Plasma metabolic profiling reveals age-dependency of systemic effects of green tea polyphenols in mice with and without prostate cancer.

Green tea polyphenols (GTP) have been widely investigated for their potential to prevent prostate cancer. However, results from epidemiological and clinical studies are equivocal. Studies in the TRAMP (TRansgenic Adenocarcinoma of the Mouse Prostate) mouse suggest that the chemopreventive efficacy of GTP is higher in young animals with early stages of carcinogenesis than in old ones. Here, effects of GTP on prostate carcinogenesis in TRAMP mice were assessed by comparing pathological changes with (1)H-NMR metabolic profiling of plasma and extracts of prostate tissue. Mice received 0.05% GTP in their drinking water for 4 or 25 weeks after weaning. Age-matched wild-type mice were included in the study in order to establish differences in GTP effects between normal and TRAMP mice. Dietary GTP did not markedly alter prostate carcinogenesis as reflected by pathology and prostate tissue metabolic profile. However, a systemic effect of GTP consumption was observed in young mice, regardless of genotype. Plasma lipid signals were decreased in 8 week old mice which received GTP compared to age-matched controls by 19, 61, 27, 34 and 15% (p

Determination of 8-oxo-2'-deoxyguanosine and creatinine in murine and human urine by liquid chromatography/tandem mass spectrometry: application to chemoprevention studies.

Urinary 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodG) represents a non-invasive biomarker for oxidative stress and may be useful for monitoring chemotherapeutic and chemopreventive interventions associated with cancer-related alterations in oxidative stress. We describe the development and validation of two separate liquid chromatography/tandem mass spectrometry (LC/MS/MS) selected reaction monitoring (SRM) methods for the determination of 8-oxodG and creatinine in both murine and human urine using stable isotope labelled internal standards. Levels of 8-oxodG were normalised to creatinine. The LC/MS/MS methods were applied to two chemoprevention studies utilising tea polyphenols in humans and TRAMP (TRansgenic Adenocarcinoma of the Mouse Prostate) mice. Patients with benign prostatic hyperplasia received 1 g/day of green tea polyphenols (GTP), 1 g/day of black tea theaflavins (BTT) or no treatment for 4 weeks. TRAMP mice received GTP (0.05% in drinking water) for 4 or 25 weeks. Prostate pathology in TRAMP mice was not affected by GTP. Levels of 8-oxodG were not altered by tea polyphenols in either mice or humans. In TRAMP mice, urinary 8-oxodG levels were elevated with increasing age (p < 0.0001) but not changed by the presence of prostate tumours. In conclusion, the LC/MS/MS SRM methods described here are ideally suited for the accurate determination of 8-oxodG and creatinine in urine samples from both clinical and pre-clinical studies.

Simultaneous determination of 8-oxo-2'-deoxyguanosine and 8-oxo-2'-deoxyadenosine in DNA using online column-switching liquid chromatography/tandem mass spectrometry.

Sensitive and reliable methods are required for the assessment of oxidative DNA damage, which can result from reactive oxygen species that are generated endogenously from cellular metabolism and inflammatory responses, or by exposure to exogenous agents. The development of a liquid chromatography/tandem mass spectrometry (LC/MS/MS) selected reaction monitoring (SRM) method is described, that utilises online column-switching valve technology for the simultaneous determination of two DNA adduct biomarkers of oxidative stress, 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodG) and 8-oxo-7,8-dihydro-2'-deoxyadenosine (8-oxodA). To allow for the accurate quantitation of both adducts the corresponding [(15)N(5)]-labelled stable isotope internal standards were synthesised and added prior to enzymatic hydrolysis of the DNA samples to 2'-deoxynucleosides. The method required between 10 and 40 microg of hydrolysed DNA on-column for the analysis and the limit of detection for both 8-oxodG and 8-oxodA was 5 fmol. The analysis of calf thymus DNA treated in vitro with methylene blue (ranging from 5 to 200 microM) plus light showed a dose-dependent increase in the levels of both 8-oxodG and 8-oxodA. The level of 8-oxodG was on average 29.4-fold higher than that of 8-oxodA and an excellent linear correlation (r = 0.999) was observed between the two adducts. The influence of different DNA extraction procedures for 8-oxodG and 8-oxodA levels was assessed in DNA extracted from rat livers following dosing with carbon tetrachloride. The levels of 8-oxodG and 8-oxodA were on average 2.9 (p = 0.018) and 1.4 (p = 0.018) times higher, respectively, in DNA samples extracted using an anion-exchange column procedure than in samples extracted using a chaotropic procedure, implying artefactual generation of the two adducts. In conclusion, the online column-switching LC/MS/MS SRM method provides the advantages of increased sample throughput with reduced matrix effects and concomitant ionisation suppression, making the method ideally suited when used in conjunction with chaotropic DNA extraction for the determination of oxidative DNA damage.

