Effects of clarithromycin on lansoprazole pharmacokinetics between CYP2C19 genotypes

Abstract

Aims

Lansoprazole is a substrate of CYP2C19 and CYP3A. The aim of this study was to compare the inhibitory effects of clarithromycin, an inhibitor of CYP3A on the metabolism of lansoprazole between CYP2C19 genotypes.

Methods

A two-way randomized double-blind, placebo-controlled crossover study was performed. Eighteen volunteers, of whom six were homozygous extensive metabolizers (EMs), six were heterozygous EMs and six were poor metabolizers (PMs) for CYP2C19, received two 6-day courses of either clarithromycin 800 mg or placebo daily in a randomized fashion with a single oral dose of lansoprazole 60 mg on day 6 in all cases. Plasma concentrations of lansoprazole and its metabolites, 5-hydroxylansoprazole and lansoprazole sulphone were monitored up to 24 h after dosing.

Results

During placebo administration, the mean AUC(0, ∞) of lansoprazole in homozygous EMs, heterozygous EMs and PMs were 4652 (95% CI, 2294, 7009) ng ml⁻¹ h, 8299 (4784, 11814) ng ml⁻¹ h and 25293 (17643, 32943) ng ml⁻¹ h (P < 0.001), respectively. Clarithromycin treatment significantly increased C_max by 1.47-fold, 1.71-fold and 1.52-fold and AUC(0, ∞) of lansoprazole by 1.55-fold, 1.74-fold, and 1.80-fold in these genotype groups, respectively, whereas elimination half-life was prolonged only in PMs. The clarithromycin-mediated percent increase in pharmacokinetic parameters such as C_max, AUC(0, ∞) or elimination half-life did not differ between the three CYP2C19 genotypes.

Conclusions

The present study indicates that there are significant drug interactions between lansoprazole and clarithromycin in all CYP2C19 genotype groups probably through CYP3A inhibition. The bioavailability of lansoprazole might, to some extent, be increased through inhibition of P-glycoprotein during clarithromycin treatment.

Introduction

Lansoprazole is a proton pump inhibitor that suppresses gastric acid secretion by inhibiting H⁺, K⁺-ATPase in the secretory membrane of gastric parietal cells [1]. Lansoprazole is effective in the treatment of various peptic diseases, including gastric and duodenal ulcer, reflux oesophagitis, and Zollinger–Ellison syndrome [2].

Lansoprazole is extensively metabolized in the liver; the two major metabolites found in plasma are lansoprazole sulphone and 5-hydroxylansoprazole [3, 4]. In vitro studies with human liver microsomes or cell lines have shown that sulfoxidation of lansoprazole is catalyzed by CYP3A, while its hydroxylation is catalyzed by CYP2C19 at lower concentrations and CYP3A at higher concentrations [4–7]. The contribution of CYP3A to lansoprazole metabolism is larger than that to omeprazole metabolism [8], though there is no quantitatively comparative study in vivo. CYP2C19 shows genetically determined polymorphism, yielding extensive metabolizers (EMs) and poor metabolizers (PMs) [9–11]. Several in vivo studies have consistently suggested that the area under the plasma lansoprazole concentration-time curve (AUC) in PMs for CYP2C19 was much greater than those in EMs [12–17]. Therefore, CYP2C19 plays an important role in lansoprazole metabolism in EMs, whereas sulfoxidation of lansoprazole catalyzed by CYP3A could be the predominant metabolic pathway in PMs.

A combination of clarithromycin and amoxicillin with proton pump inhibitors including lansoprazole has proven to be highly effective in the eradication of H. pylori[18]. As a result of its effect on the pharmacokinetics of drugs such as terfenadine, cyclosporin A, carbamazepine, and omeprazole [19–22], clarithromycin is considered to be a relevant inhibitor of CYP3A [23]. A recent in vivo study with H. pylori-positive patients with peptic ulcer disease suggested that clarithromycin co-administration increases the plasma lansoprazole concentration 3 h after lansoprazole administration [24]. However, as there was only a single blood sample in this study, a metabolic interaction between these drugs could not be confirmed. Another study has suggested that the elimination half-life of lansoprazole was significantly prolonged and the area under the concentration-time curve from 0 to 8 h was significantly increased by combination of clarithromycin and amoxicillin with lansoprazole [25]. However, there was no information about CYP2C19 genotype in this study.

