Orotidinuria induced by allopurinol.

The administration of allopurinol to patients suffering from hyperuricemia and to normal subjects results in increased excretion of the pyrimidine nucleoside orotidine. Evidence is presented for the interference by allopurinol of the de novo biosynthesis of uridine 5'-phosphate in man.

High-resolution proton nuclear magnetic resonance analysis of metastatic cancer cells.

High-resolution proton nuclear magnetic resonance (NMR) studies of intact cancer cells revealed differences between cells with the capacity to metastasize and those that produce locally invasive tumors. The NMR resonances that characterize the metastatic cells were associated with an increased ratio of cholesterol to phospholipid and an increased amount of plasma membrane-bound cholesterol ester. High-resolution NMR spectroscopy could therefore be used to assess the metastatic potential of primary tumors.

Deoxyadenosine triphosphate as a mediator of deoxyguanosine toxicity in cultured T lymphoblasts.

The mechanism by which 2'-deoxyguanosine is toxic for lymphoid cells is relevant both to the severe cellular immune defect of inherited purine nucleoside phosphorylase (PNP) deficiency and to attempts to exploit PNP inhibitors therapeutically. We have studied the cell cycle and biochemical effects of 2'-deoxyguanosine in human lymphoblasts using the PNP inhibitor 8-aminoguanosine. We show that cytostatic 2'-deoxyguanosine concentrations cause G1-phase arrest in PNP-inhibited T lymphoblasts, regardless of their hypoxanthine guanine phosphoribosyltransferase status. This effect is identical to that produced by 2'-deoxyadenosine in adenosine deaminase-inhibited T cells. 2'-Deoxyguanosine elevates both the 2'-deoxyguanosine-5'-triphosphate (dGTP) and 2'-deoxyadenosine-5'-triphosphate (dATP) pools; subsequently pyrimidine deoxyribonucleotide pools are depleted. The time course of these biochemical changes indicates that the onset of G1-phase arrest is related to increase of the dATP rather than the dGTP pool. When dGTP elevation is dissociated from dATP elevation by coincubation with 2'-deoxycytidine, dGTP does not by itself interrupt transit from the G1 to the S phase. It is proposed that dATP can mediate both 2'-deoxyguanosine and 2'-deoxyadenosine toxicity in T lymphoblasts.

Quantitation of G0 and G1 phase cells in primary carcinomas. Antibody to M1 subunit of ribonucleotide reductase shows G1 phase restriction point block.

Human cancers have an apparent low growth fraction, the bulk of cells presumed to being out of cycle in a G0 quiescent state due to the inability in the past to distinguish G0 from G1 cells. The allosteric M1 subunit of ribonucleotide reductase (M1-RR) is constitutively expressed by cycling cells (i.e., G1, S, G2-M). It is acquired during transition from G0 to G1, lost during exit to G0 and thus distinguishes G0 from G1 cells. To estimate the proportion of G0 and G1 cells in primary human breast (n = 5) and colorectal (n = 12) adenocarcinomas, we used both analytical DNA flow cytometry (ADFC) and immunoperoxidase staining of sections with the monoclonal antibody to M1-RR (MAb M1-RR). ADFC of fresh tumors revealed a low percentage of cells in the S phase (4.0 +/- 3.4%) but immunoperoxidase staining for M1-RR revealed an unexpectedly high proportion of positive cells (52.4 +/- 12.7%) in the G1, S, G2-M phases indicating a high G1 content of primary human tumors. Thus, human cancers are blocked in transition in G1 and are not predominantly in a G0 or quiescent differentiated state. This block was interpreted to mean that human cancers are responding to putative regulatory events at a restriction point in the G1 phase, such as relative growth factor deficiency, density inhibition, antiproliferative cytokines, or gene products. Using flow cytometry for both DNA and M1-RR content we found that human colon cancer cell lines arrest in the G1 but not G0 phase upon serum deprivation or density inhibition. Similarly, human breast cancer cell lines are arrested in G1 but not G0 phase by medroxyprogesterone acetate (MPA) or tamoxifen exposure. These findings match our in situ observations, and support the concept of a restriction point block in primary human tumors.