Metabolic profiling of transgenic adenocarcinoma of mouse prostate (TRAMP) tissue by 1H-NMR analysis: evidence for unusual phospholipid metabolism.

BACKGROUND: The TRansgenic Adenocarcinoma of the Mouse Prostate (TRAMP) mouse model has frequently been used in preclinical studies with chemotherapeutic/chemopreventive rationales. Here the hypothesis was tested using (1)H-NMR-based metabolic profiling that the TRAMP tumor metabolic phenotype resembles that reported for human prostate cancer. METHODS: Aqueous extracts or intact tissues of normal prostate from 8- ("young") or 28-("old") week-old C57BL/6J wild-type mice or of prostate tumor from age-matched TRAMP mice were analyzed by (1)H-NMR. Results were compared with immunohistochemical findings. Expression of choline kinase was studied at the protein and mRNA levels. RESULTS: In young TRAMP mice presenting with zonal hyperplasia, the ratio of glycerophosphocholine (GPC) to phosphocholine (PC) was 22% below that in wild-type mice (P < 0.05). In old TRAMP mice with well-defined malignancy, reduced tumor levels of citrate (49%), choline (33%), PC (57%), GPC (66%), and glycerophosphoinositol (61%) were observed relative to normal prostate (P < 0.05). Hierarchical cluster analysis of metabolite levels distinguished between normal and malignant tissue in old but not young mice. While the reduction in tissue citrate resembles human prostate cancer, low levels of choline species in TRAMP tumors suggest atypical phospholipid metabolism as compared to human prostate cancer. TRAMP tumor and normal prostate tissues did not differ in expression of choline kinase, which is overexpressed in human prostate cancer. CONCLUSION: Although prostate cancer in TRAMP mice shares some metabolic features with that in humans, it differs with respect to choline phospholipid metabolism, which could impact upon the interpretation of results from biomarker or chemotherapy/chemoprevention studies.

Evaluation of the cancer chemopreventive efficacy of silibinin in genetic mouse models of prostate and intestinal carcinogenesis: relationship with silibinin levels.

Silibinin, a flavonolignan from milk thistle seeds, possesses cancer chemopreventive properties in rodent models of carcinogenesis. We tested the hypotheses that silibinin or silipide, silibinin formulated with phospholipids, delays tumour development in TRAMP or Apc(Min) mice, genetic models of prostate or intestinal malignancies, respectively. Mice received silibinin or silipide with their diet (0.2% silibinin equivalents) from weaning. Intervention with silipide reduced the size of well differentiated TRAMP adenocarcinomas by 31%. Silipide and silibinin decreased the incidence of poorly differentiated carcinomas by 61% compared to mice on control diet. Silipide decreased plasma levels of insulin-like growth factor (IGF)-1 by 36%. Levels of circulating IGF binding protein (IGFBP)-3 in mice on silipide or silibinin were 3.9- or 5.9-fold, respectively, elevated over those in control TRAMP mice. In Apc(Min) mice silibinin, but not silipide, had only a marginal adenoma number-reducing effect. The results cautiously support the advancement of silipide to the stage of clinical investigation in prostate cancer.

Consumption of silibinin, a flavonolignan from milk thistle, and mammary cancer development in the C3(1) SV40 T,t antigen transgenic multiple mammary adenocarcinoma (TAg) mouse.