To date, however, there is no detailed published information indicating a pharmacokinetic drug interaction between lansoprazole and clarithromycin in relation to the CYP2C19 genotype status. Therefore, we examined whether clarithromycin affected the metabolism of lansoprazole and if the possible interactions might relate to the CYP2C19 genotype.

Methods

Study design

Eighteen Japanese healthy volunteers (12 males and six females) who were H. pylori negative were enrolled in the study. Their mean age (range) was 27 (22–40) years and mean body weight was 60 (40–85) kg. The Ethics Committee of Hirosaki University School of Medicine approved the study protocol, and written informed consent was obtained in advance from each participant. The mutated alleles for CYP2C19, CYP2C19*3(*3) and CYP2C19*2(*2) had been identified using the PCR-RFLP methods of de Morais et al.[26], prior to this study. The CYP2C19 genotype analyses revealed five different patterns as follows: *1/*1 in six, *1/*2 in three, *1/*3 in three, *2/*2 in five and *2/*3 in one. These were divided into three groups, homozygous EMs (*1/*1, n = 6), heterozygous EMs (*1/*2 and *1/*3, n = 6), and PMs (*2/*2 and *2/*3, n = 6).

A randomized double-blind placebo-controlled crossover study design in two phases was conducted at intervals of 2 weeks. Clarithromycin 400 (Clarith®, Taisho Pharmaceutical Co., Ltd, Tokyo, Japan) or placebo capsules of identical appearance were given orally twice a day (09.00 h, 21.00 h) for 6 days. Six volunteers within each group were allocated to either of two different drug sequences: placebo-clarithromycin or clarithromycin-placebo. On day 6, they took a single oral 60 mg dose of lansoprazole (Takepron®, Takeda Pharmaceutical Co., Ltd, Osaka, Japan) and 400 mg of clarithromycin or placebo with 240 ml of tap water at 09.00 h after overnight fasting. Compliance with the study drugs was confirmed by pill-count. No other medications were taken during the study periods. No meal was allowed until 4 h after the dosing (13.00 h). The use of alcohol, tea, coffee and cola was forbidden during the test days.

Blood sampling

Blood samples (10 ml each) for determination of lansoprazole and its metabolites lansoprazole sulphone and 5-hydroxylansoprazole were taken into heparinized tubes immediately before and 0.5, 1, 1.5, 2, 3, 4, 6, 8, 12 and 24 h after the administration of lansoprazole. Plasma was separated immediately and kept at −30 °C until analysis.

Assay

Plasma concentrations of lansoprazole, lansoprazole sulphone and 5-hydroxylansoprazole were quantified using a high performance liquid chromatography (HPLC) method developed in our laboratory [27]. In brief, after alkalization with 0.5 ml NaOH (0.001 m), 1 ml plasma was extracted with 5 ml of diethyl ether : dichloromethane (70 : 30, v : v). The organic phase was evaporated at 60 °C to dryness. Samples dissolved in the mobile phase of phosphate buffer (0.02 m, pH = 4.6), acetonitrile, and methanol (55 : 40 : 5, v : v) were injected onto the HPLC system (SHIMADZU CLASS-VP, SHIMADZU Corporation, Kyoto, Japan), with a C₁₈ STR ODS-II column as an analytical column (column II; 150 × 4.6 mm I.D., particle size 5 µm; Shinwa Chemical Industry, Kyoto, Japan). The flow rate was 0.8 ml min⁻¹ and the wavelength was set at 286 nm. The limit of quantification was 3 ng ml⁻¹ for lansoprazole, and 5 ng ml⁻¹ for lansoprazole sulphone and 5-hydroxylansoprazole, respectively. Intra- and interday CVs were less than 6.1 and 5.1% for lansoprazole, 4.4 and 5.9% for lansoprazole sulphone, 5.8 and 5.8% for 5-hydroxylansoprazole, respectively, at the lowest concentration ranges.

