Purpose: Despite the extensive clinical experience with docetaxel, unpredictable interindividual variability in efficacy and toxicity remain important limitations associated with the use of this anticancer drug. Large interindividual pharmacokinetic variability has been associated with variation in toxicity profiles. Genetic polymorphisms in drug-metabolizing enzymes and drug transporters could possibly explain the observed pharmacokinetic variability. The aim of this study was therefore to investigate the influence of polymorphisms in the CYP3A and ABCB1 genes on the population pharmacokinetics of docetaxel.

Experimental Design: Whole blood samples were obtained from patients with solid tumors and treated with docetaxel to quantify the exposure to docetaxel. DNA was collected to determine polymorphisms in the CYP3A and ABCB1 genes with DNA sequencing. A population pharmacokinetic analysis of docetaxel was done using nonlinear mixed-effect modeling.

Results: In total, 92 patients were assessable for pharmacokinetic analysis of docetaxel. A three-compartmental model adequately described the pharmacokinetics of docetaxel. Several polymorphisms in the CYP3A and ABCB1 genes were found, with allele frequencies of 0.54% to 48.4%. The homozygous C1236T polymorphism in the ABCB1 gene (ABCB1*8) was significantly correlated with a decreased docetaxel clearance (−25%; P = 0.0039). No other relationships between polymorphisms and pharmacokinetic variables reached statistical significance. Furthermore, no relationship between haplotypes of CYP3A and ABCB1 and the pharmacokinetics could be identified.

Conclusions: The polymorphism C1236T in the ABCB1 gene was significantly related to docetaxel clearance. Our current finding may provide a meaningful tool to explain interindividual differences in docetaxel treatment in daily practice.

The anticancer drug docetaxel (Taxotere) is approved for the treatment of patients with early-stage, locally advanced and/or metastatic breast cancer, non–small-cell lung cancer, and androgen-independent metastatic prostate cancer. The recommended dose ranges from 75 to 100 mg/m2 given as a 1-hour infusion once every 3 weeks. An important limitation associated with docetaxel use is the unpredictable interindividual variability in efficacy and toxicity. Potential causes for such variability in drug effects include the pathogenesis and severity of the disease being treated, the occurrence of unintended drug interactions, and impairment of hepatic and renal function (1). Despite the potential importance of these clinical variables in determining drug effects, it is recognized that inherited differences in metabolism and excretion can have an even greater effect on the efficacy and toxicity of drugs (1).

The metabolism of docetaxel consists of a CYP3A-mediated oxidation of the tert-butylpropionate side chain, which results in the formation of four metabolites with reduced cytotoxic activity (2). The elimination pathway is mediated by the membrane-localized, energy-dependent drug efflux ABC transporter, P-glycoprotein (ABCB1; MDR1; ref. 3). Several polymorphisms have been described in the CYP3A and ABCB1 genes. The exact functional significance of polymorphisms in the CYP3A4 gene is not yet known. CYP3A4*1B has been associated with an increased transcription in vitro (4). However, in vivo no effect of this polymorphism has been observed (5). CYP3A4*2 has been associated with a decreased nifedipine clearance (6); for CYP3A4*3, no effect of the genotype on metabolism of several substrates has been observed (7, 8); and CYP3A4*12 showed an altered enzyme activity for midazolam and testosterone (8). For CYP3A5, variant alleles have been described, such as CYP3A5*3, which causes alternative splicing and protein truncation, resulting in the absence of functional CYP3A5 from liver tissue (9). The polymorphism C3435T in the ABCB1 gene has been associated with decreased protein expression in homozygous mutant individuals (10). The C3435T polymorphism is often simultaneously found with C1236T and G2677T/A. These polymorphisms may also have an effect on the pharmacokinetics of substrates of ABCB1, but results of the in vivo relevance of ABCB1 polymorphisms have been contradictory (11).

In this study, we investigated the relationship between docetaxel pharmacokinetics and ABCB1 and CYP3A genotypes in more detail. For this, the presence of C1236T, G2677T/A, and C3435T in ABCB1; CYP3A4*1B, CYP3A4*2, CYP3A4*3, and CYP3A4*12; and CYP3A5*2 and CYP3A5*3 variant alleles was determined in cancer patients treated with docetaxel.

Patients and treatment. To be eligible for this study, patients were required to have a histologically or cytologically proven solid malignancy for which docetaxel treatment was indicated; of ages ≥18 years; and have adequate hematopoietic, hepatic, and renal function (which was left to the discretion of the responsible oncologist). Treatment of the patient was done according to the local protocols for standard of care. Docetaxel was administered in doses of 50 to 100 mg/m2, infused over 1 hour. When at start liver function abnormalities were present, or when significant hematologic toxicity had developed during a preceding course, the dose of docetaxel was reduced according to standard clinical practice. Patients received routine supportive treatment (5′-hydroxytryptamine antagonists and dexamethasone) prophylactically.

The study was approved by the Medical Ethical Committee of the Netherlands Cancer Institute and all patients provided written informed consent in accordance with institutional and governmental guidelines. All patients were treated between March 2002 and October 2005 at The Antoni van Leeuwenhoek Hospital/The Netherlands Cancer Institute, Amsterdam; UMC Utrecht, Utrecht; MST, Enschede; Academic Medical Hospital, Maastricht; UMC Nijmegen, Nijmegen; or Slotervaart Hospital, Amsterdam, all in the Netherlands.

