Androgen deprivation in LNCaP prostate tumour xenografts induces vascular changes and hypoxic stress, resulting in promotion of epithelial-to-mesenchymal transition.

BACKGROUND: When single-agent androgen deprivation therapy (ADT) is administered for locally advanced prostate cancer, men usually relapse within 1-2 years with more malignant castrate-resistant disease. The reason for this is currently unknown. We now hypothesise that an initial treatment response that increases tumour hypoxia drives selection of more malignant tumours. METHODS: The LNCaP prostate tumour xenografts were analysed for physiological (oxygen and vasculature) and genetic (PCR array) changes during longitudinal treatment with ADT (bicalutamide, 6 or 2 mg kg⁻¹ daily for 28 days). RESULTS: Bicalutamide caused an immediate (within 24 h) dose-dependent fall in oxygenation in LNCaP-luc prostate tumours with a nadir of ≤ 0.1% oxygen within 3-7 days; this was attributed to a significant loss of tumour microvessels (window chamber study). The hypoxic nadir persisted for 10-14 days. During the next 7 days, tumours regrew, oxygenation improved and the vasculature recovered; this was inhibited by the VEGF inhibitor B20.4.1.1. Gene expression over 28 days showed marked fluctuations consistent with the physiological changes. Accompanying the angiogenic burst (day 21) was a particularly striking increase in expression of genes associated with epithelial-to-mesenchymal transition (EMT). In particular, insulin-like growth factor 1 (IGF-1) showed increases in mRNA and protein expression. CONCLUSIONS: Hypoxic stress caused by ADT promotes EMT, providing a mechanism for the cause of malignant progression in prostate cancer.

Defining normoxia, physoxia and hypoxia in tumours-implications for treatment response.

Tumour hypoxia is increasingly recognized as a major deleterious factor in cancer therapies, as it compromises treatment and drives malignant progression. This review seeks to clarify the oxygen levels that are pertinent to this issue. It is argued that normoxia (20% oxygen) is an extremely poor comparator for "physoxia", i.e. the much lower levels of oxygen universally found in normal tissues, which averages about 5% oxygen, and ranges from about 3% to 7.4%. Importantly, it should be recognized that the median oxygenation in untreated tumours is significantly much lower, falling between approximately 0.3% and 4.2% oxygen, with most tumours exhibiting median oxygen levels <2%. This is partially dependent on the tissue of origin, and it is notable that many prostate and pancreatic tumours are profoundly hypoxic. In addition, therapy can induce even further, often unrecognized, changes in tumour oxygenation that may vary longitudinally, increasing or decreasing during treatment in ways that are not always predictable. Studies that fail to take cognizance of the actual physiological levels of oxygen in tissues (approximately 5%) and tumours (approximately 1%) may fail to identify the real circumstances driving tumour response to treatment and/or malignant progression. This can be of particular importance in genetic studies in vitro when comparison to human tumours is required.

Bicalutamide-induced hypoxia potentiates RUNX2-mediated Bcl-2 expression resulting in apoptosis resistance.

BACKGROUND: We have previously shown that hypoxia selects for more invasive, apoptosis-resistant LNCaP prostate cancer cells, with upregulation of the osteogenic transcription factor RUNX2 and the anti-apoptotic factor Bcl-2 detected in the hypoxia-selected cells. Following this observation, we questioned through what biological mechanism this occurs. METHODS: We examined the effect of hypoxia on RUNX2 expression and the role of RUNX2 in the regulation of Bcl-2 and apoptosis resistance in prostate cancer. RESULTS: Hypoxia increased RUNX2 expression in vitro, and bicalutamide-treated LNCaP tumours in mice (previously shown to have increased tumour hypoxia) exhibited increased RUNX2 expression. In addition, RUNX2-overexpressing LNCaP cells showed increased cell viability, following bicalutamide and docetaxel treatment, which was inhibited by RUNX2 siRNA; a range of assays demonstrated that this was due to resistance to apoptosis. RUNX2 expression was associated with increased Bcl-2 levels, and regulation of Bcl-2 by RUNX2 was confirmed through chromatin immunoprecipitation (ChIP) binding and reporter assays. Moreover, a Q-PCR array identified other apoptosis-associated genes upregulated in the RUNX2-overexpressing LNCaP cells. CONCLUSION: This study establishes a contributing mechanism for progression of prostate cancer cells to a more apoptosis-resistant and thus malignant phenotype, whereby increased expression of RUNX2 modulates the expression of apoptosis-associated factors, specifically Bcl-2.

