Effects of 1 alpha, 25-dihydroxyvitamin D3 on the human chronic myelogenous leukemia cell line RWLeu-4.

The effects of 1 alpha, 25-dihydroxyvitamin D3 (VD3) on proliferation, differentiation, and macromolecular synthesis in the new Philadelphia chromosome-positive chronic myelogenous leukemia cell line, RWLeu-4, were investigated. Binding of [3H]VD3 was saturable, with approximately 2000-3000 sites/cell, and half-maximal binding occurring at 0.21-0.33 nM. Treatment of RWLeu-4 cells with VD3 induced 24R-hydroxylase activity, a marker of vitamin D3 responsiveness in many tissues. Exposure of RWLeu-4 cells to VD3 also inhibited proliferation and DNA synthesis with a 50% effective dose of 3.5-10 nM within 72 h; in addition, protein and RNA synthesis were inhibited by VD3 treatment. Exposure of RWLeu-4 cells to 5 nM VD3 for 72 h caused 50% of the cells to differentiate into macrophage/monocyte type cells as judged by nitroblue tetrazolium staining and adherence to plastic. Progressive expression of cell surface maturation-specific antigens of the monocyte/macrophage lineage was induced by treatment of RWLeu-4 cells with VD3 for 24 to 72 h at doses that inhibited cellular proliferation. c-myc RNA, which is constitutively expressed in RWLeu-4 cells, increased after 0.5 h of treatment with 50 nM VD3 and then rapidly decreased to barely detectable levels after 4 h of treatment. Finally, the in vitro tyrosine kinase activity associated with the p210bcr-abl oncogene product was decreased approximately 50% by VD3 treatment. Because of the presence of a functional receptor-effector system for VD3 and multiple biological responses to the hormone, these cells provide a unique model system with which to probe the specific effects of VD3 on cell growth and differentiation in chronic myelogenous leukemia.

Therapeutic effects and pharmacokinetics of recombinant human granulocyte-macrophage colony-stimulating factor in childhood cancer patients receiving myelosuppressive chemotherapy.

Twenty-five children with refractory solid tumors were given recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF) in escalated doses of 60 to 1,500 micrograms/m2 as 2-hour intravenous infusions, beginning 24 hours after myelosuppressive treatment with cisplatin and etoposide. Tolerance to rhGM-CSF was exceptional even at dose levels that exceeded the maximum-tolerated dosage (MTD) reported for adults. The agent produced dose-related increases in platelet and neutrophil counts, resulting in significantly shorter durations of severe neutropenia and thrombocytopenia (P less than .01 for each analysis). At the higher dosages (greater than or equal to 750 micrograms/m2), treatment with rhGM-CSF reduced the median number of days of antibiotic therapy for fever and neutropenia by approximately one half. We conclude that rhGM-CSF is well tolerated by leukopenic children in doses as high as 1,500 micrograms/m2. An MTD was not reached in this study. The ability of the growth factor to reduce severe neutropenia and thrombocytopenia suggests it will have an important role in the management of childhood solid tumors.

Recombinant human interleukin-3 (rhIL-3) enhances the mobilization of peripheral blood progenitor cells by recombinant human granulocyte colony-stimulating factor (rhG-CSF) in normal volunteers.

To identify a precisely timed and safe protocol for progenitor cell mobilization, we studied the effects of rhIL-3 and rhG-CSF administration to normal volunteers. rhG-CSF 5 micrograms/kg/d was administered subcutaneously (s.c.) for 7 consecutive days either alone or preceded by rhIL-3 5 micrograms/kg/d s.c. for 4 consecutive days in sequential or partially overlapping schedules. The combined cytokines were well-tolerated--adverse effects were similar to those of the individual agents. Total white blood cell (WBC) and neutrophil counts rose briskly in response to rhG-CSF, and peak mean values were similar between treatment cohorts. Mean platelet counts were modestly elevated during rhG-CSF treatment only in the cohorts receiving rhIL-3 and rhG-CSF. Mean circulating CD34+ cells peaked on day 5 in the rhG-CSF group (38.9+/-14.3/microliter), day 6 in the sequential rhIL-3/rhG-CSF group (56.4+/-12.4/microliter), and day 6 in the partial overlap group (46.1+/-10.9/microliter). On day 3, mean CD34+ cell counts of the subjects who received sequential treatment were markedly higher than observed in the other groups (p<0.05) and were estimated to have been sufficient for collection of adequate grafts by single 10-L leukapheresis procedures in 60% of subjects. Circulating clonogenic cells (CFU-GM and/or BFU-E) were substantially higher in the sequential group than the rhG-CSF group on days 3-6 but were only minimally elevated above baseline in the partial overlap group. The numbers of circulating CD34+/Lin-/Thy-1+ cells (putative stem cells) were increased substantially, especially in the sequential group. On the basis of this pilot trial, we conclude that priming with rhIL-3 is a safe and well-tolerated method for enhancing the mobilization of human blood progenitors and stem cells by rhG-CSF.