Pharmacokinetics of docetaxel. For pharmacokinetic analysis, 5 mL of heparin blood samples were obtained before and at the end of infusion, 6 and 24 hours after start of infusion, according to the limited sampling strategy of Baille et al. (12). Blood samples were centrifuged, plasma was separated, and samples were immediately stored at −20°C until analysis. Docetaxel concentrations were determined with a validated liquid chromatography-tandem mass spectrometry assay earlier described (13). The validated range of docetaxel concentrations was 0.25 to 1,000 ng/mL and interassay accuracy and precision were between −10.2% and 1.02% and <12.8%, respectively.

Pharmacogenetic analysis. For pharmacogenetic screening, 5 mL of EDTA blood were taken before start of the docetaxel infusion. Genomic DNA was isolated according to the method of Boom et al. (14). PCR amplifications were done in 50-μL reactions with ∼100 ng of genomic DNA, 200 μmol/L deoxynucleotide triphosphates (Epicentre Technologies, Madison, WI), 10× PCR Buffer II (Applied Biosystems, Foster City, CA), MgCl2, 0.5 to 1 unit of AmpliTaq Gold (Applied Biosystems), and forward and reverse primers (Metabion, Planegg-Martinsried, Germany). The methods used for the amplification of the CYP genes have been described by Sata et al. (6) for CYP3A4 and by van Schaik et al. (15) for CYP3A5. Genetic polymorphisms in ABCB1 were all analyzed according to slightly modified methods previously described by Hoffmeyer et al. (10) and Kim et al. (16), respectively. PCR amplifications were done with a PTC-200 thermocycler system (MJ Research, Inc., Waltham, MA). Results of the PCR reaction were analyzed on a 2% agarose gel. DNA cycle sequencing was carried out essentially as described by the manufacturer (Applied Biosystems) in 20-μL reactions on a PTC-200 thermocycler (MJ Research). Residual dideoxy terminators were removed by ethanol/sodium acetate precipitation according to the protocol of the manufacturer (Applied Biosystems) and sequences analyzed on an Applied Biosystems 3100-Avant DNA sequencer. For sequence alignment, Seqscape v2.1 (Applied Biosystems) was used. Linkage disequilibrium between single-nucleotide polymorphisms on the same chromosome was done with the Graphical Overview of Linkage Disequilibrium (GOLD) software V1.1.0.0.9

Haplotype analysis was done with PHASE version 2.1 (17).

Population pharmacokinetic and pharmacogenetic data analysis. Population pharmacokinetic analysis was done using nonlinear mixed-effect modeling (NONMEM; double precision, version V, level 1.1; GloboMax LLC, Hanover, MD; ref. 18). PDx-Pop (version 1.1j Release 4; GloboMax LLC) was used as an interface for data and output processing and for modeling management. NONMEM uses a maximum likelihood criterion to simultaneously estimate population values of fixed effect (e.g., drug clearance or the influence of a certain polymorphism on drug clearance) and values of random effects (e.g., interindividual and residual variability).

The first-order conditional estimation method with interaction between interindividual, intraindividual, and residual variability was applied. Precision of variable estimates, as calculated by the Covariance Step of NONMEM, was evaluated.

A three-compartmental structural kinetic model previously described by Bruno et al. (19) was used as structural model. The relationship between polymorphisms in the CYP3A and ABCB1 genes and pharmacokinetic variables was tested by use of the log-likelihood test. For polymorphisms in the ABCB1 gene, the influence on docetaxel clearance and V1, V2 and V3 was determined; for CYP3A polymorphisms, only the influence on docetaxel clearance was determined. If the allele frequency of the polymorphisms was low (<15%), the assumption was made that individuals homozygous for this polymorphism had a two times larger effect than heterozygous individuals. For instance, a change in clearance (Cl) in CYP3A4*1B carriers was evaluated by use of the following equation:

\[\mathrm{Cl}\ =\ \mathrm{Cl}_{\mathrm{pop}}\ {\times}\ (1\ {-}\ {\theta}_{1}\ {\times}\ \mathrm{genotype})\]

where the genotype has the value of 0 (wild-type), 1 (heterozygous CYP3A4*1B), or 2 (homozygous CYP3A4*1B); Clpop is the typical value for patients with the wild-type genotype; and 𝛉1 is the fractional change in clearance in CYP3A4*1B carriers.

If the allele frequencies were high for a polymorphism (>15%), a separate fixed effect was estimated for the different genotypes (wild-type, heterozygous, and homozygous mutant). For example, a change in docetaxel clearance in C3435T ABCB1 carriers was described as follows:

\[\mathrm{Cl}\ =\ \mathrm{Cl}_{\mathrm{pop}}\ {\times}\ {\theta}_{2}^{\mathrm{heterozygous}}\ {\times}\ {\theta}_{3}^{\mathrm{homozygous}}\]

where Clpop is the fixed effect of docetaxel clearance in wild-type patients and 𝛉2 and 𝛉3 are the fractional changes in clearance for heterozygous and homozygous carriers of the C3435T ABCB1 polymorphism, respectively. Furthermore, the influence of a haplotype on pharmacokinetic variables of docetaxel was determined.

Statistical discrimination between hierarchical models was based on the log-likelihood ratio test using the minimal function of the objective value. P = 0.005 was considered statistically significant, which corresponds to a decrease in the minimal function of the objective value of 7.9 (degrees of freedom = 1) or 10.6 (degrees of freedom = 2).