Silibinin is a flavonolignan extracted from milk thistle with cancer chemopreventive activity in preclinical models of prostate and colorectal cancer. A milk thistle extract, of which silibin is a major component, has recently been shown to exacerbate mammary carcinogenesis in two rodent models. We tested the hypothesis that consumption of silibinin or silipide, a silibinin formulation with pharmaceutical properties superior to the unformulated agent, affect breast cancer development in the C3(1) SV40 T,t antigen transgenic multiple mammary adenocarcinoma mouse model. Mice received silibinin or silipide (0.2% silibinin equivalents) with their diet from weaning, and tumour development was monitored by weekly palpation and the number and weight of neoplasms at the end of the experiment. Intervention neither promoted, nor interfered with, tumour development. The result suggests that promotion of carcinogenesis is not a feature of silibinin consistent across rodent models of mammary carcinogenesis.

Androgen manipulation alters oxidative DNA adduct levels in androgen-sensitive prostate cancer cells grown in vitro and in vivo.

Intracellular reactive oxygen species (ROS) may cause oxidative DNA damage, resulting in the formation of adducts such as 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxo-dG) and the cyclic pyrimidopurinone N-1, N(2) malondialdehyde-2'-deoxyguanosine (M(1)dG). These adducts have been associated with carcinogenesis, genomic instability and clonal evolution. We tested two hypotheses in human prostate cancer cells grown in vitro and in a xenograft model: (1) treatment of androgen-sensitive cells with DHT increases levels of oxidative DNA adduct levels; (2) flutamide, a competitive androgen receptor antagonist, prevents DHT-induced changes. Levels of M(1)dG and 8-oxo-dG adducts were determined by immunoslot blot and liquid chromatography-tandem mass spectrometry. M(1)dG and 8-oxo-dG levels were significantly higher than control levels in LNCaP cells exposed to supra-physiological concentrations (25-100 nM) of DHT (both P<0.05 by ANOVA). Flutamide pre-treatment completely prevented this increase. In the xenograft model, tumour levels of M(1)dG were decreased by 46% (P=0.001 by Mann-Whitney Test) in flutamide-treated animals compared to controls. The changes demonstrated suggest that oxidative DNA adducts may serve as biomarkers of the efficacy of androgen manipulation in chemoprevention trials.

Putative cancer chemopreventive agents of dietary origin-how safe are they?

As cancer chemopreventive agents are intended for use by healthy individuals as prophylactics to prevent or retard the development of cancer, they must be amenable to ingestion over prolonged periods without toxicity. Therefore, putative chemopreventive agents need to undergo stringent testing to ensure their safety with regard to chronic exposure in humans. The diet is thought to be a source of chemopreventive agents, and dietary compounds are generally considered to be of low hazard, albeit this notion has not often been put to the test. Here the safety information available for 5 dietary putative chemopreventive compounds, indole-3-carbinol (I3C), curcumin, quercetin, epigallocatechin gallate (EGCG), and capsaicin is reviewed. For these agents, normal dietary intake, doses used in clinical trials, efficacious doses in rodents, and where available, toxic doses are compared. For curcumin, quercetin and capsaicin, toxicological data is only available from studies in rodents. Information on long-term effects in animals beyond 28 or 90 days is lacking for EGCG. Capsaicin and quercetin are suspected carcinogens. I3C and quercetin can modulate the absorption of other drugs given concomitantly. Without further investigation of their toxicology, it is difficult to recommend any of these agents for long-term use in the healthy population.

Cytotoxic action of phorbol esters on human pancreatic cancer cells.

We previously showed that phorbol esters are cytotoxic to human thyroid epithelial cells expressing a mutant RAS oncogene. Here we explore the generality of this finding using cells derived from pancreatic cancer, which, like thyroid, shows a high frequency of RAS mutation, but is a much greater cause of cancer mortality. The response to phorbol myristate acetate (PMA) and related agents was assessed on a panel of 9 pancreatic cancer cell lines, using a range of assays for cell growth and death in vitro and in vivo. In most lines, PMA induced non-apoptotic cell death which was, surprisingly, independent of its "classic" target, protein kinase C. With 24 hr exposure, the IC(50) in the most sensitive line (Aspc-1) was <1 ng/ml (1.6 nM), with survival undetectable at concentrations >/=>/=16 nM, and after only 1 hr exposure the IC(50) was still

Comparison of systemic availability of curcumin with that of curcumin formulated with phosphatidylcholine.