Retroviral marking and transplantation of rhesus hematopoietic cells by nonmyeloablative conditioning.

The ability to engraft significant numbers of genetically modified hematopoietic stem and progenitor cells without the requirement for fully myeloablative conditioning therapy is a highly desirable goal for the treatment of many nonmalignant hematologic disorders. The aims of this study were to examine, in nonhuman primates (rhesus), (1) the effects of pretreatment of host animals with cytokines (G-CSF and SCF), i.e., before nonmyeloablative irradiation, on the degree and duration of neo gene marking of circulating leukocytes after autologous cell reinfusion and (2) to compare transduction of primitive hematopoietic target cells in the presence of our standard transduction cytokine combination of IL-3, IL-6, and stem cell factor (SCF) and in the presence of an alternative combination containing SCF, G-CSF, and the thrombopoietin analog MGDF. Cytokine-mobilized rhesus peripheral blood progenitor/stem cells (PBSCs) were enriched for CD34+ cells and transduced with neo vectors (either G1Na or LNL6) for 96 hr in cultures containing rhIL-3, rhIL-6, and rhSCF or MGDF, rhSCF, and rhG-CSF and cryopreserved. Four animals underwent minimal myeloablative conditioning with 500 cGy irradiation with or without pretreatment with SCF and G-CSF, followed by reinfusion of the cryopreserved cells on the subsequent day. Neutrophil nadirs (< or =500/mm3) were 0-3 days in duration; there were no significant periods of severe thrombocytopenia. Marking of circulating granulocytes and mononuclear cells was extensive and durable in all animals (exceeding 12% in the mononuclear cells of one animal) and persisted beyond the final sampling time in all animals (up to 33 weeks). No difference in extent or duration of marking was attributable to either cytokine presensitization of recipients prior to irradiation, or to the substitution of MDGF and G-CSF for IL-3 and IL-6 during transduction.

Pharmacodynamics of daily subcutaneous recombinant human interleukin-3 in normal volunteers.

Normal volunteers received subcutaneous injections of recombinant human interleukin-3 (rhIL-3) on 4 consecutive days to characterize toxicity, pharmacokinetics, and hematopoietic effects. Dosages were 2.5, 5.0, and 7.5 micrograms/kg/day (n = 6 subjects per group). Adverse effects consisted predominantly of flu-like symptoms such as fever and headache. Mean area under the serum concentration-time curve and maximum serum concentration were linearly related to dose. Serum clearance was not apparently related to dose. Clearance increased slightly but significantly between days 1 and 4. Rapid but modest elevations in neutrophil and eosinophil counts were observed during treatment. Mean platelet counts rose modestly, peaking on day 10. Increases of CD34+ cell counts were correlated with increases of colony-forming unit-granulocyte macrophage (peak, day 7).

Pharmacodynamics of subcutaneous recombinant human interleukin-10 in healthy volunteers.

Interleukin-10 inhibits T-lymphocyte activation and proliferation and lipopolysaccharide-induced monocyte production of proinflammatory cytokines. Fifty-four healthy volunteers received single doses of recombinant human interleukin-10 (1.0, 2.5, 5.0, 10, 25, or 50 micrograms/kg) or placebo by subcutaneous injection (randomized double-blind assignment). Clinical adverse events were infrequent at doses below 50 micrograms/kg (five of six subjects had mild flu-like syndrome). Mean serum interleukin-10 concentrations were dose related. The mean terminal-phase half-life ranged from 2.7 to 4.5 hours, and the apparent volume of distribution ranged from 0.70 to 1.35 L/kg. Hematologic changes included transient mild to moderate increases of neutrophil counts, decreases of lymphocyte counts, and a delayed decrease of platelet counts. Recombinant human interleukin-10 significantly suppressed production of the proinflammatory cytokines interleukin-1 beta and tumor necrosis factor-alpha by whole blood stimulated ex vivo with Escherichia coli lipopolysaccharide.