To investigate whether the relationship between genotype and the pharmacokinetics remained significant after inclusion of known determinants of docetaxel clearance, the model previously described by Bruno et al. (19) was applied to our data set. In this model, the following relationship between clearance and covariates was found:

\[\mathrm{Cl}_{\mathrm{i}}\ =\ \mathrm{BSA}(\mathrm{Cl}_{\mathrm{pop}}\ +\ {\theta}_{10}\mathrm{AAG}\ +\ {\theta}_{11}\mathrm{AGE}\ +\ {\theta}_{12}\mathrm{ALB})(1\ {-}\ {\theta}_{13}\mathrm{HEP})\]

where body surface area (BSA), α1-acid glycoprotein (AAG), age (AGE), albumin (ALB), and hepatic function (HEP; i.e., aspartate aminotransferase and alkaline phosphatase, >60 IU and >300 IU, respectively) were the main predictors of docetaxel clearance.

In our data set, not all covariates were available for all patients. Multiple imputation from the distribution of the covariates in the population was used for the missing data (20). Therefore, six data sets were created using this multiple imputation method and the models including the genotype relationship and without this relationship were applied.

Model validation. The bootstrap resampling technique was applied as an internal validation of the developed model. Basically, bootstrap replicates were generated by sampling randomly ∼65% from the original data set with replacement. The final model was fitted in the replicate data sets using the bootstrap option in the software package Wings for NONMEM (by N. Holford, version 408B, November 2004, Auckland, New Zealand; ref. 21) and variable estimates for each of the replicate data sets were obtained. The stability of the model was evaluated by visual inspection of the distribution of the model variables. Furthermore, the median variable values and 2.5 and 97.5 percentiles of the bootstrap replicates were compared with the estimated values of the original data set.

In total, 107 patients were enrolled, of whom 92 were assessable for pharmacokinetic and pharmacogenetic analysis. From 15 patients, only pharmacogenetic data were available. Patient characteristics are listed in Table 1. The majority of patients were Caucasian, and more females were treated with docetaxel than males, which is related to the most prominent tumor type observed in this population (i.e., breast cancer). Concomitant anticancer drugs were given in 22 patients whereas 70 patients received single agent docetaxel.

Table 1.

Patient characteristics, clinical performance, and medication characteristics

(A) Patient characteristics
No. patients (total group N = 92)Mean (range)
Age (y)*  54 (31-82) 
Gender (M/F) 20/72  
Body surface area (m2)*  1.85 (1.35-2.20) 
Ethnicity   
    Caucasian 90  
    African  
Hemoglobin (mmol/L)  7.3 (4.9-8.5) 
Hematocrit  0.35 (0.24-0.42) 
WBC (×109/L)  11.9 (2.5-27) 
ANC  9.74 (0.69-21.6) 
Platelets (×109/L)  323 (114-749) 
α1-Acid glycoprotein (g/L)*  1.3 (0.49-2.49) 
Albumine (g/L)*  41 (24-52) 
ASAT (units/L)*  35 (7-291) 
ALAT (units/L)  32 (10-175) 
Alkaline phosphatase (units/L)*  174 (47-2467) 
   
(B) Clinical performance of patients treated with docetaxel
 
  

 
No. patients (total group N = 92)
 

 
Tumor types   
    Breast 64  
    Lung  
    Prostate 13  
    Other 11  
Metastasis   
    Liver 21  
    Bone 21  
    Pleural effusion  
    None 33  
    Unknown 28  
Performance status   
    0-1 80  
    2 12  
   
(C) Medication characteristics
 
  

 
No. patients (total group N = 92)
 
Mean (range)
 
Docetaxel dose (mean; mg/m2 75.0 (14.7-101.8) 
    80-100 28  
    60-80 47  
    <60 17  
Concomitant medication   
    None 70  
    Capecitabine  
    Cyclophosphamide  
    Doxorubicin  
    Trastuzumab  
    Other  
(A) Patient characteristics
No. patients (total group N = 92)Mean (range)
Age (y)*  54 (31-82) 
Gender (M/F) 20/72  
Body surface area (m2)*  1.85 (1.35-2.20) 
Ethnicity   
    Caucasian 90  
    African  
Hemoglobin (mmol/L)  7.3 (4.9-8.5) 
Hematocrit  0.35 (0.24-0.42) 
WBC (×109/L)  11.9 (2.5-27) 
ANC  9.74 (0.69-21.6) 
Platelets (×109/L)  323 (114-749) 
α1-Acid glycoprotein (g/L)*  1.3 (0.49-2.49) 
Albumine (g/L)*  41 (24-52) 
ASAT (units/L)*  35 (7-291) 
ALAT (units/L)  32 (10-175) 
Alkaline phosphatase (units/L)*  174 (47-2467) 
   
(B) Clinical performance of patients treated with docetaxel
 
  

 
No. patients (total group N = 92)
 

 
Tumor types   
    Breast 64  
    Lung  
    Prostate 13  
    Other 11  
Metastasis   
    Liver 21  
    Bone 21  
    Pleural effusion  
    None 33  
    Unknown 28  
Performance status   
    0-1 80  
    2 12  
   
(C) Medication characteristics
 
  

 
No. patients (total group N = 92)
 
Mean (range)
 
Docetaxel dose (mean; mg/m2 75.0 (14.7-101.8) 
    80-100 28  
    60-80 47  
    <60 17  
Concomitant medication   
    None 70  
    Capecitabine  
    Cyclophosphamide  
    Doxorubicin  
    Trastuzumab  
    Other  

Abbreviations: ANC, absolute neutrophil count; ALAT, alanine aminotransferase; ASAT, aspartate aminotransferase.