A study of anti-cancer effects of Funalia trogii in vitro and in vivo.

Extracts of natural products have been used for many years for health benefits. We report on an in vitro and in vivo study into the anti-tumour efficacy of an aqueous extract of the mycelial form of basidiomycete, Funalia trogii. A variety of biological assays were used to show that a 4h exposure of HT29, LNCaP, PC3, MCF-7 and MDA-MB-231 tumour cells to extract (0.5-5.0 mg/ml) resulted in significant cytotoxicity. In a clonogenic assay, IC50 values were found to range from 0.4-0.72 mg/ml; exposing fibroblast cells to the extract resulted in no cell kill. The extract resulted in significant cell kill in proliferating endothelial cells but had no toxicity to quiescent cells, this is useful in targeting tumour tissue since endothelial cells in tumours proliferate more rapidly that those found in other parts of the body. When tumours grown in immune compromised mice were injected intratumourally with extract (5 mg/ml twice a week for two weeks), a 9 day tumour growth delay was observed. The results indicate that the mycelial extract of F. trogii has a promising anti-tumour property.

Dexamethasone potentiates the antiangiogenic activity of docetaxel in castration-resistant prostate cancer.

We sought to characterise whether dexamethasone (DEX) may enhance tumour response to docetaxel in in vitro and in vivo models of metastatic prostate cancer (CaP). In vitro experiments conducted on PC3 and human bone marrow endothelial cells (hBMECs) determined that administration of DEX (10 nM) reduced constitutive nuclear factor-kappaB (NF-kappaB) activity, decreasing interleukin (IL)-8, CXCL1 and VEGF gene expression in PC3 cells. Dexamethasone also attenuated docetaxel-induced NF-kappaB and activator protein-1 transcription and reduced docetaxel-promoted expression/secretion of IL-8 and CXCL1 in PC3 and hBMECs. Although DEX failed to enhance docetaxel cytotoxicity on PC3 cells, DEX potentiated the antiangiogenic activity of docetaxel in vitro, further reducing vessel area and vessel length in developing endothelial tubes (P<0.05). Docetaxel had a potent antiangiogenic activity in the dorsal skin flap-implanted PC3 tumours in vivo. Small blood vessel formation was further suppressed in tumours co-treated with docetaxel and DEX, substantiated by an increased average vessel diameter and segment length and a decreased number of branch points in the residual tumour vasculature (P<0.001). Our data show that DEX potentiates the antiangiogenic activity of docetaxel, suggesting a putative mechanism for the palliative and survival benefits of these agents in metastatic CaP.

Bioreductive drugs: from concept to clinic.

One of the key issues for radiobiologists is the importance of hypoxia to the radiotherapy response. This review addresses the reasons for this and primarily focuses on one aspect, the development of bioreductive drugs that are specifically designed to target hypoxic tumour cells. Four classes of compound have been developed since this concept was first proposed: quinones, nitroaromatics, aliphatic and heteroaromatic N-oxides. All share two characteristics: (1) they require hypoxia for activation and (2) this activation is dependent on the presence of specific reductases. The most effective compounds have shown the ability to enhance the anti-tumour efficacy of agents that kill better-oxygenated cells, i.e. radiation and standard cytotoxic chemotherapy agents such as cisplatin and cyclophosphamide. Tirapazamine (TPZ) is the most widely studied of the lead compounds. After successful pre-clinical in vivo combination studies it entered clinical trial; over 20 trials have now been reported. Although TPZ has enhanced some standard regimens, the results are variable and in some combinations toxicity was enhanced. Banoxantrone (AQ4N) is another agent that is showing promise in early phase I/II clinical trials; the drug is well tolerated, is known to locate in the tumour and can be given in high doses without major toxicities. Mitomycin C (MMC), which shows some bioreductive activation in vitro, has been tested in combination trials. However, it is difficult to assign the enhancement of its effects to targeting of the hypoxic cells because of the significant level of its hypoxia-independent toxicity. More specific analogues of MMC, e.g. porfiromycin and apaziquone (EO9), have had variable success in the clinic. Other new drugs that have good pre-clinical profiles are PR 104 and NLCQ-1; data on their clinical safety/efficacy are not yet available. This paper reviews the pre-clinical data and discusses the clinical studies that have been reported.

p21((WAF1))-mediated transcriptional targeting of inducible nitric oxide synthase gene therapy sensitizes tumours to fractionated radiotherapy.

