The pharmacokinetic and pharmacodynamic interaction of clopidogrel and cilostazol in relation to CYP2C19 and CYP3A5 genotypes

Abstract

Aim

The primary objective of the present study was to evaluate the pharmacokinetic and pharmacodynamic interactions between clopidogrel and cilostazol in relation to the CYP2C19 and CYP3A5 genotypes.

Methods

In a randomized, three‐way crossover study, 27 healthy subjects were administered clopidogrel (300 mg), cilostazol (100 mg) or clopidogrel + cilostazol orally. Plasma concentrations of clopidogrel, cilostazol and their active metabolites (clopidogrel thiol metabolite, 3,4‐dehydrocilostazol and 4″‐trans‐hydroxycilostazol), and adenosine diphosphate‐induced platelet aggregation were measured for pharmacokinetic and pharmacodynamic assessment.

Results

The area under the plasma concentration–time curve (AUC) of the active thiol metabolite of clopidogrel was highest in the CYP2C19 extensive metabolizers (EM) and lowest in the poor metabolizers (PM). Cilostazol decreased the thiol metabolite AUC by 29% in the CYP3A5*1/*3 genotype [geometric mean ratio (GMR) 0.71; 90% confidence interval (CI) 0.58, 0.86; P = 0.020] but not in the CYP3A5*3/*3 genotype (GMR 0.93; 90% CI 0.80, 1.10; P = 0.446).

Known effects of the CYP2C19 and CYP3A5 genotypes on the exposure of cilostazol and its metabolites were observed but there was no significant difference in the AUC of cilostazol and 3,4‐dehydrocilostazol between cilostazol and clopidogrel + cilostazol.

The inhibition of platelet aggregation from 4 h to 24 h (IPA_4–24) following the administration of clopidogrel alone was highest in the CYP2C19 EM genotype and lowest in the CYP2C19 PM genotype (59.05 ± 18.95 vs. 36.74 ± 13.26, P = 0.023). However, the IPA of the CYP2C19 PM following co‐administration of clopidogrel and cilostazol was comparable with that of the CYP2C19 EM and intermediate metabolizers (IM) only in CYP3A5*3/*3 subjects.

Conclusions

The additive antiplatelet effect of cilostazol plus clopidogrel is maximized in subjects with both the CYP2C19 PM and CYP3A5*3/*3 genotypes because of a lack of change of clopidogrel thiol metabolite exposure in CYP3A5*3/*3 as well as the highest cilostazol IPA in CYP2C19 PM and CYP3A5*3/*3 subjects.

What is Already Known About this Subject

• CYP2C19 loss‐of‐function alleles produce less clopidogrel active metabolite, causing less platelet aggregation inhibition, but more cilostazol active metabolite, leading to enhanced antiplatelet effects.

• Cilostazol is a possible inhibitor of CYP3A enzymes that contribute to the formation of clopidogrel active metabolite.

• There is a possibility of a clopidogrel–cilostazol drug interaction.

What this Study Adds

• Cilostazol decreased the formation of clopidogrel active metabolite in CYP3A5*1/*3 but not in CYP3A5*3/*3 subjects. The antiplatelet effects of CYP2C19 EM, IM and PM after co‐administration of clopidogrel and cilostazol were similar in CYP3A5*3/*3 genotypes.

• Antiplatelet therapy which includes cilostazol might overcome the clopidogrel resistance caused by CYP2C19 PM in subjects with the CYP3A5*3/*3 genotype.

Introduction

Dual antiplatelet therapy with aspirin and thienopyridines such as clopidogrel is the currently recommended regimen for preventing atherothrombotic events in patients following acute myocardial infarction 1. In spite of the benefits of dual antiplatelet therapy, the risk of a thrombotic event remains an important concern 2. Triple therapy including cilostazol, clopidogrel and aspirin has been reported to reduce long‐term cardiac events after percutaneous coronary intervention (PCI), especially for patients at high risk 3. By contrast, a single study reported that triple therapy did not show any difference in reducing cardiovascular outcome compared with dual therapy 4. To understand these controversies, an intensive study of the interaction between clopidogrel and cilostazol, based on their pharmacokinetics and pharmacodynamics, is needed, but no such report has been published to date. Clopidogrel is a thienopyridine antiplatelet agent that inhibits the P2Y12 receptor, an adenosine diphosphate (ADP) receptor subtype. Clopidogrel is an inactive prodrug that must be biotransformed into its active thiol metabolite to exert antiplatelet effects; this is accomplished by hepatic cytochrome P450 isoenzymes such as CYP2B6, CYP2C9, CYP3A4 and CYP3A5, especially CYP2C19 5, 6, 7. It is well known that carriers of the CYP2C19 loss‐of‐function (LOF) alleles (*2 and *3 alleles) form less of the active thiol metabolite, causing a lack of platelet aggregation inhibition, which translates into a higher rate of subsequent cardiovascular events than in noncarriers 8, 9, 10.

It has been reported that cilostazol might ameliorate platelet responsiveness to clopidogrel in patients who have undergone primary PCI 11. Furthermore, some studies have shown that the administration of cilostazol after PCI can significantly lower the incidence of in‐stent restenosis 12, 13, 14, 15. Cilostazol is a potent inhibitor of type III phosphodiesterase and suppresses the degradation of cyclic adenosine monophosphate (cAMP), resulting in an increase in cAMP both in platelets and vascular smooth muscle cells 16, 17, 18. Cilostazol is metabolized via CYP enzymes into two major metabolites, 3,4‐dehydrocilostazol (4–7‐fold increased activity compared with cilostazol) and 4″‐trans‐hydroxycilostazol (with one‐fifth of the activity of cilostazol). 3,4‐Dehydrocilostazol is produced by CYP3A4, and 4″‐trans‐hydroxycilostazol is produced mainly by CYP3A5 and CYP2C19 19. It was reported that the single oral‐dose pharmacokinetics of cilostazol in healthy subjects was affected by CYP2C19 and CYP3A5 polymorphisms 20. Several studies have reported that addition of cilostazol to the dual antiplatelet therapy of aspirin and clopidogrel might improve nonresponsiveness in patients with CYP2C19 LOF alleles who are taking clopidogrel 21, 22.

