**Warfarin Enantiomers Pharmacokinetics by CYP2C19**

Yumiko Akamine and Tsukasa Uno *Department of Hospital Pharmacy, Faculty of Medicine, University of the Ryukyus, Okinawa, Japan* 

#### **1. Introduction**

222 Pharmacology

Swen, J., Nijenhuis, M., de Boer, A., Grandia, L., Maitland-van der Zee, A., Mulder, H.,

Teichert, M., Eijgelsheim, M., Rivadeneira, F., Uitterlinden, A., van Schaik, R., Hofman, A.,

Torn, M., Bollen, W., van der Meer, F., van der Wall, E. & Rosendaal, F. (2005). Risks of oral

U.S. Department of Health and Human Services, Food and Drug Administration, Center for

Verhoef, T., Redekop, W., Darba, J., Geitona, M., Hughes, D., Siebert, U., de Boer, A.,

Voora, D., Shah, S., Spasojevic, I., Ali, S., Reed, C., Salisbury, B. & Ginsburg, G. (2009). The

van Wieren-de Wijer, D., Maitland-van der Zee, A., de Boer, A., Kroon, A., de Leeuw, P.,

Wilffert, B., Swen, J., Mulder, H., Touw, D., Maitland-van der Zee, A., Deneer, V. & KNMP

Yukhanyan, L., Freynhofer, M., Siller-Matula, J., Schror, K. & Huber, K. (2011). Genetic

Zabalza, M., Subirana, I., Sala, J., Lluis-Ganella, C., Lucas, G., Tomas, M., Masia, R.,

update of guidelines. *Clin. Pharmacol. Ther.* 89, 5, (May 2011), 662-673. Teichert, M., van Schaik, R., Hofman, A., Uitterlinden, A., de Smet, P., Stricker, B. & Visser,

treatment period. *Clin. Pharmacol. Ther.* 85, 4, (Apr 2009), 379-386.

Guidance for Industry, Pharmacogenomic Data Submissions.

*Coll. Cardiol.* 54, 17, (Oct 2009), 1609-1616.

*Haemost.* 105 Suppl 1, (May 2011), S55-9.

clopidogrel. *Heart*, (Jun 2011).

*Pharmacoepidemiol. Drug Saf.* 18, 8, (Aug 2009), 665-671.

coumarin derivatives. *Pharmacogenomics* 11, 7, (Jul 2010), 989-1002.

(Oct 2009), 3758-86.

*Sci.* (Nov 2010).

1532.

Rongen, G., van Schaik, R., Schalekamp, T., Touw, D., van der Weide, J., Wilffert, B., Deneer, V. & Guchelaar, H. (2011). Pharmacogenetics: from bench to byte- an

L. (2009a). Genotypes associated with reduced activity of VKORC1 and CYP2C9 and their modification of acenocoumarol anticoagulation during the initial

De Smet, P., van Gelder, T., Visser, L. & Stricker, B. (2009b). A genome-wide association study of acenocoumarol maintenance dosage. *Hum. Mol. Genet.* 18, 19,

anticoagulant therapy with increasing age. *Arch Intern Med* 165, 13, (Jul 2011) 1527-

Drug Evaluation and Research (CDER), Center for Biologics Evaluation and Research (CBER) & Center for Devices and Radiological Health (CDRH). (2005).

Maitland-van der Zee, A. & EU-PACT group. (2010). A systematic review of costeffectiveness analyses of pharmacogenetic-guided dosing in treatment with

SLCO1B1\*5 genetic variant is associated with statin-induced side effects. *J. Am.* 

Schiffers, P., Janssen, R., Psaty, B., van Duijn, C., Stricker, B. & Klungel, O. (2009). Reasons for non-response in observational pharmacogenetic research.

working Group Pharmacogenetics (2010). From evidence based medicine to mechanism based medicine. Reviewing the role of pharmacogenetics. *Pharm. World* 

variability in response to clopidogrel therapy and its clinical implications. *Thromb.* 

