Advertisement
Cancer Therapy: Preclinical

Pharmacokinetics, Metabolism, and Oral Bioavailability of the DNA Methyltransferase Inhibitor 5-Fluoro-2′-Deoxycytidine in Mice

Jan H. Beumer, Julie L. Eiseman, Robert A. Parise, Erin Joseph, Julianne L. Holleran, Joseph M. Covey and Merrill J. Egorin
DOI: 10.1158/1078-0432.CCR-06-1250 Published December 2006

Abstract

Purpose:In vivo, 5-fluoro-2′-deoxycytidine (FdCyd) is rapidly and sequentially converted to 5-fluoro-2′-deoxyuridine, 5-fluorouracil, and 5-fluorouridine. The i.v. combination of FdCyd and 3,4,5,6-tetrahydrouridine (THU), a cytidine deaminase (CD) inhibitor that blocks the first metabolic step in FdCyd catabolism, is being investigated clinically for its ability to inhibit DNA methyltransferase. However, the full effects of THU on FdCyd metabolism and pharmacokinetics are unknown. We aimed to characterize the pharmacokinetics, metabolism, and bioavailability of FdCyd with and without THU in mice.

Experimental Design: We developed a sensitive high-performance liquid chromatography tandem mass spectrometry assay to quantitate FdCyd and metabolites in mouse plasma. Mice were dosed i.v. or p.o. with 25 mg/kg FdCyd with or without coadministration of 100 mg/kg THU p.o. or i.v.

Results: The oral bioavailability of FdCyd alone was ∼4%. Coadministration with THU increased exposure to FdCyd and decreased exposure to its metabolites; i.v. and p.o. coadministration of THU increased exposure to p.o. FdCyd by 87- and 58-fold, respectively. FdCyd exposure after p.o. FdCyd with p.o. THU was as much as 54% that of i.v. FdCyd with i.v. THU.

Conclusions: FdCyd is well absorbed but undergoes substantial first-pass catabolism by CD to potentially toxic metabolites that do not inhibit DNA methyltransferase. THU is sufficiently bioavailable to reduce the first-pass effect of CD on FdCyd. Oral coadministration of THU and FdCyd is a promising approach that warrants clinical testing because it may allow maintaining effective FdCyd concentrations on a chronic basis, which would be an advantage over other DNA methyltransferase inhibitors that are currently approved or in development.

  • metabolism
  • murine
  • disposition
  • bioavailability
  • pyrimidine

Epigenetic changes, such as aberrant promoter hypermethylation associated with inappropriate gene silencing, affect virtually every step in tumor progression (1). Hypermethylation takes place at the cytidine residues of CpG islands in the promoter region of genes. DNA methyltransferases, the class of enzymes that methylate 2′-deoxycytidine (dCyd) residues in DNA and thereby epigenetically silence tumor suppressor genes (1–3), are a target for two recently licensed drugs (5-azacytidine and 5-aza-2′-deoxycytidine) and a number of anticancer drugs under development. Many of these DNA methylation inhibitors are nucleoside analogues with modifications at the 5-position. After incorporation into DNA, these compounds covalently bind to, and thereby deplete, DNA methyltransferases, which allows reexpression of tumor suppressor genes (4) and, possibly, contributes to reversal of the neoplastic phenotype. Unfortunately, all of the current nucleoside DNA methyltransferase inhibitors require parenteral administration (5) because their chemical and pharmacokinetic properties preclude oral administration, which could produce sustained, adequate plasma concentrations and permit chronic use for durations that are impractical with i.v. formulations.

5-Fluoro-2′-deoxycytidine (FdCyd; NSC 48006) is a fluoropyrimidine nucleoside analogue that was previously investigated as a tumor-selective prodrug of the potent thymidylate synthase (EC 2.1.1.45)5 inhibitor 5-fluoro-2′-dUMP (FdUMP; refs. 6–8). FdUMP can be produced from FdCyd via two pathways (Fig. 1 ). The first is deamination of FdCyd to 5-fluoro-2′-deoxyuridine (FdUrd) by cytidine deaminase (CD; EC 3.5.4.5) followed by phosphorylation to FdUMP. However, FdUrd is also subjected to rapid conversion to 5-fluorouracil (5-FU) and subsequently 5-fluoro-5,6-dihydrouracil (FUH2), which is inactive. The second pathway is phosphorylation of FdCyd by deoxycytidine kinase (EC 2.7.1.74) to 5-fluoro-2′-dCMP (5-fluoro-2′-deoxycytidylate) followed by deamination by cytidylate deaminase (EC 3.5.4.12) to FdUMP. Blocking the first route with the CD inhibitor 3,4,5,6-tetrahydrouridine (THU) may increase selectivity for tumor cells, many of which are known to overexpress the enzymes involved in the second pathway (6, 8–10). The spectrum of cytotoxic activity and in vivo efficacy of FdCyd is, with few exceptions, similar to that of FdUrd and 5-FU, to which FdUrd is converted by uridine/thymidine phosphorylase (EC 2.4.2.3/EC 2.4.2.4; refs. 2, 11).

Fig. 1.
Fig. 1.

Metabolic scheme of FdCyd with its possible mechanisms of action. Compounds involved are FdCyd, FdUrd, 5-FC (FC), 5-FU (FU), 5-fluoro-5,6-dihydrouracil (FUH2), FdCMP (5-fluoro-2′-deoxycytidylate), 5-fluoro-2′-deoxycytidine-triphosphate (FdCTP), FdUMP (5-fluoro-2′-deoxyuridylate), 5-fluoro-2′-deoxyuridinetriphosphate (FdUTP), FUrd, and 5-fluorouridine-triphosphate (FUTP). THU blocks the conversion of FdCyd to FdUrd by CD.