Although combination therapy with clopidogrel and cilostazol is widely used in clinical situations, especially in East Asia 23, no studies have been conducted on the pharmacokinetic and pharmacodynamic interaction between clopidogrel and cilostazol, or the relationship between genetic polymorphisms and the interaction between clopidogrel and cilostazol.

To address these issues, we explored the potential pharmacokinetic and pharmacodynamic interactions between clopidogrel and cilostazol in relation to the CYP2C19 or CYP3A5 genotypes in healthy Korean subjects.

Method

Study subjects

Healthy Korean males, aged 20–45 years, without clinically significant abnormalities in medical history, physical examinations, ECGs or clinical laboratory measurements, were eligible for inclusion. Subjects were excluded if they had a history or evidence of hepatic, renal, gastrointestinal or haematological abnormalities, any other acute or chronic disease, any drug allergies, or if they were taking any medication that induces or inhibits drug‐metabolizing enzymes. No medications, herbal medicines, alcohol, citrus juice, grapefruit juice or beverages containing caffeine were permitted for 10 days prior to the study and for the duration thereof.

All subjects were advised of the risks and benefits of participation in the study and submitted written, signed and dated informed consent voluntarily prior to clinical trial participation. The study was conducted in compliance with the principles outlined in the Declaration of Helsinki, and the study protocol and all amendments were approved by the Institutional Review Board of Inje University Busan Paik Hospital (IRB approval number: 10–105).

Study design

The present study had a randomized, open‐label, single‐dose, three‐treatment, three‐sequence crossover design with a two‐week washout period (Supplemental Figure S1). Subjects were allocated to one of three groups in a 1 : 1 : 1 ratio according to a predesigned randomization table generated using SAS software (version 9.3; SAS Institute Inc., Cary, NC, USA). Twenty‐seven subjects, pregenotyped for CYP2C19 and CYP3A5, were administered a single oral dose of 100 mg cilostazol (Otsuka Pharmaceutical Co. Ltd, Tokyo, Japan) alone, 300 mg clopidogrel (Sanofi‐Aventis, Paris, France) alone or 100 mg cilostazol plus 300 mg clopidogrel. The dosing of cilostazol and clopidogrel was selected based on the approved maximal single dose 24, 25.

All subjects were admitted to the clinical trial centre one day prior to drug administration. Subjects fasted overnight and continued fasting until a standardized lunch was served 4 h after drug administration on day 7.

Blood samples for pharmacokinetic assessments were obtained at 0, 0.33, 0.66, 1, 1.5, 2, 3, 4, 5, 12 and 24 h for clopidogrel and its active metabolite and 0, 1, 2, 3, 4, 6, 8, 12 and 24 h for cilostazol and its active metabolites. For measurement of clopidogrel and its active metabolite, 4 ml of whole blood was collected in an Ethylenediaminetetraacetic acid (EDTA) tube pretreated with 25 μl 500 mM 2‐bromo‐3′‐methoxyacetophenone. For measurement of cilostazol and its active metabolites, 3 ml of whole blood were collected in an EDTA tube. Plasma samples were transferred immediately to eppendorf tubes and stored at −80 °C until analysis. Blood samples were also collected to measure ADP‐induced platelet aggregation. The platelet aggregation tests were performed at day −1 and 0, 1, 2, 4, 6, 12 and 24 h after drug administration.

Genotyping

The CYP2C19 and CYP3A5 genotypes were determined using the single‐base extension method. Analytical validation for the Korean population has been previously established and published by the PharmacoGenomics Research Center (PGRC, Inje University College of Medicine, Busan, Korea) 26. Genomic DNA was used to genotype for the presence of major Korean alleles, including CYP2C19*2 (rs4244285), CYP2C19*3 (rs4986893), CYP2C19*17 (rs12248560) and CYP3A5*3 (rs776746) using an ABI PRISM® genetic analyzer and its mounted GeneMapper® software according to the protocol of SNaPshot® Multiplex kit from Applied Biosystems (Foster City, CA, USA). No significant deviations from Hardy–Weinberg equilibrium were observed for any of the single nucleotide polymorphisms tested.

Drug analysis

Plasma concentrations of clopidogrel, cilostazol and their active metabolites (3,4‐dehydrocilostazol and 4″‐trans‐hydroxycilostazol) were analysed using a modified liquid chromatography–tandem mass spectrometry (LC‐MS/MS) method as reported previously 27, 28, 29.

Briefly, for clopidogrel and its thiol active metabolite, plasma samples (50 μl) were precipitated by adding 100 μl acetonitrile containing the internal standard (clopidogrel‐d4). After centrifugation, a 5 μl aliquot of the supernatant was injected into an Agilent 1200 series high‐performance liquid chromatography (HPLC) system (Agilent, Wilmington, DE, USA) coupled to a QTRAP 5500 triple quadrupole mass spectrometer (AB Sciex, Foster City, CA, USA) equipped with electrospray ionization. Chromatographic separation of the compounds was achieved using a Luna C18 column (particle size 3 μm, 100 × 2 mm², i.d.; Phenomenex, Torrance, CA, USA) with a mobile phase consisting of 50% acetonitrile in water containing 0.1% formic acid. The flow rate was 0.2 ml min⁻¹ and the retention times for clopidogrel and its thiol metabolite were 7.9 min and 8.6 min, respectively. The mass spectrometer was run in positive mode and m/z 326.2 → 216.0 for clopidogrel‐d4, m/z 322.1 → 212.0 for clopidogrel and m/z 504.3 → 353.9 for the thiol metabolite derivative were monitored. Calibration curves in the range of 0.02–20 ng ml⁻¹ for clopidogrel (r = 0.9991) and 0.5–200 ng ml⁻¹ for the thiol metabolite derivative (r = 0.9993) were established. The intraday and interday coefficients of variation were less than 11.8%. The coefficient of variation for the assay precision was <9.1%, and the accuracy was >95.9%.