Cancer gene therapy that utilizes toxic transgene products requires strict transcriptional targeting to prevent adverse normal tissue effects. We report on the use of a promoter derived from the cyclin dependent kinase inhibitor, p21((WAF1)), to control transgene expression. We demonstrate that this promoter is relatively silent in normal cells (L132, FSK, HMEC-1) compared to the almost constitutive expression obtained in tumour cells (DU145, LNCaP, HT29 and MCF-7) of varying p53 status, a characteristic that will be important in gene therapy protocols. In addition, we found that the p21((WAF1)) promoter could be further induced by both external beam radiation (up to eight-fold in DU145 cells), intracellular-concentrated radionuclides ([(211)At]MABG) (up to 3.5-fold in SK-N-BE(2c) cells) and hypoxia (up to four-fold in DU145 cells). We have previously achieved significant radiosensitization of tumour cells both in vitro and in vivo by using inducible nitric oxide synthase (iNOS) gene therapy to generate the potent radiosensitizer, nitric oxide (NO(.-)). Here, we report that a clinically relevant schedule of p21((WAF1))-driven iNOS gene therapy significantly sensitized both p53 wild-type RIF-1 tumours and p53 mutant HT29 tumours to fractionated radiotherapy. Our data highlight the utility of this p21((WAF1))/iNOS-targeted approach.

Tumor-selective drug activation: a GDEPT approach utilizing cytochrome P450 1A1 and AQ4N.

Drug metabolizing transgene products, which activate bioreductive cytotoxins, can be used to target treatment-resistant hypoxic tumors. The prodrug AQ4N is bioreduced in hypoxic cells by cytochrome P450s (CYPs) to the cytotoxin AQ4. Previously we have shown that intra-tumoral injection of CYP3A4 and CYP2B6 transgenes with AQ4N and radiation inhibits tumor growth. Here we examine the ability of other CYPs, in particular CYP1A1, to metabolize AQ4N, and to enhance radiosensitization. Metabolism of AQ4N was assessed using microsomes prepared from baculovirus-infected cells transfected with various CYP isoforms. AQ4N metabolism was most efficient with CYP1A1 (66.7 nmol/min/pmol) and 2B6 (34.4 nmol/min/pmol). Transient transfection of human CYP1A1+/-CYP reductase (CYPRED) was investigated in hypoxic RIF-1 mouse cells in vitro using the alkaline comet assay. There was a significant increase in DNA damage following transient transfection of CYP1A1 compared to non-transfected cells; inclusion of CYPRED provided no additional effect. In vivo, a single intra-tumoral injection of a CYP1A1 construct in combination with AQ4N (100 mg/kg i.p.) and 20 Gy X-rays caused a 16-day delay in tumor regrowth compared to tumors receiving AQ4N plus radiation and empty vector (P=0.0344). The results show the efficacy of a CYP1A1-mediated GDEPT strategy for bioreduction of AQ4N.

Radiogenic therapy: novel approaches for enhancing tumor radiosensitivity.

Radiotherapy (RT) is a well established modality for treating many forms of cancer. However, despite many improvements in treatment planning and delivery, the total radiation dose is often too low for tumor cure, because of the risk of normal tissue damage. Gene therapy provides a new adjunctive strategy to enhance the effectiveness of RT, offering the potential for preferential killing of cancer cells and sparing of normal tissues. This specificity can be achieved at several levels including restricted vector delivery, transcriptional targeting and specificity of the transgene product. This review will focus on those gene therapy strategies that are currently being evaluated in combination with RT, including the use of radiation sensitive promoters to control the timing and location of gene expression specifically within tumors. Therapeutic transgenes chosen for their radiosensitizing properties will also be reviewed, these include: gene correction therapy, in which normal copies of genes responsible for radiation-induced apoptosis are transfected to compensate for the deletions or mutated variants in tumor cells (p53 is the most widely studied example). enzymes that synergize the radiation effect, by generation of a toxic species from endogenous precursors (e.g., inducible nitric oxide synthase) or by activation of non toxic prodrugs to toxic species (e.g., herpes simplex virus thymidine kinase/ganciclovir) within the target tissue. conditionally replicating oncolytic adenoviruses that synergize the radiation effect. membrane transport proteins (e.g., sodium iodide symporter) to facilitate uptake of cytotoxic radionuclides. The evidence indicates that many of these approaches are successful for augmenting radiation induced tumor cell killing with clinical trials currently underway.

