Individual and combined roles of CYP3A, P-glycoprotein (MDR1/ABCB1) and MRP2 (ABCC2) in the pharmacokinetics of docetaxel



Docetaxel is one of the most widely used anticancer drugs. A major problem with docetaxel treatment, however, is the considerable interpatient variability in docetaxel exposure. Another disadvantage of the drug is that it has a very low oral bioavailability and can, therefore, only be administered intravenously. The drug-metabolizing enzyme CYP3A and the drug transporter MDR1 (P-glycoprotein) are major determinants of docetaxel pharmacokinetics. In vitro studies have indicated that docetaxel is also a substrate for the drug transporter MRP2, but the in vivo importance of MRP2 for docetaxel is currently unknown. We, therefore, investigated the role of MRP2 in the pharmacokinetics of docetaxel by utilizing Mrp2−/− mice. We also generated and characterized Cyp3a/Mdr1a/b/Mrp2−/− combination knockout mice to get more insight into how these drug-handling systems work together in determining docetaxel pharmacokinetics. The systemic exposure in Mrp2−/− mice was not significantly different from wild-type, after either oral or intravenous administration. Strikingly, however, in Cyp3a/Mdr1a/b/Mrp2−/− mice, systemic docetaxel exposure was increased 166-fold after oral administration when compared with wild-type mice, and 2.3-fold when compared with Cyp3a/Mdr1a/b−/− mice. Interestingly, this 166-fold increase was disproportionate compared with that for the separate Cyp3a (12-fold) or Mdr1a/b/Mrp2 (4-fold) knockouts. The oral bioavailability was increased to 73% in the Cyp3a/Mdr1a/b/Mrp2−/− strain, versus only 10% in wild-type mice. Our data thus indicate that in the absence of CYP3A and Mdr1a/b activity, Mrp2 has a marked impact on docetaxel pharmacokinetics. These findings could have important implications for improving the oral bioavailability and reducing the variability in docetaxel exposure.

Docetaxel is a widely used anticancer drug applied against a variety of cancer types including breast, lung and prostate cancer. A major problem with docetaxel treatment is the considerable interpatient variability in docetaxel exposure that results in a significant risk of under- or overdosing of patients.1 It is nowadays established that the drug-metabolizing enzyme system cytochrome P450 3A (CYP3A) is a major determinant of this variability in docetaxel exposure.1 Another disadvantage of docetaxel is that it has a very low oral bioavailability and can therefore only be administered intravenously. This low oral bioavailability might also be attributable to the fact that the drug is such a good substrate for CYP3A. However, docetaxel is also subject to active efflux by the drug transporter MDR1 (P-glycoprotein; ABCB1), which might further contribute to the low oral bioavailability of docetaxel.2 As there are many advantages of the oral administration route, enhancing docetaxel oral bioavailability by inhibiting CYP3A or MDR1 has received considerable interest in recent years.3–5 For example, in a recent clinical proof-of-concept study, it was demonstrated that simultaneous oral coadministration of docetaxel with the CYP3A inhibitor ritonavir resulted in a docetaxel exposure that was in the same range as that achieved after intravenous administration (without ritonavir).5

Interestingly, we have recently demonstrated that the combined absence of CYP3A and Mdr1 P-glycoprotein in mice results in a disproportionate increase in docetaxel exposure.6 In case CYP3A and Mdr1 are both absent, the oral docetaxel exposure was even increased to 760% of the values in wild-type animals that had received the drug intravenously (100%). Simultaneous inhibition of CYP3A and MDR1 might thus be a promising strategy to further improve the oral bioavailability of docetaxel.