FdCyd is not only a prodrug for FdUrd but also has other activities, as was shown by the activity of FdCyd against FdUrd-resistant cell lines (2, 12). An FdCyd-specific mechanism of action is its ability (after incorporation of FdCyd triphosphate into DNA) to inhibit DNA methyltransferase. Treatment of MCF-7 cells with 1 μmol/L FdCyd for 24 h resulted in inhibition of DNA methylation, as reflected by a 35% decrease of the methyl-dCyd/dCyd ratio (12). In that regard, the inhibitory potency of FdCyd on DNA methylation is comparable with that of 5-aza-2′-deoxycytidine, a DNA methyltransferase inhibitor recently approved for myelodysplastic syndromes (12). Therefore, FdCyd is currently being investigated clinically for its ability to inhibit DNA methyltransferases (3).

Because a number of the metabolites of FdCyd have pharmacologic activity (Fig. 1), but only FdCyd inhibits DNA methyltransferase, the cytotoxic events associated with FdCyd administration may be mediated by a combination of (a) incorporation of FdCyd into DNA, (b) inhibition of DNA methylation by FdCyd, (c) inhibition of thymidylate synthase by FdUMP, (d) incorporation of FdUrd and dUrd into DNA, and (e) incorporation of 5-fluorouridine (FUrd) into RNA (12). Coadministration of THU could delay and diminish conversion of FdCyd to potentially toxic metabolites that do not inhibit DNA methyltransferase (9, 13).

Although the combination of i.v. FdCyd and i.v. THU is currently being investigated clinically (3), the precise effects of THU on the metabolism and pharmacokinetics of FdCyd are still unknown. Furthermore, the oral bioavailability of pyrimidine nucleosides, in general, is known to be poor (14–18), which suggests poor oral bioavailability for FdCyd. Thus, clinical investigations to date have given nucleosides by i.v. or s.c. routes. This is not only inconvenient but may also limit the ability to maintain adequate drug concentrations on a chronic basis. The importance of chronic dosing is evident from preclinical studies that show the necessity of long-term exposure to DNA methylation inhibitors to prevent re-methylation and re-silencing of genes (19, 20). Until the present study, the ability of THU to augment the oral bioavailability of FdCyd had not been shown.

The current preclinical investigation was designed to support the clinical development of FdCyd as a DNA methyltransferase inhibitor by characterizing the pharmacokinetics, metabolism, and bioavailability of FdCyd. To this end, a hydrophilic-interaction high-performance liquid chromatography tandem mass spectrometry (HPLC-MS/MS) assay was developed and used to quantitate concentrations of FdCyd and its metabolites FdUrd, 5-FU, FUrd, and 5-fluorocytosine (5-FC) in plasma and urine of mice given FdCyd i.v. or p.o., alone and in combination with i.v. or p.o. THU.

Materials and Methods

Chemicals and reagents

FdCyd, FdUrd, 5-FU, FUrd, and THU were provided by the Developmental Therapeutics Program, National Cancer Institute (Bethesda, MD). 5-FC, 2-pyrimidinone, and formic acid were obtained from Sigma-Aldrich (St. Louis, MO). Acetonitrile, ethyl acetate, glacial acetic acid, sodium phosphate dibasic, and potassium phosphate monobasic were obtained from Fisher Chemicals (Fair Lawn, NJ). All reagents were of analytic grade. Water was purified using a Q-gard 1 Gradient Milli-Q system (18.2 MΩ cm; Millipore, Billerica, MA). Murine plasma for preparation of the calibration samples was obtained from control CD2F1 male mice.

Animals

Specific-pathogen-free, adult CD2F1 male mice were purchased from Taconic (Germantown, NY). Mice were allowed to acclimate to the University of Pittsburgh Cancer Institute Animal Facility for 1 week before the start of each study. To minimize infection, mice were maintained in micro-isolator cages in a separate room and handled in accordance with the Guide for the Care and Use of Laboratory Animals (National Research Council, 1996) and on a protocol approved by the Institutional Animal Care and Use Committee of the University of Pittsburgh. Ventilation and airflow were set to 12 changes per hour. Room temperatures were regulated at 22 ± 1°C, and the rooms were kept on automatic 12-h light/dark cycles. Mice received Prolab ISOPRO RMH 3000 Irradiated Lab Diet (PMI Nutrition International, St. Louis, MO) and water ad libitum, except on the evening before dosing, when all food was removed. Mice were 6 to 8 weeks old at the time of dosing. Sentinel animals were maintained in 1:5 dirty bedding in the same housing as the study mice and assayed at 3-month intervals for specific murine pathogens by mouse antibody profile testing (Charles River, Boston, MA). Sentinel animals remained free of specific pathogens, indicating that the study mice were pathogen-free.

Pharmacokinetics and metabolism of FdCyd in mice

To investigate the plasma disposition of FdCyd, mice of ∼20 g body weight were dosed with 25 mg/kg FdCyd (∼0.5 mg per mouse). FdCyd was given alone i.v. or p.o. In addition, FdCyd pharmacokinetic studies were done with coadministration of 100 mg/kg THU (∼2 mg per mouse). This resulted in the following five dosing schedules: (a) FdCyd i.v. alone, (b) FdCyd p.o. alone, (c) FdCyd i.v. and THU i.v., (d) FdCyd p.o. and THU p.o., and (e) FdCyd p.o. and THU i.v. The vehicle consisted of 5% dextrose (Baxter, Deerfield, IL). If FdCyd and THU were dosed by the same route, they were formulated together.

Mice were dosed (0.01 mL/g fasted body weight) by lateral tail vein injection or by oral gavage. Mice (three per time point) were euthanized with CO2 at 5, 10, 15, 30, 45, 60, 75, 90, 105, 120, 180, 240, 960, and 1,440 min after dosing. Blood was collected by cardiac puncture into heparinized syringes [heparin (Baxter) containing 1 mg/mL THU to prevent ex vivo FdCyd deamination] and centrifuged for 4 min at 13,000 × g to obtain plasma. The aspirated plasma was stored at −70°C until analysis. FdCyd, FdUrd, 5-FU, FUrd, and 5-FC in plasma were quantitated with a hydrophilic-interaction chromatography (21) HPLC-MS/MS assay (see below).