Studies on the coordinate activity and liability of orotidylate phosphoribosyltransferase and decarboxylase in human erythrocytes, and the effects of allopurinol administration.

A coordinate relationship between the activities of two sequential enzymes in the de novo pyrimidine biosynthetic pathway has been demonstrated in human red cells. The two enzymes, orotidylate phosphoribosyltransferase and decarboxylase are responsible for the conversion of orotic acid to uridine-5'-monophosphate. Fractionation of red cells, on the basis of increase of specific gravity with cell age, has revealed that these two enzymes have a marked but equal degree of lability in the ageing red cell. It is postulated that orotidylate phosphoribosyltransferase and decarboxylase form an enzyme-enzyme complex, and that the sequential deficiency of these two enzymes in hereditary orotic aciduria may reflect a structural abnormality in this complex. In patients receiving allopurinol, the activities of both enzymes are coordinately increased, and this increase appears to be due, at least in part, to stabilization of both orotidylate phosphoribosyltransferase and decarboxylase in the ageing red cell. Allopurinol ribonucleotide is an in vitro inhibitor of orotidine-5'-monophosphate decarboxylase and requires the enzyme hypoxanthineguanine phosphoribosyltransferase for its synthesis. However, the administration of allopurinol to patients lacking this enzyme results in orotidinuria and these patients have elevated orotidylate phosphoribosyltransferase and decarboxylase activities in their erythrocytes. Evidence is presented that the chief metabolite of allopurinol, oxipurinol, with a 2,4-diketo pyrimidine ring is capable of acting as an analogue of orotic acid. It is postulated that the in vivo formation of oxipurinol ribonucleotide, catalyzed by orotidylate phosphoribosyltransferase, after allopurinol administration, leads to inhibition of orotidine-5'-monophosphate decarboxylase. This inhibition results in the urinary excretion of excessive amounts of orotidine and orotic acid, and "pseudo-substrate" stabilization of orotidylate phosphoribosyltransferase and decarboxylase.

Mechanisms of lymphocytotoxicity induced by extracorporeal photochemotherapy for cutaneous T cell lymphoma.

Extracorporeal photochemotherapy is an effective treatment for cutaneous T cell lymphoma but its mode of action is uncertain. The reduction in viability of patients' photoirradiated buffy coat lymphocytes was correlated with a 35% increase in DNA single-strand breaks and marked decreases in cellular ATP and NAD levels (to 58 and 34% of control, respectively) immediately after photoirradiation. Complementary in vitro studies were conducted with normal human peripheral blood lymphocytes using a Therakos ultraviolet A (UVA) light box. UVA light was cytotoxic on its own but was potentiated by 8-methoxysporalen. 3-aminobenzamide, a poly (ADP-ribose) synthetase inhibitor, mitigated the cytotoxic effect of ultraviolet A light in the presence of 8-methoxypsoralen in lymphocytes and reduced the amount of nucleotide depletion they caused. 10 J/cm2 of UVA light in the presence of 300 ng/ml 8-methoxypsoralen increased the poly (ADP-ribose) synthetase activity of peripheral blood lymphocytes. Exposing lymphocytes to deoxycoformycin and deoxyadenosine was found to induce biochemical and physical effects similar to those of photochemotherapy. In summary, we have shown that the lymphocytotoxic effect of extracorporeal photochemotherapy for cutaneous T cell lymphoma is apparently mediated by DNA damage, subsequent poly (ADP-ribosyl)ation and adenine nucleotide depletion. It is not known how the DNA damage and resultant biochemical effects relate to the possible immunological mechanism of extracorporeal photochemotherapy; however, it is possible that its effects can be mimicked by other DNA-damaging agents.

Ecto-adenosine triphosphatase deficiency in cultured human T and null leukemic lymphocytes. A biochemical basis for thymidine sensitivity.