PURPOSE: Curcumin, a major constituent of the spice turmeric, suppresses expression of the enzyme cyclooxygenase 2 (Cox-2) and has cancer chemopreventive properties in rodents. It possesses poor systemic availability. We explored whether formulation with phosphatidylcholine increases the oral bioavailability or affects the metabolite profile of curcumin. METHODS: Male Wistar rats received 340 mg/kg of either unformulated curcumin or curcumin formulated with phosphatidylcholine (Meriva) by oral gavage. Rats were killed at 15, 30, 60 and 120 min post administration. Plasma, intestinal mucosa and liver were analysed for the presence of curcumin and metabolites using HPLC with UV detection. Identity of curcumin and metabolites was verified by negative ion electrospray liquid chromatography/tandem mass spectrometry. RESULTS: Curcumin, the accompanying curcuminoids desmethoxycurcumin and bisdesmethoxycurcumin, and the metabolites tetrahydrocurcumin, hexahydrocurcumin, curcumin glucuronide and curcumin sulfate were identified in plasma, intestinal mucosa and liver of rats which had received Meriva. Peak plasma levels and area under the plasma concentration time curve (AUC) values for parent curcumin after administration of Meriva were fivefold higher than the equivalent values seen after unformulated curcumin. Similarly, liver levels of curcumin were higher after administration of Meriva as compared to unformulated curcumin. In contrast, curcumin concentrations in the gastrointestinal mucosa after ingestion of Meriva were somewhat lower than those observed after administration of unformulated curcumin. Similar observations were made for curcumin metabolites as for parent compound. CONCLUSION: The results suggest that curcumin formulated with phosphatidylcholine furnishes higher systemic levels of parent agent than unformulated curcumin.

Preliminary safety evaluation of the putative cancer chemopreventive agent tricin, a naturally occurring flavone.

PURPOSE: Naturally occurring flavonoids such as quercetin and genistein possess cancer chemopreventive properties in experimental models. However, adverse effects such as their mutagenicity confound their potential clinical usefulness. Furthermore in leukaemia cells some flavonoids cleave the breakpoint cluster region of the mixed lineage leukaemia (MLL) gene as a consequence of inhibition of topoisomerase II. The choice of flavonoids to be developed as cancer chemopreventive agents depends crucially on their safety. Here, we explored safety aspects of the novel flavone tricin, a constituent of rice bran and other grass species, which has recently been found to interfere with murine gastrointestinal carcinogenesis. METHODS: Evidence of pathological or morphological changes in liver, lung, heart, spleen, kidney, adrenal gland, pancreas or thymus tissues was studied in mice which received tricin, genistein or quercetin 1,000 mg/kg daily by the oral route on five consecutive days. The ability of tricin (50 microM) to cleave the MLL gene was studied in human leukaemia cells by Southern blotting, and its effect on human topoisomerase II activity was investigated in incubations with supercoiled DNA. The mutagenicity of tricin was assessed in the Salmonella/Escherichia coli assay, and its clastogenicity was adjudged by chromosomal aberrations in Chinese hamster ovary cells and occurrence of micronuclei in bone marrow erythrocytes in Swiss-Webster mice. RESULTS: Neither tricin, quercetin, or genistein caused pathological or morphological changes in any of the murine tissues studied. Tricin (50 microM) failed to cause MLL gene breakage, and it inhibited topoisomerase II only at 500 microM, but not at 10, 50 or 100 microM. Tricin lacked genotoxic properties in the systems studied here. CONCLUSION: The results tentatively suggest that tricin may be considered safe enough for clinical development as a cancer chemopreventive agent.

A synthetic approach to the generation of quercetin sulfates and the detection of quercetin 3'-O-sulfate as a urinary metabolite in the rat.

To study the biological effects of quercetin, authentic products of quercetin metabolism are required as standards. The synthesis of quercetin sulfate standards is thus described. Quercetin was reacted with a 10-fold molar excess of sulfur trioxide-N-triethylamine, and the products were analyzed by HPLC and mass spectrometry. Four monosulfates and three disulfates were identified, and structural inferences were drawn by 1H NMR spectrometry of HPLC peak isolates. Analysis of the urine of rats that had received quercetin (1.9 g/kg po) yielded a single peak, which by comparison with the products of the reaction between quercetin and sulfur trioxide-N-triethylamine was identified as quercetin 3'-O-sulfate.