*

Variables used for covariate analysis.

Pharmacokinetics of docetaxel. The observed plasma concentration-time profiles of docetaxel were well predicted by the previously developed pharmacokinetic model of Bruno et al. (ref. 19; Fig. 1). The individual and mean pharmacokinetic variables of docetaxel were consistent with previous findings and showed high interindividual variability (Table 2).

Fig. 1.

Model predicted concentrations (A) and individual predicted concentrations (B) versus observed concentrations of docetaxel after application of the three-compartmental structural population pharmacokinetic model developed by Bruno et al. (ref. 19; without covariates).

Fig. 1.

Model predicted concentrations (A) and individual predicted concentrations (B) versus observed concentrations of docetaxel after application of the three-compartmental structural population pharmacokinetic model developed by Bruno et al. (ref. 19; without covariates).

Close modal
Table 2.

Pharmacokinetic variables of docetaxel using a three-compartmental model with NONMEM (N = 92), variable estimated of the final population pharmacokinetic model of docetaxel with covariates and C1236T polymorphisms, and the stability of the variables using bootstrap validation procedure

Pharmacokinetic variableEstimate (CV)Estimate of original data set with C1236T genotype, N = 92 (CV%)Median of the bootstrap replicates (2.5-97.5 percentiles), N = 1,300
Clearance (L/h) 54.2 (6.9) 56.5 (8.1) 56.4 (44.9-66.8) 
C1236T heterozygous mutant effect — 1.05 (8.9) 1.05 (0.9-1.3) 
C1236T homozygous mutant effect — 0.719 (15.3) 0.714 (0.5-0.9) 
Volume of compartment 1 (L) 11.0 (19.5) 10.7 (20.3) 10.3 (4.8-15.8) 
Volume of compartment 2 (L) 14.3 (31.2) 14 (35.4) 13.4 (4.6-116.6) 
Volume of compartment 3 (L) 315 (22.1) 323 (26.6) 322 (170-783) 
Intercompartmental clearance 1-2 (L/h) 6.57 (16.0) 6.8 (16.4) 7.5 (5.3-14.9) 
Intercompartmental clearance 1-3 (L/h)
 
14.0 (13.4)
 
14.3 (13.2)
 
15.2 (11.4-21.2)
 
Inter-patient variability
 
   
Clearance (%) 35.5 (17.3) 32.4 (16.0) 30.9 (0.2-0.4) 
Volume of compartment 1 (%) 30.0 (25.3) 29.7 (25.8) 30.4 (2.2 × 10−5-0.7) 
Intercompartmental clearance 1-3 (%) 11.3 (16.8) 12.6 (17.3) 14.1 (1.5 × 10−5-0.5) 
Volume of compartment 3 (%) 55.5 (44.5) 55.2 (46.9) 42.9 (4.2 × 10−5-0.9) 
Residual error 36.9 36.9 35.6 (0.276-0.437) 
Pharmacokinetic variableEstimate (CV)Estimate of original data set with C1236T genotype, N = 92 (CV%)Median of the bootstrap replicates (2.5-97.5 percentiles), N = 1,300
Clearance (L/h) 54.2 (6.9) 56.5 (8.1) 56.4 (44.9-66.8) 
C1236T heterozygous mutant effect — 1.05 (8.9) 1.05 (0.9-1.3) 
C1236T homozygous mutant effect — 0.719 (15.3) 0.714 (0.5-0.9) 
Volume of compartment 1 (L) 11.0 (19.5) 10.7 (20.3) 10.3 (4.8-15.8) 
Volume of compartment 2 (L) 14.3 (31.2) 14 (35.4) 13.4 (4.6-116.6) 
Volume of compartment 3 (L) 315 (22.1) 323 (26.6) 322 (170-783) 
Intercompartmental clearance 1-2 (L/h) 6.57 (16.0) 6.8 (16.4) 7.5 (5.3-14.9) 
Intercompartmental clearance 1-3 (L/h)
 
14.0 (13.4)
 
14.3 (13.2)
 
15.2 (11.4-21.2)
 
Inter-patient variability
 
   
Clearance (%) 35.5 (17.3) 32.4 (16.0) 30.9 (0.2-0.4) 
Volume of compartment 1 (%) 30.0 (25.3) 29.7 (25.8) 30.4 (2.2 × 10−5-0.7) 
Intercompartmental clearance 1-3 (%) 11.3 (16.8) 12.6 (17.3) 14.1 (1.5 × 10−5-0.5) 
Volume of compartment 3 (%) 55.5 (44.5) 55.2 (46.9) 42.9 (4.2 × 10−5-0.9) 
Residual error 36.9 36.9 35.6 (0.276-0.437) 

Abbreviations: NA, not applicable; CV, coefficient of variation.