Cultured leukemic T and null lymphocytes are highly sensitive to growth inhibition by thymidine, as well as the other deoxynucleosides, deoxyguanosine and deoxyadenosine. By contrast, Epstein-Barr virus-transformed B lymphocytes are relatively resistant to deoxynucleosides. Growth inhibition is associated with the development of high deoxyribotriphosphate pools after exposure to the respective deoxynucleotides. We show that malignant T and null lymphocytes are deficient in ecto-ATPase activity. We show this cell surface enzyme to be of broad specificity, capable of degrading both ribotriphosphates and deoxyribotriphosphates. High levels of this ecto-enzyme are found in deoxynucleoside-resistant, Epstein-Barr virus-transformed B lymphocytes. Ecto-ATPase deficiency may represent a mechanism for increased sensitivity to deoxynucleoside growth inhibition.

Pre-mobilization therapy blood CD34+ cell count predicts the likelihood of successful hematopoietic stem cell mobilization.

We examined pre-mobilization blood CD34+ cell count to predict ability to mobilize adequate peripheral blood progenitor cells (PBPC) in 106 cancer patients and 36 allogeneic donors. Mean pre-mobilization therapy blood CD34+ cell count was 3.1 cells x 10(6)/l (s.d. = 3.9, r = 0.3-37) and mean CD34+ cells collected were 5.3 x 10(6) cells/kg/leukapheresis procedure (s.d. = 7.0, r = 0.03-53). Yields correlated with pre-mobilization CD34+ cells x 10(6)/l (r = 0.37, P-value < 0.0001); correlation was stronger in allogeneic donors (r = 0.56, P-value = 0.0004) and males (r = 0.46, P-value < 0.0001). Based on classification and regression tree multivariate analysis, the predictive value of pre-mobilization blood CD34+ cell count was confounded by other variables, including age, gender, mobilization regimen and malignancy type. We generated an algorithm to predict a minimum PBPC yield of 1 x 10(6) CD34+ cells/kg/leukapheresis procedure after mobilization. A threshold pre-mobilization blood CD34+ cell count of 2.65 cells x 10(6)/l was the most important decision point in predicting successful mobilization. Only 2% of subjects with pre-mobilization blood CD34+ cell counts > 2.65 cells x 10(6)/l did not achieve the minimum per apheresis, whereas 24% with pre-mobilization values below threshold had inadequate mobilization. Prospectively identifying individuals at risk for mobilization failure would allow for improved treatment planning, resource utilization and time saving.

Increasing underrepresentation of elderly patients with advanced colorectal or non-small-cell lung cancer in chemotherapy trials.

BACKGROUND: Elderly patients are underrepresented in chemotherapy trials for advanced colorectal cancer (CRC) and non-small-cell lung cancer (NSCLC). However, the change in underrepresentation over time has not been documented. AIMS: This study aimed to quantify (i) the change in the median age of patients enrolled in clinical trials for metastatic CRC and NSCLC between 1982-1991 and 1992-2001 compared with the general colorectal and lung cancer population, and (ii) the proportion of trials with an upper age limit for eligibility. METHODS: A retrospective review of data from the Victorian Cancer Registry and all large published randomized chemotherapy trials for advanced CRC and NSCLC between 1982 and 2001 was conducted. RESULTS: The median age of patients with CRC enrolled in clinical trials remained constant between the two decades (62.0 and 62.2 years), whereas the median age of the CRC population increased from 68.4 to 70.2 years, increasing the median age difference from 6.4 to 8.0 years. The median age of patients with lung cancer in clinical trials increased from 59.8 to 61.8 years, whereas the median age of the lung cancer population increased from 67.4 to 70.4 years, widening the age difference from 7.6 to 8.6 years. More trials set an upper age limit for eligibility in the first decade than in the second decade for both CRC (51 vs 29%, P = 0.04) and NSCLC (68 vs 41%, P = 0.03). CONCLUSION: International clinical trials for CRC and NSCLC are becoming increasingly unsuitable for application to Australian patients because of the increasing age discrepancy, despite fewer trials restricting eligibility by age.

