Effect of the CYP2C19 oxidation polymorphism on fluoxetine metabolism in Chinese healthy subjects

Abstract

Aims

The study was designed to investigate whether genetically determined CYP2C19 activity affects the metabolism of fluoxetine in healthy subjects.

Methods

A single oral dose of fluoxetine (40 mg) was administrated successively to 14 healthy young men with high (extensive metabolizers, n =8) and low (poor metabolizers, n = 6) CYP2C19 activity. Blood samples were collected for 5–7 half-lives and fluoxetine, and norfluoxetine were determined by reversed-phase high performance liquid chromatography.

Results

Poor metabolizers (PMs) showed a mean 46% increase in fluoxetine peak plasma concentrations (C_max, P < 0.001), 128% increase in area under the concentration vs time curve (AUC(0,∞), P < 0.001), 113% increase in terminal elimination half-life (t_½) (P < 0.001), and 55% decrease in CL_o (P < 0.001) compared with extensive metabolizers (EMs). Mean ± (s.d) norfluoxetine AUC(0,192 h) was significantly lower in PMs than that in EMs (1343 ± 277 vs 2935 ± 311, P < 0.001). Mean fluoxetine C_max and AUC(0,∞) in wild-type homozygotes (CYP2C19*1/CYP2C19*1) were significantly lower than that in PMs (22.4 ± 3.9 vs 36.7 ± 8.9, P < 0.001; 732 ± 42 vs 2152 ± 492, P < 0.001, respectively). Mean oral clearance in individuals with the wild type homozygous genotype was significantly higher than that in heterozygotes and that in PMs (54.7 ± 3.4 vs 36.0 ± 8.7, P < 0.01; 54.7 ± 3.4 vs 20.6 ± 6.2, P < 0.001, respectively). Mean norfluoxetine AUC(0,192 h) in PMs was significantly lower than that in wild type homozygotes (1343 ± 277 vs 3163 ± 121, P < 0.05) and that in heterozygotes (1343 ± 277 vs 2706 ± 273, P < 0.001), respectively.

Conclusions 

The results indicated that CYP2C19 appears to play a major role in the metabolism of fluoxetine, and in particular its N-demethylation among Chinese healthy subjects.

Introduction

Fluoxetine is a potent and selective serotonin reuptake inhibitor in the central nervous system and is widely used to treat depression and obsessive-compulsive behaviour [1, 2]. The drug mainly undergoes N-demethylation leading to the formation of norfluoxetine, which has a much longer half-life and a similar potency to fluoxetine [3]. The drug is extensively metabolized by the hepatic cytochrome P450 enzymes (CYP) and less than 2.5% is excreted unchanged in urine [3]. Previous studies have shown that fluoxetine and norfluoxetine differentially inhibit the activity of CYP2D6, 3A4, and 2C19 [4]. However, little is known with regard to which CYP forms are responsible for the N-demethylation of fluoxetine. The disposition of this drug and its stereoselective metabolism are associated with the polymorphic oxidation of debrisoquine, indicating that CYP2D6 is probably a major enzyme involved in its biotransformation [5, 6]. However, von Moltke et al. [7] and our recent studies [8] found that CYP2C9 and polymorphic CYP2C19 play an important role in the N-demethylation of fluoxetine at low and high substrate concentrations, respectively, in human liver microsomes. Thus, there is some confusion as to which CYP forms are responsible for the N-demethylation of fluoxetine. The present study was designed to determine the relationship between the pharmacokinetics of fluoxetine and the genetic polymorphism of CYP2C19.

Methods

Subjects

Eight male EMs with respect to CYP2C19 and six PMs with the CYP2C19*2 or CYP2C19*3 mutation whose genotype was determined previously [9, 10] were enrolled in this study. Four of the EMs were genotyped as homozygous wild type CYP2C19*1/CYP2C19*1, four as heterozygous CYP2C19*1/CYP2C19*2. Four of the PMs were genotyped as CYP2C19*2/CYP2C19*2 and two as CYP2C19*2/CYP2C19*3. The age of the subjects ranged from 18 to 22 years (20.1 ± 1.1, mean ± s.d.) and the mean body weight from 55 to 80 kg (63.0 ± 6.4). No subject had taken any medication, or alcohol or had smoked for at least 2 weeks before the study.

All individuals were healthy as determined by medical history, physical examination, and no biochemical evidence of renal or hepatic disorders was found. The study protocol was approved by the Ethics Committee of Hunan Medical University and written informed consent was obtained from all participants.

