our data provide direct evidence for a regulatory function of NAP-2 at very low concentrations, obviously occurring through the modulation of NAP-2 and IL-8 receptor expression on PMN. Determination of cytokine plasma levels possesses many promising features concerning monitoring and studying of an array of different diseases including febrile reactions. Detailed analysis of the role of these factors is of crucial importance for the understanding of the complex cytokine network. Investigation of cytokine blood levels however is complicated by their short half-fife in circulation, the presence of soluble inhibitors and the ill-defined beginning of fever. Looking for a suitable in-vivo-model allowing a sequential and well-defined analysis of cytokine plasma levels we chose the acute toxicity after intravenous amphotericin B (Am B) application consisting of fever, chills and hypotension. These side-effects were reported to be mediated by release of pro-inflammatory cytokines such as TNF a and Interleuldn 1 (IL-1). In order to compare mutual interations and different temporal patterns of liberation we determined a panel of cytokines including TNF a, s-TNFreceptor (s-TNF-r), Interleukin-113, Interleukin-1-receptor-antagonist (IL-1-RA), Intefleukin-6 and Interleukin-8 from 15 patients suffering from acute leukemia and fungai infections. Serial EDTA-plasma samples were obtained before and up to 8 hours after start of Am B infusion. Samples were immediately centrifuged and stored at -40 ~ C until analysis by ELISA (Medgenix and R&D Systems). Patients experiencing adverse reactions showed TNF c~ peak plasma levels 90 -180 minutes after starting Am B infusion reaching maximum concentrations of 750 pg/ml. Concentrations declined m base levels within the following 2-3 hours. IL-6 as welt as IL-8 concentrations showed similar, but delayed changes of plasma concentration. Compared to TNF a s-TNF-rlevels peaked about 30 minutes later with a prolonged decrease to basal concentrations. In contrast to TNF c~ no circulating IL-I-13 could be detected, while IL-1-RA demonstrated up to 5-fold increases in concentration. We conclude that this model of a drug induced acute-phase-reaction offers wide possibilities for studying the behaviour of inflammatory cytokines and their inhibitors by means of plasma level determination. Inflammatory processes following severe trauma were found to be associated with an abnormal high secretion of inflammatory cytokines. These cytokines are discussed to be involved in neutrophil activation associated with the release of high amounts of destructive lysosomal proteases into the extracellular space. The task of our investigations was to evaluate the possible regulation of the degranulation of neutrophils by the immunostimulatory cytokine IL-6 and the immunosuppressive factor TGF-I.~. We analysed the concentration of the cqantltrypsin-complex of the lysosomal protease elastase as markers for the degranulation of neutrophils as well as the levels of IL-6 and TGF-I~ 1 in the plasma of patients with multiple trauma or after severe surgeries. The time courses of the plasma levels of IL-6 and the elastase-inhibitor-complex were found to be highly correlated, suggesting a possible regulatory role of this cytokine on the neutrophil degranulation. However the plasma concentrations of TGF-~, were not significantly altered in comparsion to the control group. In additional experiments, the effect of both cytokines on the degranulation of healthy donors was investigated in vitro. Pathological high concentrations of rh~L-6 up to 104 Ulml (as detected in several probes from the surgical area) were found to be capable to induce a significant degranulation of the azurophilic granules (56,3_+ 20,2% of the total cellular enzyme content) under serum free conditions as detected by measurement of elastase release by ELISA technique and enzymatic methods. In contrast to this, the degranulation of neutrophils was found to be uneffected by TGF-I~. In conclusion, these data suggest that the inflammatory cytokine IL-6 may contribute to the activation of neutrophil granulocytes in acute inflammatory processes following severe Irauma, whereas the immunosupresive factor TGF-fil seems to have no direct regulatory effects beside the described chemotactic effects on neutrophils. We determined serum concentrations of soluble tumor necrosis factor receptor (sTNF-Rs) in 61 HIV infected individuals. Eighty-five percent of these had increased serum concentrations of sTNF-R type I (p55) (sTNF-R55) and 95% had increased sTNF-R type II (p75) (sTNF-R75). The extent of the increase of sTNF-R75 was greater in more advanced HIV infection (p=0.046) as it was measured by dividing the 61 individuals into two groups according to the median of the CD4+ T cell count, sTNF-R-55 did not differ between these two groups. A strong correlation was found between sTNF-R75 and the soluble immune activation markers B2-microglobulin (rs=0.74, p<0.0001) and urinary neopterin rs=0.67, p<0.0001), and a less strong correlation with interferon