Purine deoxynucleoside toxicity in nondividing human lymphoid cells.

Cultured leukemic T-lymphoblasts, incubated in the presence of inhibitors of adenosine deaminase, are exquisitely sensitive to growth inhibition by deoxyadenosine. An analogy between this phenomenon and human combined immunodeficiency disease associated with inborn adenosine deaminase deficiency and the use of inhibitors of adenosine deaminase in the management of T-cell acute lymphoblastic leukemia has been noted. These phenomena are believed to reflect accumulation of high intracellular concentrations of deoxyadenosine triphosphate (dATP) following phosphorylation of deoxyadenosine, inhibiting replicating T-cells. In an attempt to extend these observations to noncultured, nonleukemic T-cells, we studied deoxyadenosine metabolism in human thymocytes. Human thymuses were separated into large replicating and small nondividing cell types by centrifugal elutriation. Both thymocyte subpopulations elevated in their dATP pools on incubation with microM concentrations of deoxyadenosine in the presence of erythro-9-[3-(2-hydroxynonyl)]adenosine, an inhibitor of adenosine deaminase. These dATP pool rises were similar in extent to those found in cultured leukemic T-lymphoblasts. However, the finding that small nonreplicating thymocytes elevate their dATP pool was unexpected. This prompted study of unstimulated peripheral blood lymphocytes. These cells (T and non-T) showed a similar elevation of their dATP pool on incubation with deoxyadenosine. Furthermore, these nondividing peripheral blood lymphocytes were killed by microM concentrations of deoxyadenosine in the presence of an inhibitor of adenosine deaminase. The biochemical mechanism of this G0-phase cell death is not known. These findings provide impetus for the investigation of adenosine deaminase inhibitors as lympholytic immunosuppressants or as agents to noncycling malignant lymphoid cells.

Mechanism of deoxycytidine rescue of thymidine toxicity in human T-leukemic lymphocytes.

Cultured malignant human lymphocytes are highly sensitive to growth inhibition by thymidine (50% inhibitory dose congruent to 10(-5) M). Growth inhibition reflects sustained elevation of the deoxythymidine 5'-triphosphate pool associated with secondary elevation of the deoxyguanosine 5'-triphosphate pool and reduction in the deoxycytidine 5'-triphosphate (dCTP) pool. Deoxycytidine was capable of partially reversing thymidine growth inhibition at a concentration of 0.5 microM, and growth recovery was virtually complete at 8 microM. The dCTP pool remained depressed until growth inhibition reversal by deoxycytidine was complete, and at a higher concentration of deoxycytidine the dCTP rose above control levels, but the deoxythymidine 5'-triphosphate and deoxyguanosine 5'-triphosphate pools remained elevated. These results support the view that thymidine growth inhibition induces a critical deficiency of dCTP via allosteric inhibition of ribonucleotide reductase rather than inhibiting DNA replication directly by elevated deoxythymidine 5'-triphosphate or deoxyguanosine 5'-triphosphate pools.

Flow cytometric analysis of adenosine analogue lymphocytotoxicity.

The induction of G1-phase arrest in T-lymphoblasts by cytostatic concentrations of 2'-deoxyadenosine (R. M. Fox, R. F. Kefford, E. H. Tripp, and I. W. Taylor, Cancer Res., 41: 5141-5150, 1981) prompted a flow cytometric analysis of the cell cycle effects of three other adenosine analogues with known effects on polyadenylated RNA metabolism in an attempt to further explore the nature of 2'-deoxyadenosine 5'-triphosphate-mediated lymphotoxicity. Cytostatic concentrations of 9-beta-D-arabinofuranosyladenine induced an S-phase block, while 3'-deoxyadenosine (cordycepin) and tubercidin (7-deazaadenosine) induced a cycle-nonspecific block. Furthermore, total cellular RNA content was unaltered by 2'-deoxyadenosine or 9-beta-D-arabinofuranosyladenine, but 3'-deoxyadenosine and tubercidin caused a marked reduction in total cellular RNA at minimally cytostatic concentrations. At concentrations of 0.3 to 20 microM, all of these nucleosides were toxic to nondividing peripheral blood lymphocytes, suggesting that in these cells their mechanism of action does not involve reactions associated with DNA replication. Inhibition of polyadenylated RNA metabolism by triphosphate derivatives of adenosine analogues may account for lymphocytotoxicity in nondividing cells, but the demonstrated diverse effects of these nucleosides on nucleic acid metabolism in dividing cells preclude elucidation of the mechanism of the unique induction of G1-phase arrest by 2'-deoxyadenosine.