Protocol

After an overnight fast, all the subjects took 40 mg fluoxetine hydrochloride (Eli Lilly and Company, USA) with 250 ml water. Subjects were confined to the research facility during the first 12 h of the study period and were not allowed to lie down until 4 h after drug administration. Venous blood samples (10 ml) were collected into heparinized tubes from a forearm vein of each subject immediately at 0, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 24, 36, 48, 72, 96, 120, 144, 168, and 192 h after administration of the drug. Plasma was obtained and kept frozen at −20°C until analysed.

Plasma fluoxetine and norfluoxetine analysis

Fluoxetine and norfluoxetine plasma concentrations were determined by a modified high performance liquid chromatographic method [11]. In brief, 0.2 ml of sodium carbonate (1.0 mol l⁻¹, pH 12) and 400 ng clomipramine (internal standard) were added to 1 ml of plasma. Samples were then briefly vortex-mixed and subsequently extracted with 6 ml of n-hexane/acetonitrile (98/2, vol:vol) on a reciprocating shaker for 5 min. After centrifugation (2000 g for 10 min), the upper organic layer was evaporated under a gentle stream of nitrogen at 37°C. The residue was dissolved in 0.1 ml mobile phase (0.05 mol l⁻¹ KH₂PO₄ containing acetonitrile, final pH adjusted to pH 2.6 using H₃PO₄) and a 20 µl aliquot was injected onto the chromatograph. This consisted of a Hewlett Packard (HP) 1050 series instrument, an Eclipse XDB-C₈ column (150 × 4.6 mm) and an ultraviolet detector set at 226 nm. Calibration curves (5–500 ng ml⁻¹) showed a good linear relationship for both fluoxetine (r = 0.999) and norfluoxetine (r = 0.999). The limit of detection was 2 ng ml⁻¹ for both fluoxetine and norfluoxetine. The average recoveries of both fluoxetine and norfluoxetine ranged from 98%-104%. The coefficients of variation (CV) for intra (n = 5) and interday (n = 11) reproducibility ranged from 3.5%-6.8% and 3.1%-6.7% for both compounds, respectively.

Pharmacokinetic analysis

A noncompartmental approach was chosen to determine fluoxetine elimination constant (λ_z) by using the best fit of concentration-time data from 7 h after administration until the end of the study [12]. The area under the plasma concentration-time profile (AUC) was calculated for fluoxetine (AUC(0,∞)) and norfluoxetine (AUC(0,192 h)) using the linear trapezoidal rule. The following equation was used to calculate the oral clearance (CL_o) of fluoxetine assuming 100% bioavailability of the drug:

CL_o = Dose/AUC

The data were analysed by the least significant difference (LSD) [13] test with P < 0.05 as the minimal level of statistical significance.

Results

PMs and EMs showed distinct differences in their plasma concentration-time profiles for fluoxetine and norfluoxetine (Figure 1). The respective pharmacokinetic parameters are summarized in Table 1. PMs showed a mean 46% increase in fluoxetine C_max (P < 0.001), a 128% increase in AUC(0,∞) (P < 0.001), a 113% increase in half-life (t_½) (P < 0.001), and a 55% decrease in CL_o (P < 0.001) compared with EMs. Mean norfluoxetine AUC(0,192 h) was significantly lower in PMs than in EMs (1343 ± 277 vs 2935 ± 311, P < 0.001). Mean fluoxetine C_max and AUC(0,∞) in wild-type homozygotes was also significantly lower than those in PMs, and mean CL_o in homozygote wild-types was significantly higher than that in heterozygotes (P < 0.01) and in PMs (P < 0.001). Mean norfluoxetine AUC(0,192 h) in PMs was significantly smaller than that in wild-type subjects (P < 0.05) and in heterozygotes (P < 0.001), respectively. The PMs had the highest mean AUC for fluoxetine and the smallest AUC for norfluoxetine. Wild-type subjects had the smallest AUC for fluoxetine and the largest AUC for norfluoxetine, with the heterozygotes having intermediate values.

Figure 1.

Mean plasma concentration of fluoxetine (a) and norfluoxetine (b) after a single oral 40 mg dose of fluoxetine in eight extensive metabolizers (•) and six poor metabolizers (○) with respect to CYP2C19. The error bars indicate s.d.

Table 1.