Marrugat, J., Brugada, R. & Elosua, R. (2011). Meta-analyses of the association between cytochrome CYP2C19 loss- and gain-of-function polymorphisms and cardiovascular outcomes in patients with coronary artery disease treated with Warfarin, a coumarin vitamin K antagonist, is the most widely prescribed anticoagulant agent for the control and prevention of atrial fibrillation-related thrombus formation, stroke, and arterial and venous thrombembolism (Hirsh J et al., 1998). The recommend warfarin therapy consists of the lowest dose required to maintain the target international normalized ratio (INR) because of the drug's narrow therapeutic window. However, there can be a 20 fold difference in the dose required by patients to achieve this target INR. It is well known that cytochrome P450 (CYP), predominantly CYP2C9, activity is an important source of variability (Kaminsky LS and Zhang ZY, 1997) . Additionally, Rieder et al. (2005) have reported that an effect of the vitamin K epooxide reductase complex subunit 1 gene (VKORC1) has an important role on dose requirement. However, Takahashi et al. (2006) shows that Caucasians and African-Americans have high frequencies of VKORC1 and CYP2C9 genotypes, which lead to either reduced metabolic activity or attenuated sensitivity to warfarin, whereas only about 20% of the Japanese population possesses these genotypes. Therefore, further study of sources of variability in warfarin dose requirements among Japanese patients is warranted.

Warfarin is administered clinically as a racemic mixture of the *S*- and *R*-enantiomer (Fig. 1), however *S*-warfarin is 3–5 times more potent than *R*-enantiomer. Both enantiomers are extensively metabolized in the liver (Chan E et al., 1994; Takahashi H and Echizen H, 2001). The more potent *S*-enantiomer is metabolized mainly to *S*-7-hydroxywarfarin by CYP2C9, whereas *R*-enantiomer is metabolized to *R*-6, *R*-7, *R*-8 and *R*-10-hydroxywarfarin by several CYPs involving CYP1A2, CYP3A4 and CYP2C19 (Kaminsky LS and Zhang ZY, 1997). Among these CYPs, it has been shown that both CYP2C9 and CYP2C19 are subject to single nucleotide polymorphisms (SNPs). In Japanese, because the heterozygous frequency of the CYP2C9 Leu359 allele is 3.5% (Takahashi H et al., 1998) and the frequency of the defective CYP2C19 alleles is 18.8% (Kubota T et al., 1996), the latter may be more closely associated with the clinical effect of warfarin. In this chapter, we therefore focus on the effect of CYP2C19 genotypes on the pharmacokinetics and pharmacodynamics of warfarin enantiomers. In addition, we characterize the impact of omeprazole, a CYP2C19 inhibitor, on the stereoselective pharmacokinetics and pharmacodynamics of warfarin between CYP2C19 genotypes.

Warfarin Enantiomers Pharmacokinetics by CYP2C19 225

RH guard column, 10 mm x 4.6 mm i.d.) for separation. The mobile phase consisted of phosphate buffer-acetonitrile (84:16 v/v, pH 2.0) for clean-up and phosphate bufferacetonitrile (45:55 v/v, pH 2.0) for separation. The peaks were monitored with an ultraviolet detector set at a wavelength of 312 nm, and total time for chromatographic separation was about 25 minutes. The retention times of *S*-7-hydoxywarfarin, *R*-warfarin, I.S. and *S*-warfarin were 17.6 min, 19.1 min, 20.0 min and 21.2 min, respectively. The validated concentration ranges of this method were 3-1000 ng/ml for *R*- and *S*-warfarin, and 3-200 ng/ml for *R*- and *S*-7-hydroxywarfarin, respectively. Intra- and inter-day coefficients of variation were less than 4.4 and 4.9% for *R*-warfarin and 4.8 and 4.0% for *S*-warfarin, and 5.1 and 4.2% for *R*-7 hydroxywarfarin and 5.8 and 5.0% for *S*-7-hydroxywarfarin at the different concentrations. The limit of quantification was 3 ng/ml for both warfarin and 7-hydroxywarfarin enantiomers.

Plasma concentrations of omeprazole and 5-hydroxyomeprazole were quantitated using HPLC method developed in our laboratory (Shimizu M et al., 2006). In brief, after alkalization with 0.1 mL of 0.5 M disodium hydrogen phosphate, 1 mL plasma was extracted with 4 mL of diethyl ether-dichloromethane (55:45, v/v). The organic phase was

μL of 50 mM disodium hydrogen phosphate buffer (pH 9.3), and then a 30-μL aliquot was injected to an HPLC system (SHIMADZU CLASS-VP, SHIMADZU Corporation, Kyoto, Japan), with a Inertsil ODS-80A column as an analytical column (particle size 5 μm; GL Science Inc, Tokyo, Japan). The mobile phase consisted of phosphate buffer-acetonitrilemethanol (65:30:5 v/v/v, pH6.5). Flow rate was 0.8 mL/min and wavelength was set at 302 nm. Limit of quantification was 3 ng/mL for omeprazole and 5-hydroxyomeprazole. Intraand inter-day coefficient variations were less than 5.1 and 6.6% for omeprazole concentrations ranging from 4 to 1600 ng/mL and 4.6 and 5.0% for 5-hydroxyomeprazole