Identification of tyrosine-phosphorylated colony-stimulating factor 1 (CSF-1) receptor and a 56-kilodalton protein phosphorylated in intact human cells in response to CSF-1.

Colony-stimulating factor 1 (CSF-1) selectively supports the survival, proliferation, and maturation of hemopoietic cells of the monocyte/macrophage lineage. Although the cellular receptor for CSF-1, (the c-fms protein) is a protein-tyrosine kinase activated by the binding of CFS-1, the role of phosphorylation of cellular proteins in CSF-1 signal transduction is poorly understood. Therefore, we examined the CSF-1-stimulated phosphorylation of cellular proteins in human BeWo choriocarcinoma cell line (known to express the c-fms protein). BeWo cells were metabolically labeled with 32Pi, stimulated with recombinant human CSF-1, and extracted with detergent. Phosphotyrosyl proteins were isolated from detergent extracts by affinity chromatography on a highly specific antibody to phosphotyrosine. Rapid phosphorylation of 170-kd protein, followed closely by the phosphorylation of a 56-kd protein, was observed in response to CSF-1. The 170-kd phosphotyrosyl protein bound to wheat germ agglutinin and was secondarily immunoprecipitated with a specific anti-fms serum, consistent with its identity as the CSF-1 receptor. Although purified human macrophages that proliferate in culture in response to CSF-1 are not generally accessible, CSF-1 did stimulate the phosphorylation of a 56-kd protein in intact mononuclear leukocytes from human peripheral blood. Thus, the BeWo cell line may represent a good model for the study of CSF-1-stimulated cellular protein phosphorylation.

Leukocytosis in response to exogenous gonadotrophin stimulation.

Leukocytosis may develop in women undergoing ovulation induction. The production of blood leukocytes and their numbers in circulation are regulated by complex interactions involving endogenous haematopoietic cytokines, such as granulocyte-colony stimulating factor (G-CSF), monocyte-colony stimulating factor (M-CSF), and interleukins. The purpose of this prospective study was to explore the presence of leukocytosis in women who receive urinary menotrophins, and to determine whether haematopoietic cytokines are changed in the stimulation process. Controls were volunteers of the same age range, not taking any medication, who received daily saline injections. Subjects underwent phlebotomy at defined points for determination of complete blood counts, G-CSF, M-CSF, and interleukin-6 concentrations. Baseline white blood cell (WBC) counts were similar in patients and controls. In menotrophin-treated patients the WBC counts rose significantly (4.19 +/- 0.28 to 6.37 +/- 0.71) during the stimulation and peaked in the luteal phase (P = 0.037). In contrast, WBC counts decreased in controls. Other leukocytic lineages were not affected. In treated patients G-CSF concentrations rose significantly (P = 0.028 versus controls), while changes in M-CSF and interleukin-6 were not significant.

Stimulation of factor-dependent myeloid cell lines with interleukin 3 induces tyrosine phosphorylation of several cellular substrates.

Interleukin 3 (IL-3) is required for the proliferation of growth factor-dependent myeloid cell lines. To determine the possible signal transduction mechanisms involved in IL-3 growth regulation, we have examined the effects of IL-3 on tyrosine phosphorylation. Using a monoclonal antibody against phosphotyrosine, IL-3 was found to specifically and rapidly induce tyrosine phosphorylation of cytoplasmic proteins of 70, 56, and 38 kDa and a membrane-associated glycoprotein of 140 kDa. Minor and/or variable detected phosphoproteins of 120, 85, 51, and 28 kDa were also seen. Oncogenes encoding tyrosine protein kinases abrogate the requirement of factor-dependent myeloid cells for IL-3. We therefore compared the phosphoprotein profiles of a transformed, IL-3-independent cell line with the IL-3-induced profile. In cells transformed with trk, the 56-, 51-, and 38-kDa cytoplasmic phosphoproteins were constitutively phosphorylated, whereas the 140-kDa phosphoprotein was only phosphorylated in the presence of IL-3. Taken together, these results support a role for tyrosine phosphorylation in the IL-3 signal transduction pathway and suggest that growth factor abrogation by oncogenes encoding tyrosine protein kinases may be due to the phosphorylation of substrates which are normally phosphorylated in response to IL-3.