To assess urinary excretion of FdCyd and its metabolites after administration of FdCyd in the various regimens, mice scheduled for euthanasia at 16 and 24 h after dosing were housed in metabolic cages. Urine was collected on ice, separately from feces. Quantitation of FdCyd and its metabolites in urine was accomplished by preparing a 10-fold, or higher, dilution of urine samples in plasma and analyzing those diluted samples with the hydrophilic-interaction chromatography HPLC-MS/MS assay used for plasma samples.

In vitro incubations of FdCyd with kidney, liver, and blood preparations

After untreated male CD2F1 mice were euthanized, their kidneys and livers were rapidly dissected, snap-frozen in liquid nitrogen, and stored at −80°C. After thawing on ice, organs were manually homogenized in a glass Potter-Elvehjem homogenizer with 4 parts (v/w) of 0.1 mol/L phosphate buffer (pH 7.4). This homogenate was centrifuged at 9,000 × g, 4°C for 20 min. The supernatant, containing microsomes and cytosol, was collected and used for incubation experiments. On ice, 12 μL of a 1 mmol/L FdCyd solution were added to 200 μL of supernatant (in triplicate) to produce a final FdCyd concentration of ∼60 μmol/L. This mixture was incubated at 37°C for 0, 5, 10, 15, or 30 min in a shaking water bath. The incubation was stopped by putting the solutions on ice, adding 5 μL of 10 mg/mL THU, vortexing, and snap-freezing in a dry ice/methanol bath. Samples were stored at −20°C until analysis. After thawing on ice, 5 μL of each sample were diluted 20-fold with murine plasma and analyzed with the hydrophilic-interaction chromatography HPLC-MS/MS assay described below. Protein concentrations in the 9,000 × g supernatants were quantitated using the Bio-Rad Protein Assay (Bio-Rad, Hercules, CA) with bovine serum albumin as the standard.

Whole-blood incubations were done with 300-μL aliquots of fresh, heparinized mouse blood that contained ∼60 μmol/L FdCyd. Aliquots (in triplicate) were incubated at 37°C for 1, 5, 15, 30, 45, 60, 90, or 120 min in a shaking water bath. The incubation was stopped by putting the samples on ice, adding 5 μL of 10 mg/mL THU, and vortexing. After centrifugation at 13,000 × g for 5 min, the resulting supernatant plasma layer was stored at −20°C until analysis.

Quantitation of FdCyd and its metabolites

Sample preparation. Initial sample preparation was done on ice. To 100 μL of plasma, or plasma-diluted sample, were added 20 μL of internal standard (5 μg/mL of 2-pyrimidinone in water) and 10 μL of glacial acetic acid, after which the mixture was vortexed for 20 sec. A liquid-liquid extraction was done by adding 1 mL of ethyl acetate and vortexing the mixture for 5 min. The vortexed samples were centrifuged for 4 min at 10,000 × g, after which each supernatant layer was aspirated and evaporated to dryness under nitrogen at 37°C. The dried residues were reconstituted in 100 μL of acetonitrile, with sonication for 5 min, and transferred to autosampler vials.

Chromatography. The HPLC system consisted of an Agilent (Palo Alto, CA) 1100 autosampler and binary pump, two Shodex Asahipak NH2-50 2D columns (5 μm, 150 mm × 2 mm inner diameter; Phenomenex, Torrance, CA) in series and a gradient mobile phase. Mobile-phase solvent A consisted of 0.1% (v/v) formic acid in acetonitrile, and mobile-phase solvent B consisted of 0.1% (v/v) formic acid in water. This resulted in a separation system based in hydrophilic interaction chromatography (21). The initial mobile phase composition was 93.5% solvent A and 6.5% solvent B pumped at a flow rate of 0.2 mL/min. From 0 to 27.0 min, solvent B was increased linearly from 6.5% to 8.5%. From 27.0 to 27.5 min, the flow was increased to 0.35 mL/min, and the gradient was increased to 60% B. From 36.5 to 36.6 min, the gradient was decreased to 0% B, and from 36.6 to 39.0 min, the flow was increased to 0.55 mL/min. From 46.6 to 46.7 min, the gradient was increased to 6.5% B, and from 48.5 to 48.6 min, the flow was decreased to 0.2 mL/min. This re-equilibration ended at 49.0 min, after which the next sample was injected.

Mass spectrometric detection. Mass spectrometric detection was carried out using a Waters (Milford, MA) Quattromicro triple-stage, bench-top quadrupole mass spectrometer with electrospray ionization alternating between positive and negative, in the multiple reaction monitoring mode. The settings of the mass spectrometer were as follows: capillary voltage, 3.0 kV; cone voltage, 10 V; source temperature, 120°C; desolvation temperature, 325°C. The cone and desolvation gas flows (nitrogen) were 110 and 550 L/h, respectively. The collision voltage used was 10 V. Quadrupoles 1 and 2 each had low mass and high mass resolution set at 12. The dwell time was 0.25 sec, and the interscan delay was 0.1 sec. The multiple reaction monitoring m/z transitions monitored for each analyte are shown in Table 1 . The HPLC system and mass spectrometer were controlled by Waters MassLynx software (version 4.0), and data were collected with the same software. The analyte-to-internal standard ratio was calculated for each standard by dividing the area of each analyte peak by the area of the respective internal standard peak for that sample. Standard curves of the analytes were constructed by plotting the analyte-to-internal standard ratio versus the known concentration of analyte in each sample. Standard curves were fit by linear regression with weighting by 1/y2 followed by back calculation of concentrations.

View this table:
Table 1.