We examined the pharmacokinetics of warfarin enantiomers by administering 10 mg of racemic warfarin to 17 healthy volunteers (Uno T et al., 2008). Blood samples were obtained before and over the course of 120 hours after dosing for the determination plasma warfarin enantiomer concentrations and prothrombin time-INR (PT-INR). Fig. 2 shows the mean plasma concentration-time curves for *R*- and *S*-warfarin between the CYP2C19 genotypes. The mean pharmacokinetic parameters of these compounds are summarized in Table 1.

In this study, the area under the plasma concentration-time curve (AUC0-∞) and the elimination half-life (t1/2) of *R*-warfarin were about 2-fold greater than those of *S*-warfarin in 17 subjects (Table 1). These values of *R*- and *S*-warfarin were in line with a previous report in which the same dose of racemic warfarin was administered (Lilja JJ et al., 1984). Additionally, AUC0-∞ and t1/2 of *R*-warfarin in PMs were significantly greater than those in hmEMs (*P* < 0.001 and *P =* 0.010, respectively). Similarly, there is a significant difference (*P =*  0.007) in the apparent oral clearance (CL) in hmEMs compared with that in PMs. The *S*/*R* ratios of AUC0-∞ of warfarin enantiomers were 0.51 in hmEMs and 0.37 in PMs (*P =* 0.005). Whereas, no difference was found in all pharmacokinetic parameters of *S*-warfarin and *S*-7-

C for 12 months.

C to dryness. The residue was dissolved with 30 μL of methanol and 100

C and analyzed within 3

Plasma samples for the pharmacokinetic study were stored at -20

months after sampling, and then were stable at -70

concentration ranging from 4 to 400 ng/mL, respectively.

hydroxywarfarin in hmEMs compared with PMs of CYP2C19.

**3. Pharmacokinetics of warfarin enantiomers** 

evaporated at 60

Fig. 1. Metabolic pathways of *R*-warfarin and *S*-warfarin.

#### **2. Analytical methods**

#### **2.1 Genotypic identification**

17 healthy Japanese volunteers (12 males and 5 females) were enrolled in this study after giving written informed consent. All subjects were enrolled in this study after giving written informed consent. Each Subject underwent a CYP2C19 genotyping test by use of a polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) method with allele–specific primer for identifying the *CYP2C19* wild-type (\*1) gene and the 2 mutated alleles, *CYP2C19*\*2 (\*2) in exon 5 and *CYP2C19*\*3 (\*3) in exon 4 (De Morais SM et al., 1994), and they were classified into 2 genotype groups as follows: homozygous extensive metabolizers (hmEMs, \*1/\*1, 10 subjects), poor metabolizers (PMs, \*2/\*2 or \*2/\*3, 7 subjects). Similarly, CYP2C9 genotyping test by use of a PCR-RFLP method with allele– specific primer was performed for identifying the *CYP2C9* wild-type (\*1) gene and the 2 mutated alleles, *CYP2C9\*2* (Arg144Cys) and *CYP2C9\*3* (Ile359Leu) (Yasar U et al., 1999). Alleles in which neither *CYP2C9\*2* nor *CYP2C9\*3* variants were identified were regarded as wild type in all subjects.

#### **2.2 Assay**

Plasma concentrations of warfarin enantiomers and *S*-7-hydoxywarfarin were determined using high performance liquid chromatography (HPLC) method developed in our laboratory (Uno T et al., 2007). In brief, warfarin enantiomers, *S*-7-hydroxywarfarin and an internal standard, diclofenac sodium, were extracted from 1 ml of plasma sample using diethyl ether-chloroform (80:20, v/v). The extract was injected onto column I (TSK precolumn BSA-C8, 5 μm, 10 mm x 4.6 mm i.d.) for clean-up and column II (Chiralcel OD-RH analytical column, 150 mm x 4.6 mm i.d.) coupled with a guard column (Chiralcel OD-

