Inhibitory effect of simvastatin on the proliferation of human myeloid leukaemia cells in severe combined immunodeficient (SCID) mice.

SCID mice were inoculated intravenously with cells from the human HL60 myeloblastic leukaemia cell line and then treated with the 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA) reductase inhibitor, simvastatin, by subcutaneous continuous infusion. The effect of the drug was measured by subsequent colony formation of surviving HL60 cells in vitro and flow cytometry. The number of clonogenic HL60 cells was reduced in the bone marrow of mice that received simvastatin compared with control mice by 65% and 68% in two separate experiments. The number of clonogenic, normal, murine, bone marrow progenitor cells concomitantly exposed to simvastatin in vivo, was not affected in either experiment. Flow cytometric analysis of bone marrow and spleen cells confirmed these results by showing that simvastatin had reduced the percentage of human leukaemia cells in these tissues by 70% and 88% respectively. The data show that the reported selective effect of simvastatin against acute myeloid leukaemia cells in vitro, can be extended to this in vivo model. HL60 bears an N-ras mutation. In further in vitro studies, ketoconazole, an inhibitor of cholesterol biosynthesis post farnesyl pyrophosphate synthesis, had a similar effect to simvastatin on HL60 colony development. Furthermore, the clonogenicity of a population of N-ras mutated, primary acute myeloid leukaemia (AML) cells was no more sensitive to simvastatin than a population without the mutation. The data suggest that the inhibition of AML cell proliferation by simvastatin may be independent of the RAS signalling pathway.

Liposomal vincristine for the treatment of human acute lymphoblastic leukaemia in severe combined immunodeficient (SCID) mice.

Non-obese diabetic NOD/SCID mice have been used to grow human leukaemia as a systemic disease. The animals were inoculated with leukaemic cells obtained from a 36-year-old male with early B-cell precursor acute lymphoblastic leukaemia and on day 15 were given the first of three weekly injections of 1 mg/kg vincristine or equimolar liposomal vincristine. The development of leukaemia in the mice was monitored by taking weekly blood samples and measuring the cell content by flow cytometry. The median time to 50% human cells in the peripheral blood of mice treated with free vincristine was 41 d from the start of treatment compared with 49 d for mice treated with liposomal vincristine (P < 0.01). The median day of death for mice treated with free vincristine was 47 d from the start of treatment and 57 d for mice receiving liposomal vincristine (P<0.01), thus providing a 21% increase in lifespan for animals treated with the liposomal preparation. There was slightly greater weight loss in mice treated with free vincristine than those given liposomal vincristine. Measurement of in vitro colony forming bone marrow progenitor cells in similarly treated, tumour-free mice, showed no difference in progenitor cell survival between mice that received either type of vincristine. We conclude that encapsulating vincristine in liposomes improves the therapeutic index of this drug measured in mice bearing human leukaemia. This may lead to use of the drug in conventional combination chemotherapy with greater safety or, in this setting, at higher dosage.

Growth of human T-cell lineage acute leukemia in severe combined immunodeficiency (SCID) mice and non-obese diabetic SCID mice.

Primary leukemic cells from patients with acute lymphoblastic leukemia (ALL) can be injected intravenously into mice with severe combined immunodeficiency (SCID) to create a model of human leukemia. Leukemic cells disseminate to murine tissues in a clinicopathologic pattern similar to that seen in humans. Thus far, reports of engraftment of lymphoid leukemia in SCID mice have mainly been from patients with B-cell lineage ALL, for which engraftment occurs more frequently with cells from high-risk patients. There are few data on the engraftment of T-cell lineage ALL in SCID mice. Leukemic cells from 19 patients (16 adult and three pediatric) with T-cell lineage ALL were injected into SCID mice, with overt engraftment of 12 cases (63%). Engraftment of leukemia in SCID mice was associated with earlier death due to leukemia of the patient donors (P < .01, log-rank test). The recently developed non-obese diabetic (NOD)/SCID mouse may expand the uses of the SCID model. Cells from the seven patients with T-cell lineage ALL that failed to cause leukemia in SCID mice were injected into NOD/SCID mice. Overt leukemia engraftment was observed in all seven cases. Thus, growth of human T-cell lineage ALL cells in SCID mice was associated with a high-risk patient group. However, this association was not observed when NOD/SCID mice were used, suggesting that this model would no longer predict patients likely to die early of leukemia, but may provide a more realistic system for studying the biology and treatment of the disease.

