Effect of an increased clopidogrel maintenance dose or lansoprazole co-administration on the antiplatelet response to clopidogrel in CYP2C19-genotyped healthy subjects


  • Clinical trial registration: ClinicalTrials.gov number NCT00413608

Jean-Sébastien Hulot, Service de Pharmacologie, Unité de Pharmacogénétique, Hôpital Pitié-Salpêtrière, 47 Bvd de l’hôpital, 75013 Paris, France.
Tel.: +33 142161661; fax: +33 142161688. E-mail: jean-sebastien.hulot@psl.aphp.fr

The loss-of-function CYP2C19*2 allele is associated with a reduced ability of clopidogrel to inhibit platelet aggregation in healthy volunteers and in patients with coronary artery disease treated with a standard maintenance dose of clopidogrel (75 mg day–1)[1–3], and with significantly higher rates of cardiovascular outcomes, including stent thrombosis [2,4–7]. Whether the negative impact of the CYP2C19*2 allele could be reversed by using higher maintenance doses of clopidogrel is not known. The implication of CYP2C19 in clopidogrel metabolism has also been investigated in drug-drug interaction studies using various proton pump inhibitors (PPI), some of them being CYP2C19 inhibitors [8,9]. We therefore conducted the CLOVIS study (ClinicalTrials.gov number_NCT00413608) to (i) determine whether a double clopidogrel daily dose allows poor responders carrying the CYP2C19*2 allele to reach a higher platelet inhibition, and (ii) assess the impact of lansoprazole, one of the most potent CYP2C19 inhibitors among PPIs, on the pharmacodynamics of clopidogrel.

Fifteen healthy male homozygous CYP2C19*1/*1 subjects and 15 heterozygous CYP2C19*1/*2 subjects were included (age 26 ± 4 years, body mass index 23 ± 2 kg m²). The study (Fig. 1) was divided into two successive phases. During the first study phase, all 30 subjects were given a standard dose of clopidogrel (75 mg day–1) for 7 days. The subjects were then classified into two groups according to their pharmacodynamic response to clopidogrel, but blinded to their genotype. Subjects with a relative inhibition to 10 μm ADP-induced platelet aggregation (IPA) <40% were classified as poor responders whereas those with an IPA ≥ 40% were classified as good responders according to our previous results [1]. IPA was calculated as the per cent decrease of the relative maximal platelet aggregation (MPA) from baseline ([(MPAbaseline–MPApost dose)/MPAbaseline] × 100). For the second phase of the study, poor responders were given a 150 mg day–1 dose of clopidogrel for 7 days after a 2-week washout period. After a 1-week washout period, good responders received lansoprazole 30 mg day–1 for 7 days followed by a 7-day combined administration (at 08.00) of 75 mg day–1 clopidogrel + lansoprazole 30 mg day–1. Genotyping and pharmacodynamic analysis (ADP-induced platelet aggregation and VASP-phosphorylation level expressed as Platelet Reactivity Index, PRI) were determined as described elsewhere [1].

Figure 1.

 Experimental design of the clopidogrel response variability study (CLOVIS). PD, pharmacodynamics evaluation.

The sample size was calculated to demonstrate that the administration of a double maintenance dose in poor responders would induce an increase of 22.5% in clopidogrel response assessed by IPA compared with the effect observed with the standard dose (alpha-risk error of 0.05, power of 80%, and a standard deviation of 18%). The pharmacodynamic response to clopidogrel between the different phases (75 mg clopidogrel vs. 150 mg clopidogrel or 75 mg clopidogrel on top of lansoprazole) was compared using a linear mixed model taking into account the within-subject correlation and a first-order autoregressive error structure (only one variance-covariance matrix estimated for the two treatment periods). Each model included treatment, time, an interaction between treatment and time, and the baseline PD parameter as covariate. Data are expressed as mean ± SD. All analyses were performed with sas statistical software (version 9.1.; SAS Institute, Cary, NC, USA). A P value <0.05 was considered to be significant.

All 30 subjects completed the study. Baseline platelet function did not significantly differ according to the CYP2C19 genotype. On day 7, IPA was 58.5 ± 11.0% in *1/*1 subjects vs. 38.9 ± 20.1% in *1/*2 subjects (P = 0.003). PRI was 42.8 ± 15.3% in *1/*1 subjects and 57.4 ± 10.5% in *1/*2 subjects (P = 0.003).

According to the predefined IPA threshold, 9 subjects were defined as poor responders (26.4 ± 11.2%; range, 11–39%) and 21 subjects were defined as good responders (IPA, 58.3 ± 11.8%; range, 40–86%). Six CYP2C19*2 carriers and all 15 CYP2C19 *1/*1 subjects were good responders. All nine poor responder subjects were CYP2C19*2 carriers (Fig. 2A).

Figure 2.

