Search

Previous studies in mice with disrupted mdr1a P-glycoprotein genes have shown that the oral bioavailability of paclitaxel is very low because of the presence of this drug-transporting protein in the intestinal wall. Additional studies with cyclosporin A have shown that this P-glycoprotein-inhibiting agent is able to increase the bioavailability of paclitaxel in mouse models and in patients. However, the potential immune-suppressive side effects of cyclosporin A renders this compound less suitable for chronic use in cancer patients. In this paper we present the results obtained with GF120918, an experimental P-glycoprotein inhibitor, on the oral bioavailability of paclitaxel in both wild-type and mdrlab knockout mice. GF120918 (25 mg/kg) was administered p.o. by gavage 15 min or 2 h before oral or i.v. dosing of paclitaxel, respectively. Paclitaxel plasma levels were quantified by high-performance liquid chromatography. GF120918 increased the plasma values for areas under the concentration-time curve of oral paclitaxel in wild-type mice by 6.6-fold from 408 to 2701 ng x ml(-1) h. Calculated relative to their respective values for area under the concentration-time curve after i.v. administration, GF120918 increased the oral bioavailability of paclitaxel in wild-type mice from 8.5 to 40.2%. The plasma pharmacokinetics of paclitaxel in mdr1ab knockout mice was not altered by GF120918, whereas the pharmacokinetics of paclitaxel in wild-type mice receiving GF120918 became comparable with mdr1ab knockout mice. This result indicates that GF120918 at this dose-level selectively and completely blocks P-glycoprotein in the intestines and does not notably interfere in the elimination of paclitaxel by metabolism or other transporters. On the basis of this result, GF120918 has been selected for additional study in humans.

We have developed and validated an analytical methodology for the quantification of docetaxel and its four major human oxidation metabolites in mouse plasma. We have used this procedure to study the pharmacokinetics and metabolism of docetaxel in female FVB mice, receiving 2.5, 10, or 33 mg/kg of docetaxel by i.v. injection. We have also studied the pharmacokinetics of polysorbate 80, because it was shown previously that the vehicle substance Cremophor EL, which is used in the formulation of paclitaxel, exerts a profound effect on the pharmacokinetics of this compound. Linear pharmacokinetics of docetaxel was observed at dose levels between 2.5 and 10 mg/kg, where plasma levels corresponded to those in patients receiving the maximum tolerated dose. At the highest dose level of 33 mg/kg, a deviation from the linear kinetics was observed. Compared with humans, mice could tolerate much higher plasma levels, suggesting that the toxic side effects are related to a certain plasma threshold concentration instead of area under the curve or Cmax. At the highest dose level, three docetaxel metabolites could be detected in the plasma samples of mice for up to 4 h after drug administration. The hydroxy metabolite of the tert-butoxy group (metabolite II) was the major metabolite, followed by the two epimeric hydroxyoxazolone-type compounds (metabolites I and III). A fourth putative metabolite (e.g., the cyclic oxazolidinedione derivative) was not detected. Because of rapid degradation of polysorbate 80 by esterases in plasma, the concentration of this vehicle substance declined very rapidly. Consequently, this substance was not able to interfere in the disposition of docetaxel.

The taxanes paclitaxel and docetaxel represent a novel class of antineoplastic agents. A major problem of both drugs is their low aqueous solubility and the design of suitable formulations has been a difficult step in the process of therapeutic development. The formulations currently used are mixtures of Cremophor EL:ethanol for paclitaxel (Taxol) and Tween 80:ethanol for docetaxel (Taxotere), but many new approaches have been tested or are under investigation. Paclitaxel and docetaxel have a similar mechanism of action, which is based on promotion of tubulin assembly and inhibition of microtubule disassembly. Pharmacokinetic studies revealed a marked non-linearity of paclitaxel in mice, which appeared to result exclusively from Cremophor EL, the major component present in the pharmaceutical formulation. An almost linear pharmacokinetic behavior was observed in the case of docetaxel. The reported plasma protein binding of both compounds ranged from 76 to 97% in different animal species. Paclitaxel and docetaxel widely distribute into most tissues of mice and rats, including tumor tissue, but only low concentrations were detected in the central nervous system. Despite the great similarity in the chemical structures of paclitaxel and docetaxel, their metabolic profile is very distinct. Furthermore, whereas paclitaxel metabolism is largely species dependent, docetaxel metabolism is similar across species in both isolated hepatic microsomal fractions and in vivo models. For both taxanes, hepatobiliary excretion is the major pathway of elimination and a major fraction of the dose is excreted in feces as parent drug or hydroxylated metabolites.