Pharmacogenetics of CYP3A and ABCB1. Polymorphisms in the CYP3A and ABCB1 genes of putative influence for docetaxel metabolism and disposition were studied (Table 3A). Several polymorphisms could not be detected in this population (i.e., CYP3A4*2, CYP3A4*3, and CYP3A4*12). Allelic frequencies of the mutant alleles varied from 0.54% to 48.4% and were in concordance with frequencies found in other (mainly Caucasian) populations (22). The polymorphisms found in this population were in Hardy-Weinberg equilibrium. Genetic linkage was observed for the polymorphisms in the ABCB1 gene and for CYP3A4*1B and CYP3A5*3 (Table 3B), with the C1236T and G2677T polymorphisms showing the strongest linkage (R2 = 0.889; P < 0.0001). The most frequent haplotypes were wild-type (frequency, 36.4%); C1236T, G2677T, and C3435T in the ABCB1 gene (frequency, 33.8%); and C3435T in the ABCB1 gene (frequency, 9.7%).

Table 3.

(A) Allele frequencies of CYP3A and ABCB1 polymorphisms in 92 patients treated with docetaxel
EnzymePolymorphismNomenclatureEffectWild-typeHeterozygous mutantHomozygous mutantAllele frequency (%)
CYP3A4 A-392G CYP3A4*1B Promoter 82 7.1 
CYP3A4 T15713C CYP3A4*2 S222P 92 
CYP3A4 T23172C CYP3A4*3 M445T 92 
CYP3A4 C21896T CYP3A4*12 L373F 92 
CYP3A5 C27289A CYP3A5*2 T398N 91 0.54 
CYP3A5 A22893G CYP3A5*3 Splicing defect 72 18 12.0 
ABCB1 C1236T ABCB1*8 G411G 30 43 19 44.0 
ABCB1 G2677T/A ABCB1*7 A893S or T 31 40/2 19 42.4/1.1 
ABCB1 C3435T ABCB1*6 E1143E 23 49 20 48.4 
        
(B) Linkage disequilibrium of polymorphisms found in the CYP3A and ABCB1 genes in this population
 
       
 
(A) Allele frequencies of CYP3A and ABCB1 polymorphisms in 92 patients treated with docetaxel
EnzymePolymorphismNomenclatureEffectWild-typeHeterozygous mutantHomozygous mutantAllele frequency (%)
CYP3A4 A-392G CYP3A4*1B Promoter 82 7.1 
CYP3A4 T15713C CYP3A4*2 S222P 92 
CYP3A4 T23172C CYP3A4*3 M445T 92 
CYP3A4 C21896T CYP3A4*12 L373F 92 
CYP3A5 C27289A CYP3A5*2 T398N 91 0.54 
CYP3A5 A22893G CYP3A5*3 Splicing defect 72 18 12.0 
ABCB1 C1236T ABCB1*8 G411G 30 43 19 44.0 
ABCB1 G2677T/A ABCB1*7 A893S or T 31 40/2 19 42.4/1.1 
ABCB1 C3435T ABCB1*6 E1143E 23 49 20 48.4 
        
(B) Linkage disequilibrium of polymorphisms found in the CYP3A and ABCB1 genes in this population
 
       
 

NOTE: Linkage disequilibrium between single-nucleotide polymorphisms in the CYP3A and ABCB1 genes was done with GOLD software. A pairwise two-dimensional map between single-nucleotide polymorphisms was obtained; in the lower part of the figure, ρ2 values, and in the upper right of the figure, P values between single-nucleotide polymorphisms are depicted.

Pharmacogenetic influence on pharmacokinetics of docetaxel. The hypothesis that polymorphisms could influence pharmacokinetic variables of docetaxel was evaluated in the population pharmacokinetic model. The effects of the polymorphisms found in this population on the pharmacokinetic variable and subsequently on the model are depicted in Table 4. The pharmacokinetic model significantly improved when the C1236T polymorphism in the ABCB1 gene was implemented on the docetaxel clearance (Fig. 2). The typical values of clearance for wild-type, heterozygous, and homozygous patients were 58.0, 61.6, and 43.5 L/h, respectively. Consequently, this polymorphism resulted in a 25% increase in the area under the concentration-time curve (AUC) in homozygous mutant patients. The mean AUC values for wild-type, heterozygous, and homozygous mutant patients were 2.4, 2.4, and 3.0 mmol h/L, respectively. No statistically significant associations were observed among other variants of the CYP3A and ABCB1 genes and any of the pharmacokinetic variables of docetaxel (Table 4).

Table 4.

Effect of polymorphisms in the CYP3A and MDR1 genes on pharmacokinetic variables of docetaxel in the used NONMEM model (N = 92)

PolymorphismPharmacokinetic variablePValue of pharmacokinetic variable
Wild-typeHeterozygous mutantHomozygous mutant
CYP3A4*1B Cl (L/h) 0.41 54.2 54.2 54.2 
CYP3A5*3 Cl (L/h) 0.76 55.0 53.4 51.7 
C1236T Cl (L/h) 0.0039 58.0 61.6 43.5 
 V1 (L) 0.95 9.4 7.3 10.8 
 V2 (L) 14.3 14.3 14.3 
 V3 (L) 0.66 340 322 248 
G2677T Cl (L/h) 0.24 57.8 55.7 47.3 
 V1 (L) 0.77 11.2 10.1 11.9 
 V2 (L) 14.3 14.3 14.3 
 V3 (L) 0.74 342 280 356 
C3435T Cl (L/h) 0.31 61.2 52.7 51.7 
 V1 (L) 0.12 13.3 9.0 13.6 
 V2 (L) 14.3 14.3 14.3 
 V3 (L) 0.53 442 300 287 
PolymorphismPharmacokinetic variablePValue of pharmacokinetic variable
Wild-typeHeterozygous mutantHomozygous mutant
CYP3A4*1B Cl (L/h) 0.41 54.2 54.2 54.2 
CYP3A5*3 Cl (L/h) 0.76 55.0 53.4 51.7 
C1236T Cl (L/h) 0.0039 58.0 61.6 43.5 
 V1 (L) 0.95 9.4 7.3 10.8 
 V2 (L) 14.3 14.3 14.3 
 V3 (L) 0.66 340 322 248 
G2677T Cl (L/h) 0.24 57.8 55.7 47.3 
 V1 (L) 0.77 11.2 10.1 11.9 
 V2 (L) 14.3 14.3 14.3 
 V3 (L) 0.74 342 280 356 
C3435T Cl (L/h) 0.31 61.2 52.7 51.7 
 V1 (L) 0.12 13.3 9.0 13.6 
 V2 (L) 14.3 14.3 14.3 
 V3 (L) 0.53 442 300 287 
*

