Summary
Background
Several studies have suggested that proton-pump inhibitors (PPIs), mostly omeprazole, interact with clopidogrel efficacy by inhibiting the formation of its active metabolite via CYP2C19 inhibition. Whether this occurs with all PPIs is a matter of debate. As rabeprazole is a less potent CYP2C19 inhibitor than other PPIs, we studied the interaction between rabeprazole and the antiplatelet actions and pharmacokinetics of clopidogrel.
Aim
To demonstrate the non-inferiority of rabeprazole over placebo using change in platelet reactivity index (PRI; vasodilator-stimulated phosphoprotein [VASP] assay) in a predefined population of good clopidogrel responders. Omeprazole was used as the positive control.
Methods
In this randomized three-period crossover study in healthy volunteers, 36 healthy men received clopidogrel (75 mg/day for 7 days) with placebo, omeprazole (20 mg/day) or rabeprazole (20 mg/day). Clopidogrel antiplatelet effects and disposition kinetics were assessed on day 7 of combination therapy. Non-inferiority threshold was predefined as an upper limit of the 90% confidence interval for the difference in change in PRI between placebo and rabeprazole of < 10% in good clopidogrel responders.
Results
In good clopidogrel responders (inhibition of VASP index > 30%), the clopidogrel antiplatelet effect remained non-inferior to placebo during rabeprazole (difference 3.4% [–1.7; 8.5]) but not omeprazole (difference 7.5% [2.5; 12.6]) co-administration. The AUC 0–24 and C max of active clopidogrel metabolite decreased with both omeprazole and rabeprazole, and conditions of bioequivalence were not met, except for AUC 0–24 with rabeprazole.
Conclusions
Rabeprazole does not interact with clopidogrel to the same extent as omeprazole. However, under our experimental conditions and proton-pump inhibitor doses, there was no significant pharmacodynamic interaction between rabeprazole or omeprazole and clopidogrel, despite a significant decrease in the formation of clopidogrel active metabolite.
Résumé
Contexte
Plusieurs études, principalement menées avec l’oméprazole, ont suggéré une interaction entre inhibiteurs de la pompe à protons (IPPs) et clopidogrel, via l’inhibition du CYP2C19 impliqué dans la transformation de la pro-drogue clopidogrel en metabolite actif. L’importance de cette interaction avec les autres inhibiteurs de pompes à protons est discutée. Cette étude avait pour objectif l’analyse de l’interaction pharmacodynamique et pharmacocinétique entre le rabéprazole, un inhibiteur plus faible du CYP2C19 que l’oméprazole, et le clopidogrel.
Objectif
L’objectif primaire était de démontrer la non-infériorité du rabéprazole par comparaison au placebo en utilisant l’index de réactivité plaquettaire (test VASP) dans une population de volontaires sains bon répondeurs au clopidogrel. L’oméprazole a été utilisé comme contrôle positif.
Méthodes
Étude croisée, randomisée, en trois périodes, menée chez 36 hommes volontaires sains recevant du clopidogrel (75 mg/jour pendant 7 jours) avec du placebo, de l’oméprazole (20 mg/jour) ou du rabéprazole (20 mg/jour). L’effet anti-plaquettaire du clopidogrel et ses données pharmacocinétiques ont été mesurés au 7 e jour de traitement. Le seuil de non-infériorité a été défini a priori comme une limite supérieure de l’intervalle de confiance à 90 % < 10 % pour la différence entre la diminution de l’index de réactivité plaquettaire (test VASP) entre le placebo et le rabéprazole chez les bons répondeurs au clopidogrel.
Résultats
Dans le groupe de bons répondeurs (inhibition du VASP PRI > 30 %), l’effet antiplaquettaire du clopidogrel était non inférieur à celui du placebo avec le rabéprazole (différence 3,4 % [–1,7; 8,5]) contrairement à l’oméprazole (différence 7,5 % [2,5; 12,6]). Toutefois, l’AUC 0–24 et la C max du métabolite actif du clopidogrel étaient significativement diminuées avec l’oméprazole et le rabéprazole et les conditions de bioéquivalence n’étaient pas remplies, excepté pour l’AUC 0–24 avec le rabéprazole.