For analysis of plasma concentrations of cilostazol, 3,4‐dehydrocilostazol and 4″‐trans‐hydroxycilostazol, a 0.3 ml aliquot of plasma was spiked with an internal standard (propyphenazone 5 μg ml⁻¹) and extracted with 3 ml methyl tert‐butyl ether. After vortex mixing for 5 min and centrifugation at 1910 g for 10 min, the organic layer was separated and evaporated to dryness. The residue was reconstituted with 150 μl 50% methanol. A 2 μl aliquot was injected into the LC‐MS/MS system. Chromatographic separation was performed on a Luna C18 column (particle size 3 μm, 100 × 2 mm², i.d.; Phenomenex, Torrance, CA, USA). The mobile phases were as follows: mobile phase A, 1 mM ammonium acetate (pH 4.5); mobile phase B, acetonitrile containing 0.1% formic acid. A gradient program was used for the HPLC separation, with a flow rate of 0.2 ml min⁻¹. The initial composition of solvent B was 25%, increased to 60% after 8 min and then maintained for 4 min, followed by re‐equilibriation to the initial condition for 6 min. Quantification was performed using a mass spectrometer in Multiple‐Reaction Monitoring (MRM) with positive electrospray ionization at m/z 370.2 → 288.3 for cilostazol, m/z 368.3 → 286.2 for 3,4‐dehydrocilostazol, m/z 386.2 → 288.3 for 4″‐trans‐hydroxycilostazol and m/z 231.2 → 189.1 for the internal standard, respectively. Calibration curves in the range of 2–3000 ng ml⁻¹ for cilostazol (r² = 0.9976) and 2–1000 ng ml⁻¹ for 3,4‐dehydrocilostazol and 4″‐trans‐hydroxycilostazol) (r² = 0.9992 and 0.9995, respectively) were established. The coefficient of variation for the assay precision was <5.6%, and the accuracy was >92.5%. No relevant crosstalk and matrix effect were observed.

Measurement of platelet aggregation

Platelet aggregation was measured using a Chrono‐log Lumi Aggregometer (model 700‐4DR; Chrono‐log Corp., Havertown, PA, USA) equipped with the AggroLink software package, and using a turbidometric method, as described previously 9, 29.

Pharmacokinetic and pharmacodynamic data analysis

The pharmacokinetic parameters of clopidogrel, cilostazol and their metabolites were calculated using WinNonlin® version 6.1 (Pharsight Co., Mountain View, CA, USA). The maximum concentration (C_max) and the time to C_max (T_max) were obtained from concentration–time curves. The area under the concentration–time curve from zero to the last observation (AUC_t) was calculated using the linear trapezoidal rule. AUC_t molar ratios of metabolite to parent drug were estimated, to evaluate metabolic status.

For the pharmacodynamic evaluation, platelet reactivity [absolute value of the observed maximal platelet aggregation (MPA)] was used and the inhibition of platelet aggregation (IPA) value was calculated from the observed MPA as follows:

$$\mathrm{IPA}\left(\%\text{inhibition}\right)=\frac{\mathit{MP}{A}_0-\mathit{MP}{A}_t}{\mathit{MP}{A}_0}\times 100\%$$

where MPA₀ is the average value of the platelet aggregation response at baseline (on day −1 and 0 h before treatment on day 1), and MPA_t is MPA at each scheduled time t after each treatment. The maximum effect of IPA (IPA_max) was obtained from IPA–time curves. The area under the effect–time curve from 0 h to 24 h (AUEC) was calculated using the linear trapezoidal rule. The average value of the IPA from 4 h to 24 h after dosing (IPA_4–24) was estimated.

Statistical analysis

Continuous variables were expressed as means ± standard deviation. Geometric mean ratios (GMRs) of natural log‐transformed AUC_t and C_max, and arithmetic mean ratios of T_max for clopidogrel, cilostazol and their metabolites were calculated in order to evaluate the magnitude of the drug interaction between clopidogrel and cilostazol, along with estimated 90% confidence intervals (CIs). Log‐transformed AUC_t and C_max, other pharmacokinetic and pharmacodynamic parameters of clopidogrel and cilostazol, alone and co‐administered were analysed using a mixed‐model analysis of variance with fixed effects for sequence, period and treatment, and a random effect for subject within sequence. The pharmacokinetic and pharmacodynamic parameters among the various CYP2C19 genotype groups and between different CYP3A5 genotype groups were compared using the Kruskal–Wallis or Wilcoxon rank‐sum tests after testing for normality.

Values of P < 0.05 were considered to indicate statistical significance. All statistical analyses were performed using SAS version 9.2 (SAS institute Inc.).

Results

Subjects

Twenty‐seven healthy male subjects were enrolled in and completed the study. All subjects were classified into three genotype groups according to the number of LOF alleles of the CYP2C19 genotypes: CYP2C19 extensive metabolizers (CYP2C19 EM; CYP2C19*1/*1; n = 9), CYP2C19 intermediate metabolizers (CYP2C19 IM; CYP2C19*1/*2 or CYP2C19*1/*3; n = 9), and CYP2C19 poor metabolizers (CYP2C19 PM; CYP2C19*2/*2, CYP2C19*2/*3 or CYP2C19*3/*3; n = 9). In addition, two genotype groups, CYP3A5*1/*3 (n = 10) and CYP3A5*3/*3 (n = 17), were compared to evaluate the effect of the CYP3A5 genotype on the pharmacokinetics and pharmacodynamics of clopidogrel and cilostazol. No subjects with the CYP2C19*17 allele or CYP3A5*1/*1 genotype were identified in the present study.

The average age, weight and height of all subjects were 24.9 (±2.4) years , 68.0 (±7.2) kg and 174.5 (±5.0) cm, respectively. There was no significant difference in age, weight or height among the CYP2C19 EM, IM and PM genotypes or between CYP3A5*1/*3 and CYP3A5*3/*3.

Pharmacokinetic results of clopidogrel and its active thiol metabolite

There was no significant difference in the AUC_t or C_max of clopidogrel between administration of clopidogrel with and without cilostazol (Table 1 and Figure 1A). The T_max of clopidogrel was delayed from 0.66 h to 1 h. after the co‐administration of clopidogrel and cilostazol. However, the T_max of the thiol metabolite was no different between administration of clopidogrel with and without cilostazol. The AUC_t and C_max of the thiol metabolite were significantly decreased after co‐administration of clopidogrel and cilostazol compared with administration of clopidogrel alone (GMR 0.82; P = 0.015 and GMR 0.84; P = 0.028, respectively; Table 1 and Figure 1B).

Table 1.