Potential use of the alkaline comet assay as a predictor of bladder tumour response to radiation.

Bladder tumours show a variable response to radiotherapy with only about 50% showing good local control; currently there is no test to predict outcome prior to treatment. We have used five bladder tumour cell lines (T24, UM-UC-3, TCC-SUP, RT112, HT1376) to investigate the potential of the alkaline comet assay (ACA) to predict radiosensitivity. Radiation-induced DNA damage and repair were compared to clonogenic survival. When the five cell lines were irradiated and initial DNA damage was plotted against cell survival, at all doses (0-6 Gy), a significant correlation was found (r2=0.9514). Following 4 Gy X-irradiation, all cell lines, except T24, showed a correlation between SF2 vs half-time for repair and SF2 vs residual damage at 5, 10, 20 and 30 min. The T24 cell line showed radioresistance at low doses (0-2 Gy) and radiosensitivity at higher doses (4-6 Gy) using both cell survival and ACA end points, explaining the lack of correlation observed for this cell line. These data indicate that initial DNA damage and residual damage can be used to predict for radiosensitivity. Our data suggest that predictive tests of radiosensitivity, appropriate to the clinical situation, may require the use of test doses in the clinical range.

Relationship between clonogenic radiosensitivity, radiation-induced apoptosis and DNA damage/repair in human colon cancer cells.

The intrinsic radiation sensitivity of normal and tumour tissue is a major determinant of the outcome of radiotherapy. There is currently no established test that can be used routinely to measure the radiosensitivity of the cells in an individual patient's cancer in a manner that can inform treatment planning. The purpose of this study was to evaluate, in four human colorectal adenocarcinoma cell lines, two possible end points as surrogate markers of radiation response--apoptosis and induction of DNA single-strand breaks--and to compare the results with those of a conventional clonogenic assay. Cell lines (SW707 SW480, SW48 and HT29) known to differ in radiosensitivity were exposed to single doses of X-rays ranging from 0.5 to 5 Gy and cell survival was measured using the clonogenic assay. Apoptosis was determined on the basis of morphology under fluorescent microscopy and DNA damage/repair was measured, as tail moment, using an adaptation of the alkaline comet assay. The relationship between surviving fraction at 2 Gy (SF2) and the percentage of apoptotic cells 24 h after the same dose was complex, but apoptosis accurately predicted the order of radiosensitivities as measured by SF2. Initial damage measured after 2 Gy using the alkaline comet assay gave a close correlation with SF2 (r2=0.95), whereas there was no correlation between initial DNA damage repair rate and SF2.

The chemopotentiation of cisplatin by the novel bioreductive drug AQ4N.

AQ4N is a bioreductive drug that can significantly enhance the anti-tumour effect of radiation and cyclophosphamide. The aim of this study was to examine the ability of AQ4N to potentiate the anti-tumour effect of cisplatin and to compare it to the chemopotentiation effect of tirapazamine. In the T50/80 murine tumour model, AQ4N (50-100 mg/kg) was administered 30 min, 2.5 or 6 h prior to cisplatin (4 mg/kg or 8 mg/kg); this produced an anti-tumour effect that was approximately 1.5 to 2 times greater than that achieved by a single 4 or 8 mg/kg dose of cisplatin. Tirapazamine (25 mg/kg) administered 2.5 h prior to cisplatin (4 mg/kg) resulted in a small increase in anti-tumour efficacy. AQ4N was also successful in enhancing the anti-tumour effect of cisplatin in the SCCVII and RIF-1 murine tumour models. This resulted in an increased cell kill of greater than 3 logs in both models; this was a greater cell kill than that observed for tirapazamine with cisplatin. Combination of cisplatin with AQ4N or tirapazamine resulted in no additional bone marrow toxicity compared to cisplatin administered alone. In conclusion, AQ4N has the potential to improve the clinical efficacy of cisplatin.

AQ4N: a new approach to hypoxia-activated cancer chemotherapy.

