Duloxetine pharmacokinetics are similar in Japanese and Caucasian subjects

Abstract

What is already known about this subject

• The pharmacokinetics of duloxetine have been assessed in a number of clinical studies.

• Duloxetine is eliminated through oxidative metabolism via CYP1A2 and, to a lesser degree, CYP2D6.

• There is strong evidence that the prevalence of CYP2D6 phenotypes and the activity of CYP1A2 enzyme activity differ between Japanese and Caucasians.

• Given the characteristics of duloxetine metabolism, there is good reason to assess pharmacokinetic differences between Japanese and Caucasians.

What this study adds

• Duloxetine pharmacokinetics in Japanese or Caucasian subjects is not meaningfully different after single or multiple doses of duloxetine.

• The magnitude of pharmacokinetic differences between groups is small relative to the pharmacokinetic variability in either group, and these small differences can be accounted for by differences in body weight.

• The result of this study suggests that different dose recommendations for Caucasian or Japanese patients are not likely to be necessary.

Aims

To compare single- and multiple-dose duloxetine pharmacokinetics between healthy Japanese and Caucasians.

Methods

Twenty-four subjects of each race were given single oral doses of duloxetine (20, 40 and 60 mg) in a randomized, double-blind study. Another 20 subjects of each race received 20, 40 mg or placebo (2 : 2 : 1) twice-daily for 5 days.

Results

Following single doses, the mean duloxetine C_max and AUC were approximately 20% greater in Japanese. This difference could be explained by the 15% lower average body weight in Japanese. Similar results were observed following multiple dosing.

Conclusion

Duloxetine pharmacokinetics are not meaningfully different between Japanese and Caucasians.

Introduction

Duloxetine, a potent dual inhibitor of norepinephrine and serotonin uptake, is used to treat depression [1], diabetic peripheral neuropathic pain and stress urinary incontinence [2]. Duloxetine is well absorbed after oral dosing, reaching a peak concentration in approximately 6 h in fasted subjects and declining with a half-life of approximately 12 h [3]. Duloxetine is extensively metabolized by cytochrome P450 (CYP) 1A2 and to a lesser degree by CYP2D6 to form oxidative metabolites, which are then conjugated before being renally excreted [4]. The circulating metabolites of duloxetine are inactive.

In clinical studies, fluvoxamine, a strong CYP1A2 inhibitor, increased duloxetine peak concentrations 2.5-fold and overall exposure six fold [1]. Paroxetine, a potent CYP2D6 inhibitor, increased duloxetine exposure only 60%[3], to a level similar to duloxetine exposure in CYP2D6 poor metabolizers (PM). Steady-state duloxetine concentrations in PMs do not exceed the upper range of typical duloxetine exposure range.

Racial differences in genotypic and phenotypic distributions of CYP2D6 have been reported. In Caucasians, 5–10% of the population are CYP2D6 PM, whereas only 1% of the Oriental population has this phenotype [5]. Oriental CYP2D6 extensive metabolizers (EM) have lower CYP2D6 activity due to the higher allele frequency of CYP2D6*10 [6]. Similarly, evidence suggests that CYP1A2 activity in Orientals is lower than in Caucasians [5]. A trimodal distribution in CYP1A2 activity has been suggested for the Caucasian population, whereas a bimodal distribution may exist for Japanese [6]. Genetic polymorphism and enzyme activities are factors that may cause racial differences in drug responsiveness and adverse events, which in substantial cases may require different dosage recommendations for certain racial groups.

Given the characteristics of duloxetine metabolism, there is good reason to assess pharmacokinetic differences between Japanese and Caucasians. The aim of this study was to compare rigorously the single- and multiple-dose pharmacokinetics of duloxetine in Japanese and Caucasians.

Methods

Subjects and clinical protocol

Healthy adult Japanese and Caucasian subjects in this study self-reported having an ancestral background from a single racial group for all ancestors for at least two prior generations. At the time of the study, the subjects resided in Australia. The study was approved by the Research Ethics Committee of the Royal Adelaide Hospital, Australia, and performed in accordance with the Declaration of Helsinki. Subjects could withdraw from the study at any time.

For the single-dose study, healthy Japanese and Caucasian subjects, as determined by medical history and physical examination, gave written informed consent. Participation of 24 subjects in each group provided 80% power to show a 30% difference in duloxetine exposure between the racial groups. Using a double-blind, three-period design, subjects were randomized to single oral doses of 20, 40 and 60 mg duloxetine (Eli Lilly and Co., Indianapolis, IN, USA) whilst fasted. In each period, serial venous blood samples (5 ml) were collected predose (periods 2 and 3 only) and at 1, 2, 4, 6, 8, 10, 12, 24, 36, 48 and 72 h postdose. There were at least 4 days of wash-out between study periods.