Studies on Limulus amoebocyte lysate II. Purification of the coagulogen and the mechanism of clotting.

A coagulogen in Limulus lysate has been purified to apparent homogeneity as judged by sodium dodecyl sulfate-gel electrophoresis. The purified coagulogen identified by its ability to clot with either trypsin or the purified Limulus clotting enzyme (Tai, J.Y., and Liu, T. Y. (1976) Fed. Proc. 35, 1486) has a molecular weight of 24,500. It consists of a single polypeptide chain of about 220 amino acids with glycine and serine as its NH2- and COOH-terminal residues; respectively. When acted upon by the Limulus clotting enzyme, the coagulogen releases a soluble C-peptide of about 45 amino acids and an insoluble coagulin of about 170 amino acids. The latter interacts in a noncovalent fashion to form the clot. Amino acid analyses together with the results of NH2- and COOH-terminal analyses suggest that the clot formation involves the cleavage of an--Arg--Lys--bond. Trypsin acts on the coagulogen to cause clotting by splitting the same peptide bond.

Effects of single intravenous doses of recombinant human interleukin-10 on subsets of circulating leukocytes in humans.

Recombinant human interleukin-10 (rhIL-10) is a potent and specific immunomodulatory agent which inhibits endotoxin-stimulated pro-inflammatory cytokine production by monocytes, blocks T-lymphocyte activation by antigen presenting cells, and modulates T(H)1/T(H)2 balance in immune responses. In previous clinical trials, rhIL-10 administered to healthy volunteers induced rapid and transient elevations of neutrophil and monocyte counts and reductions of lymphocyte counts in addition to suppression of endotoxin-stimulated whole blood cytokine synthesis. We sought to better characterize the effects of rhIL-10 on immunophenotypically defined subsets of circulating leukocytes that could be relevant to its immunomodulatory effects. Healthy volunteers were given single doses of 10 microg/kg rhIL-10 (n = 8) or equivalent placebo (n = 4) by intravenous injection. Significant changes of circulating leukocytes included transiently increased neutrophils and monocytes with parallel increases of CD33+ and CD14+ cells. Total lymphocytes as well as total CD3+, CD3+/CD4+ and CD3+/CD8+ cells transiently decreased. Mean fluorescence intensity of CD11a (integrin alpha-chain subunit of lymphocyte function antigen-1, LFA-1) on lymphocytes transiently but significantly decreased, suggesting a mechanism for transient alteration of lymphocyte trafficking. In addition, mean fluorescence intensity of HLA-DR (major histocompatibility class II) on CD14+ cells (predominantly monocytes) transiently but significantly decreased, implying a possible alteration of antigen presenting function. Further study will be required to elucidate the immunomodulatory roles and potential clinical significance of these hematologic changes in therapeutic trials of rhIL-10 in patients with chronic inflammatory and autoimmune diseases.

Pharmacokinetics of intravenous recombinant human granulocyte colony-stimulating factor (rhG-CSF) in children receiving myelosuppressive cancer chemotherapy: clearance increases in relation to absolute neutrophil count with repeated dosing.

Limited evidence suggests increased efficacy of rhG-CSF by subcutaneous (SQ) compared with intravenous (IV) administration. To examine the possibility that rapid elimination of IV rhG-CSF could substantially shorten the duration of systemic exposure and could explain a difference in pharmacodynamics, we characterized the pharmacokinetic profile of IV rhG-CSF for comparison to that previously reported for SQ administration. Twelve children were randomly assigned to receive 10 or more days of IV rhG-CSF at dosages of 5 or 10 microg/kg a day beginning 24 hr after chemotherapy. Enzyme-linked immunosorbent assay (ELISA) was used to measure rhG-CSF concentrations in timed serum samples on days 1 and 10. Pharmacokinetic parameters were estimated by nonlinear, least squares regression. All serum concentration-time profiles were best described by a two-compartment model of elimination. Mean t1/2beta values ranged from 3.68 +/- 0.86 to 22.4 +/- 12.0 hr. ANC was correlated with log CLT (r = 0.72, P < 0.05), and inversely with log dose-adjusted AUC (r = 0.75, P < 0.05) and log dose-adjusted Cmax (r = -0.65, P < 0.05). Estimated duration of serum rhG-CSF concentrations above 1 ng/ml exceeded 24 hr for all but the 5 microg/kg cohort on day 1. Pharmacokinetic parameters of IV rhG-CSF are similar to those previously reported for SQ administration in children treated with myelosuppressive cancer chemotherapy. Daily IV administration should be a suitable alternative route of administration in this patient population.