G-CSF mobilization of haemopoietic cell populations in SCID mice engrafted with human leukaemia.

We have studied granulocyte colony-stimulating factor (G-CSF)-induced mobilization of haemopoietic cells in severe combined immune-deficient (SCID) mice engrafted with human leukaemia. Three leukaemia cell lines were investigated: the HL60 myeloblastic cell line, a chronic myeloid leukaemia (CML) xenograft cell line and an acute myeloid leukaemia (AML) xenograft line. Engraftment was detected using immunofluorescent staining of class I human leukocyte antigens and flow cytometry. All the tumours grew as disseminated disease with engraftment of bone marrow preceding involvement of peripheral blood (PB). After treatment with G-CSF (250 microg/kg/day) for 5 days, mobilization of haemopoietic progenitor cells (HPCs) was observed in non-engrafted SCID mice (40-fold) and in mice engrafted with human leukaemia (20-fold). G-CSF stimulated increases in PB HPCs and total numbers of host nucleated cells in leukaemia-bearing mice but did not induce rises in numbers of circulating HL60 colony-forming cells. Similarly, in mice engrafted with human CML or AML, G-CSF did not increase the number of malignant cells in the PB. These results provided evidence that the migration of normal and malignant haemopoietic cells into the PB are controlled by different mechanisms, and that contamination of PBSC harvests with leukaemic cells in SCID-human chimaeric mice is not enhanced by G-CSF-stimulated mobilization.

A comparison of the effect of the 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors simvastatin, lovastatin and pravastatin on leukaemic and normal bone marrow progenitors.

Simvastatin is an inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase and also selectively inhibits the growth of leukaemic progenitor cells. The antileukaemic action of simvastatin was compared in vitro with that of lovastatin and pravastatin, chemically related compounds which are also competitive inhibitors of HMG-CoA reductase. After 18 hours incubation with 2.5-20 microM of inhibitor, no effect was observed by any of the compounds on the subsequent clonogenic growth of normal bone marrow (BM) progenitor cells from 4 donors and BM cells from one patient in remission. However, simvastatin and lovastatin produced inhibition of acute myeloid leukaemia (AML) progenitor cell growth of between 25% and 100% in 5 populations tested (4 primary AMLs and the HL60 cell line). Pravastatin showed similar growth inhibitory effects to simvastatin and lovastatin in 2 out of 3 primary AMLs but was less active against one primary AML cell population and HL60 cells. These results indicate that, in addition to simvastatin, lovastatin and pravastatin are also selective inhibitors of leukaemic cell growth, however simvastatin was chosen for clinical trial in patients with leukaemia.

Interleukin-4 enhances the survival of severe combined immunodeficient mice engrafted with human B-cell precursor leukemia.