Pharmacokinetic parameters of clopidogrel and its thiol metabolite in healthy subjects after a single oral administration of clopidogrel alone or co‐administration of clopidogrel and cilostazol

	Clopidogrel	Clopidogrel + cilostazol	GMR (90% CI)	P value
Clopidogrel
AUCt (ng h ml−1)	6.80 ± 6.09	7.28 ± 5.46	1.11* (0.99 ~ 1.26)	0.139
Cmax (ng ml−1)	3.76 ± 3.81	4.12 ± 4.22	1.07* (0.93 ~ 1.23)	0.435
Tmax (h)	0.66 (0.33 ~ 2)	1.00 (0.66 ~ 3)	1.35 (1.19 ~ 1.52)	0.028
Thiol metabolite
AUCt (ng h ml−1)	33.00 ± 18.84	28.10 ± 15.78	0.82* (0.72 ~ 0.93)	0.015
Cmax (ng ml−1)	27.83 ± 13.95	23.41 ± 12.13	0.84* (0.74 ~ 0.95)	0.028
Tmax (h)	1.00 (0.33 ~ 2)	1.00 (0.33 ~ 3)	1.15 (0.98 ~ 1.31)	0.374
AUCthiol/AUCclopidogrel †	4.82 ± 3.92	3.73 ± 3.71	0.87 (0.69 ~ 1.04)	0.070

Data represent as arithmetic mean ± standard deviation, except for data for T_max, which represent the median (range).

AUC, area under the concentration–time curve; AUC_t, area under the concentration–time curve from zero to the last observation; CI, confidence interval; C_max, maximum concentration; T_max, time to C_max.

^*GMR, geometric mean ratio of clopidogrel + cilostazol vs. clopidogrel.

^†AUC_thiol/AUC_clopidogrel, AUC molar ratio of thiol metabolite vs. clopidogrel.

Figure 1.

Mean concentration–time profiles of clopidogrel (A) and its thiol metabolite (B) after a single oral administration of 300 mg clopidogrel alone and co‐administration of 300 mg clopidogrel and 100 mg cilostazol, and comparison of the thiol metabolite area under the plasma concentration–time curve stratified by CYP2C19 and CYP3A5 genotypes. clopidogrel, clopidogrel+cilostazol

The pharmacokinetic parameters of the thiol metabolite were further compared between clopidogrel with and without cilostazol, stratified into CYP2C19 or CYP3A5 genotype groups (Table 2). The AUC_t and C_max of the thiol metabolite were significantly highest in the CYP2C19 EM genotype, and lowest in the CYP2C19 PM genotype group after administration of clopidogrel alone (51.83 ± 18.00 vs. 15.91 ± 7.94 for AUC_t, P < 0.001; 39.43 ± 12.00 vs. 15.92 ± 7.59 for C_max, P = 0.001). Differences in thiol metabolite pharmacokinetic parameters among the CYP2C19 genotype group were also observed after co‐administration of clopidogrel and cilostazol. However, these parameters of the thiol metabolite were not statistically different between administration of clopidogrel with and without cilostazol in any CYP2C19 genotype groups.

Table 2.

The pharmacokinetic parameters of clopidogrel thiol metabolite according to CYP2C19 or CYP3A5 genotypes in healthy subjects after a single oral administration of clopidogrel alone or co‐administration of clopidogrel and 100 mg cilostazol

		Clopidogrel	Clopidogrel + cilostazol	GMR* (90% CI)	P value
CYP2C19 †	AUCt (ng h ml−1)
	EM	51.83 ± 18.00	46.33 ± 8.89	0.93 (0.81 ~ 1.06)	0.330
	IM	31.27 ± 5.92	24.65 ± 6.17	0.78 (0.63 ~ 0.96)	0.057
	PM	15.91 ± 7.94	13.31 ± 7.69	0.77 (0.52 ~ 1.13)	0.229
	P value	<0.001	<0.001
	Cmax (ng ml−1)
	EM	39.43 ± 12.00	35.33 ± 10.54	0.90 (0.69 ~ 1.17)	0.464
	IM	28.13 ± 10.96	20.46 ± 6.30	0.75 (0.55 ~ 1.04)	0.142
	PM	15.92 ± 7.59	14.45 ± 8.22	0.87 (0.67 ~ 1.13)	0.334
	P value	0.001	0.001
CYP3A5	AUCt (ng h ml−1)
	*1/*3	37.70 ± 23.58	27.81 ± 18.30	0.71 (0.58 ~ 0.86)	0.020
	*3/*3	30.24 ± 15.55	28.26 ± 14.70	0.90 (0.76 ~ 1.06)	0.280
	P value	0.407	0.707
	Cmax (ng ml−1)
	*1/*3	30.71 ± 14.09	22.54 ± 13.97	0.69 (0.56 ~ 0.87)	0.029
	*3/*3	26.13 ± 14.01	23.93 ± 11.34	0.93 (0.80 ~ 1.10)	0.446
	P value	0.328	0.633

CI, confidence interval; EM, extensive metabolizer; IM, intermediate metabolizer; PM, poor metabolizer.

^*GMR, geometric mean ratio of clopidogrel + cilostazol vs.clopidogrel.

^†CYP2C19 EM includes the CYP2C19*1/*1 genotype subset; CYP2C19 IM includes the CYP2C19*1/*2 and CYP2C19*1/*3 genotype subsets; CYP2C19 PM, includes the CYP2C19*2/*2, CYP2C19*2/*3 and CYP2C19*3/*3 genotype subsets.

Unlike the CYP2C19 genotypes, no significant difference was observed in the pharmacokinetic parameters of the thiol metabolite between subjects with the CYP3A5*1/*3 and CYP3A5*3/*3 genotypes after administration of clopidogrel with and without cilostazol. However, the AUC_t and C_max of the thiol metabolite were significantly decreased in subjects with the CYP3A5*1/*3 but not CYP3A5*3/*3 genotype after co‐administration of clopidogrel and cilostazol compared with clopidogrel alone (GMR 0.71 and 0.69 for CYP3A5*1/*3 vs. GMR 0.90 and 0.93 for CYP3A5*3/*3; Table 2).