After a pre-operative 1-h i.v. infusion of 150 mg/m2 of teniposide (Vumon; VM26), the drug levels were determined in resected brain tumor specimens from three patients with malignant glioma and from three patients with brain metastases. Tissue dissections were performed within 0-2.5 h after drug administration in three patients and after 24 h in the other three patients. Teniposide was quantified by high-performance liquid chromatography and the levels of albumin in the resected tissue samples were quantified by radial immunodiffusion. In addition, albumin levels were quantified in normal brain tissue, in malignant glioma and in metastatic brain tumor tissue obtained post mortem from deceased patients. The albumin levels indicated that a substantial fraction (range: 0.16-0.50) of the resected brain tumor specimens consisted of blood. As the plasma concentration of teniposide during the first hours after infusion is high, the major part of the drug measured in the tumor specimens collected within 2.5 h after drug administration originated from the blood compartment. At 24 h after drug administration, when the plasma level of teniposide had declined to approximately 0.20 microgram/ml, we could discern a real tissue uptake of teniposide ranging from 0.15-0.27 microgram/g wet tissue weight in the resected tumor. Although the number of patients in this study is small, this work clearly illustrates that an accurate determination of the tissue concentration of teniposide is hindered by the high concurrent plasma levels. It is therefore essential that future tissue distribution studies also include a suitable procedure that establishes the contribution of drug originating from the blood compartment.

Background: P-glycoprotein (Pgp) is a membrane protein that acts as an extrusion pump for many cytotoxic drugs. Pgp is expressed in normal tissues, and its (over)expression in tumor cells contributes to their drug resistance. Human Pgp is encoded by the MDR1 gene, In mice, two Pgps (encoded by the mdr1a and mdr1b genes) appear to perform the same function as the single human protein. The simultaneous use of cytotoxic drugs and agents that block Pgp function has raised questions of safety, since a blockade of Pgp in normal tissues could alter drug pharmacokinetics and change the spectrum of toxic side effects. Analysis of the consequences of Pgp blockade has been facilitated by the generation of mice with disrupted mdr1a genes [mdr1a(-/-)]., Purpose: We studied the plasma pharmaco-kinetics, tissue distribution, and excretion of the cytotoxic drug vinblastine (VBL) and its metabolites in mdr1a (-/-) mice and in wild-type [mdr1a(+/+)] mice., Methods: VBL was administered to mice in bolus doses of either 1 or 6 mg/kg body weight by intravenous injection. VBL and its metabolites were quantified in tissue specimens, plasma, feces, and urine by use of high-performance liquid chromatography. Liquid scintillation counting was used to measure radioactivity in specimens from animals that had received [3H]VBL. Pharmacokinetic parameters were calculated by use of noncompartmental methods. Only two-sided P values are reported., Results: The half-life (t1/2) of VBL during its terminal phase of elimination was longer in mdr1a (-/-) mice than in wild-type mice. The t1/2 values with a 1-mg/kg dose were 3.6 hours +/- 0.3 hour (mean +/- standard error) and 2.1 hours +/- 0.3 hour, respectively (P < .05); with a 6-mg/kg dose, the values were 8.6 hours +/- 1.8 hours and 4.2 hours +/- 0.2 hour, respectively (P = .058). Fecal excretion of nonmetabolized VBL was reduced from 20%-25% of the administered dose (either 1 or 6 mg/kg) in wild-type mice to 9.3% (1-mg/kg dose) or 3.4% (6-mg/kg dose) in mdr1a(-/-) mice (both P < .05); the cumulative urinary excretion of VBL was low (< 6% of the administered dose) and not substantially different in the two types of mice. The metabolism of VBL to hydrophilic compounds, a primary mechanism involved in its elimination, was not altered in mdr1a(-/-) mice. The brains of mdr1a(-/-) mice accumulated substantially more VBL than the brains of wild-type mice. In mdr1a(-/-) mice, a few other tissues, such as the heart and the liver, accumulated increased amounts of VBL, but the relative levels of accumulation were lower than those found in the brain., Conclusions: Mice lacking the Pgp encoded by the mdr1a gene exhibit reduced fecal excretion of VBL, leading to a prolonged elimination t1/2 for this drug. Intact mdr1a function appears to protect the brain against high plasma levels of VBL, but most other tissues are not similarly protected., Implications: Enhanced drug accumulation in nonmalignant tissues after Pgp blockade should be carefully considered in future clinical trials of Pgp modulation.

We have developed a procedure suited for the isolation of metabolites of docetaxel (Taxotere) from human feces. The compounds were extracted from the feces with diethyl ether and further purified by (semipreparative) HPLC. Four metabolic products were obtained in submilligram quantities. Analytical HPLC with photodiode array detection showed that the purity of each compound was higher than 98%. There structures have been characterized by UV absorption and FAB/MS. All four compounds were oxidation products of the tert-butyl group attached to the C13-side chain, and corresponded to structures identified previously. The purified products were used for evaluating their cytotoxic activities against a human ovarian cancer (A2780) and a rat colon cancer (CC531) cell line, and their myelosuppressive effects in a hematopoietic progenitor toxicity assay. Although distinctions in biological activities between the compounds were evident, all metabolites showed a marked reduction in both cytotoxic and myelotoxic properties.