The pharmacokinetic model significantly improved after the inclusion of variables for the heterozygous and homozygous mutation C1236T in the MDR1 gene.

Fig. 2.

Effect of ABCB1 C1236T polymorphism on docetaxel clearance (N = 92). The docetaxel clearance (Cl) was significantly lower for C1236T homozygous mutant patients. The typical values for clearance were 58.0, 61.6, and 43.5 L/h for wild-type, heterozygous, and homozygous patients, respectively (P = 0.0039).

Fig. 2.

Effect of ABCB1 C1236T polymorphism on docetaxel clearance (N = 92). The docetaxel clearance (Cl) was significantly lower for C1236T homozygous mutant patients. The typical values for clearance were 58.0, 61.6, and 43.5 L/h for wild-type, heterozygous, and homozygous patients, respectively (P = 0.0039).

Close modal

Because several polymorphisms were linked, a haplotype analysis was done. The effect of the haplotype was also evaluated in the population model. However, no statistically significant effect of haplotypes on docetaxel clearance could be identified.

The C1236T polymorphism in the ABCB1 gene was tested in the covariate model developed by Bruno et al. (19). Multiple imputation from the distribution of the covariates in our population was used to create values for the missing covariates, thereby developing six replicate data sets. The missing covariates ranged from 5% for BSA to 52% for AAG. The C1236T polymorphism remained statistically significant in all six replicate data sets obtained after multiple imputation (data not shown).

In this study, the following relationship between clearance and covariates was found:

\[\mathrm{Cl}_{\mathrm{i}}\ =\ \mathrm{BSA}(\mathrm{Cl}_{\mathrm{pop}}\ {\times}\ {\theta}_{14}^{\mathrm{H}}\ {\times}\ {\theta}_{15}^{\mathrm{M}}\ +\ {\theta}_{10}\mathrm{AAG}\ +\ {\theta}_{11}\mathrm{AGE}\ +\ {\theta}_{12}\mathrm{ALB})(1\ {-}\ {\theta}_{13}\mathrm{HEP})\]

where, besides the covariates of Bruno et al. (19), also the C1236T polymorphism was included (H, heterozygous mutant; M, homozygous mutant).

Model validation. From the original data set, 1,300 replicate data sets were generated and used for the evaluation of the stability of the final model. The mean values of the original data set were in the 2.5 to 97.5 percentiles of the bootstrap procedure (Table 2), indicating that all pharmacokinetic variables could be estimated with acceptable precision.

In this study, we investigated the influence of polymorphisms in the CYP3A and ABCB1 genes on pharmacokinetic variables of docetaxel. The C1236T polymorphism in the ABCB1 gene significantly decreased the docetaxel clearance by 25%. Other polymorphisms in the ABCB1 gene and in the CYP3A4 and CYP3A5 genes did not influence pharmacokinetic variables of docetaxel.

The pharmacokinetic variables of docetaxel were well described by the previously published model of Bruno et al. (19). Overall, in the presented limited sampling strategy model, the pharmacokinetic variables of docetaxel could be estimated with acceptable precision. The mean clearance of docetaxel in this study was 54.2 L/h, which is in the range of docetaxel clearance values reported earlier (19, 23). The C1236T polymorphism is an additional determinant of docetaxel clearance, besides the known determinants of docetaxel clearance described by Bruno et al. (19). The AUC of docetaxel increased by 25% in homozygous mutant patients. Consequently, in these patients the docetaxel dose should be reduced by 25%.

The pharmacogenetic effect of C1236T on docetaxel clearance found in this study was explicit for homozygous mutant patients; heterozygous patients showed no difference in docetaxel clearance in comparison with wild-type patients. This might be explained by up-regulation of the wild-type allele in heterozygous patients. Mathijssen et al. (24) also showed that the polymorphism C1236T in the ABCB1 gene in homozygous mutant patients, and not in heterozygous mutant patients, caused a decreased clearance of irinotecan, an anticancer agent used in colorectal therapy.

Furthermore, Baker et al. (25) showed that the exposure to docetaxel was related to hematologic toxicity. Prospective identification of patients with a decreased docetaxel clearance, and thereby an increased risk of developing severe toxicity, may be important to decrease toxicity.