Pharmacokinetic results of cilostazol and its active metabolites

The AUC_t of cilostazol and 3,4‐dehydro cilostazol (which is 4–7 times more potent than cilostazol) was not significantly different between the administration of clopidogrel with and without cilostazol, although a small decrease in C_max was observed after co‐administration of clopidogrel and cilostazol (Table 3, and Figure 2A,B). The AUC_t and C_max of 4″‐trans‐hydroxycilostazol, which has one‐fifth the activity of cilostazol, were decreased after co‐administration of clopidogrel and cilostazol compared with cilostazol alone (Table 3 and Figure 2C). A small decrease in the AUC_t molar ratio of 4″‐trans‐hydroxycilostazol vs. cilostazol was observed (GMR 0.91, P = 0.005).

Table 3.

Pharmacokinetic parameters of cilostazol, 3,4‐dehydrocilostazol and 4″‐trans‐hydroxycilostazol in healthy subjects after a single oral administration of cilostazol alone or co‐administration of clopidogrel and cilostazol

	Cilostazol	Clopidogrel + cilostazol	Ratio (90% CI)	P value
Cilostazol
AUCt (ng h ml−1)	8235.34 ± 2142.97	7763.40 ± 2116.56	0.94* (0.88 ~ 1.00)	0.100
Cmax (ng ml−1)	666.91 ± 176.21	594.20 ± 166.92	0.88* (0.81 ~ 0.97)	0.031
Tmax (h, median, range)	4 (2 ~ 6)	4 (2 ~ 6)	1.25 (0.99 ~ 1.51)	0.491
3,4‐dehydrocilostazol
AUCt (ng h ml−1)	2217.72 ± 743.78	2110.96 ± 830.95	0.93* (0.86 ~ 1.00)	0.104
Cmax (ng ml−1)	133.46 ± 43.47	119.40 ± 44.62	0.88* (0.79 ~ 0.97)	0.030
Tmax (h, median, range)	8 (4 ~ 24)	8 (4 ~ 24)	1.05 (0.94 ~ 1.16)	0.574
4″‐trans‐hydroxycilostazol (M2)
AUCt (ng h ml−1)	616.36 ± 153.37	521.64 ± 118.04	0.85* (0.80 ~ 0.90)	<0.001
Cmax (ng ml−1)	50.06 ± 20.02	37.64 ± 11.99	0.77* (0.70 ~ 0.85)	<0.001
Tmax (h, median, range)	4 (2 ~ 8)	4 (3 ~ 6)	1.16 (1.00 ~ 1.32)	0.396
AUCM1/AUCcilostazol ‡	0.268 ± 0.038	0.265 ± 0.041	0.99 (0.96 ~ 1.02)	0.557
AUCM2/AUCcilostazol §	0.077 ± 0.031	0.069 ± 0.027	0.91 (0.88 ~ 0.93)	0.005

AUC, area under the concentration–time curve; CI, confidence interval; C_max, maximum concentration; T_max, time to C_max.

^*GMR, geometric mean ratio of clopidogrel + cilostazol vs. cilostazol).

^‡AUC_M1/AUC_cilostazol, AUC molar ratio of 3,4‐dehydrocilostazol vs. cilostazol.

^§AUC_M2/AUC_cilostazol, AUC molar ratio of 4″‐trans‐hydroxycilostazol vs. cilostazol.

Figure 2.

Mean concentration–time profiles of (A) cilostazol, (B) 3,4‐dehydrocilostazol and (C) 4″‐trans‐hydroxycilostazol after a single oral administration of 100 mg cilostazol and co‐administration of 300 mg clopidogrel and 100 mg cilostazol. cilostazol, clopidogrel+cilostazol

The AUC_t of cilostazol and 3,4‐dehydro cilostazol was highest in the CYP2C19 PM genotype after administration of cilostazol both with and without clopidogrel, and that of 4″‐trans‐hydroxycilostazol was lowest for the CYP2C19 PM genotype (Table 4). No significant differences in the AUC_t of cilostazol or 3,4‐dehydro cilostazol were observed in any of the CYP2C19 genotype groups between cilostazol with and without clopidogrel, whereas the AUC_t of 4″‐trans‐hydroxycilostazol was decreased after co‐administration of cilostrazol and clopidogrel.

Table 4.

Comparison of the area under the concentration–time curve for cilostazol, 3,4‐dehydrocilostazol and 4″‐trans‐hydroxycilostazol according to CYP2C19 or CYP3A5 genotypes in healthy subjects after a single oral administration of cilostazol alone or co‐administration of clopidogrel and cilostazol

		Cilostazol	Clopidogrel + cilostazol	GMR† (90% CI)	P value
CYP2C19 ‡	Cilostazol
	EM	7919.84 ± 2420.78	7184.35 ± 2286.82	0.90 (0.81 ~ 0.99)	0.085
	IM	6885.45 ± 1501.44	6966.04 ± 1584.03	1.01 (0.91 ~ 1.12)	0.883
	PM	9900.75 ± 1236.22	9139.79 ± 1896.59	0.91 (0.80 ~ 1.04)	0.220
	P value	0.006	0.075
	3,4‐dehydro cilostazol
	EM	2110.55 ± 775.76	1945.68 ± 827.25	0.89 (0.79 ~ 1.01)	0.127
	IM	1860.82 ± 680.19	1791.53 ± 645.93	0.96 (0.84 ~ 1.08)	0.508
	PM	2681.78 ± 578.77	2595.43 ± 856.33	0.94 (0.78 ~ 1.12)	0.498
	P value	0.023	0.188
	4″‐trans‐hydroxycilostazol
	EM	719.4 ± 183.40	599.31 ± 142.23	0.79 (0.72 ~ 0.87)	0.006
	IM	614.10 ± 89.51	512.29 ± 72.70	0.86 (0.75 ~ 0.98)	0.060
	PM	515.57 ± 106.86	453.34 ± 87.36	0.87 (0.77 ~ 0.99)	0.104
	P value	0.013	0.024
CYP3A5	Cilostazol
	*1/*3	7043.78 ± 1936.67	5859.26 ± 1206.72	0.85 (0.73 ~ 0.98)	0.066
	*3/*3	8936.26 ± 1985.79	8883.48 ± 1691.61	1.00 (0.94 ~ 1.06)	0.993
	P value	0.017	<0.001
	3,4‐dehydro cilostazol
	*1/*3	1670.74 ± 578.18	1318.13 ± 389.91	0.80 (0.66 ~ 0.97)	0.062
	*3/*3	2539.47 ± 643.38	2577.20 ± 642.56	1.01 (0.96 ~ 1.07)	0.675
	P value	0.002	<0.001
	4″‐trans‐hydroxycilostazol
	*1/*3	684.33 ± 171.58	513.85 ± 139.43	0.75 (0.53 ~ 0.78)	0.004
	*3/*3	576.38 ± 130.68	526.17 ± 107.92	0.92 (0.78 ~ 0.94)	0.023
	P value	0.065	0.980

CI, confidence interval; EM, extensive metabolizer; IM, intermediate metabolizer; PM, poor metabolizer.