Assay variables for FdCyd and its metabolites

Assay performance. Assay performance (precision and accuracy) was assessed for each compound at four concentrations spanning the dynamic calibration range, and extraction recovery was assessed at one concentration (one third of the highest calibration point). Precision was calculated as the relative SD of six replicates; accuracy was calculated as the back-calculated concentration relative to the nominal concentration; recovery was calculated as the relative response of an extracted sample versus the response of an extracted blank with the compounds added after reconstitution. Dynamic calibration ranges were FdCyd, 0.039 to 11.6 μmol/L; FdUrd, 0.019 to 5.77 μmol/L; 5-FU, 0.0077 to 2.30 μmol/L; FUrd, 0.036 to 10.9 μmol/L; and 5-FC, 0.023 to 23.3 μmol/L.

Pharmacokinetic analysis

The area under the plasma concentration versus time curve (AUC) was calculated noncompartmentally using WinNonlin Professional, version 4.1 (Pharsight Corp., Mountain View, CA) and the linear trapezoidal rule. Pharmacokinetic variables for FdCyd and its metabolites were calculated if the plasma concentration could be quantitated for a sufficient period of time to allow reliable estimation of the half-life (3-4 half-lives).

Results

The assay proved linear over the concentration ranges used. A representative chromatogram is displayed in Fig. 2 . Assay performance and retention times are displayed in Table 1.

Fig. 2.
Fig. 2.

Representative chromatogram [intensity versus retention time (min)] of a plasma sample containing 5-FU (8.8 min), FdUrd (9.4 min), internal standard (12.8 min), FUrd (18.1 minutes), 5-FC (24.3 min), and FdCyd (25.6 min), analyzed with the newly developed HPLC-MS/MS method.

The plasma concentration versus time profiles of the individually quantitated analytes after administration of FdCyd, i.v. or p.o. with or without THU, are shown in Fig. 3A to E . Pharmacokinetic variables associated with these profiles are displayed in Table 2 . Peak concentrations (Cmax) of FdCyd were observed at the first time sampled (5 min) after i.v. administration and at later time points (30 or 45 min) after p.o. administration. After administration of FdCyd, with or without coadministered THU, FdCyd concentrations in plasma declined monoexponentially with half-lives ranging from 10 to 73 min. Coadministration of i.v. THU with i.v. FdCyd decreased FdCyd clearance by a factor of 5 relative to i.v. administration of FdCyd alone (2.1-0.4 mL/min). The volume of distribution was ∼30 mL for FdCyd alone or with i.v. THU. The absolute oral bioavailability of FdCyd alone was 4.3%. The absolute oral bioavailability of FdCyd, with coadministered i.v. THU, was 80.7%. Coadministration of p.o. FdCyd with p.o. THU resulted in a plasma exposure to FdCyd that was 2.5-fold that of i.v. FdCyd given alone, 54% that of i.v. FdCyd with i.v. THU, and 67% that of p.o. FdCyd with i.v. THU.

Fig. 3.
Fig. 3.

Mean ± SD plasma concentration (n = 3) versus time curves of FdCyd (▪), FdUrd (□), FU (▵), FUrd (⋄), and FC (▴) in male CD2F1 mice after (A) i.v. administration of 25 mg/kg FdCyd, (B) p.o. administration of 25 mg/kg FdCyd, (C) i.v. administration of 25 mg/kg FdCyd with i.v. coadministration of 100 mg/kg THU, (D) p.o. administration of 25 mg/kg FdCyd with i.v. coadministration of 100 mg/kg THU, and (E) p.o. administration of 25 mg/kg FdCyd with p.o. coadministration of 100 mg/kg THU.

View this table:
Table 2.

Pharmacokinetic variables of FdCyd and metabolites after administration of 25 mg/kg FdCyd to male CD2F1 mice, either i.v. or p.o. and with or without 100 mg/kg i.v. or p.o. THU

The peak plasma concentrations of FdCyd and its metabolites and the time to reach those peak concentrations are compatible with the sequential, metabolic conversion of FdCyd to FdUrd, 5-FU, and FUrd. FUrd was only detected after i.v. administration of FdCyd alone. 5-FC was only observed after p.o. administration of FdCyd, with or without THU coadministration, and plasma 5-FC concentrations increased with time. Coadministration of THU, whether i.v. or p.o, drastically altered the ratio of FdCyd Cmax to FdUrd Cmax, although it did not significantly increase the Cmax of FdCyd given i.v. After 240 min, concentrations of FdCyd and all of its metabolites fell below the lower limit of quantitation of the assay.

Renal excretion of FdCyd and metabolites, expressed as percentage of the dose recovered in 0- to 16- and 0- to 24-h urine, is displayed in Table 2. The recovery of metabolites in urine varied greatly, depending on the presence of THU and the route of administration; i.v. coadministered THU increased the recovery of FdCyd in urine.

In vitro, FdCyd was rapidly metabolized to FdUrd and 5-FU by liver homogenate and, even more so, by kidney homogenate (Fig. 4A and C ). As expected, this catabolism of FdCyd was abrogated by THU (Fig. 4A and B). Although the incubation of FdCyd in blood resulted in a steady production of FdUrd, the rate was much lower than those observed in incubations of FdCyd with kidney or liver homogenates (Fig. 4D).

Fig. 4.
Fig. 4.

Mean ± SD (n = 3) concentrations of FdCyd (▪), FdUrd (□), and FU (▵) after incubation of FdCyd in supernatant of liver homogenate without THU (A) and with THU (B; 51 μg protein/μL), and in supernatant of kidney homogenate (C; 39 μg protein/μL). Concentrations of FdUrd in plasma after incubation of FdCyd in whole blood are in (D). Note differences in scales among (A), (B), (C), and (D).