Conclusions
L’interaction pharmacodynamique entre le rabéprazole et le clopidogrel n’a pas le même degré d’intensité que celle entre l’oméprazole et le clopidogrel. Cependant, dans nos conditions expérimentales, l’interaction entre rabéprazole ou oméprazole et le clopidogrel n’était pas significative malgré une inhibition significative de la génération du métabolite actif du clopidogrel.
Background
Dual antiplatelet therapy with aspirin and clopidogrel is associated with a significant reduction in cardiovascular ischemic events after acute coronary syndromes or percutaneous coronary interventions and is recommended in guidelines from the USA and Europe . Clopidogrel is an inactive prodrug that undergoes two oxidative steps involving multiple cytochrome P-450 (CYP) enzymes in its bioactivation to its pharmacologically active metabolite. Among them, CYP2C19, a CYP enzyme whose activity is determined genetically, contributes predominantly to this bioactivation and modulates the antiplatelet and therapeutic response to clopidogrel. Patients with loss of function polymorphism in the CYP2C19 gene are less responsive to clopidogrel , although the importance of this phenomenon remains controversial and may be limited to the risk of stent thrombosis .
Proton-pump inhibitors (PPIs) are recommended in patients treated with dual antiplatelet therapy who are at high risk of gastrointestinal bleeding . PPIs are metabolized primarily via the CYP2C19 and CYP3A4 isoenzymes and are competitive inhibitors of CYP2C19 activity . However, the contribution of the CYP2C19 isoenzyme to PPI biotransformation and H. pylori eradication rates and the potency of inhibition of CYP2C19 activity vary among different PPIs. CYP2C19 activity appears to affect the response to omeprazole, esomeprazole and lansoprazole and to be inhibited by these PPIs . This does not seem to be the case, at least not to the same extent, with pantoprazole and rabeprazole .
Concerns about PPI and clopidogrel interaction were raised when omeprazole was found to inhibit the antiplatelet effect of clopidogrel in an in vivo study of 124 patients undergoing elective coronary stent implantation . Several studies have suggested that omeprazole interacts with clopidogrel efficacy by inhibiting the formation of its active metabolite via CYP2C19 inhibition . Whether this occurs with all PPIs or is even of significant amplitude with omeprazole remains a matter of debate . However, it was recently demonstrated that generation of clopidogrel active metabolite and inhibition of platelet function are reduced less by the co-administration of dexlansoprazole or lansoprazole with clopidogrel than by the co-administration of esomeprazole or omeprazole .
As rabeprazole is a less potent CYP2C19 inhibitor than other PPIs , we performed a pharmacodynamic antiplatelet activity study of the interaction between standard recommended repeated doses of rabeprazole and clopidogrel in CYP2C19-genotyped healthy male subjects. Omeprazole and placebo were used as controls. Our primary objective was to demonstrate non-inferiority of rabeprazole over placebo using the change in platelet reactivity index (ΔPRI%) in good clopidogrel responders as derived from the vasodilator-stimulated phosphoprotein (VASP) assay as the primary endpoint.
Methods
Study design
This was a prospective, placebo- and active-controlled, open-label, blinded-evaluation, randomized, three-way crossover study. The study assessed the influence of rabeprazole (20 mg/day for 7 days) and omeprazole (20 mg/day for 7 days) on the antiplatelet effects and pharmacokinetics of clopidogrel (75 mg/day for 7 days) in 36 CYP2C19-genotyped non-smoking healthy Caucasian male subjects with normal basal platelet aggregation testing (> 50% aggregation to 1 μg/mL collagen, 1–2 mmol/L arachidonic acid and 10 μM adenosine diphosphate [ADP]), platelet count, complete blood count and prothrombin time. Subjects gave written informed consent to participate and to have CYP2C19 genotyping (but were not selected on the basis of their genotype) and the protocol was approved by the Committee for Protection of Human Subjects Île-de-France II and the French Medicine Agency.