^†GMR, geometric mean ratio of clopidogrel + cilostazol vs. clopidogrel.

^‡CYP2C19 EM includes the CYP2C19*1/*1 genotype subset; CYP2C19 IM includes the CYP2C19*1/*2 and CYP2C19*1/*3 genotype subsets; CYP2C19 PM includes the CYP2C19*2/*2, CYP2C19*2/*3 and CYP2C19*3/*3 genotype subsets.

Similar to the CYP2C19 genotype, the AUC_t values of cilostazol and 3,4‐dehydro cilostazol were higher in the CYP3A5*3/*3 than in the CYP3A5*1/*3 genotype group after administration of cilostazol both with and without clopidogrel (Table 4). The AUC_t of 4″‐trans‐hydroxycilostazol was lower in the CYP3A5*3/*3 compared with the CYP3A5*1/*3 genotype group, but without statistical significance (Table 4). No significant difference in the AUC_t of cilostazol and 3,4‐dehydro cilostazol was observed between cilostazol with and without clopidogrel in the CYP3A5*1/*3 and CYP3A5*3/*3 genotype groups, whereas the AUC_t of 4″‐trans‐hydroxycilostazol was decreased after co‐administration of cilostrazol and clopidogrel.

Pharmacodynamic results

The platelet reactivity–time profile and IPA–time profile after co‐administration of clopidogrel and cilostazol were similar to the those after administration of clopidogrel alone (Supplemental Figures S2A and Figure 3A). The IPA_max, IPA_4–24 values and AUEC after co‐administration of clopidogrel and cilostazol were significantly higher than those after administration of cilostazol alone (64.87 ± 14.98 vs. 20.45 ± 15.26 for IPA_max, P < 0.001; 54.18 ± 13.37 vs. 10.21 ± 9.71 for IPA_4‐24, P < 0.001; 1217 ± 299 vs. 212 ± 204 for AUEC, P < 0.001) but not significantly different following administration of clopidogrel alone (Table 5). Moreover, IPA_max and IPA_4–24 mean difference values for clopidogrel vs. clopidogrel + cilostazol were less than half of the IPA_max and IPA_4–24 values after administration of cilostazol alone (6.60 vs. 20.45 for IPA_max; 4.79 vs. 10.21 for IPA_4–24). A similar trend was observed when stratified by CYP2C19 or CYP3A5 genotype.

Figure 3.

Mean inhibition of platelet aggregation (IPA)–time profiles after a single oral administration of 300 mg clopidogrel alone, 100 mg cilostazol alone or co‐administration of 300 mg clopidogrel and 100 mg cilostazol in all subjects (A), and comparison of mean IPA–time profiles among CYP2C19 extensive metabolizers (EM), intermediate metabolizers (IM) and poor metabolizers (PM) in CYP3A5*1/*3 genotypes (B) and CYP3A5*3/*3 genotypes (C) after co‐administration of clopidogrel and cilostazol. (A) Clopidogrel, Cilostazol, Clopidogrel+Cilostazol. (B) CYP2C19 EM (n=3), CYP2C19 IM (n=4), CYP2C19 PM (n=3). (C) CYP2C19 EM (n=16), CYP2C19 IM (n=5), CYP2C19 PM (n=6)

Table 5.

Pharmacodynamic parameters estimated from inhibition of platelet aggregation (IPA) after a single oral administration of clopidogrel alone, cilostazol alone and co‐administration of clopidogrel and cilostazol

		Clopidogrel (A)	Clopidogrel + cilostazol (C)	C : A		Cilostazol (B)	C : B
				Difference (95% CI)	P value		Difference (95% CI)	P value
All subjects	IPAmax (%)	58.27 ± 17.41	64.87 ± 14.98	6.60 (−3.03,16.23)	0.096	20.45 ± 15.26	44.41 (34.79,54.04)	<0.001
	IPA4–24 (%)	49.40 ± 17.84	54.18 ± 13.37	4.79 (−3.42,12.99)	0.155	10.21 ± 9.71	43.97 (35.76,52.17)	<0.001
	AUEC (%*h)	1111 ± 420	1217 ± 299	147 (3, 292)	0.346	212 ± 204	1005 (868, 1143)	<0.001
CYP2C19	IPA4–24 (%)
	EM	59.05 ± 18.95	62.95 ± 8.62	3.91 (−9.17,16.99)	0.430	6.82 ± 7.00	56.14 (43.06,69.22)	<0.001
	IM	52.40 ± 14.30	53.13 ± 12.48	0.74 (−15.40,16.87)	0.903	11.04 ± 12.28	42.10 (25.96,58.23)	<0.001
	PM	36.74 ± 13.26	46.46 ± 13.99	9.72 (−2.76,22.19)	0.053	12.78 ± 9.25	33.68 (21.20,46.15)	<0.001
	P value	0.023	0.039			0.289
	AUEC (%*h)
	EM	1356 ± 431	1411 ± 190	55 (−247, 357)	0.895	172 ± 173	1239 (1035, 1443)	<0.001
	IM	1147 ± 359	1193 ± 267	174 (−148, 497)	0.923	213 ± 245	980 (699, 1261)	<0.001
	PM	834 ± 310	1046 ± 328	212 (−13, 439)	0.200	251 ± 205	796 (583,1008)	<0.001
	P value	0.039	0.049			0.613
CYP3A5	IPA4–24 (%)
	*1/*3	52.23 ± 21.54	56.12 ± 15.47	5.06 (−9.57,19.70)	0.369	9.95 ± 6.60	46.46 (31.82,61.10)	<0.001
	*3/*3	47.73 ± 15.76	53.04 ± 12.33	5.34 (−5.13,15.81)	0.206	10.37 ± 11.34	42.51 (32.04,52.98)	<0.001
	P value	0.530	0.564			0.782
	AUEC (%*h)
	*1/*3	1213 ± 506	1232 ± 351	19 (−182, 221)	0.940	201 ± 125	1031 (758, 1305)	<0.001
	*3/*3	1047 ± 359	1208 ± 276	223 (20, 426)	0.249	219 ± 243	989 (816, 1162)	<0.001
	P value	0.343	0.581			0.581