Similar to MDR1, the apical drug transporter multidrug resistance protein 2 (MRP2/ABCC2) can have a profound impact on the pharmacokinetics of drugs.7 We have previously demonstrated that docetaxel is a good substrate for MRP2 in vitro.8 Yet, little in vivo evidence is currently available on how important MRP2 actually is in the pharmacokinetics of docetaxel. Earlier studies with the closely related drug paclitaxel have revealed an increase in systemic exposure in Mrp2−/− mice after intravenous, though not after oral, administration.9 In addition, the paclitaxel exposure in Mdr1a/b/Mrp2−/− mice was significantly higher compared with the single Mdr1a/b and Mrp2 knockout mice, both after oral and intravenous administration.9 In this study, we aimed to get more insight into the importance of MRP2 in the pharmacokinetics of docetaxel. In addition, we investigated the effect of a possible interplay between MRP2, MDR1 and CYP3A on docetaxel pharmacokinetics.

Material and Methods


Mice were housed and handled according to institutional guidelines complying with Dutch legislation. Mdr1a/b−/− Mrp2−/−, Mdr1a/b/Mrp2−/−, Cyp3a−/− and Cyp3a/Mdr1a/b−/− mice have been described.6, 9–12 Mdr1a/b/Mrp2−/− and Cyp3a−/− mice were crossed to obtain Cyp3a/Mdr1a/b/Mrp2−/− combination knockout mice. Genotypes of mice were evaluated by PCR. All mice used in this study had a >99% FVB genetic background and were between 8 and 14 weeks of age. All experiments were done using male mice. Animals were kept in a temperature-controlled environment with a 12-hr light/12-hr dark cycle and received a standard diet (AM-II, Hope Farms, Woerden, The Netherlands) and acidified water ad libitum.

Docetaxel plasma pharmacokinetics

Docetaxel (10 mg/ml) formulated in polysorbate 80/ethanol/water [20:13:67 (v/v/v)] (Taxotere, Aventis, Gouda, The Netherlands) was diluted with saline (0.9% NaCl) and administered by oral gavage or by injection into the tail vein of male mice. To minimize variation in absorption, mice were fasted for 2 hr before docetaxel was administered by gavage into the stomach using a blunt-ended needle. Multiple blood samples (∼40 μl) were collected from the tail vein at 15 and 30 min and 1, 2, 4 and 8 hr using heparinized capillary tubes (Oxford Labware, St. Louis, MO). In case of intravenous administration, the first time point was 7.5 min instead of 15 min. Blood samples were centrifuged at 2,100g for 10 min at 4°C, and the plasma fraction was collected, supplemented to 200 μl with human plasma and stored at −20°C until analysis.

Drug analysis

Docetaxel concentrations in plasma samples were determined using a previously described sensitive and specific liquid chromatograpy-mass spectrometry/mass spectrometry assay.13

Clinical-chemical analysis of serum

Standard clinical chemistry analyses on serum of wild-type and Cyp3a/Mdr1a/b/Mrp2−/− mice (n = 9–11, males) were performed on a Roche Hitachi 917 analyzer (Roche Diagnostics, Basel, Switzerland) to determine levels of bilirubin, alkaline phosphatase, aspartate aminotransaminase, alanine aminotransaminase, γ-glutamyl transferase, lactate dehydrogenase, creatinine, ureum, Na+, K+, Ca2+, phosphate, total protein, albumin and cholesterol.

Oligoarray analysis of Cyp3a/Mdr1a/b/Mrp2−/− mice

Illumina MEEBO 40K oligo microarrays were hybridized with Cy-dye labeled pooled liver and intestinal amplified RNA (n = 4) of adult wild-type and Cyp3a/Mdr1a/b/Mrp2−/− males, using the TECAN HS4800 hybridization station. The original data and detailed protocols for RNA isolation, amplification, labeling, hybridization and gene ID list are available at ( and the data are deposited at ArrayExpress, EBI (

Pharmacokinetic calculations and statistical analysis

Pharmacokinetic variables were calculated by noncompartmental methods using the software package PK Solutions 2.0.2 (Summit Research Services, Ashland, OH). The area under the plasma concentration–time curves (AUCs) were calculated using the trapezoidal rule, with extrapolation to infinity. Elimination half-lives (t1/2) were calculated by linear regression analysis of the log-linear part of the plasma concentration–time curves. Plasma clearance after i.v. docetaxel administration was calculated by the formula clearance = dose/AUCi.v. and the oral bioavailability (F) was calculated by the formula F = AUCoral/AUCi.v. × 100%. Data obtained with single and combination knockout mice were compared with each other and with data obtained with wild-type mice. Differences were considered statistically significant when p < 0.05. Data are presented as means ± SD.