 (A) Distribution of IPA (inhibition of platelet aggregation) among CYP2C19 *1/*1 (closed circles, n = 15) and CYP2C19 *1/*2 (open circles, n = 15) after 7 days of clopidogrel 75 mg day–1. The dashed line represents the predefined threshold of IPA (40%) to separate good and poor responders. (B) Impact of a 150 mg o.d. clopidogrel maintenance dose in poor responders to 75 mg o.d. dosing. Scatter plot of IPA after 7 days of clopidogrel 75 mg day–1 (left) or 150 mg day–1 (right) in the nine poor responders (all of them carrying one 2C19*2 allele). (C) Impact of lansoprazole on clopidogrel response in the 21 good responders, according to CYP219 genotype. VASP PRI (%) under clopidogrel 75 mg day–1 alone (left) or clopidogrel 75 mg day–1 + lansoprazole 30 mg day–1 (right) according to CYP2C19 genotype.

Doubling the clopidogrel dose fully reversed the poor-responsiveness pattern: IPA significantly increased from 26.4 ± 11.2% on the 75 mg day–1 dose to 54.5 ± 10.6% (P < 0.0001) on the 150 mg day–1 dose. All poor responder subjects became good responders with an IPA > 40% (range, 40–72%; Fig. 2B), similar to that measured in the 21 good responders under the clopidogrel 75 mg day–1 dose. A similar pattern was observed with PRI, which significantly decreased from 58.1 ± 11.3% to 48.7% ± 8.9%.

In the 21 good responders, after a 7-day combined lansoprazole + clopidogrel administration, mean IPA was slightly lower (52.1 ± 17.6%) than that measured after 7 days of clopidogrel 75 mg day–1 given alone (58.3 ± 11.8%, P = 0.094). Ten among the 21 subjects had a decrease in their IPA responsiveness to clopidogrel 75 mg day–1 when combined with lansoprazole (range, 1% to 35.9%) and three of them became poor responders (IPA < 40%). The inhibitory impact of lansoprazole was more marked on PRI: 46.9 ± 15.1% on clopidogrel alone to 55.3 ± 13.6% on clopidogrel + lansoprazole (P = 0.001, Fig. 2C). We observed a non-significant trend for a less pronounced influence of lansoprazole in CYP2C19*2 carriers (PRI, 10.8 ± 10.5% in the 15 CYP2C19*1/*1 subjects vs. 3.6% ± 5.5% in the 6 CYP2C19*1/*2 subjects; P = 0.11).

Our study shows that doubling the daily clopidogrel dosage from 75 mg to 150 mg overcomes the effect of CYP2C19*2 polymorphism in healthy subjects displaying a poor-response profile. Doubling of the clopidogrel maintenance dose may represent an appealing alternative compared with a switch to prasugrel in subjects having both a poor-response profile and carrying the loss-of-function CYP2C19*2 allele. However, although the administration of a systematic increased clopidogrel dose to all patients to overcome resistance in some of them is beneficial to reduce stent thrombosis after angioplasty (preliminary results of CURRENT-OASIS 7, ESC congress 2009), it might expose the responders to the lowest dose to an increased bleeding risk. Our study was conducted in a limited number of healthy subjects and it may be difficult to ascertain that increasing clopidogrel maintenance dose will provide a systematic improvement in all poor responders, especially in patients with cardiovascular risk factors associated with circulating activated platelets such as diabetic patients or CYP2C19*2 patients with prior myocardial infarction who subsequently developed stent thrombosis [10].

We hypothesize that the “normalized” antiplatelet activity obtained by the administration of a double dose in subjects with impaired CYP2C19 activity results from an adequate active metabolite production through CYP2C19-independent pathways. In contrast, the few CYP2C19*2 allele carriers displaying a good responsiveness to the 75 mg day–1 clopidogrel dose may have more efficient secondary metabolic pathways than the non-responders.

Finally, lansoprazole, which inhibits CYP2C19 activity, decreased the antiplatelet effect of clopidogrel 75 mg day–1 in good responder subjects, as revealed by the VASP assay, consistent with previous findings with omeprazole [8]. This effect was less marked than the effect of the CYP2C19*2 polymorphism.

The extent of interaction between lansoprazole and clopidogrel 75 mg day–1 was highly variable between subjects. However, dramatic inhibition of platelet response was found in some subjects. Our data suggest that the CYP2C19 genotype does not influence the interaction, but the small sample size might have affected the power to detect such influence. Finally, whether higher doses of lansoprazole could affect more markedly the response to clopidogrel and whether the doubling of the dose of clopidogrel could also reverse the partial neutralization of its antiplatelet effects by lansoprazole remain to be evaluated.

In conclusion, the decrease in platelet response to standard maintenance dose of clopidogrel (75 mg day–1) associated with the CYP2C19*2 loss-of-function polymorphism can be reversed by doubling the clopidogrel maintenance dose. We confirm that the PPI lansoprazole also decreases the effect of clopidogrel.


The authors thank the nursing staff of the Clinical Investigation Center for their assistance during the study. We also thank V. Remones, F. Dali-Ali and C. Vrignaud for excellent technical assistance

Funding sources

This work was funded by grants from Contrat d’Initiation à la Recherche Clinique (Ministère chargé de la Santé, CIRC 06082-P060309, sponsor: AP-HP), Programme National de Recherche Cardiovasculaire (INSERM, PNRC2006), and the Leducq TransAtlantic Network of Excellence on Atherothrombosis Research (Grant 04CVD01).

Disclosure of Conflict of Interests

The authors declare that they have no conflict of interest.