Subjects were randomized based on a Latin square design to receive clopidogrel 75 mg/day in the morning in the fasting state for 7 days during three study periods separated by a drug-free period of 2–3 weeks, together with placebo, 20 mg of rabeprazole or 20 mg of omeprazole, given at the same time as clopidogrel. Platelet function evaluation (pharmacodynamics) was performed on day 1 before dosing (D1H0) and on day 7 before and 4 hours after the last intake of study drugs (D7H0 and D7H4, respectively). The pharmacokinetics of clopidogrel, its inactive carboxylic acid metabolite and the active metabolite were determined from blood samples taken before (H0) and at various times after administration of the last dose of clopidogrel with the concomitant drug (either placebo or PPI). Additional blood samples for determination of omeprazole, 5-hydroxyomeprazole, rabeprazole and rabeprazole thioether plasma concentrations were taken 3 and 4 hours postdose on day 7 to confirm proper exposure to PPIs.
Pharmacodynamic evaluations
The primary test to assess platelet function was based on the VASP phosphorylation level measured in whole blood using a flow cytometric assay (Platelet VASP ® ; Diagnostica Stago, Biocytex, Asnières, France) and a FACScan flow cytometer (Becton Dickinson, Le Pont de Claix, France). Results were expressed as platelet reactivity index (PRI%), calculated from the mean fluorescence intensity (MFI) of samples incubated with prostaglandin E1 (PGE1) alone or with both PGE1 and ADP simultaneously, using the following formula: (MFI PGE1 –MFI PGE1+ADP /MFI PGE1 ) × 100, as previously described . This test – also referred to as the VASP index – specifically assesses the activity of the P2Y12 receptor (the target of clopidogrel antiplatelet action), and is widely used for monitoring the responsiveness to clopidogrel . The percentage change in PRI on study day 7 just before the last administration of study drugs relative to baseline, i.e. prior to drug administrations (percentage change in ΔPRI [%] D7H0), was used as the primary study endpoint. ΔPRI (%) relative to day 1 was also calculated for D7H4.
Platelet aggregation was determined at the same time points as those used for VASP phosphorylation level assessments, with ADP-induced platelet optical aggregometry (Biopool, Ventura, CA, USA; ADP 10 and 20 μM) using platelet-rich plasma adjusted to 250 × 10 9 /L. Inhibition of platelet aggregation (IPA%) induced by ADP was calculated as: (MPA[day 1] – MPA[day 7]/MPA[day 1] × 100), where MPA is the maximal platelet aggregation induced by ADP. Platelet aggregation tests were performed on a TA-8 V optical platelet aggregometer (Soderel Medical, Heillecourt, France) within 3 hours of sampling in all subjects.
Pharmacodynamic evaluations were performed blind to the study period and the CYP2C19 genotype.
Pharmacokinetic evaluations
Blood samples for the clopidogrel assay were collected in 6 mL ethylenediaminetetraacetic acid (EDTA) vials stored at 4 °C, to which 38 μL of 2-bromo-3′-methoxyacetophenone (500 mM in acetonitrile) were added within 30 seconds of sampling to stabilize the active metabolite. Blood samples were centrifuged at 4 °C within 30 minutes and stored at −80 °C until assay. Clopidogrel, clopidogrel carboxylic acid, clopidogrel active metabolite, omeprazole, 5-hydroxyomeprazole, rabeprazole and rabeprazole thioether were extracted from plasma on a solid phase OASIS HLB cartridge (10 mg/1 mL; Waters SAS, Milford, MA, USA). Chromatographic separation and detection of all compounds was performed on a YMC–UltraHT Pro C18 analytical column (YMC, Dinslaken, Germany), using ultra-high-performance liquid chromatography coupled to a tandem mass spectrometry system (UPLC-Acquity-TQD; Waters SAS, Milford, MA, USA). Limits of quantification were 0.1 ng/mL for clopidogrel and clopidogrel active metabolite, 5 ng/mL for rabeprazole and rabeprazole thioether, 10 ng/mL for clopidogrel carboxylic acid and 50 ng/mL for omeprazole and 5-hydroxyomeprazole.