Evidence for a new mechanism of cytotoxicity of 1-beta-D arabinofuranosylcytosine.

Inhibition of DNA synthesis by a pulse of 1-beta-D-arabinofuranosylcytosine (ara-C) results in reinitiation of DNA replication in DNA segments replicated earlier in that S phase and hence double replication of some DNA segments. Experiments were with tissue culture cell lines of human origin. DNA replicated early in S phase was labeled with a pulse of [3H]deoxycytidine with the cells pulsed later in S phase with ara-C. The DNA replicated after the time of the ara-C pulse was density labeled with 5-bromodeoxyuridine. Reinitiation of DNA replication in the already replicated [3H]DNA segments was demonstrated using CsCl density gradient analysis by an increase in the 3H label present in the light-heavy peak of semiconservatively replicated DNA. Also, in DNA of control cells, all of the 3H was in the same strand of the light-heavy DNA duplex as was the 5-bromodeoxyuridine, as shown by alkaline CsCl density gradient analysis of purified light-heavy DNA. However, after a pulse of ara-C, utilization of [3H]DNA strands as template strands was demonstrated by the presence of 3H label at the density of unsubstituted DNA chains in alkaline CsCl gradients of the purified light-heavy DNA. This double replication phenomenon can explain certain chromosomal abnormalities induced by ara-C.

Characterization of transformed cells and tumors by proton nuclear magnetic resonance spectroscopy.

Cultured acute lymphoblastic leukemic cells give a well-resolved proton nuclear magnetic resonance spectrum characteristic of isolated plasma membranes. We demonstrate that the signals, in the spectrum of whole cells, arise predominantly from the plasma membrane and that cells transformed by pokeweed mitogen have membranes which are significantly less rigid than are normal human peripheral blood lymphocytes. Normal thymus, malignant thymus, and a leukemic T-cell line have been compared by proton nuclear magnetic resonance spin echo experiments, and the normal thymus was found to differ. Cells transformed by the Epstein-Barr virus can also be characterized and shown to differ from the leukemically transformed cells by spin echo experiments. Since no probe molecule was required to obtain these results, this is the first definitive evidence that the structure and fluidity of the plasma membranes change as a result of transformation of lymphocytes. Proton nuclear magnetic resonance spectroscopy can now be used to compare the effect of different mitogens on T- and B-lymphocytes as well as to monitor the effects of drugs, metals, etc., on the plasma membrane of transformed lymphocytes.

G1-phase arrest of cultured human leukemic T-cells induced by deoxyadenosine.

Cultured human T-cell leukemia lymphocytes have enhanced sensitivity to growth inhibition by deoxyadenosine. We have used flow cytometry to investigate the mechanism of deoxyadenosine toxicity in cultured T-leukemic cells. Comparative studies on deoxyadenosine-resistant Epstein-Barr virus-transformed B-lymphocyte cell lines were also performed. After exposure of T-cells to low concentrations of deoxyadenosine (3 microM), in the presence of an adenosine deaminase inhibitor (erythro-9-[3-(2-hydroxynonyl)]adenosine), accumulation of cells of cells with a G1 DNA content was demonstrated. In contrast, B-cell lines showed a similar degree of growth inhibition after exposure to 200 to 400 microM deoxyadenosine but were blocked in S phase. The T-cell G1 block was associated with a rise in the deoxyadenosine triphosphate pool, and both these phenomena were prevented by the addition of deoxycytidine. The biochemical mechanism of this G1 block induced by deoxyadenosine in T-cells is not understood.