Data analyses of pharmacokinetics

The peak concentration (C_max) and the time of the peak concentration (t_max) were obtained directly from the original data. The terminal rate constant (λ_z) used for the extrapolation was determined by regression analysis of the log-linear part of the concentration-time curve for each subject. The elimination half-life was determined by 0.693/λ_z. The area under the plasma concentration-time curve (AUC(0,24 h)) was calculated with use of the trapezoidal rule. AUC from zero to infinity (0, ∞) was calculated by AUC(0,last) + C_last/λ_z, where C_last is the last detectable plasma drug concentration.

Statistical analyses

One-way anova and Fisher's exact test were used for comparisons between the three CYP2C19 genotypes and clinical profiles such as age, body weight and gender. A paired t-test for the comparison of placebo vs clarithromycin was conducted on pharmacokinetic parameters. Wilcoxon signed-rank test was performed on the parameter t_max. Percent changes in pharmacokinetic parameters including C_max, AUC and elimination half-life during clarithromycin treatment between the three genotype groups were compared using one-way anova followed by Scheffe's test. AUC ratios of either 5-hydroxylansoprazole or lansoprazole sulphone to lansoprazole between the three genotype groups were compared using one-way anova followed by Scheffe's test. A P value of 0.05 or less was regarded as significant. SPSS 8.0.1 for Windows (SPSS Japan Inc., Tokyo) was used for these statistical analyses.

Results

Although none of the subjects needed to be withdrawn from this study, mild to moderate side-effects were observed during clarithromycin administration: mild diarrhoea in two subjects and mild abdominal disturbance in six subjects. These side effects continued until day 6 and ameliorated the day after discontinuation of clarithromycin. No adverse events were reported during placebo administration or after lansoprazole plus placebo administration.

No differences between the groups with homozygous EM, heterozygous EM and PM genotypes were found in subject profiles, including age [mean and 95% confidence interval, 25 (22, 28), 28 (21,35) and 27 (24, 31) years, nonsignificant (NS)] body weight [62 (45, 79), 57 (48, 66) and 61 (49, 73) kg, NS] and genders (M : F; 4 : 2, 4 : 2 and 4 : 2, NS). Plasma concentration-time curves of lansoprazole during the two phases in each genotype group for CYP2C19 are shown in Figure 1, and the pharmacokinetic parameters are summarized in Table 1. During placebo administration, significant changes were found in C_max (P < 0.001), AUC(0, ∞) (P < 0.001) and elimination half-life (P < 0.001), but not in t_max of lansoprazole between the different CYP2C19 genotypes (Table 1). The C_max and AUC ratio to parent drug for 5-hydroxylansoprazole and all parameters for lansoprazole sulphone differed between CYP2C19 genotypes (Tables 2 and 3).

Figure 1.

Mean plasma concentration-time curves of lansoprazole during placebo and clarithromycin treatment in homozygous extensive metabolizers (EMs) (n = 6), heterozygous EMs (n = 6), and poor metabolizers (PMs) (n = 6) for CYP2C19. Data are shown as mean and bars are SDs. Open circles and closed squares indicate data during placebo and clarithromycin treatments, respectively.

Table 1.

Pharmacokinetic parameters of lansoprazole during placebo or clarithromycin treatment in homozygous EMs, heterozygous EMs and PMs for CYP2C19