Results and Discussion

To allow investigation of the combined impact of CYP3A, MDR1 and MRP2 on the pharmacokinetics of docetaxel, we first generated a mouse model by cross-breeding Cyp3a knockout mice12 with Mdr1a/b/Mrp2 knockout mice.9 In spite of missing 3 important detoxifying systems, homozygous Cyp3a/Mdr1a/b/Mrp2−/− mice were viable, fertile and had normal life spans and body weights. Similar to Mrp2−/−11 and Mdr1a/b/Mrp2−/−9 mice, Cyp3a/Mdr1a/b/Mrp2−/− mice had an increased liver weight (7.6 ± 1.2% of body weight vs. 4.9 ± 0.4% in wild-type, n = 9–11, p < 0.001). No other macroscopic or microscopic anatomic abnormalities were evident in these mice. Standard clinical-chemical analysis of serum of adult males revealed several alterations, including modest but significant increases in bilirubin, creatinine and urea (Supporting Information data 1). The increased serum bilirubin levels are also seen in Mrp2−/− and Mdr1a/b/Mrp2−/− mice.9, 11

To evaluate possible alterations in gene expression, an Illumina mouse microarray analysis was performed to compare RNA expression in liver and small intestine of wild-type and Cyp3a/Mdr1a/b/Mrp2−/− adult males. A number of genes was significantly (fold > 2) upregulated (n = 53) or downregulated (n = 18) in the livers of Cyp3a/Mdr1a/b/Mrp2−/− mice (data deposited at ArrayExpress, EBI). Also in the small intestine, a number of genes was upregulated (n = 80) or downregulated (n = 46). Among the genes that were differentially expressed, several detoxifying systems genes were upregulated (Supporting Information data 2), consistent with earlier studies in CYP3A deficient mice.6, 12, 14, 15 It is important to realize that upregulations of other detoxifying systems can for some drugs affect the results obtained, as we have demonstrated previously.15 Still, we have no indications that these specific upregulations (Supporting Information data 2) are detrimental for the evaluation of docetaxel pharmacokinetics and thus considered the Cyp3a/Mdr1a/b/Mrp2−/− mouse model suitable for the pharmacokinetic analysis of docetaxel.

To address the importance of MRP2 in the pharmacokinetics of docetaxel, we administered docetaxel orally and intravenously to wild-type, Mrp2−/−, Mdr1a/b/Mrp2−/− and Cyp3a/Mdr1a/b/Mrp2−/− mice and subsequently determined docetaxel plasma levels at several time points (Fig. 1). To allow good comparisons between the different knockout strains, we also included previously reported data6 for Mdr1a/b−/−, Cyp3a−/− and Cyp3a/Mdr1a/b−/− mice in Figure 1 and Table 1. Despite that docetaxel has been found to be a good substrate for both human and mouse MRP2/Mrp2 in vitro,8, 16 the systemic exposure in single Mrp2−/− mice was not significantly different from wild-type, after both oral and intravenous administration (Fig. 1; Table 1). The docetaxel exposure was significantly higher in Mdr1a/b/Mrp2−/− mice in comparison to wild-type mice (3.6- and 1.4-fold after oral and intravenous administration, respectively). Yet, there was no significant difference between Mdr1a/b/Mrp2−/− and Mdr1a/b−/− mice, indicating that the increase in Mdr1a/b/Mrp2−/− mice is primarily caused by the absence of Mdr1a/b and that Mrp2 has apparently no effect. These results are qualitatively different from those observed with paclitaxel, where we did see a significant, albeit modest, effect of Mrp2 absence.9

Figure 1.