Preclinical studies demonstrate that in vivo AQ4N enhances the anti-tumour effects of radiation and chemotherapeutic agents with a dose-modifying factor of approximately 2.0. With careful scheduling no, or very little, additional normal tissue toxicity should be observed. AQ4N is a bioreductive prodrug of a potent, stable, reduction product which binds non-covalently to DNA, facilitating antitumour activity in both hypoxic and proximate oxic tumour cells. AQ4N is clearly different in both its mechanism of action and potential bystander effect compared to previously identified bioreductive drugs. In particular AQ4N is the only bioreductive prodrug topoisomerase II inhibitor to enter clinical trials. Targeting this enzyme, which is crucial to cell division, may help sensitize tumours to repeated (fractionated) courses of radiotherapy. This is because in principle, the bioreduction product of AQ4N can inhibit the topoisomerase activity of hypoxic cells as they attempt to re-enter the cell cycle.

Regulation of FOS by different compartmental stresses induced by low levels of ionizing radiation.

We irradiated different cellular compartments and measured changes in expression of the FOS gene at the mRNA and protein levels. [(3)H]Thymidine and tritiated water were used to irradiate the nucleus and the whole cell, respectively. (125)I-Concanavalin A binding was used to irradiate the cell membrane differentially. Changes in FOS mRNA and protein levels were measured using semi-quantitative RT-PCR and SDS-PAGE Western blotting, respectively. Irradiation of the nucleus or the whole cell at a dose rate of 0.075 Gy/h caused no change in the level of FOS mRNA expression, but modestly (1.5-fold) induced FOS protein after 0.5 h. Irradiation of the nucleus at a dose rate of 0.43 Gy/h induced FOS mRNA by 1.5-fold after 0.5 h, but there was no significant effect after whole-cell irradiation. FOS protein was transiently induced 2.5-fold above control levels 0.5 h after a 0. 43-Gy/h exposure of the nucleus or the whole cell. Irradiation of the cell membrane at a dose rate of 1.8 Gy/h for up to 2 h caused no change in the levels of expression of FOS mRNA or protein, but a dose rate of 6.8 Gy/h transiently increased the level of FOS mRNA 3-fold after 0.5 h. These data demonstrate the complexity of the cellular response to radiation-induced damage at low doses. The lack of quantitative agreement between the transcript and protein levels for FOS suggests a role for post-transcriptional regulation.

Increased repair and cell survival in cells treated with DIR1 antisense oligonucleotides: implications for induced radioresistance.

PURPOSE: To determine whether repression of a recently isolated, X-ray-responsive gene, DIR1, using antisense oligonucleotides could affect clonogenic cell survival and repair of DNA strand breaks and have a possible role in the mechanism underlying the phenomenon of 'induced radioresistance' (IRR). MATERIALS AND METHODS: Three cell lines, V79, RT112 and UM-UC-3, which are known to exhibit low-dose hypersensitivity (HRS) and induced radioresistance (IRR), and the radiosensitive cell line ATBIVA, were transfected with antisense oligonucleotides directed towards the DIR1 gene. Scrambled oligonucleotides were used as controls. DNA single-strand break (ssb) repair, using the alkaline comet assay, and cell survival using a standard clonogenic assay was measured after exposure to X-rays. RESULTS: Following treatment with 4Gy X-rays, the V79, RT112 and UM-UC-3 cell lines all exhibited significantly increased rates of ssb repair after transfection with DIR1 antisense oligonucleotides compared with cells transfected with scrambled oligonucleotides. They also demonstrated significantly enhanced survival after exposure to 2 Gy X-rays; the radiosensitive ATBIVA cells did not show these effects. CONCLUSIONS: Repression of the DIR1 gene product leads to an increase in the rate of repair and cell survival in three radioresistant cells lines but not in the radiosensitive ATBIVA cell line. Because DIR1 is repressed by X-rays in the dose range where IRR is observed, it may represent a candidate gene involved in the IRR phenomenon.

Enhancement of chemotherapy and radiotherapy of murine tumours by AQ4N, a bioreductively activated anti-tumour agent.