Human interleukin-4 (huIL-4) has been shown to inhibit the growth in vitro of cells from patients with acute lymphoblastic leukemia (ALL). With the aim of determining whether this cytokine might be useful in the treatment of patients with ALL, the effects of huIL-4 on human B-cell precursor ALL engrafted in severe combined immunodeficient (SCID) mice were examined. The inhibition of [3H] thymidine uptake of primary ALL cells by huIL-4 was maintained following engraftment and passage of leukemia in SCID mice. Five of seven xenograft leukemias showed significant inhibition in vitro by huIL-4 at concentrations as low as 0.5 ng/mL; furthermore, huIL-4 counteracted the proliferative effects of IL-7. When used to treat two human leukemias engrafted in SCID mice, huIL-4 200 microgram/kg/d, as a continuous 14-day subcutaneous infusion, suppressed the appearance of circulating lymphoblasts and extended survival of mice by 39% and 108%, respectively, the first demonstration of IL-4 activity against human leukemia in vivo. The mean steady-state huIL-4 level in mouse plasma during the infusion was 1.46 ng/mL (SEM +/- 0.14 ng/mL), which was similar to concentrations found to be effective in vitro. ALL cells obtained from mice relapsing after huIL-4 treatment continued to show inhibition by the cytokine in vitro. These data suggest that IL-4 may be useful in the treatment of patients with ALL.

The sensitivity of leukemic bone marrow to simvastatin is lost at remission: a potential purging agent for autologous bone marrow transplantation.

BACKGROUND: Autologous bone marrow transplantation (ABMT) is frequently used in the treatment of malignant disease but carries the risk of reintroducing tumor cells into the patient. Methods are required for removing malignant cells from harvested bone marrow (BM) without impairing hematopoietic reconstitution. We have shown that simvastatin is toxic to leukemic progenitor cells at a concentration that conserves normal BM progenitors and may be of use clinically as a novel BM purging agent. METHODS: A two-stage culture system was used to compare the effects of simvastatin on both normal BM progenitor and primary acute myeloid leukemia (AML) cells. AML cells and normal BM mononuclear cells were incubated for 18 hours in suspension culture with 10 micrograms per mL simvastatin and the numbers of surviving clonogenic progenitor cells assayed in semisolid agar culture. RESULTS: Following simvastatin treatment of 18 AML cell populations, the mean surviving fraction of progenitor cells was 21.3 +/- 4.8% ( +/- standard error of the mean [SEM]). In contrast, the mean survival of normal BM progenitors from 16 donors was 89.6 +/- 8.6% ( +/- SEM). Samples were taken from 6 AML patients before treatment and after remission of disease had been induced by chemotherapy. In 5 of these cases the AML sample was significantly more sensitive to simvastatin than the remission sample, 4 of the 5 showed > 80% difference in progenitor cell survival. CONCLUSIONS: AML progenitor cells are sensitive to a short-term exposure to simvastatin that spares normal BM hematopoietic progenitor cells. We conclude that simvastatin may be an effective in vitro purging agent in ABMT for AML.

Deletions and rearrangement of CDKN2 in lymphoid malignancy.

Recurrent abnormalities of the short arm of chromosome 9, including translocations and interstitial deletions, have been reported in both leukemia and lymphoma. The pathologic consequences of these abnormalities remain unknown. The cyclin-dependent kinase 4 inhibitor (CDKN2) gene, which maps to 9p21, has been implicated by the finding of a high frequency of biallelic deletions in leukemic cell lines. We have determined the incidence of structural abnormalities affecting CDKN2 by DNA blot in a panel of 231 cases of leukemia and lymphoma and 66 cell lines derived from patients with lymphoid malignancies with defined cytogenetic abnormalities. Structural alterations of CDKN2 were seen in 20 (8.3%) of all fresh cases and 10 (15.1%) of all cell lines. Biallelic CDKN2 deletions were seen in 11 of 53 (21%) cases of B-cell precursor acute lymphoblastic leukemia (BCP-ALL). There was no association with any particular cytogenetic abnormality. Biallelic deletions were also found in high-grade and transformed non-Hodgkin's lymphoma (NHL) of both B- and T-cell lineages. In two cases of transformed NHL, analysis of sequential samples showed loss of CDKN2 with transformation. Neither deletions nor rearrangements of the CDKN2 gene were seen in any of the 119 leukemias of mature B or T cells analyzed. Biallelic deletions of CDKN2 were observed in 6 of 13 NHL cell lines. Three of the 6 cases had undergone transformation from low- to high-grade disease: in 2 of these cases it was possible to show that the CDKN2 deletions were present in fresh material from the patient and were therefore not an artifact of in vitro culture. Rearrangements of CDKN2 were seen in 2 cases (4%) of BCP-ALL, in 1 case of B-NHL, and in 1 Burkitt's lymphoma cell line and suggest the presence of a "hot spot" for recombination in the vicinity of the CDKN2 gene. These data indicate that the loss of CDKN2 expression may be involved in the pathogenesis of a subset of BCP-ALL, some high-grade NHL, and in the transformation of NHL from low- to high-grade disease. CDKN2 deletions and rearrangements occurred in the absence of detectable cytogenetic changes of chromosome 9p in 25 of 30 (83%) cases. Finally, of 10 cases of BCP-ALL that produced overt, transplantable leukemia in mice with severe combined immunodeficiency (SCID), seven showed biallelic CDKN2 deletions. In contrast, none of 11 cases that failed to engraft showed biallelic CDKN2 deletions. BCP-ALL cases that lack CDKN2 expression may have a particular propensity to grow in SCID mice.