Plasma concentration–time curves of docetaxel in male FVB wild-type, Mdr1a/b−/−, Cyp3a−/−, Cyp3a/Mdr1a/1b−/−, Mrp2−/−, Mdr1a/b/Mrp2−/−, Cyp3a/Mdr1a/b/Mrp2−/− mice after oral (a) and intravenous (b) administration of docetaxel (10 mg/kg). Data are shown as the mean concentration, and error bars represent the SD (n = 5–8 per time point). Inset shows semilog plot of the data. Previously published data6 have been included to allow good comparison between the strains.

Table 1. Plasma pharmacokinetic parameters after oral and intravenous administration of docetaxel at 10 mg/kg
inline image

Interestingly, however, in Cyp3a/Mdr1a/b/Mrp2−/− mice, the systemic docetaxel exposure after oral administration was 166-fold higher compared with wild-type mice, and also substantially higher (2.3-fold, p < 0.001) than in the Mrp2-proficient Cyp3a/Mdr1a/b−/− strain (Fig. 1; Table 1). The 166-fold increase in docetaxel AUCoral in Cyp3a/Mdr1a/b/Mrp2−/− compared with wild-type mice is clearly disproportionate in comparison to the increases in Cyp3a (12-fold) or Mdr1a/b/Mrp2 (4-fold) knockout mice (Supporting Information data 3). We have previously discussed that such a disproportionate increase in exposure demonstrates effective collaboration, but not necessarily true functional synergism between CYP3A and MDR1.6 Note that we define synergism here as the situation that one detoxifying system (e.g., metabolizing enzyme) works more efficiently in the presence of another active detoxifying system (e.g., drug transporter) than in its absence. Scrutiny of our data indicates that such synergy does not occur for docetaxel in our mouse models (Supporting Information data 3).

The increase in docetaxel exposure in Cyp3a/Mdr1a/b/Mrp2−/− mice was less pronounced after intravenous administration, but still 24-fold when compared with wild-type, and again significantly higher (1.4-fold, p < 0.01) than in the Cyp3a/Mdr1a/b−/− strain (Table 1). Accordingly, whereas in wild-type mice, the oral docetaxel bioavailability was 10%, this was increased to 73% in the Cyp3a/Mdr1a/b/Mrp2−/− strain. Note that this oral bioavailability was also substantially higher than the 44.4% seen in Cyp3a/Mdr1a/b−/− mice (Table 1). Moreover, the absolute systemic exposure in Cyp3a/Mdr1a/b/Mrp2−/− mice after oral docetaxel administration was even 17-fold higher than achieved in wild-type mice after intravenous administration, further illustrating the pronounced impact of the simultaneous loss of CYP3A, Mdr1a/b and Mrp2 activity on docetaxel pharmacokinetics (Table 1). Together, these data indicate that Mrp2 does have a noticeable impact on reducing docetaxel exposure in the absence of CYP3A and Mdr1a/b, upon both oral and intravenous drug administration. The results also demonstrate the value of combination knockout mouse models to investigate the individual and combined impact of CYP3A and drug transporters on the pharmacokinetics of drugs. The insights obtained from these studies can have implications for improving the oral bioavailability and reducing the toxicity of docetaxel and many other drugs.

Given the advantages of the oral administration route, enhancing docetaxel oral bioavailability by inhibiting CYP3A or MDR1 has received considerable interest in recent years.3–5 We here found a substantial increase in systemic exposure and oral bioavailability in Cyp3a/Mdr1a/b/Mrp2−/− mice, when compared with wild-type mice but also when compared with Cyp3a/Mdr1a/b−/− mice. Yet, we could not observe an effect of Mrp2 at the relatively low docetaxel levels achieved in a CYP3A-proficient situation. This suggests that Mrp2 is a low affinity transporter for docetaxel, compared with the other detoxifying systems that predominate at the lower plasma levels. On the other hand, the impact of Mrp2 at high plasma concentrations suggests that Mrp2 has a relatively high capacity for transporting docetaxel in vivo. It is important to note, however, that extrapolating data from mice to humans should always be done with caution and that based on different kinetic parameters between mouse and human Mrp2/MRP2 a relatively more (or less) pronounced impact of MRP2 on docetaxel pharmacokinetics in humans is possible.