AQ4 (1,4-Bis-[[2-(dimethylamino-N-oxide)ethyl]amino]5,8-dihydroxyanthrace ne-9, 10-dione) is a prodrug designed to be excluded from cell nuclei until bioreduced in hypoxic cells to AQ4, a DNA intercalator and topoisomerase II poison. Thus, AQ4N is a highly selective bioreductive drug that is activated in, and is preferentially toxic to, hypoxic cells in tumours. Five murine tumours (MAC16, MAC26, NT, SCCVII and RIF-1) have been used to investigate the anti-tumour effects of AQ4N. In only one tumour (MAC16) was AQ4N shown to be active as a single agent. However, when combined with methods to increase the hypoxic tumour fraction in RIF-1 (by physical clamping) and MAC26 tumours (using hydralazine) there was a substantial enhancement in anti-tumour effect. Notably, RIF-1 tumours treated with AQ4N (250 mg kg(-1)) followed 15 min later by physically occluding the blood supply to the tumour for 90 min, resulted in a 13-fold increase in growth delay. When combined with radiation or chemotherapy, AQ4N substantially increased the effectiveness of these modalities in a range of in vivo model systems. AQ4N potentiates the action of radiation in both a drug and radiation dose-dependent manner. Further the enhancement observed is schedule-independent with AQ4N giving similar effects when given at any time within 16 h before or after the radiation treatment. In combination with chemotherapy it is shown that AQ4N potentiates the activity of cyclophosphamide, cisplatin and thiotepa. Both the chemotherapeutic drugs and AQ4N are given at doses which individually are close to their estimated maximum tolerated dose (data not included) which provides indirect evidence that in the combination chemotherapy experiments there is some tumour selectivity in the enhanced action of the drugs.

Induction and rejoining of DNA double-strand breaks in bladder tumor cells.

The induction and rejoining of radiation-induced double-strand breaks (DSBs) in cells of six bladder tumor cell lines (T24, UM-UC-3, TCC-SUP, RT112, J82, HT1376) were measured using the neutral comet assay. Radiation dose-response curves (0-60 Gy) showed damage (measured as mean tail moment) for five of the cell lines in the same rank order as cell survival (measured over 0-10 Gy), with the least damage in the most radioresistant cell line. Damage induction correlated well with clonogenic survival at high doses (SF10) for all six cell lines. At the clinically relevant dose of 2 Gy, correlation was good for four cell lines but poor for two (TCC-SUP and T24). The rejoining process had a fast and slow component for all cell lines. The rate of these two components of DNA repair did not correlate with cell survival. However, the time taken to reduce the amount of DNA damage to preirradiated control levels correlated positively with cell survival at 10 Gy but not 2 Gy; radioresistant cells rejoined the induced DSBs to preirradiation control levels more quickly than the radiosensitive cells. Although the results show good correlation between SF10 and DSBs for all six cell lines, the lack of correlation with SF2 for TCC-SUP and T24 cells would suggest that a predictive test should be carried out at the clinically relevant dose. At present the neutral comet assay cannot achieve this.

Enhancement of the anti-tumour effect of cyclophosphamide by the bioreductive drugs AQ4N and tirapazamine.

The ability of the bioreductive drugs AQ4N and tirapazamine to enhance the anti-tumour effect of cyclophosphamide was assessed in three murine tumour models. In male BDF mice implanted with the T50/80 mammary carcinoma, AQ4N (50-150 mg kg(-1)) in combination with cyclophosphamide (100 mg kg(-1)) produced an effect equivalent to a single 200 mg kg 1 dose of cyclophosphamide. Tirapazamine (25 mg kg(-1)) in combination with cyclophosphamide (100 mg kg(-1)) produced an effect equivalent to a single 150 mg kg(-1) dose of cyclophosphamide. In C3H mice implanted with the SCCVII or RIF-1 tumours, enhancement of tumour cell killing was found with both drugs in combination with cyclophosphamide (50-200 mg kg(-1)); AQ4N (50-200 mg kg(-1)) produced a more effective combination than tirapazamine (12.5-50 mg kg(-1)). Unlike tirapazamine, which showed a significant increase in toxicity to bone marrow cells, the combination of AQ4N (100 mg kg(-1)) 6 h prior to cyclophosphamide (100 mg k(-1)) resulted in no additional toxicity towards bone marrow cells compared to that caused by cyclophosphamide alone. In conclusion, AQ4N gave a superior anti-tumour effect compared to tirapazamine when administered with a single dose of cyclophosphamide (100 mg kg(-1)).