		Homozygous EMs (n = 6)	Heterozygous EMs (n = 6)	PMs (n = 6)
Cmax (ng ml–1)	With placebo	2190 (1086, 3295)###	2690 (1419, 3961)###	3649 (2865, 4432)
	With clarithromycin	2782 (1920, 3644)#*	4490 (2267, 6714)*	5458 (4578, 6338)***
tmax (h)1	With placebo	1.5 (1.0–2.0)	1.75 (1.0–3.0)	2.5 (1.5–3.0)
	With clarithromycin	1.5 (1.5–3.0)	1.5 (1.0–3.0)	1.5 (1.0–2.0)
AUC(0, ∞) (ng ml–1 h)	With placebo	4652 (2294, 7009)###	8299 (4784, 11814)###	25293 (17643, 32943)
	With clarithromycin	6411 (4294, 8527)###*	14584 (7428, 21739)###*	45741 (30618, 60863)**
Elimination half-life (h)	With placebo	0.86 (0.67, 1.05)###	1.7 (1.2, 2.2)##	4.3 (3.3, 5.2)
	With clarithromycin	1.1 (0.86, 1.3)###	1.9 (1.5, 2.4)###	6.6 (4.9, 8.3)**

Data are shown as mean and 95% confidence interval.

¹t_max is given as median (range).

^*P < 0.05,

^**P < 0.01,

^***P < 0.001, compared with placebo,

^#P < 0.05,

^##P < 0.01,

^###P < 0.001, compared with PMs.

Table 2.

Pharmacokinetic parameters of 5-hydroxylansoprazole during placebo or clarithromycin treatment in homozygous EMs, heterozygous EMs and PMs for CYP2C19

		Homozygous EMs (n = 6)	Heterozygous EMs (n = 6)	PMs (n = 6)
Cmax (ng ml–1)	With placebo	257 (150, 364)###	197 (155, 329)#	71 (36, 103)
	With clarithromycin	366 (187, 545)##	304 (175, 434)#	85 (44, 126)
tmax (h)1	With placebo	1.5 (1.5–2.0)	1.75 (1.0–3.0)	2.0 (1.5–3.0)
	With clarithromycin	1.5 (1.5–2.0)	1.75 (1.0–3.0)	2.25 (1.5–4.0)
AUC (0, ∞) (ng ml–1 h)	With placebo	541 (210, 871)	665 (317, 1013)	301 (93, 509)
	With clarithromycin	891 (346, 1436)	967 (602, 1332)*	458 (131, 785)*
AUC ratio to lansoprazole	With placebo	0.11 (0.085, 0.14)##	0.093 (0.038, 0.15)##	0.015 (−0.001, 0.031)
	With clarithromycin	0.13 (0.095, 0.17)###	0.075 (0.041, 0.11)#	0.014 (−0.003, 0.031)

Data are shown as mean and 95% confidence interval.

¹t_max is given as median (range).

^*P < 0.05, compared with placebo,

^#P < 0.05,

^##P < 0.01,

^###P < 0.001, compared with PMs.

Table 3.

Pharmacokinetic parameters of lansoprazole sulphone during placebo or clarithromycin treatment in homozygous EMs, heterozygous EMs and PMs for CYP2C19

		Homozygous EMs (n = 6)	Heterozygous EMs (n = 6)	PMs (n = 6)
Cmax (ng ml–1)	With placebo	161 (9, 313)##	238 (92, 384)##	743 (300, 1185)
	With clarithromycin	18 (5, 31)###	28 (8, 48)###·*	160 (99, 220)*
tmax (h)1	With placebo	1.5 (1.0–1.5)##	2.5 (1.0–3.0)#	3.5 (3.0–4.0)
	With clarithromycin	1.5 (1.5–2.0)	1.5 (1.25–3)	8.0 (2.0–12)*
AUC (0, ∞) (ng ml–1 h)	With placebo	266 (51, 481)###	557 (26, 1088)##	7789 (3372, 12207)
	With clarithromycin	47 (13, 82)###	80 (25, 136)###	2471 (1239, 3703)*
AUC ratio to lansoprazole	With placebo	0.068 (0.013, 0.12)###	0.066 (0.019, 0.11)###	0.28 (0.18, 0.39)
With clarithromycin	0.0083 (0.0003, 0.017)###*	0.0061 (0.002, 0.010)###*	0.051 (0.038, 0.063)***

Data are shown as mean and 95% confidence interval.

¹t_max is given as median (range).