We have previously demonstrated that the combined absence of CYP3A and Mdr1a/b already causes a substantial increase in oral bioavailability, resulting in a systemic exposure that was even higher than normally achieved in wild-type after intravenous administration (Fig. 1; Table 1).6 As such, the combined inhibition of CYP3A and MDR1 will already suffice to get a very substantial improvement of the oral bioavailability. Importantly, however, based on our data, one has to be aware that in regimens where CYP3A and MDR1 are both inhibited, variation in MRP2 activity levels may still result in considerable interpatient variation in docetaxel exposure.

While individuals without any MRP2 activity are known (Dubin-Johnson patients), no null alleles for CYP3A and MDR1 have been found so far. Yet, there can be potentially very large variations in expression and activity levels of both CYP3A and MDR1, in particular also due to drug–drug and drug–food interactions, which can drastically alter the docetaxel pharmacokinetics with subsequent risks for either toxicity or underdosing of patients. In a recent pharmacogenetic study, it was found that polymorphisms in the CYP3A genes can have a profound effect on docetaxel exposure, whereas polymorphisms in MDR1 or MRP2 could not be correlated to altered docetaxel exposure.17 Accordingly, it was concluded that the involvement of the genotype status of these drug transporters in docetaxel pharmacokinetics is relatively unimportant.17 Our present study indeed shows a relatively modest increase in exposure when both drug transporters are completely absent, in contrast to the strong effect of CYP3A deficiency. Yet, we note that the combined absence of CYP3A, MDR1 and MRP2 results in a disproportionate increase in docetaxel AUC, and that in particular cases polymorphisms in MDR1 and MRP2 may thus become of clinical relevance in docetaxel therapy. For example, in strategies that focus on improving the oral bioavailability by solely inhibiting CYP3A,5 patients with polymorphisms in MDR1 and/or MRP2 could be at risk of reaching toxic docetaxel levels.

The increased docetaxel oral bioavailability and the altered pharmacokinetics (e.g., longer half-life, Fig. 1) when CYP3A and MDR1 are both inhibited would allow more chronic treatment regimens. A more chronic exposure, accompanied by a much lower maximal plasma concentration (Cmax) than achieved after intravenous administration, could result in a more effective treatment with lower toxicity. Indeed, in vitro studies have indicated a better response of tumor cells to an increased exposure duration rather than to an increased docetaxel concentration.18 Also, recent clinical trials have demonstrated that patients who are on a weekly schedule with a lower docetaxel dose (33–40 mg/m2, intravenous) have a similar overall survival as with the standard treatment, which is a single administration every 3 weeks with a higher dose (75 mg/m2, intravenous).19 Importantly, however, significantly less and less severe febrile neutropenia, the most dangerous docetaxel-related toxicity, was reported in the weekly dosing schedules.19

Despite the interesting opportunities, it should be noted that there are also potential risks of the combined inhibition of CYP3A and drug transporters. The oral administration route but also the significantly altered pharmacokinetics can give rise to toxicities that have not been previously encountered. Indeed, when we administered docetaxel orally to Cyp3a/Mdr1a/b−/− mice, we did not only find hematotoxicity but also severe, even lethal, intestinal toxicity,6 a toxicity that was not observed in previous safety studies where wild-type mice received the drug intravenously.20 Intestinal toxicity should therefore receive special attention during clinical trials that focus on improving docetaxel oral bioavailability by simultaneously inhibiting CYP3A and MDR1. Another important issue is the selectivity profile of the inhibitor used to boost the oral bioavailability of docetaxel. It would, for example, be highly undesirable if also (unknown) transporter proteins responsible for the uptake of docetaxel in tumor cells would be inhibited. Great caution should, therefore, be exercised in the design and execution of efforts to translate these insights from mouse models to the human situation.


The authors thank Mr. Rob Lodewijks and Mr. Enver Delic for clinical chemistry analysis, Mr. Martin van der Valk and Ms. Nadine Meertens for histological analysis and Dr. Ron Kerkhoven and coworkers for performing the oligoarray analysis.