Selective inhibition of primary acute myeloid leukaemia cell growth by simvastatin.

Primary human acute myeloid leukaemic (AML) cells from bone marrow (BM) and peripheral (PB), the human myeloblastic leukaemia cell line (HL60) and normal human BM mononuclear cells were cultured in serum-free medium. The survival of progenitor cells from normal BM, HL60 and AML cell populations was reduced over a range of concentrations of simvastatin. This dose response relationship was more pronounced in HL60 and AML cell cultures, indicating greater sensitivity of AML progenitor cells compared with normal BM progenitors. Short-term exposure (18 h) to a range of concentrations of simvastatin showed the same differential response between leukaemic and normal BM cells in terms of clonogenicity. At a concentration of 10 micrograms/ml progenitor cell survival remained above 65% for normal BM while at this concentration leukaemia progenitor cell survival fell below 25% of the untreated values. The differential effect of simvastatin on normal and leukaemic progenitor cells may have value in the clinical management of AML. The possible use of simvastatin, or related drugs, as adjuvants to conventional chemotherapy including in vitro BM purging, merits consideration.

Selective inhibition of primary acute myeloid leukaemia cell growth by lovastatin.

Primary human acute myeloid leukaemic (AML) cells from bone marrow (BM) and peripheral blood (PB), the human myeloblastic leukaemia cell line (HL60) and normal human BM mononuclear cells were cultured in serum-free medium. The survival of progenitor cells from normal BM, HL60 and AML cell populations was reduced over a range of concentrations of lovastatin. This dose response relationship was more pronounced in HL60 and AML cell cultures, indicating greater sensitivity of AML progenitor cells compared with normal BM progenitors. Short-term exposure (18 h) to a range of concentrations of lovastatin showed the same differential response between leukaemic and normal BM cells in terms of clonogenicity. At a concentration of 10 micrograms/ml progenitor cell survival remained above 65% for normal BM while at this concentration leukaemia progenitor cell survival fell below 25% of the untreated values. The differential effect of lovastatin on normal and leukaemic progenitor cells may have value in the clinical management of AML. The possible use of lovastatin, or related drugs, as adjuvants to conventional chemotherapy including in vitro BM purging, merits consideration.

Human Philadelphia chromosome-positive chronic myeloid leukemia: a potential model for antisense therapy.

We have developed an in vivo model of human chronic myeloid leukemia (CML). A peripheral blood (PB) sample of Philadelphia (Ph) chromosome-positive CML cells in lymphoid blast crisis was transplanted intravenously (IV) into sublethally irradiated severe combined immunodeficient (SCID) mice, and this resulted in engraftment with systemic proliferation. Growth of leukemia was monitored by PB cell morphology and by flow cytometric analysis of murine PB cells labelled with an anti-human leukocyte antigen (HLA) monoclonal antibody. Human cells were first detected in the PB at 4 weeks and comprised a mean of 57% of the total nucleated cells in the PB of these mice by 15 weeks. The Ph chromosome was retained and the population has been successfully passaged. BCR/ABL fusion gene expression was detected in a subsequent passage. Experiments are underway to use this in vivo model to assess the antileukemic activity of BCR/ABL antisense oligonucleotides.