gamma (rs=0.51, P=0.0001). The correlations observed for sTNF-R55 were also significant but were always weaker than that for sTNF-R75. A weak inverse correlation was found between the number of CD4+ T cells and sTNF-R75 (rs=-0.33, p=0.012), no such correlation was observed with sTNF-R55. Our findings suggest that increased concentrations of serum sTNF-Rs in HIV infection are linked to immune activation where synergistic action of interferongamma and the TNF-alpha system are likely to play an important role. To answer the question whether IL-6 is of hematogenous origin, immunohistochemistry and in situ hybridization with IL-6 specific S 35 labeled RNA probes was established to study the frequencies of IL-6 expressing peripheral blood mononuclear cells. A two color immunofluorescence assay with antibodies to CD68 and IL-6 was utilized to correlate IL-6 mRNA expression and IL-6 protein production in monocytes of patients and control individuals. 60 -70% of circulating monocytes spontaneously expressed IL-6 mRNA compared to 15% in normal individuals. A strong correlation of IL-6 protein production and IL-6 mRNA was found in monocytes of patients and controls. Semiquantification of IL-6 mRNA and IL-11I mRNA by PCR demonstrated that about 10 to 60 fold higher amounts of mRNA were found in untreated patients compared to normal individuals. In contrast, we did not find evidence for excessive synthesis of TNFa mRNA. The overproduction of IL-6 and IL-111 in the absence of detectable TNFn mRNA establishes a specific cytokine pattern in circulating monocytes of PMR and GCA patients. We analyzed the IL-6 mRNA expression in biopsy specimens from GCA patients applying in situ hybridization. Autoradiographs were analyzed by visual examination and by using an image-analysis system. Tissue infiltrating macrophages expressed IL-6 mRNA, however, the proportion of IL-6 mRNA + CD68 T was lower than in the peripheral blood. Also, immunohistochemistry with an IL-6 specific antibody demonstrated that only a small fraction of macrophages produced this cytokine. These data suggest that in PMR and GCA, circulating monocytes are activated and produce IL-1 and IL-6. Tissue infiltration of macrophages is not accompanied by a local activation, in contrast, cytokine production is downregulated. Our objective was to evaluate the prognostic value of p21 protein levels in patients with advanced HIVinfection during cytokine therapy (IFN or IFN/combination). Methods: p21 serum levels were measured every two months by ELISA in 29 HIV-infected patients during a period of 2 to 15 months (median 8,6) and compared with clinical stage (WR), disease progression and response to therapy. 18 patients suffered from an infection or an inflammation disease during therapy and 6 from Kaposi sarcoma (KS). Results: In 15 patients with stable disease constant p21 levels (• 37 ng/ml) were observed. Changes of disease correlated with distinct increasing levels of p21 (in 16 patients with acute infection and/or inflammation disease, in 2 patients with progress of KS). After start of AZT-therapy, chemotherapy and antibiotic/ antimycotic therapy p21 level decreased. The highest amount of p21 (up to 165 ng/ml) was found in phases with opportunistic infections. The lowest amount of p21 (31 ng/ml) was observed in a clinically stable patient without an evident progress of disease. Conclusions: Here, we detected increased levels of p21 in HIV-I infected patients under IFN-therapy during acute opportunistic infection. Protein p21 may be a useful marker for the activation of the immunesystem and a prognostic marker during the course of IFN-therapy. Previous studies have shown that the diacetylated synthetic pentapeptide splenopentin (DAc-SP-5) accelerates hemopoietic recovery following sublethal irradiation and nutologons bone marrow transplantation in mice, but the effects of the peptide on human bone marrow cells were still unknown. In this study, human granulocyte-macrophage (rhGM-CSF), macrophage (rhM-CSF), and granulocyte colonystimulating factor (rhG-CSF) as well as interleukin-l~ (rhIL-l~) and interleukin-3 (rhlL-3) were compared for their stimulatory activity on human granulocytemacrophnge colony-forming cells (CFC-GM) alone and in combination with DAc-SP-5. After depletion of accessory cells from bone marrow mononuclear cells (BMMNC) in semi-liquid cultures the combination of rhGM-CSF plus splenopentin stimulated the growth of CFC-GM/-M in dose-depended manner. Furthermore, similar effects were seen in combination of DAc-SP-5 plus rhIL-la and and rhIL-3, but not in rhG-CSF and rhM-CSF plus splenopentin. In unseparated as well as in BMMNC enriched for CD34 + cells comparable stimulatory effects of SP-5 alone were missed. Additionally, in BMMNC enriched for CD33+/CD34 + population, the preculture with DAc-SP-5 for l0 days enhanced the ability of rhGM-CSF, rhIL-lct and rhlL-3 to induce colony formation from this cell source. However, in all these combinations, mainly differentiation to macrophage lineage has been observed. Pheuotypie analysis