^*P < 0.05,

^***P < 0.001, compared with placebo;

^#P < 0.05,

^##P < 0.01,

^###P < 0.001, compared with PMs.

Clarithromycin treatment significantly increased C_max of lansoprazole in homozygous EMs (1.47-fold, P < 0.05), heterozygous EMs (1.71-fold, P < 0.05) and PMs (1.52-fold, P < 0.001) and AUC(0, ∞) of lansoprazole in homozygous EMs (1.55-fold, P < 0.01), heterozygous EMs (1.74-fold, P < 0.05) and in PMs (1.80-fold, P < 0.01). Elimination half-life was prolonged by 1.54-fold (P < 0.05) only in PMs. Although values of pharmacokinetic parameters of 5-hydroxylansoprazole were not changed in homozygous EMs, AUC(0, ∞) (P < 0.05) of 5-hydroxylansoprazole was increased in heterozygous EMs and PMs. Clarithromycin treatment significantly decreased C_max (P < 0.05) of lansoprazole sulphone in heterozygous EMs and PMs, and AUC(0, ∞) (P < 0.05) of lansoprazole sulphone only in PMs. The AUC ratio of lansoprazole : lansoprazole sulphone was significatly increased by clarithromycin in homozygous EMs (P < 0.05), heterozygous EMs (P < 0.05) and PMs (P < 0.001). As with placebo treatment, several pharmacokinetic parameters of lansoprazole and its metabolites were significantly different between CYP2C19 genotypes during clarithromycin treatment (Tables 1, 2 and 3).

The clarithromycin-mediated percent increase in pharmacokinetic parameters such as C_max, AUC (0–8) and elimination half-life did not differ between the groups with the three CYP2C19 genotypes (Figure 2). There were also no differences in clarithromycin-mediated percent increase in AUC(0, ∞) of lansoprazole : lansoprazole sulphone ratios (Figure 3).

Figure 2.

Effects of CYP2C19 genotypes on the mean clarithromycin-mediated percent increase in peak concentration (C_max), the area under the concentration-time curve from zero to infinity (AUC(0, ∞)) and elimination half-life of lansoprazole. Error bars indicate SD.

Figure 3.

Effects of CYP2C19 genotypes on the mean clarithromycin-mediated percent change in the area under concentration-time curve from zero to infinity (AUC(0, ∞)) ratio of 5-hydroxylansoprazole to lansoprazole and the AUC(0, ∞) ratio of lansoprazole sulphone to lansoprazole, respectively. Error bars indicate SD

Discussion

In this study significant differences were shown in AUC(0, ∞) of lansoprazole between different CYP2C19 genotypes during placebo treatment (P < 0.001). The relative values of the AUC(0, ∞) of lansoprazole in homozygous EMs, heterozygous EMs and PMs were 1 : 1.8 : 5.4. This is in line with several previous studies [12, 13, 14, 15, 16, 17] and indicates that the CYP2C19 genotype is the major determinant of lansoprazole disposition.

Although clarithromycin significantly increased the C_max and AUC(0, ∞) of lansoprazole in all genotype groups, clarithromycin significantly prolonged the elimination half-life of lansoprazole only in PMs. Meanwhile the AUC ratio of lansoprazole to lansoprazole sulphone, which could be regarded as CYP3A activity, was significantly increased in all genotype groups. Therefore, these findings suggest that clarithromycin has an inhibitory effect sufficient to suppress sulphoxidation of lansoprazole in all genotypes. However, the sulphoxidation metabolic pathway is not the major determinant of overall lansoprazole metabolism in EMs, whereas this pathway is critical for PMs. These findings strongly suggest that CYP2C19 is predominantly involved in lansoprazole metabolism in EMs and CYP3A is predominant in PMs.