Growth of xenografted human bone marrow: comparison with hemopoietic reconstitution in patients after allogeneic bone marrow transplant and response to granulocyte macrophage colony stimulating factor.

Normal human bone marrow was grown as xenografts in mice immune-suppressed by thymectomy and total body irradiation. Mononuclear cell fractions isolated from marrow harvests from 17 donors all gave rise to subcutaneous nodules which grew to a variable maximum size and then regressed. Human granulocyte/macrophage progenitors (CFU-GM) were recovered from xenografts up to 20 days postimplantation. Xenograft growth, measured by maximum nodule volume, area under the growth curve, and rate of regression, did not correlate with the speed of neutrophil or platelet recovery in bone marrow transplant patients infused with the same marrow. Assay of numbers of stromal fibroblastoid colony forming cells (CFU-F) in donor marrow was also not predictive of subsequent hemopoietic recovery in recipients. Treatment of host animals with daily intraperitoneal injections of 100 micrograms/kg human recombinant granulocyte/macrophage colony stimulating factor produced a more rapid growth of subcutaneous nodules. This technique may therefore be of use in determining the in vivo efficacy of human hemopoietic regulatory factors.

Effects of verapamil and alcohol on blood flow, melphalan uptake and cytotoxicity, in murine fibrosarcomas and human melanoma xenografts.

Verapamil had previously been shown to increase cellular melphalan uptake and cytotoxicity in fibrosarcomas, and increased the area under the blood concentration versus time curve (AUC) for melphalan in CBA mice. Verapamil (10 mg kg-1 i.p.) had no effect on the fractional distribution of cardiac output (FDCO), measured with 86Rb-rubidium chloride, to subcutaneous fibrosarcomas. 14C-Melphalan uptake by FS13 fibrosarcomas was increased 60 min after verapamil (10 mg kg-1 i.p.), but not after lower doses which did not affect the AUC. Flunarizine (5 mg kg-1 i.p.) also had no effect on FDCO to FS13 fibrosarcomas, and tended to increase 14C-melphalan content of blood and the fibrosarcomas and to promote growth delay by melphalan. Alcohol increased FDCO to FS13 fibrosarcomas, maximally at a 1:20 dilution in saline, but had no effect on 14C-melphalan uptake or growth delay. Thus, melphalan cytotoxicity correlated with tumour melphalan uptake, and both followed changes in the AUC for melphalan but not changes in FDCO. In these murine fibrosarcomas melphalan uptake and cytotoxicity were not limited by blood flow. In subcutaneous human melanoma HX46 xenografts, verapamil had no effect on the FDCO, nor on 14C-melphalan uptake, and did not affect blood 14C-melphalan levels, suggesting absence of effects on the AUC and on cellular uptake. Alcohol did not increase the FDCO to HX46 xenografts, providing evidence for a different vascular supply.

Verapamil potentiation of melphalan cytotoxicity and cellular uptake in murine fibrosarcoma and bone marrow.