of precultured BMMNC with splenopentin leads to the suggestion that this compound may recruit a cell population being more sensitive to GM-CSF, IL-la and IL-3, because the percentage of CD34 + cells decreased rapidly, whereas the expression of HLA-DR + was enhanced. Therefore, this potentially interresting molecule might be a candidate as a therapeutic adjuvant with hemopoietic growth factors. Although, the examination of GM-CSF, IL-lcc and 11. The role of T-lymphocyte subsets in the development of aplastie anaemia (AA) remains poorly understood. Therefore we analysed by cytophomotry the contribution and proportion of lymphocyte subpopulationa in the peripheral blood (PB) and bone maxrow (IBM3 of patients with AA before, after 6 weeks of treatment and additionally after 12 weeks of therapy with anti-lymphocyte globulin (ALG), methylprednisolon, and eyelosporine A (CSA). For double labeling immtmofluorescence studies mouoelonal antibodies directed against the following specifities were used: TCRct~-, TCR~5, 5TCS 1, CD2, CD3, CD4, CDS, CD16, CD56, CD38, and HLA-DR. hi 13 patients with AA a sigmficant decrease of CD3 positive T-cells was observed after 6 weeks of therapy (PB 35,94.12,0; BM 16,2• as compared with normal controls (PB 75,7• BM 39,8+3,1). Before therapy, the CD4/CD8 ratio in PI3 and lqM did not differ ~om the ratio in the control population; however a reversed ratio (<1) was present in PB and BM after 6 and 12 weeks of therapy. The number of activated T-cells defined by the antigens CD38, HLA-DR and CD56 were low or in the normal range,and did not fu~er decrease during therapy in contrast to the non-activated T-cells. ?5-T-cells were significantly decreased betbre and after 6 weeks of therapy as compared with healthy controls. However, the proportion of the ~/~-subpopulation ~TCSI was markedly increased betbre (PB 39+_3,5; BM 31+_3,7 I) after 6 weeks (PB 58, 7+2, 9; BM 45, 1+_3, 4) and esspeciaUy after 12 weeks of therapy (PB 94,.t~L7.1) as compared with that in normal subjects (PB 17_+2,0; BM 9,0_+_0,8). These data indicate that TCS1-T-celIs are a T-cell population not affected by treatment with ALG, methylpredrtisolone and CSA . Ftwther studies have to show whether tiffs subpopulafion has an effect on the hematopoiesis of patients with AA. IL-1J3, IL-2, TNF-I3, G-CSF and other cytokines are characterised to have proline in the second position of the N-terminal peptide sequence. From this they could be potential substrates of the dipeptidyl peptidase IV (DP IV). Using the method of capillary electrophoresis, here we show that purified soluble DP IV is capable of hydrolysing oligopeptides with sequences analogous to the N-terminal part of human IL-lb, IL-2, TNF-b and mouse IL-6 up to a length of 12 amino acids. Furthermore, it could be demonstrated that hydrolysis rates are negatively correlated the with chain length of the oligopeptides. Glycosylation of threonine in the third position of the IL-2 hexapeptide sequence has no effect on the hydrolysis rate of this peptide by the DP IV. In contrast to these results, no degradation was found in the case of rlL-lp, rlL-2, natural IL-2, and rG-CSF, using up to 1000-fold higher DPIV concentrations than in the experiments with oligopeptides. After incubation with both, DP IV and aminopeptidase N, also no cytokine degradation was found. Possible explanations for this results will be discussed. Hemopd~'etic growth factors are now being tested in several institutions, in an effort to reduce the duration of neutropenia after bone marrow trasnplantation (BMT). In the past 3 years, we have conducted or participated to several double blind multicentric studies using GM-CSF and G-CSF (Schering-Plough/Sandoz, Behring/Hoechst and Roche lab.) in patients (pts) with NHL. We wish to summarize these studies and also report on our own observations. 1-GM-CSF post-ABMT: double blind international studies and the French national trial that we conducted (Blood 1992 (Blood , 180, 1149 (Blood -1157 have clearly demonstrated the GM-CSF infusion post-ABMT significantly accelerates neutrophil recovery by 4 to 7 days, both in pts receiving unpurged marrow or marrow purged by Mafosfamide. The duration of hospitalization is reduced by 4 days with a possible cost benefit. 2-In pts with documented CMV infection requiring DHPG treatment, myelosuppression was not seen in those who received GM-CSF and DHPG concomitantly while in contrast a severe neutropenia developped in those who received DHPG after administration and discontinuation of GM-CSF (Lancet, 1989 , 1273 . 