Discussion

The present investigation was designed to characterize the pharmacokinetics and metabolism of FdCyd in mice, and use those data to enhance the use of FdCyd as a DNA methyltransferase inhibitor in patients with cancer (3). A critical first step involved development of a highly sensitive HPLC-MS/MS assay that was capable of quantitating FdCyd and its metabolites FdUrd, 5-FU, FUrd, and 5-FC in murine plasma. The assay, which allowed the simultaneous quantitation of these closely related, polar, low-molecular-weight compounds, was precise and accurate over a >300-fold range of concentrations that spanned low ng/mL to μg/mL. After validation in human plasma, this assay should be an important resource in supporting further clinical development of FdCyd.

The pharmacokinetics of FdCyd are similar to those of other cytidine compounds in that they are characterized by a relatively short half-life. The 10-minute half-life of FdCyd after i.v. administration agrees well with the short half-lives of other cytidine analogues, such as cytosine arabinoside [12 min in mice (18) and 6 min in monkeys (22)], gemcitabine (17 min in both mice and humans; ref. 18), and 5-chloro-2′-deoxycytidine (10 min in mice; ref. 18). The ability of coadministered i.v. THU to increase the FdCyd plasma half-life 5-fold is consistent with reports of the effects of THU on the half-lives of 5-chloro-2′-deoxycytidine (2-fold; ref. 18) and cytosine arabinoside (2-fold; ref. 22) in mice. As expected, the peak plasma concentration of i.v. FdCyd was relatively unaffected by THU. However, peak plasma concentrations produced by p.o. FdCyd were increased 26- and 43-fold when THU was coadministered p.o. and i.v., respectively. Similarly, the AUC of i.v. FdCyd was increased 4.6-fold by i.v. THU coadministration. The volume of distribution of i.v. FdCyd, alone and with i.v. THU, was very similar, indicating that THU did not affect the distribution characteristics of FdCyd. This is in contrast to 5-chloro-2′-deoxycytidine in mice, for which coadministration of THU has been reported to cause a decrease in volume of distribution from 39 to 30 mL (18), and cytosine arabinoside in humans, for which coadministration of THU has been reported to change the cytosine arabinoside plasma concentration versus time profile from biphasic to monophasic (23).

In general, pyrimidine nucleosides have a low oral bioavailability (14–18). Because deamination is the first metabolic step in the catabolism of most cytidine analogues, blocking CD with THU is expected to decrease first-pass metabolism. Our results show that the oral bioavailability of FdCyd alone is only 4.3%, which is compatible with literature values for other nucleosides. The FdCyd AUC ratio of p.o. FdCyd with i.v. THU to i.v. FdCyd with i.v. THU is an indication of what fraction of an FdCyd dose is actually absorbed from the gut lumen (80.7%). Of this fraction, only 5.3% (5.3% × 80.7% = 4.3% of the initial dose) reaches the systemic circulation when THU is absent, with the remaining 94.7% being subjected to first-pass metabolism in the gastrointestinal epithelium or liver. Obviously, increasing systemic availability of p.o. FdCyd can only be effected by inhibiting the first-pass metabolism. Apparently, p.o. THU is absorbed in sufficient quantities to inhibit the CD-mediated metabolism of FdCyd, as is evidenced by the FdCyd AUC after concomitant p.o. administration of FdCyd and THU. Although the apparent FdCyd half-life with the p.o. combination is substantially longer than that of the i.v. combination (indicating a more extensive inhibition of CD), the exposure of the p.o. combination is only 54% that of the i.v. combination. Possibly, in the initial absorption phase, the THU plasma and tissue concentrations were not yet high enough to inhibit CD fully, and later, the p.o. administration of THU ensured a greater inhibition of liver CD activity. This hypothesis is supported by the fact that the time to reach the maximum plasma concentration (Tmax) of FdUrd was earlier than that of FdCyd after the p.o. combination of FdCyd and THU. Therefore, an even greater FdCyd exposure might be achieved if p.o. THU were dosed slightly before p.o. FdCyd. These results clearly indicate that p.o. administration of FdCyd in combination with p.o. THU is feasible. Clinical investigations of DNA methyltransferases inhibitors to date have employed the i.v. or s.c. administration route, which may limit the ability to maintain effective drug concentrations on a chronic basis. An orally given DNA methyltransferase inhibitor would avoid the logistical restrictions related to i.v. delivery and greatly expand the flexibility and applicability of epigenetic treatment. The results of the current study warrant clinical exploration of the combination of FdCyd and THU coadministered p.o.

Dogs have low CD activity in serum and tissues, and rat tissues have little CD activity (22, 24). However, CD activity in humans and other primates is relatively high, which could explain the reported interspecies differences in the half-lives of cytidine analogues (22). Because of their relatively high CD activity, mice are the best rodent species in which to model the pharmacokinetic behavior of FdCyd in humans. Major human sites of CD are the liver and spleen (25). In mice, CD is reported to be mainly present in kidney and much less so (18-fold on a weight basis; ref. 18), or even absent (26), in liver. This conflicts with our results, which suggest that FdCyd undergoes an extensive first-pass metabolism by CD. However, our hypothesis was supported by the in vitro incubations with FdCyd, which confirmed the relatively high CD activity in kidney but also showed the presence of substantial CD activity in liver, which could be abolished by THU. In addition, CD in murine gastrointestinal epithelium may play a role in the first-pass metabolism of FdCyd (26, 27).