The homozygous synonymous C1236T polymorphism might probably have an indirect effect on the stability of mRNA (26). Shen et al. (27) suggested that allele-specific differences in RNA folding may influence downstream mRNA splicing, processing, or translational control and regulation. It is also possible that the glycine encoded by the synonymous single-nucleotide polymorphism has reduced translational activity. More complex mechanisms, such as gene-gene interactions, may also play a role. Recently, Wang et al. (28) identified the synonymous C3435T polymorphism as a main factor in allelic variation of mRNA expression of ABCB1. The wild-type allele of C3435T resulted in significantly higher mRNA expression than the mutant allele. The mutant alleles of C1236T and C3435T have different mRNA foldings, possibly causing a less efficient translation of the RNA. It was also shown that the C1236T polymorphism may not be the functional polymorphism causing allelic expression imbalance because this polymorphism did not affect mRNA expression (28). However, these results need to be confirmed in studies with larger number of samples. Evidently, more research is warranted to identify the molecular biological characteristics of C1236T that lead to altered P-glycoprotein function.

In our study, neither the C3435T nor the G2677T/A polymorphism was associated with altered pharmacokinetics of docetaxel. These polymorphisms were linked to the C1236T polymorphism, although the linkage was not very strong for the C1236T and C3435T polymorphisms. The linkage between C1236T and G2677T was strong, but the effect of the G2677A polymorphism could obscure the influence of the G2677T polymorphism on docetaxel pharmacokinetic variables. Consequently, haplotype analysis did not influence docetaxel clearance.

Other attempts to unravel the effect of polymorphisms in metabolizing enzymes and drug transporters on the pharmacokinetics of docetaxel have been undertaken. Goh et al. (29) could not detect an influence of the polymorphisms C3435T in ABCB1, CYP3A4*1B and CYP3A5*3, in an Asian population of 31 primarily non–small-cell lung cancer patients. It should, however, be noticed that in the study of Goh et al. (29), no patients were included who used substrates, inhibitors, or inducers for CYP3A. In our study, all patients received dexamethasone, which is a weak inducer of CYP3A.

No effect of polymorphisms in the CYP3A gene on the pharmacokinetics of docetaxel could be observed in this study. The allelic frequencies of mutations in the CYP3A gene in this population are too low to point out differences in docetaxel clearance. The clearance of docetaxel was related to the CYP3A activity measured by midazolam pharmacokinetics (29). Hirth et al. (30) and Yamamoto et al. (31) also showed a correlation between the CYP3A4 activity and docetaxel clearance using the erythromycin breath test and urinary metabolite of exogenous cortisol, respectively. Unfortunately, not all interindividual variation could be explained by measuring the CYP3A activity in patients.

It is important to screen whole genes, and not only coding regions, to identify possible relationships between polymorphisms or genes. Besides, to identify such effects, it is necessary to screen for polymorphisms in larger populations with defined ethnicities. Furthermore, details about the extent of the genomic region and allele frequencies of variants should be provided. A quantitative estimate of the prior probability of genes considered relevant should also be provided and a correction for multiple comparisons should be made due to multiple gene studies (32). First, the screening of polymorphisms and their effect on pharmacokinetics and pharmacodynamics of anticancer drugs have to be elucidated. Thereafter, high-throughput technology will find application in the simultaneous identification of polymorphisms in drug-metabolizing enzymes and drug transporters in large clinical settings.

In conclusion, the polymorphism C1236T in the ABCB1 gene was significantly related to docetaxel clearance and adds up to earlier defined determinants of docetaxel clearance. It is of interest to test in prospective trials whether dose-adaptation based on characterization of the C1236T status of ABCB1 will result in reduced interindividual variation of pharmacokinetics of docetaxel.

Grant support: Sanofi-Aventis B.V., the Netherlands.

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.