Ribonucleotide reductase M1 subunit in cellular proliferation, quiescence, and differentiation.

Ribonucleotide reductase catalyzes the first unique, rate-limiting step in DNA synthesis; both its large (M1) and small (M2) subunits are necessary for activity. While direct studies of M2 expression have previously shown a tight correlation with S phase, the kinetic features of M1-expressing cells have remained ill defined. Therefore we have, using immunofluorescence flow cytometry, analyzed changes in whole cell M1 levels and DNA content during various cell cycle and differentiation events. In asynchronous cultures M1 levels are sustained throughout the cell cycle, including G1 phase when M2 levels and ribonucleotide reductase catalytic activity are known to be very low. In contrast M1 is virtually absent from quiescent lymphocytes but is expressed following mitogen stimulation, shortly before S phase cells appear. M1 declines to low levels in "plateau phase" cultures, the major reduction occurring in cells with 2n (G0/G1) DNA content. HL-60 promyelocytic leukemia cells, induced into either myeloid or monocyte-macrophage differentiation, show a similar marked decrease in M1 levels concomitant with the cessation of cell division. We conclude that the M1 subunit of ribonucleotide reductase is constitutively expressed by cycling cells. It is acquired during stimulated transition from G0 to G1 and is lost during exit to G0 or terminal differentiation. This pattern of expression suggests that determination of cellular M1 content may be useful in distinguishing proliferating (including G1) and quiescent (including G0) cells in vivo.

Sensitivity of leukemic human null lymphocytes to deoxynucleosides.

The growth of cultured leukemic T-lymphocytes is readily inhibited by deoxynucleosides, particularly thymidine, deoxyguanosine, and deoxyadenosine. By contrast, Epstein-Barr virus-transformed B-lymphocytes are relatively resistant to deoxynucleosides. Growth inhibition correlates with the development of high deoxyribonucleoside triphosphate pools following exposure to deoxynucleosides. Leukemic T-lymphocytes are deficient in ecto-5'-nucleotidase (5'-ribonucleotide phosphohydrolase, EC 3.1.3.5) activity, and it has been postulated that deficiency of this enzyme decreases the capacity of these cells to degrade deoxyribonucleotides, rendering them sensitive to deoxynucleosides. We find that the sensitivity of cultured null-type leukemic lymphocytes to growth inhibition of deoxynucleosides is similar to that of T-cells. However, the null cells contain normal levels of ecto-5'-nucleotidase. We infer that ecto-5'-nucleotidase deficiency does not have a central role in determining the deoxynucleoside sensitivity of leukemic lymphocytes.

Use of aminoglutethimide as second-line endocrine therapy in metastatic breast cancer.

Sixty-five patients with advanced breast cancer, progressive despite prior endocrine therapy in all cases and prior chemotherapy in most cases, were treated with aminoglutethimide, 250 mg four times a day. At present, 52 are evaluable for response assessment, and of these 10 (19%) showed an overall objective response, major sites of response being soft tissue and lung. A further 12 patients (23%) had stable disease during aminoglutethimide therapy, while a total of 9 patients with bone metastases reported marked relief of pain without objective evidence of response. Forty-nine patients had received prior treatment with tamoxifen, and of the 10 tamoxifen responders 4 (40%) responded to subsequent aminoglutethimide, while of the 20 tamoxifen failures only 2 (10%) responded to subsequent aminoglutethimide. Aminoglutethimide was reasonably well tolerated, although 6 patients (0%) discontinued treatment because of intolerable side effects. Six of the 10 responding patients have subsequently relapsed, with a mean duration of response of 17 weeks, but 4 continued to respond at 24, 32, 55, and 111 weeks, respectively. The median survival from the start of aminoglutethimide therapy is in excess of 41 weeks for responders and 11 weeks for nonresponders, while the median survival from first relapse is 48 months for aminoglutethimide responders and 28 months for aminoglutethimide nonresponders. These results confirm that aminoglutethimide can offer a useful alternative form of endocrine therapy for advanced breast cancer, but the response rates obtained in heavily pretreated patients are inferior to those obtained when aminoglutethimide is used earlier in sequential treatment. For optimal results, particularly in terms of quality of life, aminoglutethimide should generally be used prior to chemotherapy.