In contrast, although no change in the elimination half-life of lansoprazole was seen in EMs, clarithromycin significantly increased the C_max and AUC(0, ∞) in EMs as well as in PMs. This finding suggests that clarithromycin increased the bioavailability of lansoprazole through a mechanism independent of lansoprazole metabolism. A recent in vitro study has shown some involvement of P-glycoprotein in lansoprazole transport [28], while an in vitro study with cell lines has recently demonstrated that the inhibitory effect of clarithromycin on P-glycoprotein is intermediate [29–33]. Therefore, it is possible that the bioavailability of lansoprazole was increased through inhibition by clarithromycin of P-glycoprotein mediated transport back to the intestinal lumen after absorption. The effect of clarithromycin on lansoprazole pharmacokinetics appears to be via inhibition of both CYP3A4 and P-glycoprotein in PMs, compared with only P-glycoprotein inhibition in EMs. In addition, there may be some involvement of P-glycoprotein inhibition in the interactions between lansoprazole and clarithromycin in all genotype groups, contributing to the overall clarithromycin-mediated percent increase in pharmacokinetic parameters such as C_max, AUC(0, ∞) and elimination half-life of lansoprazole between CYP2C19 genotypes.

A previous study with a design similar to ours has shown that clarithromycin significantly increases the AUC of omeprazole by 2.1-fold in homozygous EMs, 2.1-fold in heterozygous EMs, and 2.3-fold in PMs, respectively [34]. The corresponding values with lansoprazole in the present study were slightly lower: 1.5-fold in homozygous EMs (NS), 1.7-fold in heterozygous EMs, and 1.8-fold in PMs. These findings are somewhat surprising because in vitro data suggest that the contribution of sulphoxidation catalyzed by CYP3A in lansoprazole metabolism is larger than that of omeprazole [8].

Several studies have suggested that the CYP2C19 genotype influences the cure rate of gastric acid-related disorders including eradication rate of H. pylori. PMs for CYP2C19 have significantly higher eradication rates of H. pylori following treatment with such proton pump inhibitors as omeprazole [35], lansoprazole [36] and rabeprazole [37] than do EMs. During clarithromycin treatment, there were significant differences in AUC(0, ∞) of lansoprazole between different CYP2C19 genotypes. The relative values of the AUC(0, ∞) of lansoprazole in homozygous EMs, heterozygous EMs and PMs were 1 : 2.3 : 7.1. This difference in the AUC (0–8) may account for the different cure rates of gastric acid-related disorders including eradication rate of H. pylori[36].

A limitation of our study was the small number of subjects and the single dosing of lansoprazole. There were several inconsistent findings regarding lansoprazole disposition and 5-hydroxylansoprazole between CYP2C19 genotypes. These might be reflected by limited power to detect a significant difference. In PMs, detectable plasma concentrations of lansoprazole 24 h after administration suggest there may be accumulation of lansoprazole at steady state, and hence a larger difference in the steady-state AUC(0, 24 h). However several studies suggest no differences in pharmacokinetic parameters between single dosing and repeated dosing [38, 39]. The possibility of some saturation in lansoprazole metabolism cannot be excluded because omeprazole increases its own relative AUC following repeated dosing [40]. A previous study by Ushiama et al.[24] reported that the mean 3 h plasma lansoprazole concentrations elevated in proportion to the doses of clarithromycin: 385 ± 338 ng ml⁻¹ for the control subjects, 696 ± 797 ng ml⁻¹ for the H. pylori-positive patients given 400 mg day⁻¹ clarithromycin, and 947 ± 806 ng ml⁻¹ for the H. pylori-positive patients given 800 mg day⁻¹ clarithromycin (P < 0.05 vs the control subjects). This is a larger effect of clarithromycin on lansoprazole concentration than in our study. Therefore, lansoprazole concentrations in H. pylori-positive patients who are PMs for CYP2C19 and are given 800 mg day⁻¹ clarithromycin are expected to be much higher than those in patients who are EMs for CYP2C19 and are given 400 mg day⁻¹ clarithromycin.

In conclusion, this study indicates that there are significant drug interactions between lansoprazole and clarithromycin in all genotype groups, probably largely through CYP3A inhibition. In addition the bioavailability of lansoprazole might be increased to some extent, through inhibition of P-glycoprotein during clarithromycin treatment.