Depending on the routes of administration of FdCyd and THU, 10% to 40% of the FdCyd dose was recovered in urine. The low recovery after p.o. FdCyd (1.3-3.5% of dose) reflects the high first-pass effect in the absence of THU, resulting in faster conversion to compounds more downstream in the metabolic pathway than the analytes monitored in the present study. Accordingly, coadministration of THU resulted in an increase in the percentage of FdCyd dose recovered in urine. In contrast to what might be expected, the urinary excretion of FdCyd did not consistently reflect the observed plasma AUC. This may be attributed to the high CD activity present in kidney, causing FdCyd deamination to FdUrd after glomerular filtration. When THU was coadministered i.v. with i.v. or p.o. FdCyd, excretion of FdCyd and FdUrd reasonably reflected their corresponding plasma AUC. Apparently, the same degree of renal CD inhibition is not achieved using p.o. THU, as evidenced by the unexpectedly low renal excretion of FdCyd after coadministration of p.o. FdCyd with p.o. THU. The occasional discrepancies between urinary excretion of FdCyd and its metabolites by the 0- to 16-h and 0- to 24-h mice may be attributable to experimental error or circadian rhythms of enzymes involved in nucleoside metabolism, such as dihydropyrimidine dehydrogenase, uridine phosphorylase, and thymidine phosphorylase (28–30). The 0- to 24-h mice were dosed early in the morning, whereas the 0- to 16-hour mice were dosed in the afternoon. Consequently, the enzymatic activities of the FdCyd metabolic pathway between both groups could be different, resulting in different rates of metabolism and, subsequently, different amounts of metabolites excreted in urine.

Our results clearly show that FdUrd, FUrd, and 5-FU are deaminated, metabolic products of FdCyd in mice, and that THU coadministration substantially reduces their plasma concentrations. Often, deamination of cytidine analogues constitutes an inactivation step (25, 31). However, whereas FdUrd and the other downstream metabolites of FdCyd are incapable of inhibiting DNA methyltransferase, they do have potent cytotoxic activity, and 5-FU and FdUrd are frequently used therapeutically (32, 33). In addition, their cellular effects result in cell cycle arrest, which inherently prevents incorporation of FdCyd into DNA (34). There still exists the possibility of metabolite formation via deoxycytidylate deaminase (dCMP deaminase; Fig. 1), a pathway that theoretically could be blocked by deoxytetrahydrouridine (27). We have recently shown that FdUrd and 5-FU concentrations in plasma of patients treated i.v. with the combination of FdCyd and THU are, at most, 10% of the concentrations reported when FdUrd or 5-FU are given as chemotherapeutic agents by continuous infusion (35). Consequently, addition of THU to FdCyd treatment likely results in therapy that is more selectively targeted to inhibition of DNA methylation. An increased incorporation of FdCyd into DNA after coadministration with THU has been reported (9), and future investigations should confirm the increased efficacy in down-regulating cellular levels of DNA methyltransferases.

THU seems to be an important adjuvant in FdCyd-based epigenetic therapy. It has been given i.v. in doses of up to 1,000 mg/kg to monkeys, and i.v. doses of up to 50 mg/kg have been given to humans for 20 days, without toxicity (22, 36). THU has been reported not to influence cytidine and nucleotide pools (18, 37). Based on urinary excretion of radioactivity (without further speciation), the oral bioavailability of THU is ∼10% (38). Although this may seem low, the oral route of dosing implies that the THU that does get absorbed will immediately reach its primary target organ, the liver, to inhibit CD.

Because uridine/thymidine phosphorylase is present in tissues and not in plasma (in contrast to CD), 5-FU and FUrd are metabolites produced intracellularly only and released back into plasma. FUrd, as a more distant downstream metabolite of FdCyd, was only detected after i.v. administration of FdCyd without THU, which provides the conditions allowing passage through the metabolic cascade to proceed at a maximum rate. It is very likely that FUrd was also produced with the other FdCyd/THU dosing combinations, but that it did not reach detectable plasma concentrations.

5-FC was only observed in plasma after p.o. administration of FdCyd. This is consistent with the ability of gut bacteria to break the N-glycosidic bond, which mammalian cells can not do (39). Furthermore, mammalian cells, as opposed to bacteria and fungi, do not express cytosine deaminase (EC 3.5.4.1) and therefore cannot convert 5-FC to 5-FU and ammonia (40). Consequently, 5-FC is not subjected to hepatic metabolism, and 5-FC clearance is dependent entirely on renal elimination (41), which would explain why plasma 5-FC concentrations increased over time, whereas plasma concentrations of all other compounds declined.

In conclusion, FdCyd is rapidly metabolized to cytotoxic metabolites that do not contribute to the inhibitory effect of FdCyd on DNA methyltransferase, but that do have their own distinct cytotoxic effects. Combining FdCyd with THU drastically increases and prolongs exposure to FdCyd and decreases exposure to FdCyd metabolites that lack DNA methyltransferase inhibitory activity. Oral coadministration of FdCyd and THU results in an FdCyd plasma exposure of more than half that after i.v. coadministration of FdCyd and THU and could therefore be a feasible way of administering FdCyd. Full exploitation of DNA methyltransferase inhibition may well necessitate long-term exposure to low concentrations of the inhibitor, as implied by recent investigations with 5-azacytidine and 5-aza-2′-deoxycytidine (5), which are currently given i.v. or s.c. However, these administration routes are not only inconvenient but may also limit the ability to maintain effective drug concentrations on a chronic basis. The inability to deliver drugs on a daily or twice daily basis may result in suboptimal utilization of DNA methyltransferase inhibition as a therapeutic approach. This limitation could be overcome by an orally bioavailable DNA methyltransferase inhibitor, which would enable chronic dosing. The preclinical data presented in the current article argue strongly that clinical investigation of the oral combination of FdCyd and THU might allow the strategy of inhibiting DNA methylation to be explored clinically to its fullest extent.

Acknowledgments

We thank Jeremy A. Hedges for excellent secretarial assistance; the University of Pittsburgh Cancer Institute, Hematology/Oncology Writing Group for constructive suggestions regarding the article; Diane Mazzei et al. in the University of Pittsburgh Animal Facility, without whose expert assistance these studies would not have been possible; and the Superbowl Champion Pittsburgh Steelers for numerous hours of entertainment during the preparation of this article.

Footnotes

  • ↵5 BRENDA: The Comprehensive Enzyme Information System (http://www.brenda.uni-koeln.de/index.php4).

  • Grant support: RAID project #266 (Dr. Edward Newman, City of Hope National Medical Center) by National Cancer Institute grant NO1-CM-52202.

  • The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

  • Note: WinNonlin software was provided as part of the Pharsight Academic Licensing Program.