Terminal incorporation of 2'-deoxyadenosine into polyadenylate segments of polyadenylated RNA in G1-phase-arrested human T-lymphoblasts.

In the presence of the adenosine deaminase inhibitor erythro-9-[3(2-hydroxynonyl)]adenine microM concentrations of 2'-deoxyadenosine (dAdo) are toxic to nondividing human lymphoid cells and induce G1-phase arrest in T-leukemic lymphoblasts, effects which appear to be independent of ribonucleotide reductase inhibition by accumulated 2'-deoxyadenosine 5'-triphosphate. We sought to determine if 2'-deoxyadenosine 5'-triphosphate had effects similar to those of other cytotoxic adenosine analogues which are incorporated into polyadenylated RNA [poly(A)+ RNA]. In the presence of erythro-9-[3-(2-hydroxynonyl)]adenine, 8-14C]dAdo, at minimal cytostatic concentrations, was incorporated into the polyadenylate segments of cytoplasmic poly(A)+ RNA in the human T-leukemic lymphoblast line CCRF-CEM, and 70% of incorporated dAdo was in the 3'-terminal position. No DAdo was found in enzyme hydrolysates of nonpolyadenylated regions of poly(A)+ RNA or of poly(A)-RNA. Enzymic hydrolysis of polyadenylated segments from labeled poly(A)+ RNA yielded adenosine:dAdo ratios of approximately 55:1.

Phase I study of intravenously administered bacterially synthesized granulocyte-macrophage colony-stimulating factor and comparison with subcutaneous administration.

A Phase I study of bacterially synthesized recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF) was undertaken in 21 patients with advanced malignancy or neutropenia. rhGM-CSF was administered once daily by i.v. bolus injection (0.3 to 3 micrograms/kg/day) or 2-h i.v. infusion (3 to 20 micrograms/kg day) for 10 days. rhGM-CSF at all i.v. doses caused an immediate transient decrease in circulating neutrophils, eosinophils, and monocytes. By 6 h after rhGM-CSF, circulating leukocyte levels were restored. Daily i.v. bolus dosing (0.3 to 3 micrograms/kg/day) did not elevate leukocyte levels except in one neutropenic patient. Daily 2-h i.v. infusions (10 to 20 micrograms/kg/day) caused a dose-dependent leukocytosis with increased levels of neutrophils (up to 4.3-fold), eosinophils (up to 18-fold), and monocytes (up to 3.5-fold). Marrow aspirates showed increased proportions of promyelocytes and myelocytes during rhGM-CSF administration. Retreatment after 10 days without rhGM-CSF resulted in a more marked leukocytosis at doses greater than or equal to 10 micrograms/kg/day. Platelet levels decreased for the first 3 days and then increased during the first course of rhGM-CSF administration. Two patients with chronic lymphocytic leukemia had a transient reduction in lymphocytosis. Serum cholesterol and albumin levels decreased, and vitamin B12 levels increased during rhGM-CSF treatment. At doses of up to 15 micrograms/kg/day, rhGM-CSF was relatively well tolerated by the patients, but adverse effects included bone pain, lethargy, fever, rash, and weight gain. A first dose reaction characterized by hypoxia and hypotension was identified at dose levels greater than or equal to 1 microgram/kg. Dosing i.v. was less potent at inducing a leukocytosis than previously observed for equivalent s.c. doses and was associated with a higher incidence of generalized rash and first dose reactions. The maximal tolerated dose of i.v. rhGM-CSF was 15 micrograms/kg/day. Phase II studies in which the derived effect is to raise leukocyte levels should be undertaken at rhGM-CSF doses of 3 to 15 micrograms/kg/day.