Dietary agent indole-3-carbinol protects female rats against the hepatotoxicity of the antitumor drug ET-743 (trabectidin) without compromising efficacy in a rat mammary carcinoma.

ET-743, an experimental antitumor drug with promising activity in sarcoma, breast and ovarian carcinoma, is currently under phase 2 clinical evaluation. It is hepatotoxic in animals and patients. We tested the hypothesis that indole-3-carbinol (I3C), the hydrolysis product of glucosinolates occurring in cruciferous vegetables, may protect against ET-743-induced hepatotoxicity in the female Wistar rat, the animal species with the highest sensitivity toward the adverse hepatic effect of this drug. Hepatotoxicity was adjudged by measurement of plasma levels of bilirubin, alkaline phosphatase (ALP) and aspartate aminotransferase (AST) and by liver histopathology. The effect of I3C on the kinetics of ET-743 in rat plasma and liver was investigated by high-pressure liquid chromatography. The effect of I3C on the antitumor efficacy of ET-743 was explored in rats bearing the 13762 mammary carcinoma. ET-743 (40 microg/kg i.v.) alone caused an elevation of plasma bilirubin, ALP and AST levels and degeneration and patchy focal necrosis of bile duct epithelial cells. Addition of I3C to the diet (0.5%) for 6 days prior to ET-743 administration almost completely abolished manifestations of hepatotoxicity. In contrast, a dietary concentration of 0.1% I3C did not protect, nor did dietary diindolylmethane (0.2%), an acid-catalyzed condensation product of I3C. Ingestion by rats of I3C for 6 days prior to ET-743 (40 microg/kg i.v.) decreased plasma but not hepatic concentrations of ET-743 compared to animals that received ET-743 alone. I3C did not interfere with the antitumor efficacy of ET-743. The results suggest that ingestion of I3C may counteract the unwanted effect of ET-743 in the liver. I3C should be investigated as a hepatoprotectant in patients who receive ET-743 therapy.

Pharmacokinetics and tissue disposition of indole-3-carbinol and its acid condensation products after oral administration to mice.

Indole-3-carbinol (I3C) and 3,3'-diindolylmethane (DIM) are promising cancer chemopreventive agents in rodent models, but there is a paucity of data on their pharmacokinetics and tissue disposition. The disposition of I3C and its acid condensation products, DIM, [2-(indol-3-ylmethyl)-indol-3-yl]indol-3-ylmethane (LTr(1)), indolo[3,2b]carbazole (ICZ) and 1-(3-hydroxymethyl)-indolyl-3-indolylmethane (HI-IM) was studied, after oral administration of I3C (250 mg/kg) to female CD-1 mice. Blood, liver, kidney, lung, heart, and brain were collected between 0.25 and 24 h after administration and the plasma and tissue concentrations of I3C and its derivatives determined by high-performance liquid chromotography. I3C was rapidly absorbed, distributed, and eliminated from plasma and tissues, falling below the limit of detection by 1 h. Highest concentrations of I3C were detected in the liver where levels were approximately 6-fold higher than those in the plasma. Levels of DIM, LTr(1), and HI-IM were much lower, although they persisted in plasma and tissues for considerably longer. DIM and HI-IM were still present in the liver 24 h after I3C administration. Tissue levels of DIM and LTr(1) were found to be in equilibrium with plasma at almost every time point measured. In addition to acid condensation products of I3C, a major oxidative metabolite (indole-3-carboxylic acid) and a minor oxidative metabolite (indole-3-carboxaldehyde) were detected in plasma of mice after oral administration of I3C. ICZ was also tentatively identified in the liver of these mice. This study shows for the first time that, after oral administration to mice, I3C, in addition to its acid condensation products, is absorbed from the gut and distributed systemically into a number of well-perfused tissues, thus allowing the possibility for some pharmacological activity of the parent compound in vivo.

Articular chondromatosis and chrondroid metaplasia in transgenic TAg mice.