Pharmacokinetic variable values were calculated using WinNonlin ® Professional, version 5.2 or higher (Pharsight Corp., Mountain View, CA, USA). The maximum plasma concentration ( C max ) and the time of its occurrence ( T max ) were obtained from observed values. The area under the concentration-time curve (AUC) in the sampled matrix during a dosing interval was calculated by linear up/log down trapezoidal summation. The apparent terminal rate constant (λz) after multiple dosing (1/hour), was determined by linear regression of the terminal points of the log-linear concentration-time curve. The apparent terminal half-life after multiple dosing (hours) was determined as (ln2/λz).
CYP2C19 genotyping and activity
The loss-of-function CYP2C19 variants *2 (rs4244285) and *3 (rs4986893) were tested using polymerase chain reaction (PCR)-based specific probe hybridization and single base extension. 681G>A and 636G>A comprise the two common reduced functional variants CYP2C19*2 and CYP2C19*3 , respectively. Subjects with the CYP2C19*1/*1 genotype were designated as CYP2C19 extensive metabolizer (EM) subjects.
The molar omeprazole/5-hydroxyomeprazole metabolic ratio in plasma samples at 3 hours was calculated as an index of CYP2C19 activity . In one EM subject, this ratio was calculated from the blood sample taken at 4 hours because 5-hydroxyomeprazole was not detectable at 3 hours.
Statistical analyses
Sample size was calculated with the assumption that approximately 66% of subjects would be good antiplatelet responders, defined as subjects in whom the VASP index on study day 7 relative to study day 1 would decrease by ≥ 30%, with an expected intrasubject standard deviation of differences in ΔPRI of ≤ 14% or a PRI value at day 7 below a cut-off value of 60%, as recently proposed for clopidogrel 75 mg daily maintenance dose . With these assumptions, 36 subjects are sufficient to conclude non-inferiority of rabeprazole to placebo with 10% ΔPRI as the limit of non-inferiority with > 95% power when true difference in treatment means is equal to 2%. Pharmacodynamic analyses were first performed on good antiplatelet responders as defined above, then on all 36 subjects.
Mixed-effect models were fitted to the ΔPRI% data as the dependent variable, with sequence, treatment and period as factors and subject as a random effect. Ninety percent confidence intervals (CIs) were calculated for the difference in means between rabeprazole versus placebo. Non-inferiority was concluded if the upper limit of the 90% CI fell below 10%. This non-inferiority limit was chosen because it represents the difference between omeprazole and placebo reported by Gilard et al. (10.7% in absolute value, 13.4% in relative value), which prompted the US Food and Drug Administration’s warning on the interaction of PPIs with clopidogrel.
Additional post-hoc analyses were performed to compare the change in VASP index on study day 7 relative to study day 1 with omeprazole and rabeprazole relative to placebo, using the Wilcoxon signed-rank test in good antiplatelet responders. Post-hoc correlation analyses were performed using Pearson’s correlation.
A linear mixed-effects model suitable for three-way crossover design was fitted to log-transformed pharmacokinetic variables, and 90% CIs for the ratio of the mean pharmacokinetic variables of clopidogrel were constructed using least-square means and intrasubject variance from the model. The above analysis was performed for clopidogrel active metabolite and clopidogrel major carboxylic acid metabolite. Bioequivalence was considered as demonstrated if the 90% CIs of the ratios for AUC 0–24 and C max between the placebo and PPI study periods fell in the range 80–125%.
Results
Thirty-six subjects completed the three study periods. Mean age, body weight and body mass index were 33.6 ± 7.9 years, 74.1 ± 8.7 kg and 23.6 ± 2.3 kg/m 2 , respectively. Of these 36 subjects, 23 were CYP2C19*1/*1 EMs, 12 were heterozygous CYP2C19*1/*2 and one was a poor metabolizer with the CYP2C19*2/*2 genotype.