3-13 pts. in our institution received GM-CSF in a compassionate use for delayed enraftment or engraftment failure after autologous (ABMT : 12) or allogeneic bone marrow transplantation (BMT : 1). The pretransplant regiment included total body irradiation (TBI) in 10 and consisted of high dose polychemotherapy only in 2. Interestingly 9 pts were transplanted for acute myelocytic leukemia (AML), a disease in which the presence of receptors to GM-CSF has been detected in vitro in about 50% of cases, 2 pts hat NHL, and 2 ALL. Of the 12 ABMT, 10 pts received marrow heavily treated in vitro by Mafosfamide and 1 marrow purged by long term marrow culture (LTC). In 7 (AML 4, ALL 2, NHL 1), granulopoietic recovery occured within 3 days (1-16). This included a pt with refractory ALL who received LTC marrow, developped cytomegalovirus (CMV) infection and had not engraftment by day 31 : in this pt the absolute nucleated count (ANC) peaked to 2.7 10 9/I, 3 days later. An additional pt with AML had not engrafted by day 63 when GM-CSF was administered for 12 days with no efficacy. Infusion of back up marrow (which has never been effective in our past experience) combined with GM-CSF and Cyclosporine A was follwoed by sustained engraftment occuring 17 days later. 1 Dt (AML) had a minor and transient response and 4 failed. 6 additional pts in the context of entgraftment failure have since received in various sequences GM-CSF and Cyclosporine A + back up marrow after no response to GM-CSF. 3 have recovered sufficient hemopdfesis strongly suggesting a role of Cycloporine A in engraftment failure (manuscript in prepration). We conclude that GM-CSF is now the first line treatment for poor engraftment and/or engraftment failure. We do not freeze any longer back up marrow in pts autografted after pretransplant regimens not containing TBI. In pts with engraftment failure post TBI, who do not respond to first line GM-CSF, we suggest that the combination of back up marrow and Cyclosporine A on top of GM-CSF should be further evaluated. 4-We administered GM-CSF at a dose of 250 #g/m2/day to 5 pts with resistant NHL and bone marrow involvement after the BEAM myeloablative regimen; reconstitution occurred within the same delays than observed after ABMT in 3. We propose that GM-CSF may replace ABMT in highly selected cases of non-Hodgkin's lymphoma with progressive disease and bone marrow involvement (Lancet 1991, 338, 601) . 5-We have transplanted since September 1992 8 NHL pts with CD 34 purified stem cells followed by infusion of GM-CSF. Successful engraftment was obtained in all with recovery to 0.5x109 PMN/I by day 14 (11-25) and to 50x10 9 plts/l by day 23 (13-32). The expansion of CD 34+ cells in short term liquid culture with 6 cytokines (including GM and G-CSF) is another approach toward transplantation with total abrogation of neutropenia. Overall the introduction of hemopoietic growth factors in transplant units has considerably changed the situation in several aspects. Further studies will combine GM-CSF and G-CSF to other cytokines. Dept. of Hematology, Bone Marrow Transplant Unit -H6pital St-Antoine, Paris, France Alter a pre-phase of VCR 1.4rag/@ (maximum 2rag) i.v. days 1,8 and pradnisolone 60rng/n~ p.o. days 1-10 the high-dose chemotherapy consisted of prednisolone 60rag/n? p.o. days 1-4, ifosfamide days 1-4, methotrexate 5,000 mg/m 2 day 1 as a 24 hour infusion, cytosine-arabinoside 1,000 mg/m 2 i.v. days 3 + 4, and etopeside i.v. days 3 + 4. Etoposide has been escalated from 170 mg/rn 2 to 500 rng/ nf at the present time. The dose of itostamide is currently escalated from 1,500 rag/ rlf to 2,500 mg/rTF and finally 3,500 mg/rn 2 as a continuous 24 hour infusion. The high-dose chemotherapy is repeated for a maximum of four times. During the prephase 12pg/kg fiigrastim (recombinant G-CSF) are given twice daily. Apheresis of APBSC is done on days 5-7. APBSC are reinfused after the high-dose chemotherapy and fiigrastim is given at a dose of 5p.g/kg. With WBC rising to >1,000/p.I aphereses are performed repeatedly. The pre-phase for induction of PBSC was excellently tolerated and a median of 10 We report here the effects of long-term G-CSF subcutaneous administration in 48 patients (congenital n=32, cyclic n=4, idiopathic n= 12) treated for 3-5 years. A sustained ANC response was seen in 30/32 congenital patients, in 4/4 cyclic patients, and in 10/12 idiopathic patients. The G-CSF doses needed to maintain these responses ranged between 1 and 60 #g/kg/d. The ANC responses were associated with a significant decrease in the incidence of severe infections and the need for intravenous antibiotics. G-CSF has been well tolerated in the majority of patients and resulted in a dramatic improvement in the quality of life. The adverse events noted included: osteopenia (n= 12), vasculitis (n=2), mesangioproliferative glomerulonephritis (n=2), and the development of MDS/leukemia (n=2). These adverse events are most likely associated with the underlying disease and not caused by G-CSF treatment We have evaluated recombinant human granulocyte-macrophage colonystimulating factor (rhGM-CSF) (Sandoz-Schering/Plough) as an adjunct for COP-BLAM in the primary treatment of high grade malignant non-Hodgkin's lymphomas (NHL). Patients (n = 182, stage II-lV, age 15-73 years), were randomized to rhGM-CSF (400 #g) or placebo for 7 days s.c. following chemotherapy. Efficacy was analyzed for patients receiving at least 70% of study medication (n = 125). The frequency of clinical relevant infection was reduced by rhGM-CSF (28 