17 healthy Japanese volunteers (12 males and 5 females) were enrolled in this study after giving written informed consent. All subjects were enrolled in this study after giving written informed consent. Each Subject underwent a CYP2C19 genotyping test by use of a polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) method with allele–specific primer for identifying the *CYP2C19* wild-type (\*1) gene and the 2 mutated alleles, *CYP2C19*\*2 (\*2) in exon 5 and *CYP2C19*\*3 (\*3) in exon 4 (De Morais SM et al., 1994), and they were classified into 2 genotype groups as follows: homozygous extensive metabolizers (hmEMs, \*1/\*1, 10 subjects), poor metabolizers (PMs, \*2/\*2 or \*2/\*3, 7 subjects). Similarly, CYP2C9 genotyping test by use of a PCR-RFLP method with allele– specific primer was performed for identifying the *CYP2C9* wild-type (\*1) gene and the 2 mutated alleles, *CYP2C9\*2* (Arg144Cys) and *CYP2C9\*3* (Ile359Leu) (Yasar U et al., 1999). Alleles in which neither *CYP2C9\*2* nor *CYP2C9\*3* variants were identified were regarded as

Plasma concentrations of warfarin enantiomers and *S*-7-hydoxywarfarin were determined using high performance liquid chromatography (HPLC) method developed in our laboratory (Uno T et al., 2007). In brief, warfarin enantiomers, *S*-7-hydroxywarfarin and an internal standard, diclofenac sodium, were extracted from 1 ml of plasma sample using diethyl ether-chloroform (80:20, v/v). The extract was injected onto column I (TSK precolumn BSA-C8, 5 μm, 10 mm x 4.6 mm i.d.) for clean-up and column II (Chiralcel OD-RH analytical column, 150 mm x 4.6 mm i.d.) coupled with a guard column (Chiralcel OD-

Fig. 1. Metabolic pathways of *R*-warfarin and *S*-warfarin.

**2. Analytical methods 2.1 Genotypic identification** 

wild type in all subjects.

**2.2 Assay** 

RH guard column, 10 mm x 4.6 mm i.d.) for separation. The mobile phase consisted of phosphate buffer-acetonitrile (84:16 v/v, pH 2.0) for clean-up and phosphate bufferacetonitrile (45:55 v/v, pH 2.0) for separation. The peaks were monitored with an ultraviolet detector set at a wavelength of 312 nm, and total time for chromatographic separation was about 25 minutes. The retention times of *S*-7-hydoxywarfarin, *R*-warfarin, I.S. and *S*-warfarin were 17.6 min, 19.1 min, 20.0 min and 21.2 min, respectively. The validated concentration ranges of this method were 3-1000 ng/ml for *R*- and *S*-warfarin, and 3-200 ng/ml for *R*- and *S*-7-hydroxywarfarin, respectively. Intra- and inter-day coefficients of variation were less than 4.4 and 4.9% for *R*-warfarin and 4.8 and 4.0% for *S*-warfarin, and 5.1 and 4.2% for *R*-7 hydroxywarfarin and 5.8 and 5.0% for *S*-7-hydroxywarfarin at the different concentrations. The limit of quantification was 3 ng/ml for both warfarin and 7-hydroxywarfarin enantiomers. Plasma samples for the pharmacokinetic study were stored at -20 C and analyzed within 3 months after sampling, and then were stable at -70 C for 12 months.

Plasma concentrations of omeprazole and 5-hydroxyomeprazole were quantitated using HPLC method developed in our laboratory (Shimizu M et al., 2006). In brief, after alkalization with 0.1 mL of 0.5 M disodium hydrogen phosphate, 1 mL plasma was extracted with 4 mL of diethyl ether-dichloromethane (55:45, v/v). The organic phase was evaporated at 60 C to dryness. The residue was dissolved with 30 μL of methanol and 100 μL of 50 mM disodium hydrogen phosphate buffer (pH 9.3), and then a 30-μL aliquot was injected to an HPLC system (SHIMADZU CLASS-VP, SHIMADZU Corporation, Kyoto, Japan), with a Inertsil ODS-80A column as an analytical column (particle size 5 μm; GL Science Inc, Tokyo, Japan). The mobile phase consisted of phosphate buffer-acetonitrilemethanol (65:30:5 v/v/v, pH6.5). Flow rate was 0.8 mL/min and wavelength was set at 302 nm. Limit of quantification was 3 ng/mL for omeprazole and 5-hydroxyomeprazole. Intraand inter-day coefficient variations were less than 5.1 and 6.6% for omeprazole concentrations ranging from 4 to 1600 ng/mL and 4.6 and 5.0% for 5-hydroxyomeprazole concentration ranging from 4 to 400 ng/mL, respectively.