AUEC, area under the effect (IPA)–time curve from time 0 to 24 h after dosing; CI, confidence interval; EM, extensive metabolizer; IM, intermediate metabolizer; IPA_4–24, average value of IPA 4–24 h after treatment; IPA_max, maximum effect of inhibition of platelet aggregation obtained from IPA–time curves; PM, poor metabolizer. CYP2C19 EM includes the CYP2C19*1/*1 genotype subset; CYP2C19 IM includes the CYP2C19*1/*2 and CYP2C19*1/*3 genotype subsets; CYP2C19 PM includes the CYP2C19*2/*2, CYP2C19*2/*3 and CYP2C19*3/*3 genotype subsets.

There was no significant difference in IPA_4–24 and AUEC between CYP3A5*1/*3 and CYP3A5*3/*3 genotypes after administration of clopidogrel both with and without cilostazol. However, when analysed according to CYP2C19 genotypes, the IPA_4–24 and AUEC values were highest in the CYP2C19 EM and lowest in the CYP2C19 PM genotype after administration of clopidogrel alone (59.05 ± 18.95 vs. 36.74 ± 13.26 for IPA_4‐2–4, P = 0.023; 1356 ± 431 vs. 834 ± 310 for AUEC, P = 0.039). By contrast, IPA_4–24 and AUEC values after administration of cilostazol alone tended to be lowest in the CYP2C19 EM and highest in the CYP2C19 PM genotype, without statistical significance. The IPA_4–24 and AUEC after co‐administration of clopidogrel and cilostazol were higher compared with clopidogrel only in all CYP2C19 genotype groups (Table 5). The difference in IPA_4–24 and AUEC between clopidogrel with and without cilostazol was highest in the CYP2C19 PM genotype, although this did not show statistical significance.

IPA after co‐administration of clopidogrel and cilostazol was compared in the CYP2C19 genotype group stratified by CYP3A5 genotype. In the CYP3A5*1/*3 genotype subgroup, the IPA–time profile for CYP2C19 IM but not CYP2C19 PM was similar to that of CYP2C19 EM (Figure 3B; P = 0.041 for EM/IM vs. PM), whereas in the CYP3A5*3/*3 genotype subgroup, the IPA–time profile of CYP2C19 PM was comparable to those of CYP2C19 EM and IM (Figure 3C; P = 0.327 for EM/IM vs. PM). The addition of cilostazol to clopidogrel in CYP3A5*3/*3 carriers seemed fully to overcome the CYP2C19 genotype effects on the clopidogrel pharmacodynamic response. The platelet reactivity–time profile showed a similar trend in the subset analysis of the CYP3A5 genotype (Supplemental Figure S2B and S2C).

Discussion

Triple therapy including cilostazol has been known to be more effective at reducing major adverse cardiovascular events, without increasing the risk of bleeding after PCI, than dual therapy with aspirin and clopidogrel, although this is controversial 2. It has been reported that adding cilostazol improves nonresponsiveness in patients taking clopidogrel with CYP2C19 LOF alleles because the formation of 3,4‐dehydro cilostazol, the metabolite with the most potent activity is increased in patients with CYP2C19 LOF alleles 30. It has been suggested that triple therapy including cilostazol is a better pharmacotherapeutic option than increasing the dosage of clopidogrel in patients with CYP2C19 LOF‐related high platelet reactivity 24, 31. However, no report has been published to explain why the antiplatelet effect of triple therapy is not significantly increased in noncarriers of CYP2C19 LOF alleles, while cilostazol alone has an antiplatelet effect. The pharmacokinetic and pharmacodynamic interaction between clopidogrel and cilostazol is not yet fully understood.

The present study is the first to provide evidence for pharmacokinetic and pharmacodynamic drug interactions between clopidogrel and cilostazol. These results advance the mechanistic understanding of the interaction between clopidogrel and cilostazol in relation to CYP2C19 and CYP3A5.

In the present study, we observed that the concentration of the clopidogrel thiol metabolite was decreased on co‐administration of cilostazol, particularly in subjects with the CYP3A5*1/*3 genotype, compared with administration of clopidogrel alone, but no such difference was evident for the CYP3A5*3/*3 genotype. We found that the additive IPA value of co‐administered cilostazol was highest in subjects with the CYP2C19 PM genotype, particularly those with both the CYP2C19 PM and CYP3A5*3/*3 genotypes, with no decrease in the concentration of the thiol metabolite.

Clopidogrel is metabolized into the thiol metabolite by the enzymes CYP2B6, CYP2C19, CYP3A4 and CYP3A5 5, 32. It is well known that CYP2C19 LOF alleles cause a reduction in the formation of the active thiol metabolite, resulting in a lack of platelet aggregation inhibition, but other metabolic enzymes, including CYP3A5, have been found to have no effect on the formation of the active thiol metabolite 8, 9. We also observed this effect of CYP2C19 on the active thiol metabolite and no effect of CYP3A5 after administration of clopidogrel alone. Interestingly, after co‐administration of clopidogrel and cilostazol, the C_max and AUC of the thiol metabolite were decreased, especially in subjects with the CYP3A5*1/*3 genotype, compared with administration of clopidogrel alone. It was reported that CYP3A inhibitors decrease the formation of the clopidogrel thiol metabolite 7, leading to decreased IPA 33. Bramer et al. reported that the concentration of lovastatin and its β‐hydroxy acid metabolite, CYP3A substrate, was increased by up to twofold after multiple co‐administrations of cilostazol 34. This suggests that cilostazol or its metabolites might have a mild inhibitory effect on CYP3A. Thus, cilostazol and/or its metabolites might inhibit the formation of the clopidogrel thiol metabolite through the CYP3A pathway. Some in vitro and in vivo data indicate that CYP3A5 expressers (CYP3A5*1/*1 and CYP3A5*1/*3) have higher metabolic activity and are significantly more susceptible to CYP3A inhibitors than are CYP3A5 nonexpressers (CYP3A5*3/*3) 2, 30. In line with these reports, decreased Cmax and AUC of the thiol metabolite after co‐administration of clopidogrel and cilostazol was observed only in subjects with the CYP3A5*1/*3 genotype and not in those with the CYP3A5*3/*3 genotype. However, the CYP3A5*1/*1 genotype was not observed in the present study. We therefore could not evaluate the differences in the pharmacokinetic parameters of the thiol metabolite between clopdigorel with and without cilostazol in subjects with the CYP3A5*1/*1 genotype. In addition, we did not evaluate the effects of CYP3A4 because the clinical significance of the CYP3A4 variant allele is minimal 35. Further exploration of the effect of cilostazol on clopidogrel thiol metabolite pharmacokinetics in subjects with the CYP3A5*1/*1 genotype, and on the effects of CYP3A4 on the clopidogrel–cilostazol interaction might be necessary.