vs 69 infections, 16 vs 30 patients, P = .02) with a cumulative probability of remaining infection free in 70% vs 48% (P = .05, log rank test at 190 days). Periods of neutropenia (P _< .01 in 5/8 courses), days with fever (2.1 vs 4.0, P = .04) and days of hospitalization for infection (3.5vs8.0 days, P = .01) were significantly reduced. Complete response (CR) rates, assessed by prognostic risk 28 patients (pts) entered an outpatient protocol designed to test the tolerance, and the clinical and biological effects of extended administration of SC low-dose IL-2 following high dose chemotherapy with autologous bone marrow rescue (BMT). Two IL-2 regimens were used. Protocol A consisted of a once daily dose administered in 5-day cycles of 3 millions IU/m z SC rIL-2 (RU49637 kindly provided by Roussel-Uclaf) every other week. The pts were treated at home and the treatment was scheduled for 6 months (i.e. 12 cycles for 6 months) in the absence of major disease progression or sideeffects. Protocol B consisted of a once daily dose of 6 millions IU/m 2 SC IL2 in 5 days/cycles for 6 consecutive weeks (i.e. 6 cycles for 6 weeks). IL-2 was started 60 to 90 days following BMT. Blood lymphocyte subsets and NK/LAK cytotoxic activity were determined monthly. 17 pts received regimen A and 11 pts were given regimen B. In both regimens, inflammatory skin reactions were the main side effect, leading to interruption of treatment in 1 case. No capillary leak was observed. Flue like syndrom was occasionny observed in protocol B. Hematological parameters were not adversely affected by IL-2. At the lowest dose levels (Regimen A), long term administration of IL-2 did not produce any changes of blood lymphocyte subsets. On the other hand, the administration of 6 Millions IU/m 2 IL-2 resulted in a significant increase of CD3+, CD25+ cells and CD3-CD56+ ceils, and NK/LAK cytotoxic activity of fresh Pbls. This study confirms the feasability of long-term administration of SC lowdose IL2 following autologous ABMT. A dose of 6 Millions IU/m2/d resulted in detectable activation of circulating lymphocytes. Further studies are needed to assess the clinical impact of prolonged low-dose iL-2 in this clinical setting. AIM: To study the incidence of infectious episodes (IE) in a single centre series of 7 autologous BMT patients who received IL-2 in an attempt to prevent relapse. PATIENTS. MEtHODS: IL-2 was given as a oontinuous i.v. infusion through a Hickman catheter with a portable pump: 200.000 IU/m2/d for the first week and 400.000 IU/m2/day thereafter during 12 weeks. IL-2 was started when full neutrophil engraftment was achieved and platelet counts remained stable over 50.000 platelets/mm 3 (median: day +64. Range: 38-118).Median age was 39 years (5-46). 6 were mate. BMT was performed for Acute Lymphoblastic Leukemia (3), Lymphoma (2), and Myeloma (2). RESULTS: IL-2 planned therapy was completed in 5 patients (1 early relapse, 1 refusal) in a median of 103 days (90-154). Median intensity dose of IL-2 was 2.8 x 105 IU/m2/day (1.2 x 105-3.82 x 105).10 infectious episodes who precised admission to the hospital were observed in 5 patients. 7 ol these infections were microbiologically documented: 56% of isolations corresponded to gram negative and 44% to gram positive. First IE was detected at a median of 11 days (4-56) after the starting of IL-2. Median cumulative tL-2 dose until first infectious episode was 3 x 106 IU/m 2 (0.8 x 10 ~' -15 x lOe). Catheter were removed in 3 of them. In our experience, low dose continuous i.v. administration ol IL-2 is associated with a high incidence of infectious episodes. SR 29009 (Sanofi Recherche) is a CHO-derived r-IL2, glycosylated on the threonine in position 3. Its specific activity is identical to native IL2. Dose escalation studies were performed using SR29009 either as IV bolus or IV continuous infusion in advanced-stage cancer patients. Treatment schedule consisted in three 5-day cycles with 9-day rest in between. 22 pts received IV bolus Q8-12 hrs at 4 dose levels (1;3;6;9 Millions IU/m-'/bolus) and 19 pts received IV continuous infusion at 5 dose levels (0.3;0.5; 1;2;3 Millions IU/m2/day). MTD were found to be 6 Millions IU/m 2 Q8 hrs/day and 2 Millions IU/m2/day for bolus and continuous IV infusion respectively. Toxieities were similar in nature to those described under treatment with non-glycosylated rlL-2. DLT were mainly related to capillary leak. 1PR (NHL) was obtained. A significant rise of T-cell and NK cell subsets as well as NK/LAK cytotoxicity in blood was observed at day 8. Stimulation was already maximum for the lowest dose levels. At 0.5 Millions IU/mZ/day, the mean number of CD3+ CD25+ cells increased from 113 to 486/mm -~ on day 8. Similarly, CD56+ cells increased from 461 to 1345/mm 3, and cytotoxicity of fresh Pbl against K562 increased from 10% to 40% (E/T ratio = 25 : 1). In vivo, SR29009 appears to exhibit effects similar to those observed with higher doses (x 4-6 fold) of non glycosylated rlL2. Although intravesical therapy with Bacillus Calmette-Gutrin (BCG) against superficial bladder cancer recurrences and carcinoma in situ is highly effective, its mode of action is still unclear. The bladder tumor cell lines BT-A, BT-B (grade 3 transitional cell carcinoma), SBC2, and SBC7 (grade 1 transitional cell carcinoma) are nearly resistant to natural killer cell activity in vitro. We could demonstrate that these cell lines are susceptible to lymphokine-activated killer cells generated by interleukin (IL)-2 or interferon (IFN)-% We have now investigated whether BCG itself is able to activate a LAK cell-like reaction against bladder tumor cells. Our results show that activation of PBMC with viable BCG is a potent way of generating cytotoxic effector cells against bladder tumor cells. In contrast, killing of the NK and LAK cell-sensitive K562 cell line was not enhanced by BCG-induced PBMC. Because of their different target cell pattern and their, compared to LAK ceils, distinct way of activation these cytotoxic effector cells were termed "BCG-activated killer (BAK) cells". Antibodies neutralizing IFN-T activity blocked the induction of BAK cell cytotoxicity, indicating that this cytokine is playing a crucial role during this process. With respect to the phenotype of BAK cells, our data show that the effector cells belong to the CD8+/CD56 § lymphocyte subset. Depletion of either CD8 + or CD56 + cells from BCG-induced PBMC led to a decrease of cytotoxicity against bladder tumor cells. Furthermore, positively selected CD8 + cells could maintain the level of cytotoxicity exerted by BCG-induced PBMC whereas the depletion of CD56 § cells from this population also eliminated the cytotoxic effect. A direct involvement of CD4 + cells or macrophages in the killing of bladder tumor cells was not observed. In patients with superficial bladder cancer LAK and BAK cells might play an important role in the maintenance of the relapse free state. In this report we summarize our experience with high-dose therapy and peripheral blood stem cell (PBSC) autografting in advanced malignant lymphoma. Since May 1991,27 patients (19 male / 8 female) were included into this study. The median age was 37 years (range 19-58). 11 patients had Hodgkin's disease and 16 non-HodgkJn's lymphoma (9 low-grade / 7 highgrade NHL). Four patients received recombinant G-CSF (filgrastim) (5 I~g/m2/ day s.c.) during steady-state hematopoiesis, while in the remaining 23 patients recombinant G-CSF (filgrastim) was started 24 hours after conventional chemotherapy. For all patients, a target quantity of 0.4 x 109/kg total nucleated cells (TNC) was reached. A wide interindividual range with respect to the level of circulating hematopoietic progenitor cells (325-fold for CFU-GM/ml and 221fold for CD34+ cells/id) was observed. With a median number of 5 leukaphereses (range 2-11), a median of 12.7 x 104 CFU-GM/kg bw and 4.7 x 106 CD34+ celts/kg bw could be harvested, respectively. High-dose therapy consisted of either TBI (14.4 Gy, hyper-fractionated)/Cyclophosphamide (200 mg/kg) or the BEAM-protocol. With the exception of 1 patient who died of a respiratory distress syndrome 32 days following autografting, transplant-related toxicity was moderate. The low toxicity is reflected by the kinetics of hematological reconstitution: 15 (median) days for 1.0 x 109/1 PMN and 11 days for 20 x 10~/I platelets. Platelet reconstitution was closely related to the number of CD34+ cells reinfused. Twenty-six patients are evaluable for response: 5 patients (19%) relapsed after a median of 5 months (range 2-6) posttransplant. The remaining 21 patients are alive in remission with a median follow-up of 9 months (range 3-19). In summary, recombinant G-CSF (filgrastim) is highly efficient in mobilizing PBSC capable of restoring hematopoiesis after myeloablative conditioning therapy. The quantity of CD34+ cells harvested was inversely related to the amount of previous cytotoxic chemotherapy. Therefore, high-dose therapy with stem cell support should be considered as an upfront treatment modality in poor prognosis NHL patients.Department of Internal Medicine, University of Heidelberg, 6900 Heidelberg, Germany. Many of the in vivo effects of the haemopoietic cell growth factors may well have been anticipated based on their known ability to stimulate proliferation and development of multipotent and lineage-restricted progenitor ceils in vitro. What was not predicted from these in vitro studies, however, was the observed ability of at least some of these growth factors to mobillse large numbers of haemopoietic progenitor cells from the bone marrow into the peripheral blood. This was an added "oonus" effect of growth factors and one that is now being widely exploited in a variety of clinical situations.Initially these mobilised PBPC were used, in combination with bone marrow cells, to facilitate the more rapid recovery of myeloid cells following transfer into patients receiving high dose cytotoxic therapy. But when experimental studies showed that the mobilised PBPC also contained primitive (repopulating) stem cells -this