Platelet function assays
Baseline VASP index before administration of clopidogrel was not significantly different across study periods ( P = 0.60). As expected, there was considerable interindividual variability in platelet function inhibition, as measured by use of the VASP index (VASP ΔPRI%) on day 7 of the clopidogrel plus placebo study period prior to last drug administration (D7H0) ( Fig. 1 ). The decrease in VASP index was < 30% in 18 subjects while the other 18 subjects were classified as good clopidogrel responders (change of VASP index ≥ 30%). Table 1 shows the results of platelet aggregation studies on day 7 (D7) of each study period before (H0) and 4 hours after (H4) administration of the last dose of clopidogrel together with placebo, omeprazole and rabeprazole.
| Treatment | n | Least square | Pairwise comparisons | |||||
|---|---|---|---|---|---|---|---|---|
| Mean (%) | 95% CI | Pair | Difference (%) | 90% CI | P a | |||
| VASP ΔPRI (%) | ||||||||
| Good VASP antiplatelet responders b | ||||||||
| Day 7/hour 0 | RABE | 18 | –47.3 | (–52.5; –42.1) | RABE/OME | –4.1 | (–9.2; 1.0) | 0.18 |
| RABE/PLBO | 3.4 | (–1.7; 8.5) | 0.26 | |||||
| OME | 18 | –43.2 | (–48.4; –38.0) | OME/PLBO | 7.5 | (2.5; 12.6) | 0.017 | |
| PBO | 18 | –50.7 | (–55.9; –45.6) | |||||
| Day 7/hour 4 | RABE | 18 | –56.2 | (–62.5; –49.9) | RABE/OME | –4.9 | (–10.5; –0.8) | 0.15 |
| RABE/PLBO | 4.4 | (–1.2; 9.9) | 0.2 | |||||
| OME | 18 | –51.3 | (–57.6; –45.1) | OME/PLBO | 9.2 | (3.6; 14.8) | 0.0087 | |
| PBO | 18 | –60.3 | (–66.8; –54.3) | |||||
| All subjects | ||||||||
| Day 7/hour 0 | RABE | 36 | –32.1 | (–38.8; –25.5) | RABE/OME | –1.6 | (–5.1; 1.8) | 0.44 |
| RABE/PLBO | 0.4 | (–3.1; 3.8) | 0.85 | |||||
| OME | 36 | –30.5 | (–37.5; –23.9) | OME/PLBO | 2 | (–1.5; 5.5) | 0.34 | |
| PBO | 36 | –32.5 | (–39.2; –25.9) | |||||
| Day 7/hour 4 | RABE | 36 | –39.8 | (–47.4; –32.1) | RABE/OME | –3.7 | (–7.2; –0.1) | 0.089 |
| RABE/PLBO | 0.5 | (–3.1; 4.0) | 0.82 | |||||
| OME | 36 | –36.1 | (–43.8; –28.5) | OME/PLBO | 4.2 | (0.6; 7.7) | 0.056 | |
| PBO | 36 | –40.3 | (–47.9; –32.6) | |||||
| Inhibition of platelet aggregation induced by ADP day 7/hour 0 | ||||||||
| All subjects | ||||||||
| ADP 20 μM | RABE | 36 | 39.4 | (32.4; 46.4) | RABE/OME | 4.3 | (–0.1; 8.8) | 0.11 |
| RABE/PLBO | –0.8 | (-5.3; 3.7) | 0.77 | |||||
| OME | 36 | 35.1 | (28.1; 42.1) | OME/PLBO | –5.1 | (–9.6; –0.6) | 0.063 | |
| PBO | 35 | 40.2 | (33.1; 47.2) | |||||
| ADP 10 μM | RABE | 36 | 39.8 | (32.7; 46.9) | RABE/OME | 0.6 | (–4.7; 5.8) | 0.86 |
| RABE/PLBO | –6.3 | (–11.6; –1.1) | < 0.05 | |||||
| OME | 36 | 39.2 | (32.1; 46.3) | OME/PLBO | –6.9 | (–12.2; –1.6) | 0.033 | |
| PBO | 35 | 46.1 | (39.0; 53.3) | |||||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