For the multiple-dose study, another 20 healthy subjects of each racial group, who had not participated in the single-dose study, were enrolled. The number of subjects was chosen to provide adequate safety data for each treatment group. Using a double-blind design, subjects received 20 or 40 mg of duloxetine or placebo (2 : 2 : 1), twice-daily for a total of nine doses over 5 days. A blood sample was collected before the morning dose on days 3 and 4. On day 5, a series of blood samples was collected using time intervals similar to those in the single-dose study.

A blood sample was collected from all subjects for CYP2D6 genotyping. Throughout the study, all subjects were monitored for clinical and laboratory safety.

Assays

Plasma harvested from centrifuged blood samples was stored at −70 °C until analysis. Duloxetine concentrations were assayed at Prevalere Life Sciences, Inc. (Whitesboro, NY, USA) using a previously described validated LC/MS/MS method [4]. The lower and upper limits of quantification were 0.5 and 100 ng ml⁻¹, respectively. The assay’s overall relative standard deviation (RSD) was ≤4.32% and the overall relative error (RE) was ≤3.58%.

CYP2D6 was genotyped using a commercially available method (DNA Sciences Laboratories, Morrisville, NC, USA), and subjects were classified as ultrarapid (UM), extensive (EM), intermediate (IM) or poor metabolizers (PM).

Pharmacokinetic analysis

Data were analysed using noncompartmental pharmacokinetic methods (WinNonlin Professional version 3.1). Maximum plasma duloxetine concentrations (C_max and C_max,ss), and the time to maximum (T_max and T_max,ss) were from the observed data. The elimination half-life (t_1/2), apparent volume of distribution (V_z/F and V_ss/F), area under the plasma concentration–time curve from dosing extrapolated to infinity (AUC) following a single dose, and during multiple dosing on day 5 during a 24-h period (AUC_τ,ss), were calculated. Dose and body weight-normalized estimates of AUC (AUC_norm and AUC_τ,ss,norm) and C_max (C_max,norm and C_max,ss,norm) were derived by dividing the corresponding values by mg kg⁻¹ dose.

Statistical analysis

Statistical analysis used a mixed-effects model (SAS System ® for Windows Version 8.2; SAS Inc., Cary, NC, USA). For the single-dose study, log-transformed pharmacokinetic parameter estimates were fitted with log-dose as a covariate. Race, period and sequence were fixed effects, and subject was fitted as a random effect in the model. The ratio of geometric means between racial groups and its two-sided 90% confidence interval (CI) were estimated.

For the multiple-dose study, pharmacokinetic parameter estimates were log-transformed and fitted to an anova model with race, dose and race × dose as fixed effects.

Results

Subject demographics are presented in Table 1. Despite similar body mass index (BMI) values, Japanese subjects had a lower mean body weight in the single- (17% lower, P< 0.001) and multiple-dose (13% lower; P= 0.002) studies.

Table 1.

Demographic description of the subjects

					Mean (CV%)
Study description	Race	n	Gender	Age, years	Weight (kg)	BMI (kg m−2)	CYP2D6 status
Single dose	Japanese	24	14 Female 10 Male	25 (20–31)	58.7 (15.2)	21.9 (9.6)	17 EM, 6 IM, 1 PM
Single dose	Caucasian	24	13 Female 11 Male	25 (20–47)	70.7 (15.5)	24.2 (11.7)	21 EM, 3 PM
Multiple doses	Japanese	16*	4 Female 12 Male	23 (20–34)	63.1 (9.2)	22.2 (9.3)	10 EM, 6 IM
Multiple doses	Caucasian	16*	5 Female 11 Male	25 (19–48)	72.5 (12.7)	24.4 (9.3)	15 EM, 1 UM

Age shown as median (range).

^*Number of subjects assigned to active drug. CV, Coefficient of variation; UM, ultrarapid metabolizer; EM, extensive metabolizer; IM, intermediate metabolizer (reduced activity with CYP2D6 *10 genotype); PM, poor metabolizer.

Following single doses, no statistically significant racial difference was detected for C_max, AUC, t_1/2 or V_z/F, although the C_max (18%) and AUC (22%) tended to be higher in the Japanese group. This trend disappeared after normalization for dose and body weight (Table 2).

Table 2.

Statistical comparison of the least squares (LS) geometric mean pharmacokinetic parameters following single and multiple oral doses of duloxetine in the Japanese and Caucasian racial groups