  • Received May 23, 2006.
  • Revision received August 4, 2006.
  • Accepted August 28, 2006.

References

  1. ↵
    Jones PA, Baylin SB. The fundamental role of epigenetic events in cancer. Nat Rev Genet 2002;3:415–28.
    OpenUrlPubMed
  2. ↵
    Newman EM, Santi DV. Metabolism and mechanism of action of 5-fluorodeoxycytidine. Proc Natl Acad Sci U S A 1982;79:6419–23.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    Newman EM, Longmate J, Lenz H, et al. Phase I and clinical pharmacokinetic evaluation of the DNA methyltransferase inhibitor 5-fluoro-2′-deoxycytidine: a California Cancer Consortium study. Proc Am Soc Clin Oncol 2002;21:108a.
    OpenUrl
  4. ↵
    Yoo CB, Jones PA. Epigenetic therapy of cancer: past, present and future. Nat Rev Drug Discov 2006;5:37–50.
    OpenUrlCrossRefPubMed
  5. ↵
    Lyko F, Brown R. DNA methyltransferase inhibitors and the development of epigenetic cancer therapies. J Natl Cancer Inst 2005;97:1498–506.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    Boothman DA, Briggle TV, Greer S. Metabolic channeling of 5-fluoro-2′-deoxycytidine utilizing inhibitors of its deamination in cell culture. Mol Pharmacol 1985;27:584–94.
    OpenUrlAbstract
  7. Van Triest B, Pinedo HM, Giaccone G, Peters GJ. Downstream molecular determinants of response to 5-fluorouracil and antifolate thymidylate synthase inhibitors. Ann Oncol 2000;11:385–91.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    Boothman DA, Briggle TV, Greer S. Protective, tumor-selective dual pathway activation of 5-fluoro-2′-deoxycytidine provided by tetrahydrouridine in mice bearing mammary adenocarcinoma-755. Cancer Res 1987;47:2344–53.
    OpenUrlAbstract/FREE Full Text
  9. ↵
    Boothman DA, Briggle TV, Greer S. Tumor-selective metabolism of 5-fluoro-2′-deoxycytidine coadministered with tetrahydrouridine compared to 5-fluorouracil in mice bearing Lewis lung carcinoma. Cancer Res 1987;47:2354–62.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    Mekras JA, Boothman DA, Perez LM, Greer S. Use of 5-fluorodeoxycytidine and tetrahydrouridine to exploit high levels of deoxycytidylate deaminase in tumors to achieve DNA- and target-directed therapies. Cancer Res 1984;44:2551–60.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    el Kouni MH, el Kouni MM, Naguib FN. Differences in activities and substrate specificity of human and murine pyrimidine nucleoside phosphorylases: implications for chemotherapy with 5-fluoropyrimidines. Cancer Res 1993;53:3687–93.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    Kaysen J, Spriggs D, Kufe D. Incorporation of 5-fluorodeoxycytidine and metabolites into nucleic acids of human MCF-7 breast carcinoma cells. Cancer Res 1986;46:4534–8.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    Stoller RG, Myers CE, Chabner BA. Analysis of cytidine deaminase and tetrahydrouridine interaction by use of ligand techniques. Biochem Pharmacol 1978;27:53–9.
    OpenUrlCrossRefPubMed
  14. ↵
    van Groeningen CJ, Peters GJ, Nadal JC, Laurensse E, Pinedo HM. Clinical and pharmacologic study of orally administered uridine. J Natl Cancer Inst 1991;83:437–41.
    OpenUrlAbstract/FREE Full Text
  15. Holleran JL, Parise RA, Joseph E, et al. Plasma pharmacokinetics, oral bioavailability, and interspecies scaling of the DNA methyltransferase inhibitor, zebularine. Clin Cancer Res 2005;11:3862–8.
    OpenUrlAbstract/FREE Full Text
  16. Asif G, Hurwitz SJ, Gumina G, Chu CK, McClure HM, Schinazi RF. Pharmacokinetics of the antiviral agent β-l-3′-fluoro-2′,3′-didehydro-2′,3′-dideoxycytidine in rhesus monkeys. Antimicrob Agents Chemother 2005;49:560–4.
    OpenUrlAbstract/FREE Full Text
  17. Hernandez-Santiago BI, Chen H, Asif G, et al. Pharmacology and pharmacokinetics of the antiviral agent β-d-2′,3′-dideoxy-3′-oxa-5-fluorocytidine in cells and rhesus monkeys. Antimicrob Agents Chemother 2005;49:2589–97.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    Hale JT, Bigelow JC, Mathews LA, McCormack JJ. Analytical and pharmacokinetic studies with 5-chloro-2′-deoxycytidine. Biochem Pharmacol 2002;64:1493–1502.
    OpenUrlCrossRefPubMed
  19. ↵
    Cheng JC, Weisenberger DJ, Gonzales FA, et al. Continuous zebularine treatment effectively sustains demethylation in human bladder cancer cells. Mol Cell Biol 2004;24:1270–8.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    Herranz M, Martin-Caballero J, Fraga MF, et al. The novel DNA methylation inhibitor zebularine is effective against the development of murine T-cell lymphoma. Blood 2006;107:1174–7.
    OpenUrlAbstract/FREE Full Text
  21. ↵
    Olsen BA. Hydrophilic interaction chromatography using amino and silica columns for the determination of polar pharmaceuticals and impurities. J Chromatogr A 2001;913:113–22.
  22. ↵
    Dareer SM, Mulligan LT, Jr., White V, Tillery K, Mellett LB, Hill DL. Distribution of [3H]cytosine arabinoside and its products in mice, dogs, and monkeys and effect of tetrahydrouridine. Cancer Treat Rep 1977;61:395–407.