Studies in humans and mice have demonstrated a nonlinear pharmacokinetic behavior of paclitaxel. Because of its poor water solubility, the drug is formulated in a mixture of Cremophor EL and ethanol (1:1, v/v; Taxol). We hypothesized that the substantial amounts of concurrently administered Cremophor EL on the disposition of paclitaxel, female FVB mice received paclitazel by i.v. injection at does levels of 2, 10, and 20 mg/kg by appropriate (standard) dilution of the commercially available formulation of paclitaxel (Taxol) with saline. The drug was also given at 2 mg/kg with supplemental Cremophor EL-ethanol to achieve the same amount of vehicle as by standard administration of 10 mg/kg. Furthermore, paclitaxel formulations in Tween 80-ethanol (1:1, v/v) and dimethylacetamide were tested. Plasma samples were collected between 5 min and 48 h, and tissue specimens were sampled at 1, 4, and 8 h after drug administration. Paclitaxel and metabolites were quantified by high-performance liquid chromatography. Cremophor EL levels were determined by a novel high-performance liquid chromatography procedure. For comparative reasons, Cremophor EL was also assayed in plasma samples from three patients receiving a 3-h i.v. infusion of 175 mg/m2 of paclitaxel. A marked nonlinear pharmacokinetic behavior of paclitaxel was observed when the drug was formulated in Cremophor EL-ethanol. The clearance of 2.37 L/h/kg at 2 mg/kg was reduced to 0.33 and 0.15 L/h/kg at 10 and 20 mg/kg, respectively. When 2 mg/kg were given with an amount of Cremophor EL-ethanol matching that of the 10-mg/kg dose level, the clearance was 0.56 L/h/kg. If administered at 10 mg/kg in Tween 80-ethanol or at 2 and 10 mg/kg in dimethylacetamide, the clearances were 2.66, 2.57, and 2.62 L/h/kg, respectively. Despite the fact that much higher plasma levels of paclitaxel are reached when given in the Cremophor EL-ethanol formulation, the tissue levels were essentially similar with all tested drug preparations. The Cremophor EL levels in patients were in the same order of magnitude as those observed in mice after administration of 2 and 10 mg/kg. These data demonstrate that Cremophor EL has a profound effect on the pharmacokinetics of paclitaxel im mice. Because Cremophor EL also contributes substantially to the nonlinear pharmacokinetic behavior of paclitaxel observed in humans.

So far, all animal pharmacokinetic studies of paclitaxel, which used analytical procedures based on HPLC, have not been sensitive enough to quantify drug levels below 500 ng/ml. Consequently, the interpretation of the results is restricted because drug levels of paclitaxel as low as at least 50 nM (43 ng/ml) are relevant for the pharmacology of this drug. We recently described an accurate and very sensitive method based on HPLC for the determination of paclitaxel and the metabolites 3'-p-hydroxypaclitaxel (I), 6 alpha-hydroxypaclitaxel (II) and 6 alpha,3'-p-dihydroxypaclitaxel (III) in a wide variety of biological matrices. We have now implemented this methodology in a comprehensive pharmacokinetic study in female FVB mice. Previous pharmacokinetic studies in humans demonstrated a large steady-state volume of distribution, indicating that the drug is widely distributed into tissues. Comprehensive tissue distribution studies may, therefore, be helpful in providing more insight into possible relationships between plasma levels, drug levels in tissues and toxicity. Paclitaxel, formulated in Cremophor EL and ethanol (1:1, v/v), was given as a single i.v. bolus dose of 2, 10 and 20 mg/kg to female FVB mice. Except for the brain, the distribution of paclitaxel to all other tissues in the female mice was substantial and maximum drug levels were achieved within 0.5 or 1 h. A marked non-linear increase in the area under the concentration-time curve (AUC) in plasma was observed, which was not paralleled by a proportional increase in the tissue AUC levels. It is postulated that this effect may be related to the substantial amounts of Cremophor EL administered concurrently. The recovery of paclitaxel in the feces (0-96 h) was reduced from 58% at the 2 mg/kg dose level to 44% at the 20 mg/kg dose level. Small amounts of metabolites I and II were detected in the gut, liver and gall bladder, but not in the systemic circulation or any other tissue. Metabolite III was not detected. Metabolites I and II are likely excreted directly into the bile, and since their recovery in the feces accounts for about 25% of the administered dose, their formation thus represents an important pathway of detoxification.

The plasma pharmacokinetics of vinblastine and N-(deacetyl-O-4-vinblastoyl-23)-L-ethyl isoleucinate (VileE) in mice have been studied as part of the preclinical investigations of VileE, a new investigational semisynthetic Vinca alkaloid. Groups of animals received the test compounds through i.v. bolus injection at LD10, 0.5 x LD10, and 0.1 x LD10 doses. VileE has also been administered p.o. Drug plasma levels have been analyzed with a sensitive and selective method using liquid-liquid extraction for sample clean-up and high-performance liquid chromatography combined with fluorescence detection for quantification. Following i.v. injection, plasma kinetics of both vinblastine and VileE can be described adequately by a three-compartment open model. VileE demonstrates nonlinear pharmacokinetics with decreasing clearance and increasing terminal half-lives at increasing doses. Comparison of the plasma concentration versus time curves for vinblastine in humans and mice indicates that the toxicity of these compounds may not be directly related to the drug exposure expressed by the area under curve in plasma but by the terminal half-life and the time that a toxic threshold level is attained. Pharmacokinetically guided dose escalation in coming phase I trials of VileE is, therefore, discouraged.