encouraged the use of PBPC alone as a source of cells for transplantation. In most of these studies however, collection of the PBPC was performed over several cycles of apheresis -and this was somewhat limiting to their widespread use.Because of this, our approach has been to define the optimal growth factor regimen for mobillsation of PBPC such that apheresis is not required, i.e. to use whole blood and then to identify groups of patients who may benefit from receiving these cells either for autologous or for allogeneic marrow recenstitution. Using a sensitive immunohistochemical technique, 1.3->5.0 logs of breast cancer cell depletion was documented in the positively-selected fractions of marrow (10 patients) and 2->5 logs from the PBPC fractions (3 patients), in whom tumor was initially detected. To date no tumor has been detected in the CD34+ PBPC fractions. The engraftment rates for cohorts 4 and 5 were significantly faster than those of the other cohorts. The preliminary data suggest that since CD34+ PBPCs alone are capable of restoring hematopoiesis following high-dose therapy, a marrow fraction may no longer be needed for this purpose. Longer follow-up will be required to assess the ultimate therapeutic effect of the entire treatment program. The principal morbidity and mortality of high dose chemotherapy with autologous bone marrow support (ABMT) relates to the infectious complications which occurs during 3-4 week aplasia until the marrow autografi recovers. Progenitor cells can be mobilized into the peripheral blood compartment by hematopoietic growth factors, alone or used after chemotherapy. We describe four trials using cytokine-mobilized peripheral blood progenitor cells (PBPC). In the first trial, PBPC collected after GM-CSF administration are used to augment marrow. Reconstitution of tririneage marrow function occurred quickly, resulting in short hospital stays and fewer platelet transfusions. In a second study, GM-CSF/chemotherapy-mobilized PBPC were used as the sole hematopoietic support during high dose chemotherapy. Granulocyte and platelet reconstitution was rapid. Time to hematopoietie recovery, transfusion requirements and duration of hospital stay were all significantly improved for the patients receiving PBPC compared with similar patients receiving marrow alone. However, some patients had poor platelet engraftment. Two recent trials were designed to explore multiple high-dose therapy. In the third trial, PBPC with and without marrow made it feasible to deliver two sequential cycles of high dose therapy. The fourth trial utilizes PBPC in addition to cytokines to deriver four cycles of dose-intensive therapy utilizing doses of chemotherapy that could not be given with cytokine support alone. PBPC appears to make multiple course combination high dose therapy feasible, is particularly useful to support platelets, and may enhance the safety, tolerance and cost of high dose therapy. Severe neutropenia and functJonal defects of neutrophils are an increasing problem in treating advanced HIV infected P. Neutropenia may result from bone marrow failure due to HIV by itself, or maybe.secondary to drug toxicity (mainly Zidovudine, ganciclovir, chemotherapy) or bone marrow infiltration (Mycobacterial infection, neoplasms). P. with severe neutropenia are at a higher risk of developing bacterial infections. Since both GM-CSF and G-CSF have been-successfully used in cancer to optimalize dose intensity and to prevent neutropenia and secondary infections in P. receiving myelotoxic agents, clinical trials have recently been conducted in ARC and AIDS P. with neutropenia to determine their potential benefits and toxicity profile. Our previous experience in treating P. with non Hodgkin's lymphoma (NHL) with a 6-cycle third generation regimen (ProMnCE-CYTABOM), with GM-CSF given at alternate cycles, clearly demonstrated that doses of chemotherapy following the adjunction of GM-CSF were significantly higher than doses given without prior GM-CSF administration. GM-CSF was largely well tolerated in those P., with no sign of active viral replication as measured by HIV-1 P24 Antigenemia. We have also performed an open clinical trial in P. with severe neutTopenia to assess the efficacy and tolerance of GM-CSF with a three weekly subcutaneous administration in P. who initially responded to a daily dose. 24 out the 28 enrolled P. (median duration : 40 days) showed a rise of the absolute neutrophil count (ANC) above the target value of 1000/mm 3 after a median period of 3 day at 3 mcg/kg/d. Only 5/28 P. were given antiretrovirals while other concomitant drugs included mainly ganciclovir (8) and pyrimethamine (10). Of the 21 P. who entered the maintenance phase at 3weekly doses, 12 (57 %) showed a sustained increase of the ANC >1000, while in the remaining 9 P, a new episode of neutropenia was observed.In contrast to P. with NHL, adverse reactions were commonly observed, mostly fever (68%) and flu-like symptoms (50 %). Furthermore among the 22P. without antiretroviral therapy who were P24 Ag negative