Parameter	Dose (mg)	LS geometric mean Japanese (90% CI)	LS geometric mean Caucasian (90% CI)	Ratio of geometric mean (J/C) (90% CI)	P-value for difference in geometric mean
Single-dose study
Cmax (ng ml−1)	20	14.7 (12.8, 17.0)	12.5 (10.8, 14.4)
	40	30.0 (26.2, 34.5)	25.5 (22.2, 29.3)
	60	45.6 (39.6, 52.5)	38.7 (33.6, 44.6)	1.18 (0.97, 1.43)	0.164
AUC (ng ml−1 h)	20	207 (172, 248)	169 (141, 203)
	40	452 (377, 541)	370 (308, 443)
	60	714 (595, 857)	584 (486, 702)	1.22 (0.95, 1.58)	0.194
t1/2 (h)	20	10.5 (9.72, 11.2)	9.67 (8.98, 10.4)
	40	10.8 (10.0, 11.5)	9.95 (9.27, 10.7)
	60	10.9 (10.2, 11.8)	10.1 (9.41, 10.9)	1.08 (0.98, 1.19)	0.195
Vz/F (l)	20	1459 (1261, 1689)	1648 (1423, 1909)
	40	1374 (1192, 1583)	1551 (1346, 1789)
	60	1326 (1148, 1532)	1497 (1295, 1731)	0.89 (0.72, 1.08)	0.314
Cmax,norm (ng ml−1)/(mg kg−1)	20	43.0 (37.6, 49.1)	43.8 (38.3, 50.1)
	40	43.7 (38.5, 49.7)	44.6 (39.2, 50.7)
	60	44.2 (38.7, 50.4)	45.0 (39.5, 51.4)	0.98 (0.82, 1.18)	0.861
AUCnorm (ng ml−1 h)/(mg kg−1)	20	601 (508, 710)	591 (499, 699)
	40	657 (557, 774)	646 (548, 762)
	60	692 (586, 817)	681 (576, 804)	1.02 (0.81, 1.28)	0.907
Multiple-dose study
Vss/F (l)	20	1548 (1156, 2073)	1784 (1332, 2389)	0.87 (0.57, 1.31)	0.564
	40	1175 (877, 1573)	1681 (1256, 2252)	0.70 (0.46, 1.06)	0.151
Cmax,ss,norm (ng ml−1)/(mg kg−1)	20	81.4 (59.3, 112)	72.2 (52.6, 99.1)	1.13 (0.72, 1.76)	0.654
	40	108 (78.4, 148)	80.7 (58.8, 111)	1.33 (0.85, 2.08)	0.284
AUCτ,ss,norm (ng ml−1 h)/(mg kg−1)	20	668 (481, 927)	604 (435, 840)	1.10 (0.69, 1.76)	0.718
	40	966 (695, 1341)	693 (499, 963)	1.39 (0.88, 2.22)	0.235

During the multiple-dose study, steady state was reached by the third dosing day. Mean concentration–time profiles on day 5 were predictable based upon nonparametric superposition of the single-dose data. The median T_max,ss was 6 h and was similar to the single-dose median T_max. There were no differences between the two racial groups for any multiple-dose pharmacokinetic parameters calculated (Table 2).

Japanese subjects consistently reported approximately 40% fewer adverse events than the Caucasian subjects. The most common adverse events related to duloxetine were nausea, headache and diarrhoea after a single dose, and headache, dizziness and fatigue during multiple dosing. There was a trend toward a higher incidence of adverse events for the larger doses. No apparent differences in severity of adverse events were observed between racial groups.

Discussion

This study compared the single- and multiple-dose pharmacokinetics of duloxetine between healthy Japanese and Caucasians. The observed duloxetine pharmacokinetics and variability study are comparable to those previously described [3]. While the mean exposure was generally 20% higher in the Japanese group, the differences were not statistically significant and are unlikely to be relevant, especially considering duloxetine’s typical pharmacokinetic variability in both racial groups. After normalizing pharmacokinetic parameter estimates for dose and body weight, any semblance of a pharmacokinetic difference between the groups was resolved. These analyses suggest that any pharmacokinetic differences are attributable to the known differences in body weight between Japanese and Caucasians.

Differences in hepatic drug metabolism have been commonly cited as a factor influencing the pharmacokinetics of drugs in various racial groups [5, 6]. Duloxetine is cleared from the body by multiple enzymatic pathways. Therefore, even if there are intrinsic differences in a specific CYPP450 activity between the two racial groups, the contribution of each metabolic pathways is difficult to detect in this clinical study.

Examining the pharmacokinetic and genotype data together provides further evidence that duloxetine metabolism is not limited to a single pathway. Only four subjects (one Japanese, three Caucasians) from the single-dose groups were classified as CYP2D6 PM Although two of the PMs (one Japanese, one Caucasian) had one- to threefold higher duloxetine exposure compared with the mean of subjects who were not PMs, the two other PMs had pharmacokinetic parameter estimates comparable to those in the other subjects. Furthermore, two subjects from the 40-mg multiple-dose treatment groups had two- to five fold higher exposure compared with the respective treatment mean, yet both were genotyped as EM. The high duloxetine concentrations observed in these subjects suggests that exposure cannot be predicted by knowledge of CYP2D6 metabolizer status alone, and other factors, such as the degree of expression of CYP1A2 activity, appear to affect duloxetine pharmacokinetics more substantially.

In conclusion, although the mean exposure tends to be approximately 20% higher in Japanese subjects because of their lower body weight, duloxetine pharmacokinetics in Japanese and Caucasian subjects are not meaningfully different and are unlikely to necessitate different dose recommendations for Caucasian and Japanese patients.