Concurrent presentation of erythrocytic and megakaryocytic aplasia.

A case of a patient presenting with idiopathic concurrent erythrocytic and megakaryocytic aplasia is reported. The patient's response to immunosuppressive therapy and her bone marrow pathology clearly suggest an immune mechanism. Based on the lack of suppression of erythroid colony growth, several mechanisms are postulated. Well-established molecular and genetic evidence, along with clinical observations, suggests that a relationship exists between the erythrocytic and megakaryocytic cell lines. This may be related to a common bipotential stem cell or common cell surface markers. This case provides strong clinical evidence to support this relationship.

Evaluation of granulocyte-macrophage colony-stimulating factor for treatment of pancytopenia in children with fanconi anemia.

Fanconi anemia is a congenital syndrome characterized by multiple specific physical anomalies, progressive marrow failure, and a predisposition to acute leukemia. We studied the toxicity and efficacy of daily subcutaneous administration of recombinant human granulocyte-macrophage colony-stimulating factor (GM-CSF) in patients with Fanconi anemia and pancytopenia. The toxicity of GM-CSF at the doses and schedule used was minimal. Six of seven patients entered had an increase in the neutrophil count of 7- to 25-fold, which was maintained during the course of study. Despite increases in the reticulocyte count, increases in hemoglobin concentration were rare. No improvement in platelet count was evident in any patient. No patient has evidence of leukemia after up to 19 months of continuous GM-CSF exposure, and all five surviving patients remain responsive to treatment. Although the optimal dose, schedule, and choice of cytokine for patients with marrow failure and Fanconi anemia are not established by this preliminary study, the data indicate that (1) GM-CSF may be able to palliate at least the neutropenia and potentially the neutropenic complications of the disease, (2) this effect can be sustained for more than 1 year, and (3) rapid evolution of acute leukemia is unlikely to be a frequent outcome of such treatment. The clinical impact of GM-CSF or other cytokines in patients with Fanconi anemia and pancytopenia remains to be established by further studies.

Cell lines and peripheral blood leukocytes derived from individuals with chronic myelogenous leukemia display virtually identical proteins phosphorylated on tyrosine residues.

An aberrant p210BCR-ABL protein that possesses constitutive protein-tyrosine kinase activity is presumed to be involved in the development of the neoplastic phenotype in chronic myelogenous leukemia (CML). Using a highly specific antibody against phosphotyrosine, we have isolated the tyrosine-phosphorylated p210BCR-ABL and several other proteins containing phosphotyrosine from a variety of CML cell lines. p210BCR-ABL isolated by the monoclonal anti-phosphotyrosine antibody possessed protein-tyrosine kinase activity in vitro comparable to that of the p210BCR-ABL isolated by antibody to a specific peptide sequence in the ABL protein-tyrosine kinase. Other prominent proteins containing phosphorylated tyrosine residues were observed at 185, 150, 120, 105, 63, 56, 36, and 32 kDa, and less prominent proteins were observed at 195, 155, 94, 53, 40, and less than 29 kDa. Staphylococcal V8 peptide mapping indicated that proteins of similar molecular weights were highly homologous to each other across cell lines, despite the diverse hematopoietic lineages of these cells and the genetic heterogeneity of the patients from whom the CML cell lines were derived. Phosphopeptide mapping also revealed that these proteins were distinct from each other as well as from p210BCR-ABL. Because virtually identical phosphotyrosine-containing proteins were found in peripheral blood leukocytes taken directly from CML patients, these proteins are not an artifact of long-term tissue culture but appear to be an integral part of the CML phenotype.

Pharmacokinetics and immunomodulatory properties of intravenously administered recombinant human interleukin-10 in healthy volunteers.