1
Evans WE, McLeod HL. Pharmacogenomics—drug disposition, drug targets, and side effects.
N Engl J Med
2003
;
348
:
538
–49.
2
Bruno R, Sanderink GJ. Pharmacokinetics and metabolism of Taxotere (docetaxel).
Cancer Surv
1993
;
17
:
305
–13.
3
Shirakawa K, Takara K, Tanigawara Y, et al. Interaction of docetaxel (“Taxotere”) with human P-glycoprotein.
Jpn J Cancer Res
1999
;
90
:
1380
–6.
4
Amirimani B, Ning B, Deitz AC, Weber BL, Kadlubar FF, Rebbeck TR. Increased transcriptional activity of the CYP3A4*1B promoter variant.
Environ Mol Mutagen
2003
;
42
:
299
–305.
5
Ball SE, Scatina J, Kao J, et al. Population distribution and effects on drug metabolism of a genetic variant in the 5′ promoter region of CYP3A4.
Clin Pharmacol Ther
1999
;
66
:
288
–94.
6
Sata F, Sapone A, Elizondo G, et al. CYP3A4 allelic variants with amino acid substitutions in exons 7 and 12: evidence for an allelic variant with altered catalytic activity.
Clin Pharmacol Ther
2000
;
67
:
48
–56.
7
Dai D, Tang J, Rose R, et al. Identification of variants of CYP3A4 and characterization of their abilities to metabolize testosterone and chlorpyrifos.
J Pharmacol Exp Ther
2001
;
299
:
825
–31.
8
Eiselt R, Domanski TL, Zibat A, et al. Identification and functional characterization of eight CYP3A4 protein variants.
Pharmacogenetics
2001
;
11
:
447
–58.
9
Kuehl P, Zhang J, Lin Y, et al. Sequence diversity in CYP3A promoters and characterization of the genetic basis of polymorphic CYP3A5 expression.
Nat Genet
2001
;
27
:
383
–91.
10
Hoffmeyer S, Burk O, von Richter O, et al. Functional polymorphisms of the human multidrug-resistance gene: multiple sequence variations and correlation of one allele with P-glycoprotein expression and activity in vivo.
Proc Natl Acad Sci U S A
2000
;
97
:
3473
–8.
11
Marzolini C, Paus E, Buclin T, Kim RB. Polymorphisms in human MDR1 (P-glycoprotein): recent advances and clinical relevance.
Clin Pharmacol Ther
2004
;
75
:
13
–33.
12
Baille P, Bruno R, Schellens JH, et al. Optimal sampling strategies for bayesian estimation of docetaxel (Taxotere) clearance.
Clin Cancer Res
1997
;
3
:
1535
–8.
13
Kuppens IE, van Maanen MJ, Rosing H, Schellens JH, Beijnen JH. Quantitative analysis of docetaxel in human plasma using liquid chromatography coupled with tandem mass spectrometry.
Biomed Chromatogr
2005
;
19
:
355
–61.
14
Boom R, Sol CJ, Salimans MM, Jansen CL, Wertheim-van Dillen PM, van der NJ. Rapid and simple method for purification of nucleic acids.
J Clin Microbiol
1990
;
28
:
495
–503.
15
van Schaik RH, van der Heiden I, van den Anker JN, Lindemans J. CYP3A5 variant allele frequencies in Dutch Caucasians.
Clin Chem
2002
;
48
:
1668
–71.
16
Kim RB, Leake BF, Choo EF, et al. Identification of functionally variant MDR1 alleles among European Americans and African Americans.
Clin Pharmacol Ther
2001
;
70
:
189
–99.
17
Stephens M, Smith NJ, Donnelly P. A new statistical method for haplotype reconstruction from population data.
Am J Hum Genet
2001
;
68
:
978
–89.
18
Beal SL, Boeckman AJ, Sheiner LB. NONMEM user's guide. San Francisco: University of California at San Francisco; 1988.
19
Bruno R, Vivler N, Vergniol JC, De Phillips SL, Montay G, Sheiner LB. A population pharmacokinetic model for docetaxel (Taxotere): model building and validation.
J Pharmacokinet Biopharm
1996
;
24
:
153
–72.
20
Rubin DB, Schenker N. Multiple imputation in health-care databases: an overview and some applications.
Stat Med
1991
;
10
:
585
–98.
21
Parke J, Holford NH, Charles BG. A procedure for generating bootstrap samples for the validation of nonlinear mixed-effects population models.
Comput Methods Programs Biomed
1999
;
59
:
19
–29.
22
Bosch TM, Doodeman VD, Smits PHM, Meijerman I, Schellens JH, Beijnen JH. Pharmacogenetic screening for polymorphisms in drug-metabolising enzymes and drug transporters in a Dutch population.
Mol Diagn Ther
2006
;
10
:
175
–85.
23
Clarke SJ, Rivory LP. Clinical pharmacokinetics of docetaxel.
Clin Pharmacokinet
1999
;
36
:
99
–114.
24
Mathijssen RH, Marsh S, Karlsson MO, et al. Irinotecan pathway genotype analysis to predict pharmacokinetics.
Clin Cancer Res
2003
;
9
:
3246
–53.
25
Baker SD, Li J, ten Tije AJ, et al. Relationship of systemic exposure to unbound docetaxel and neutropenia.
Clin Pharmacol Ther
2005
;
77
:
43
–53.
26
Frittitta L, Ercolino T, Bozzali M, et al. A cluster of three single nucleotide polymorphisms in the 3′-untranslated region of human glycoprotein PC-1 gene stabilizes PC-1 mRNA and is associated with increased PC-1 protein content and insulin resistance-related abnormalities.
Diabetes
2001
;
50
:
1952
–5.
27
Shen LX, Basilion JP, Stanton VP, Jr. Single-nucleotide polymorphisms can cause different structural folds of mRNA.
Proc Natl Acad Sci U S A
1999
;
96
:
7871
–6.
28
Wang D, Johnson AD, Papp AC, Kroetz DL, Sadee W. Multidrug resistance polypeptide 1 (MDR1, ABCB1) variant 3435C>T affects mRNA stability.
Pharmacogenet Genomics
2005
;
15
:
693
–704.
29
Goh BC, Lee SC, Wang LZ, et al. Explaining interindividual variability of docetaxel pharmacokinetics and pharmacodynamics in Asians through phenotyping and genotyping strategies.
J Clin Oncol
2002
;
20
:
3683
–90.
30
Hirth J, Watkins PB, Strawderman M, Schott A, Bruno R, Baker LH. The effect of an individual's cytochrome CYP3A4 activity on docetaxel clearance.
Clin Cancer Res
2000
;
6
:
1255
–8.
31
Yamamoto N, Tamura T, Murakami H, et al. Randomized pharmacokinetic and pharmacodynamic study of docetaxel: dosing based on body-surface area compared with individualized dosing based on cytochrome P450 activity estimated using a urinary metabolite of exogenous cortisol.
J Clin Oncol
2005
;
23
:
1061
–9.
32
Wacholder S, Chanock S, Garcia-Closas M, El Ghormli L, Rothman N. Assessing the probability that a positive report is false: an approach for molecular epidemiology studies.
J Natl Cancer Inst
2004
;
96
:
434
–42.