The C3(1)/SV40 T antigen transgenic mouse model for which rapid mammary and prostate tumor development has been documented uses the FVB/N mouse as a background strain. In this study, where the background strain used was the C57BL/6J mouse, neither mammary nor prostate tumors developed over periods of up to 40 weeks. However, a disturbance of hyaline cartilage in joints was observed similar to that found in synovial chondromatosis in humans. In addition, cartilage thickening in the external ears and cartilaginous metaplasia of the ascending aorta also occurred. This suggests that rearrangement of the transgene occurred in breeding on the C57BL background, thus modifying its expression. It raises the possibility that the genetic changes induced by the SV40 T antigen transforming sequence are important in cartilage homeostasis.

Comparison of four modulators of drug metabolism as protectants against the hepatotoxicity of the novel antitumor drug yondelis (ET-743) in the female rat and in hepatocytes in vitro.

PURPOSE: Yondelis (ET-743), a tetrahydroisoquinoline alkaloid isolated from a marine tunicate, is a novel drug with demonstrated anticancer activity in early clinical trials against sarcoma, breast and ovarian carcinoma. Yondelis has myelotoxic and hepatotoxic side effects, the latter reflected by reversible transaminitis and cholangitis. In the female rat pretreatment with high-dose dexamethasone has been shown to abrogate yondelis-mediated hepatotoxicity, an effect tentatively linked to its ability to induce cytochrome P450 CYP3A isoenzymes, which metabolize yondelis. Here we tested the hypothesis that pretreatment of rats with modulators of hepatic drug metabolism, beta-naphthoflavone, phenobarbitone or N-acetylcysteine, protect rat livers against the effects of yondelis. METHODS: Female rats received yondelis (40 microg/kg intravenously) and liver damage in vivo was assessed in terms of changes in plasma levels of bilirubin, alkaline phosphatase (ALP) and aspartate aminotransferase (AST) and by histopathology. In order to investigate yondelis toxicity in vitro, hepatocytes isolated from untreated rats or from rats pretreated with dexamethasone, beta-naphthoflavone or phenobarbitone were maintained in culture and exposed to yondelis. RESULTS: Pretreatment with beta-naphthoflavone and phenobarbitone ameliorated yondelis-mediated hepatotoxicity in vivo. The former abrogated plasma indicators on day 3, but hardly on day 6, and the latter suppressed elevation of bilirubin, but not of ALP or AST. Pretreatment with N-acetylcysteine did not protect from, but slightly exacerbated, yondelis-induced liver changes. Hepatocytes from naive animals or from pretreated rats did not differ in their susceptibility towards yondelis-induced cytotoxicity in vitro. Nor did inclusion of N-acetylcysteine (1 m M) in the cellular incubation medium affect yondelis-induced hepatocytotoxicity. CONCLUSIONS: The results suggest that certain inducers of cytochrome P450 enzymes such as dexamethasone and beta-naphthoflavone can protect rat liver against the unwanted effects of yondelis, but such protection cannot be mimicked in in vitro experiments using liver cells in culture.

Determination of the flavone tricin in human plasma by high-performance liquid chromatography.

Tricin is a flavone constituent of brown rice and rice bran, which interferes potently with the survival of human-derived breast and colon cancer cells in vitro. A specific and simple high-performance liquid chromatographic (HPLC) method was developed for the determination of tricin in human plasma with UV-visible detection. HPLC separation on Hypersil-BDS C(18) (4.6 x 250 mm) was carried out with an isocratic mobile phase of 52% methanol in 0.1 m ammonium acetate, pH 5.10, containing 0.27 mm disodium ethylenediamine tetraacetic acid and detection at 355 nm. The retention times of tricin and quercetin (internal standard) were 14.2 and 7.8 min, respectively. The assay was linear in the range 1-100 microg/mL (r(2 ) > or = 0.995). Tricin in plasma was efficiently extracted with 0.1 m acetic acid in acetone, and the recoveries were in the range 92.6-102.8% (n = 6) with relative standard deviation below 10% for three concentrations of tricin, 5, 10 and 100 microg/mL. The lower limit of quantitation (relative standard deviation <20%) was 1 microg/mL.