Normal volunteers received single doses of recombinant human interleukin-10 (rhIL-10; n = 6 per group) or placebo (n = 3 per group) by intravenous injection to characterize pharmacokinetics, tolerability, and immunomodulatory effects. Dosages were 0.1, 0.5, 1.0, 2.5, 5.0, 10.0, 25.0, 50.0, and 100.0 micrograms/kg. Dose-related adverse effects consisted of a mild-to-moderate flu-like syndrome characterized by fever with chills, headache, and myalgias at the highest dose. The mean terminal phase t1/2 ranged from 2.3 +/- 0.5 to 3.7 +/- 0.8 hours. Dose-related effects of rhIL-10 included transient increases of circulating neutrophils and monocytes and decreases of lymphocytes. rhIL-10 markedly suppressed, in a time- and dose-dependent manner, the synthesis of the inflammatory cytokines IL-1 beta and tumor necrosis factor alpha by whole blood stimulated ex vivo with bacterial lipopolysaccharide. Circulating numbers of CD14+/HLA-DR+ cells at 24 hours after the dose were increased in a dose-dependent manner. Effects on expression of HLA-DR by CD14+ cells were variable. There was no apparent effect on HLA-DR expression by CD20+ cells. The immunomodulatory effects of rhIL-10 merit further clinical investigation.

Pharmacokinetics and leukocyte responses of recombinant human interleukin-10.

PURPOSE: To study the pharmacokinetics and ex vivo leukocyte responses of recombinant human IL-10 (rHuIL-10) following single s.c. and i.v. dosing. METHODS: A randomized two-way cross-over study was undertaken in 17 healthy volunteers in which rHuIL-10 was administered as 25 microg/kg s.c. and i.v. doses. Blood samples were collected for 48 hr after dosing to determine serum IL-10 concentrations. Inhibitory activity of IL-10 on ex vivo production of inflammatory cytokines (TNF-alpha and IL-1beta) by LPS-treated peripheral blood cells were measured over 96 hr. RESULTS: A physiologically-relevant modeling approach was developed to determine the pharmacokinetics for two routes of administration (s.c. and i.v.). The i.v. dose showed polyexponential disposition with CL of 65 mL/kg/hr, Vss of 70 mL/kg, and t1/2 of 1.94 hr. Absolute bioavailability averaged 42% for s.c. dosing which produced lower but sustained concentrations. Substantial and prolonged suppression of TNF-alpha and IL-1beta production was achieved during IL-10 treatment. The Hill Function was used to account for the joint concentration-dependent immunosuppressive action of rHuIL-10 after both i.v. and s.c. doses. The IC50 values were about 0.03 ng/ml and Imax values were about 0.85 for both TNF-alpha and IL-1beta suppression. The degree of change as well as the duration of leukocyte response was greater after s.c. administration than after i.v. administration. CONCLUSION: rHuIL-10 shows favorable PKPD characteristics especially by the s.c. route of administration which produced prolonged suppression of cytokine production (ex vivo) which may be applicable in various immune-related disorders.

Effect of montelukast on single-dose theophylline pharmacokinetics.

The effect of montelukast (MK-0476), a cysteinyl leukotriene receptor antagonist in development for treatment of asthma, on single-dose theophylline plasma concentrations was studied in three separate clinical trials. Montelukast was evaluated at 10 mg once daily (the clinical dosage), 200 mg once daily, and 600 mg (200 mg three times daily). At the clinical dosage, montelukast did not change single-dose theophylline plasma concentration in a clinically important manner. The geometric mean ratios for theophylline area under the plasma concentration versus time curve (AUC0-->infinity ) (0.92) and maximal plasma concentration (Cmax ) (1.04) were well within the predefined and generally accepted bioequivalence range of 0.80 and 1.25. Montelukast decreased theophylline Cmax by 12% and 10%, AUC0-->infinity by 43% and 44%, and elimination half-time by 44% and 39% at 200 mg/d (oral and intravenous, respectively), and at 600 mg/d, montelukast decreased theophylline Cmax by 25%, AUC0-->infinity by 66%, and elimination half-time by 63%. These results show that montelukast at the clinical dosage did not change theophylline pharmacokinetics in a clinically important manner, but at 20- to 60-fold higher dosages, montelukast significantly reduced the theophylline pharmacokinetics parameters; an apparent dosage dependence is suggested.