Growth delay by melphalan of two fibrosarcomas in CBA mice was prolonged by intraperitoneal (i.p.) verapamil, 10 mg kg-1. Verapamil also increased the area under the blood concentration time curve and the gastrointestinal toxicity of melphalan. Verapamil promoted melphalan cytotoxicity to murine bone marrow both in vivo, by CFU-S assay, and in vitro, by CFU-GM assay. In 1 microgram ml-1 [14C]-melphalan, verapamil (10 micrograms ml-1) increased by 1.5 times the [14C]-melphalan accumulation by murine bone marrow, reversibly and independently of external calcium. Efflux of [14C]-melphalan from murine bone marrow was retarded by verapamil. Verapamil increased [14C]-melphalan uptake by disaggregated fibrosarcoma cells but had no effect on melphalan accumulation and cytotoxicity in human bone marrow. Although verapamil affected melphalan pharmacokinetics, enhancement of cellular melphalan uptake by verapamil in murine fibrosarcoma and bone marrow appeared to account for much of the increase in melphalan cytotoxicity. The lack of potentiation of melphalan by verapamil in human marrow suggests differences in melphalan transport or in verapamil membrane interactions in mouse and man.

Epidermal growth factor (hEGF) has no effect on murine intestine epithelial damage and regeneration after melphalan.

The effect of epidermal growth factor (hEGF) on intestinal epithelial damage by melphalan was explored in CBA mice. Human EGF was administered in doses of 100 micrograms kg-1 or 1000 micrograms kg-1 using a variety of schedules. Mucosal damage was assessed 4, 8 and 13 days later, by [14C]-xylose uptake and by microcolony survival of jejunum, ileum and colon. The only regimen to show enhanced jejunal crypt survival was administration of hEGF, 100 micrograms kg-1, i.p., 8 hourly, beginning 24 h before melphalan treatment. Oral administration of hEGF had no effect on melphalan induced damage nor on subsequent recovery of intestinal mucosa. Activity of hEGF in mice was confirmed by demonstration of precocious eyelid opening in newborn mice. No consistent protective or restorative effect of hEGF on melphalan-induced intestinal epithelial damage could be demonstrated with the doses and schedules used.

Studies on the development of human acute myeloid leukaemia xenografts in immune-deprived mice: comparison with cells in short-term culture.

Human AML cells from the blood of a series of patients have been implanted subcutaneously into mice immune-suppressed by thymectomy and total-body irradiation. Solid tumours resulted from 18 out of 19 samples and their growth was compared with the proliferation of AML cells in culture. In 17 cases tumours grew to a maximum size and then spontaneously regressed. Cells from one patient produced tumours which did not regress and could be retransplanted into freshly immune-suppressed mice. Cells from a human promyelocytic cell line (HL60) also produced nonregressing and retransplantantable tumours. Normal human mononuclear bone marrow cells implanted s.c. produced a growth pattern similar to that of the majority of AML cells. A second inoculum of AML cells into animals with regressing tumours also produced tumours and thus regression cannot be accounted for on the basis of returning immunity. AML cells placed into short-term suspension culture invariably matured to monocyte/macrophage type cells and/or granulocytic cells as identified by cytochemical staining. However, no correlation was observed between proliferation or maturation of cells in culture, and tumour growth in vivo. Cells derived from disaggregated AML tumours also showed evidence of myeloid differentiation suggesting that tumour regression is due to maturation of leukaemic cells.

Prevention of acute deaths in mice after very high dose cyclophosphamide by divided dose schedule.

Very high dose cyclophosphamide (Cy) (500-600 mg kg-1) given by single bolus i.p. injection in mice caused acute deaths in all animals within 48 h of treatment (0/10 survivors). These acute deaths were abolished or very significantly reduced if Cy was administered in divided dosage over 8 h (10/10 survivors) or 12 h (14/15 survivors). The effect was maintained at doses of up to 600 mg kg-1 administered in divided dosage over 24 h (15/15 survivors). In 2 human small cell carcinoma xenografts anti-tumour efficacy was not diminished by divided dosage. In both xenografts tumour growth delay was enhanced, although not significantly so, when treated with divided dosage compared with single dose, and in one of the xenografts 3 complete remissions were achieved with divided dosage compared with none after single dosage. It is postulated that the underlying mechanism concerns diminished cardiotoxicity. These results may have significance in clinical studies investigating very high dose Cy.