The concentration of 4″‐trans‐hydroxycilostazol following co‐administration with clopidogrel was significantly decreased compared with cilostazol alone. 4″‐Trans‐hydroxycilostazol is formed via CYP2C19 and CYP3A 19, and clopidogrel is a known CYP2C19 inhibitor 36. We speculate that the lower concentration of 4″‐trans‐hydroxycilostazol following co‐administration with clopidogrel was caused by clopidogrel's inhibition of CYP2C19 activity. However, there were no differences in the pharmacokinetic parameters of cilostazol and 3,4‐dehydrocilostazol. 4“‐trans‐hydroxycilostazol has the lowest pharmacologic activity and the lowest plasma level among cilostazol and its two metabolites. Cilostazol has the immediate potency and the highest plasma level and 3,4‐dehydrocilostazol has the most potent activity and the immediate plasma level. The change in the low plasma level of 4″‐trans‐hydroxycilostazol might not have affected the high plasma levels of cilostazol and 3,4‐dehydrocilostazol. Thus, the pharmacokinetic changes in cilostazol and its metabolites might have had only a minimal effect on the pharmacodynamic changes after co‐administration of clopidogrel.

In the present study, IPA mean difference values for clopidogrel vs. clopidogrel + cilostazol were less than half of the IPA value for cilostazol alone. This might have been caused by the decreased concentration of the clopidogrel thiol metabolite after co‐administration of cilostazol. In addition, the IPA–time profile of CYP2C19 IM after co‐administration of clopidogrel and cilostazol was similar to that of CYP2C19 EM in the CYP3A5*1/*3 genotype subset, but in the CYP3A5*3/*3 genotype subset the IPA–time profile of CYP2C19 PM was comparable to that of CYP2C19 EM and IM. This might have resulted from the lack of decreased exposure to the clopidogrel thiol metabolite in CYP3A5*3/*3 after co‐administration of clopidogrel and cilostazol. The lower IPA value of clopidogrel in the CYP2C19 PM genotype was overcome only in the CYP3A5*3/*3 genotype subset after co‐administration of clopidogrel and cilostazol.

In conclusion, the additive antiplatelet effects of cilostazol and clopidogrel was maximized in subjects with both CYP2C19 PM and CYP3A5*3/*3 owing to the lack of decrease in the concentration of the thiol metabolite by cilostazol in CYP3A5*3/*3 carriers as well as the highest IPA of cilostazol in CYP2C19 PM. The present study was conducted in a small number of healthy subjects following the single administration of clopidogrel and cilostazol. A larger cohort study, with a long‐term follow‐up is needed to validate the CYP2C19 and CYP3A5 genotype‐based pharmacokinetic and pharmacodynamic interactions between clopidogrel and cilostazol at steady state.

Competing Interests

All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: support from Korea Science and Engineering Foundation (KOSEF) funded by the Ministry of Education, Science and Engineering (MOEST) (No. R13‐2007‐023‐00 000‐0) for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work; no other relationships or activities that could appear to have influenced the submitted work.

We appreciate the contribution of the study participants. We thank Yoon‐Mi Lee and Yeon‐ha Kim for assistance in conducting the clinical trial, Eun‐Young Cha for genotype analysis and Yune‐jung Yoon for drug analysis. This study was supported by a grant from the Korean Health Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (HI14C0067) and by the Korea Science and Engineering Foundation (KOSEF) grant funded by the Ministry of Education, Science and Engineering (MOEST) (No. R13‐2007‐023‐00000‐0).

Contributors

HSK and YL contributed equally to the study concept, design, interpretation of data, and critical writing and revision of the intellectual content. MO contributed to data analysis. JLG and EYK contributed to the study design and clinical trial. DHK contributed to drug analysis and the interpretation of data. JGS was the principal investigator and was responsible for the study concept, design, interpretation of data and critical writing. All authors approved the final version of this manuscript.

Supporting information

Figure S1 Design of three‐way crossover study, in which an oral single dose of 100 mg cilostazol alone, 300 mg clopidogrel alone or 100 mg cilostazol plus 300 mg clopidogrel were administered at each period, with a 2‐week washout period. PD, pharmacodynamics; PK, pharmacokinetics

Figure S2 Mean platelet reactivity–time profiles after a single oral administration of 300 mg clopidogrel alone, 100 mg cilostazol alone or co‐administration of 300 mg clopidogrel and 100 mg cilostazol in all subjects (A), and comparison of mean IPA–time profiles among CYP2C19 extensive metabolizers (EM), intermediate metabolizers (IM) and poor metabolizers (PM) in CYP3A5*1/*3 genotypes (B) and CYP3A5*3/*3 genotypes (C) after co‐administration of clopidogrel and cilostazol

Supporting info item

Supporting info item

Kim, H.‐S. , Lim, Y. , Oh, M. , Ghim, J.‐l. , Kim, E.‐Y. , Kim, D.‐H. , and Shin, J.‐G. (2016) The pharmacokinetic and pharmacodynamic interaction of clopidogrel and cilostazol in relation to CYP2C19 and CYP3A5 genotypes. Br J Clin Pharmacol, 81: 301–312. doi: 10.1111/bcp.12794.