Pharmacokinetic parameters of oral fluoxetine in EMs and PMs with respect to CYP2C19

Genotype					P value (Mean difference; 95% CI)
Parameters	CYP2C19*1/ CYP2C19*1 (A) n = 4	CYP2C19*1/ CYP2C19*2 (B) n = 4	CYP2C19*1/ CYP2C19*1+ CYP2C19*1/ CYP2C19*2 (C) n = 8	CYP2C19*2/ CYP2C19*2+ CYP2C19*2/ CYP2C19*3 (D) n = 6	A vs B	A vs D	B vs D	C vs D
Fluoxetine	732	1156	944	2152	NS	< 0.001	< 0.001	< 0.001
AUC(0,∞) (µg l−1 h)	±42	±244	±278	±492	(−425; −912, 63)	(−1420; −1865, −975)	(−966; −1440, −550)	(−1208; −1580, −836)
CL (l h−1)	54.7	36.0	45.4	20.6	<0.01	< 0.001	< 0.05	< 0.001
	±3.4	±8.7	±11.7	±6.2	(18.8; 5.7, 31.9)	(34.6; 22.7, 46.6)	(15.8; 3.9, 27.8)	(25.3; 15.3, 35.2)
Cmax (µg l−1)	22.4	27.9	25.2	36.7	NS	< 0.001	< 0.05	< 0.001
	±3.9	±3.1	±4.1	±8.9	(−5.6; −13.6, 2.4)	(−14.3; −21.6, −6.9)	(−8.7; −15.9, −1.4)	(−11.5; −17.6, −5.4)
t1/2 (h)	28.7	28.0	28.4	62.3	NS	< 0.001	< 0.001	< 0.001
	±7.9	±3.3	±6.1	±16.4	(0.75; −14.4, 15.9)	(−33.6; −47.5, −19.7)	(−34.3; −48.2, −20.5)	(−33.9; −45.6, −22.4)
Norfluoxetine	3163	2706	2935	1343	< 0.05	< 0.001	< 0.001	< 0.001
AUC(0,192 h) (µg l−1 h)	±121	±273	±311	±277	(455; 54, 857)	1820; 1453, 2186)	(1365; 978, 1731)	(1592; 1285, 1899)

Data are mean values ± s.d.; NS, Not statistically significant (P > 0.05); AUC, area under the concentration-time curve; C_max, peak plasma concentration; t_½, half-life, CL, clearance.

Discussion

In the present study, we found that PMs showed a markedly higher mean fluoxetine C_max, AUC, a longer t_½, and a small CL_o compared with EMs. Furthermore, norfluoxetine AUC(0,192 h) was significantly lower in PMs than in EMs. Furthermore fluoxetine C_max and AUC(0,∞) in homozygous wild-type subjects were significantly smaller than those in PMs, and CL_o in wild-types was significantly higher than that in heterozygotes and in PMs. Norfluoxetine AUC(0,192 h) in PMs was significantly lower than that in the other genotypic groups. These data suggest that polymorphic metabolism by CYP2C19 is likely to be one of the major factors causing interindividual differences in the steady-state plasma concentrations of fluoxetine and norfluoxetine. They also suggest that metabolism of fluoxetine by CYP2C19 is gene dose-dependent.

Recently, an in vivo study has shown that CYP2D6 is likely to be a major human cytochrome P450 enzyme involved in the disposition of fluoxetine after a single 20 mg dose [5]. However, an in vitro study by von Moltke et al. [7] reported that CYP2C9 appear to be the principal form of human cytochrome P450 mediating fluoxetine N-demethylation, with minor contributions from CYP2C19 and CYP3A4, but their data suggested that CYP2D6 is likely to be of minor importance. Our recent in vitro study [8] confirmed that of von Moltke et al. [7] showing that CYP2C9 is likely to be a major form of CYP catalysing fluoxetine N-demethylation in human liver microsomes at a low substrate concentration, and that polymorphic CYP2C19 may make a significant contribution to this reaction at a high substrate concentration. Moreover, our current in vivo results indicated that CYP2C19 may play an important role in the N-demethylation of fluoxetine after a single oral 40 mg dose of fluoxetine in Chinese healthy subjects. Thus, there are some discrepancies in which CYP isoforms are responsible for the N-demethylation of fluoxetine which have yet to be resolved.

In summary, we investigated the disposition of fluoxetine and its major metabolites in 14 healthy subjects with either high or low genetically determined CYP2C19 activity. We observed significant differences in the pharmacokinetics of fluoxetine and norfluoxetine among the different genotype subjects. These data suggest that formation of norfluoxetine from fluoxetine in humans appears to be mediated primarily by CYP2C19 in Chinese subjects in a gene dose dependent manner.