at baseline, a detectable P24 antigenemia was found in 9 P. These preliminary data on this original schedule of GM-CSl = administration was effective in preserving ANC levels at > 1000 in 57 % of the responders. Such an approach could enhance P.'s compliance and have a direct impact on the cost/benefit evaluation of such therapy. In addition, antiretrovirals should always be associated with GM-CSF to reduce the increased risk of viral replication. Attempts to further improving the curative treatment for AML have to overcome the problem of residual disease represented by leukemic cells surviving in a therapy resistant state. A novel approach aims at recruiting those cells into a sensitive state by the use of GM-CSF before and together (priming) with chemotherapy. In vitro, the recruitment of leukemic blasts to colony forming cells in presence of GM-CSF (Blood 67:1448 , 1986 ) is documented by numerous reports. Given to patients with newly diagnosed AML GM-CSF priming decreased the proportion of leukemic cells in Go and increased cells in sensitive cycle phases within 24-48 h (Blood 77:700, 1991) . A similar priming with preinfusion of GM-CSF for a variable period of 0-8 days before chemotherapy started resulted in significantly inferior outcome and more persistent leukemias than in historical controls suggesting protective effects of GM-CSF (Blood 79:2246 , 1992 . In a first study we showed that GM-CSF following chemotherapy in high risk AML effectively reduced both neutrophiI recovery and early mortality and had no adverse impact on leukemic regrowth and remission duration (Blood 78:1190 (Blood 78: , 1991 . In current randomized study in patients with newly diagnosed AML we give GM-CSF from 24 h before chemotherapy and then on to neutrophil recovery which is repeated in each of the initial 5 treatment courses and is compared to chemotherapy alone. 2 1/2 years from study start and after entering 72 patients (median age 50, range 17-75 years) this update in the GM-CSF group and controls shows 78% and 81% CR, 2 and 5 persistent leukemias and a clearance of day 16 b.m. blasts to < 5% in 59% and in 38%, and remission duration shows a trend in favour of the GM-CSF group. In addition to recruitment, other effects of GM-CSF like enhancement of Ara-C cytotoxicity (Leukemia 3:789, 1989; Ch. Router et al. Leukemia in press) may contribute to this strategy. Thus, GM-CSF appears not to antagonize antileukemic chemotherapy. Whether GM-CSF priming and longterm administration ultimately improves the cure rate in AML should be shown some later from the multiple course strategy used in this study.Medizinische Universit~itsldinikA (H~imatologie/Onkologie), Albert-Schweitzer-Str. 33, D-4400 Mfinster In this multicenter trial was evaluated, whether rhu GM-CSF given concomitant and after chemotherapy (CT) can improve the outcome of AML patients by increasing the cytotoxic effect of CT and by reducing the rate of infectious complications. Induction and early consolidation therapy included cytarabine (Ara-C, 100 mg/m 2, day 1 -8 CIVI), daunorubicin (60 mg/m 2, IV, day 3 -5,) and etoposide (100 mg/m 2, day 4 -8, 2 h IV infusion) with reduced dosages in the second induction and early consolidation course. Late consolidation included one cycle with high-dose Ara-C (3g/m 2, 12 doses) and daunorubicin (30 mg/m 2, day 7 -9) for patients aged 50 years and younger, whereas patients over 50 years received a reduced dose of Ara-C (0.6g/m 2, 12 doses). Patients were randomized after the first induction course to receive either rhu GM-CSF (E. coil, 250 pg/m2/day, s.c.) or placebo starting 48 hours prior to the second induction and the subsequent courses and given throughout chemotherapy until absolute neutrophil count had recovered > 500/pl. Eighty out of 82 patients (median age = 50 years) included into the study could be randomized to receive either GM-CSF (n = 41, median age = 50 years) or placebo (n = 39, median age = 50 years). 64 out of 82 patients (78%) achieved a complete remission (CR). 15 patients (18.5%) were treatment failures. Two patients died and in one patient the response to therapy was not evaluable. There was no statistically significant difference in CR rate between patients of the GM-CSF (81%) and placebo group (79%) nor between patients aged 50 years or less and those over 50 years old. The proportion of relapse free survival at a median follow up of 12 months is 48 % in the GM-CSF versus 33% in the placebo group (p = 0.64). The proportion of survival of the GM-CSF group at 22 months is 38% versus 56% in the placebo group (p = 0.64). GM-CSF did not significantly shorten the period of critical neutropenia and prolonged the period of critical thrombocytopenia especially in patients aged over 50 years. The overall incidence of infectious complications as well as the non-hematological toxicity was similiar in boths groups. Thus GM-CSF therapy is feasible in AML therapy while its influence on remission quality and survival remains still open.Dept. Int. Med. III, University of UIm, 7900 UIm/D, FRG.Al17