    OpenUrlPubMed
  23. ↵
    Kreis W, Chan K, Budman DR, et al. Effect of tetrahydrouridine on the clinical pharmacology of 1-β-d-arabinofuranosylcytosine when both drugs are coinfused over three hours. Cancer Res 1988;48:1337–42.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    Camiener GW, Smith CG. Studies of the enzymatic deamination of cytosine arabinoside. I. Enzyme distribution and species specificity. Biochem Pharmacol 1965;14:1405–16.
    OpenUrlCrossRefPubMed
  25. ↵
    Laliberte J, Marquez VE, Momparler RL. Potent inhibitors for the deamination of cytosine arabinoside and 5-aza-2′-deoxycytidine by human cytidine deaminase. Cancer Chemother Pharmacol 1992;30:7–11.
    OpenUrlPubMed
  26. ↵
    Kelly CJ, Gaudio L, Yesair DW, Schoenemann PT, Wodinsky I. Pharmacokinetic considerations in evaluating the effects of tetrahydrouridine on 5-azacytidine chemotherapy in L1210 leukemic mice. Cancer Treat Rep 1978;62:1025–32.
    OpenUrlPubMed
  27. ↵
    Mekras JA, Boothman DA, Greer SB. Use of 5-trifluoromethyldeoxycytidine and tetrahydrouridine to circumvent catabolism and exploit high levels of cytidine deaminase in tumors to achieve DNA- and target-directed therapies. Cancer Res 1985;45:5270–80.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    Hrushesky WJ. Circadian pharmacodynamics of anticancer therapies. Clin Chem 1993;39:2413–8.
    OpenUrlAbstract/FREE Full Text
  29. Hrushesky WJ. More evidence for circadian rhythm effects in cancer chemotherapy: the fluoropyrimidine story. Cancer Cells 1990;2:65–8.
    OpenUrlPubMed
  30. ↵
    Scheving LE, Tsai TH, Scheving LA. Chronobiology of the intestinal tract of the mouse. Am J Anat 1983;168:433–65.
    OpenUrlCrossRefPubMed
  31. ↵
    Nabhan C, Krett N, Gandhi V, Rosen S. Gemcitabine in hematologic malignancies. Curr Opin Oncol 2001;13:514–21.
    OpenUrlCrossRefPubMed
  32. ↵
    Kelly H, Goldberg RM. Systemic therapy for metastatic colorectal cancer: current options, current evidence. J Clin Oncol 2005;23:4553–60.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    Tsuruta K, Okamoto A, Egawa N, Kamisawa T, Karasawa K, Takahashi T. Survival benefits of adjuvant chemotherapy with oral doxifluridine (5′-DFUR) following radiotherapy in patients with unresectable pancreatic cancer. J Surg Oncol 2001;78:202–7.
    OpenUrlCrossRefPubMed
  34. ↵
    Kamoshida S, Matsuoka H, Shiogama K, et al. Immunohistochemical analysis of thymidylate synthase, p16(INK4a), cyclin-dependent kinase 4 and cyclin D1 in colorectal cancers receiving preoperative chemotherapy: significance of p16(INK4a)-mediated cellular arrest as an indicator of chemosensitivity to 5-fluorouracil. Pathol Int 2004;54:564–75.
    OpenUrlCrossRefPubMed
  35. ↵
    Beumer JH, Parise RA, Newman EM, et al. Concentrations of the DNA methyltransferase inhibitor 5-fluoro-2′-deoxycytidine (FdCyd) & its metabolites 5-fluoro-2′-deoxyuridine (FdUrd), 5-fluorouracil (FU), 5-fluorouridine (FUrd), & 5-fluorocytosine (FC) in plasma of patients treated with FdCyd. Proceedings of the American Society of Clinical Oncology 2006;25:84.
  36. ↵
    Xiang TX, Niemi R, Bummer P, Anderson BD. Epimer interconversion, isomerization, and hydrolysis of tetrahydrouridine: implications for cytidine deaminase inhibition. J Pharm Sci 2003;92:2027–39.
    OpenUrlCrossRefPubMed
  37. ↵
    Heinemann V, Plunkett W. Modulation of deoxynucleotide metabolism by the deoxycytidylate deaminase inhibitor 3,4,5,6-tetrahydrodeoxyuridine. Biochem Pharmacol 1989;38:4115–21.
    OpenUrlCrossRefPubMed
  38. ↵
    Kreis W, Woodcock TM, Gordon CS, Krakoff IH. Tetrahydrouridine: physiologic disposition and effect upon deamination of cytosine arabinoside in man. Cancer Treat Rep 1977;61:1347–53.
    OpenUrlPubMed
  39. ↵
    De Verdier CH, Potter VR. Alternative pathways of thymine and uracil metabolism in the liver and hepatoma. J Natl Cancer Inst 1960;24:13–29.
    OpenUrlPubMed
  40. ↵
    Rooseboom M, Commandeur JNM, Vermeulen NPE. Enzyme-catalyzed activation of anticancer prodrugs. Pharmacolo Rev 2004;56:53–102.
  41. ↵
    Vermes A, Guchelaar HJ, Dankert J. Flucytosine: a review of its pharmacology, clinical indications, pharmacokinetics, toxicity and drug interactions. J Antimicrob Chemother 2000;46:171–9.
    OpenUrlAbstract/FREE Full Text
Back to top
Clinical Cancer Research: 12 (24)
December 2006
Volume 12, Issue 24

Sign up for alerts

View this article with LENS

Open full page PDF
Article Alerts
Email Article
Citation Tools
Pharmacokinetics, Metabolism, and Oral Bioavailability of the DNA Methyltransferase Inhibitor 5-Fluoro-2′-Deoxycytidine in Mice
Jan H. Beumer, Julie L. Eiseman, Robert A. Parise, Erin Joseph, Julianne L. Holleran, Joseph M. Covey and Merrill J. Egorin
Clin Cancer Res December 15 2006 (12) (24) 7483-7491; DOI: 10.1158/1078-0432.CCR-06-1250
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

Advertisement