Complete protection by high-dose dexamethasone against the hepatotoxicity of the novel antitumor drug yondelis (ET-743) in the rat.

Yondelis (ET-743) is a promising antitumor drug with hepatotoxic properties in animals and humans. Here the hypothesis was tested that dexamethasone can ameliorate manifestations of yondelis-induced hepatotoxicity in the female Wistar rat, which is the animal species with the highest sensitivity toward the adverse hepatic effect of yondelis. Hepatotoxicity was adjudged by measurement of plasma levels of alkaline phosphatase, aspartate aminotransferase, and bilirubin, and by liver histopathology. Yondelis (40 micro g/kg i.v.) alone caused a dramatic elevation of plasma alkaline phosphatase, aspartate aminotransferase, and bilirubin levels, and degeneration and patchy focal necrosis of bile duct epithelial cells. Pretreatment of rats with dexamethasone (5-20 mg/kg, p.o.) 24 h before yondelis ameliorated or abrogated the biochemical and histopathological manifestations of yondelis-induced liver changes. In contrast, when dexamethasone was administered simultaneously with yondelis, its toxicity was not reduced. Pretreatment with dexamethasone (10 mg/kg) also reversed the gene expression changes induced by yondelis in rat liver. However, dexamethasone pretreatment did not interfere with the antitumor efficacy of yondelis in rats bearing the 13762 mammary carcinoma or in four murine models. Dexamethasone (10 mg/kg) administered 24 h before yondelis decreased hepatic levels of yondelis dramatically compared with those obtained after administration of yondelis alone, whereas yondelis plasma levels after the drug combination were not markedly different from those in rats on yondelis alone. The results suggest that pretreatment with high-dose dexamethasone effectively protects rats against yondelis-mediated hepatic damage by decreasing hepatic exposure to yondelis, perhaps linked to induction of metabolism by cytochrome P450 enzymes. Pretreatment with high-dose dexamethasone should be investigated in patients who receive yondelis to ameliorate its unwanted effect on the liver.

Hepatobiliary damage and changes in hepatic gene expression caused by the antitumor drug ecteinascidin-743 (ET-743) in the female rat.

Ecteinascidin-743 (ET-743) is a novel marine-derived anticancer drug with clinical activity in soft tissue sarcoma and ovarian cancer. Reversible transaminitis and subclinical cholangitis have frequently been described in patients who receive ET-743. To facilitate understanding of this adverse effect and help design suitable therapeutic rescue strategies, we characterized the hepatic effects of ET-743 in rats. Female rats received ET-743 (single dose, 40 microg/kg) i.v., and liver changes were assessed from 6 h up to 3 months after dosing by histopathology, immunohistochemistry, electron microscopy, hepatic and plasma biochemistry, and DNA microarray analysis. At 24 h posttreatment and beyond, livers displayed degeneration and patchy focal necrosis of bile duct epithelial cells associated with mild inflammation followed by fibrosis. Sporadic and focal zones of hepatic necrosis and hemorrhage were observed from day 2 onward, although the majority of hepatocytes appeared normal as judged by electron microscopy. Pathological alterations persisted up to 3 months after dosing. Plasma levels of total bilirubin were elevated up to 7-fold over those in untreated rats from day 2 onward and returned to control values by day 24. Activities of alkaline phosphatase and aspartate aminotransferase in plasma were elevated for 2 and 3 months, respectively. Activities of the hepatic microsomal drug-metabolizing enzymes cytochrome P-450 A1/2, CYP2E1, and CYP3A2 were decreased. DNA microarray analysis of livers from ET-743-treated animals showed a dramatic increase in the expression of ATP binding cassette transport genes Abcb1a and Abcb1b, which impart resistance to anticancer drugs, and of Cdc2a and Ccnd1, the rodent homologues of human cell cycle genes CDC2 and cyclin D1, respectively. The cell cycle gene expression changes mirrored ET-743-induced increases in liver weight and Ki-67 labeling of liver nuclei. The results suggest that the toxicity exerted by ET-743 in the rat liver is a consequence of biliary rather than hepatocellular damage and that it is accompanied by a wave of mitogenic activity, which may be driven by the transcriptional increase in Cdc2a expression.