Drug Information Center

• Drug Information Group
• Department of Pharmacy Practice
• About Us
• People
• Potential Partnerships
• The RxPress
• Residency Program
• Frequently Asked Questions
• Contact Us
• Feedback

What is the recent data regarding the interaction between proton pump inhibitors and the thienopyridines (clopidogrel and prasugrel)?

The thienopyridines, clopidogrel (Plavix) and prasugrel (Effient) inhibit platelet activation and aggregation. They work by directly inhibiting adenosine diphosphate receptor binding on platelets and the subsequent activation of the platelets.^1,2 Both of these agents are considered prodrugs and must be converted into active forms by the cytochrome P450 (CYP) enzyme system.³ Proton pump inhibitors (PPIs) are frequently prescribed with thienopyridines in an effort to reduce the risk of gastrointestinal (GI) complications. In fact, an expert consensus document recommends PPI therapy to minimize the risk of GI bleeding, adding that it becomes increasingly cost-effective in the presence of more risk factors.⁴ However, concerns regarding this practice have been raised with the publication of data suggesting that PPIs may reduce the anti-platelet effectiveness of clopidogrel.³ One hypothesis is that PPIs, particularly omeprazole, may inhibit the conversion of clopidogrel to its active metabolite by inhibiting CYP2C19. Data are conflicting; however, as another report did not find an association between the drug combination and a risk of adverse outcomes. Therefore, an analysis was undertaken by O’Donoghue et al in an attempt to resolve the inconsistent results and understand the clinical implications of concomitant PPIs and thienopyridines by examining 2 clinical trials: PRINCIPLE (Prasugrel In Comparison to Clopidogrel for Inhibition of Platelet Activation and Aggregation)-TIMI 44 and TRITON (Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition with Prasugrel)-TIMI 38.^3,5,6

Methods
PRINCIPLE-TIMI 44: In this double-blind, crossover, randomized controlled trial, 201 subjects undergoing cardiac catheterization with planned percutaneous coronary intervention (PCI) were assigned to prasugrel 60 mg loading dose then 10 mg daily (n=102) vs. clopidogrel 600 mg loading dose then 150 mg daily (n=99), for 28 days.^3,5 At 15 days following PCI, patients were crossed over to the alternate therapy. Inhibition of platelet aggregation was measured at 30 minutes (±5), 2 hours (±10 minutes), and 6 hours (±30 minutes) after the loading dose and at 18 to 24 hours, day 15, and day 29 after randomization. However, O’Donoghue et al restricted analysis to platelet function results prior to the crossover and defined platelet hypo-responsiveness as <20% inhibition of platelet aggregation. Inhibition of platelet aggregation = 1 – (maximal platelet aggregation at time after loading dose/maximal platelet aggregation at baseline) x 100. Patients who received a glycoprotein IIb/IIIa inhibitor were excluded from the analysis.

TRITON-TIMI 38: In this double-blind, randomized controlled trial, 13,608 subjects with acute coronary syndrome undergoing planned PCI were assigned to prasugrel 60 mg loading dose then 10 mg daily (n=6813) vs. clopidogrel 300 mg loading dose then 75 mg daily (n=6795).^3,6 The primary endpoint was a composite of cardiovascular (CV) death, non-fatal myocardial infarction (MI), or non-fatal stroke. In order to minimize confounding by indication and detect PPI use at different time points, the analysis was restricted to patients who did not have an efficacy or bleeding event after randomization and before landmark dates of 3 days, 3 months, and 6 months.

All concomitant medications were recorded; PPI use was not restricted in either study.³ Sensitivity analyses were performed in the O’Donoghue paper to assess differences in the risk of adverse outcomes at different times and with different follow-up durations, as well as the effect of specific PPIs and other gastric acid suppressive medications (i.e., H₂-receptor antagonists). Although both PRINCIPLE-TIMI 44 and TRITON-TIMI 38 were sponsored by the study drug manufacturers, no external funding was received for the O’Donoghue analysis.

Results
PRINCIPLE-TIMI 44: At randomization, 53/201 (26.4%) subjects were on a PPI. Platelet aggregation was similar between patients on a PPI vs. not on a PPI at baseline in either treatment arm.³ In each group (clopidogrel and prasugrel), there were no significant differences in baseline characteristics between patients treated with a PPI vs. not treated with a PPI, except that the clopidogrel-PPI group had more smokers than the clopidogrel-no PPI group and the prasugrel-PPI group had more patients with a previous MI than the prasugrel-no PPI group. In the clopidogrel group, mean inhibition of platelet aggregation was significantly lower in subjects on a PPI vs. not on a PPI at 2 hours (p=0.003), 6 hours (p=0.02), and 18 to 24 hours (p=0.03) after the loading dose; there was no significant difference at 30 minutes (p=0.98). After 15 days on daily maintenance doses, the trend continued (p=0.06). In the prasugrel group, mean inhibition of platelet aggregation was significantly lower in subjects on a PPI vs. not on a PPI at 30 minutes (p=0.009) after the loading dose; there was no significant difference at the other time points. After 15 days on daily maintenance doses, mean inhibition of platelet aggregation was significantly lower in subjects on a PPI vs. not on a PPI (p=0.01). Hypo-responsiveness to clopidogrel occurred in 50.0% of those on a PPI vs. 18.2% not on a PPI (p=0.009) 24 hours after the loading dose, and in 50.0% vs. 7.9% (p=0.012), respectively, after 15 days on maintenance therapy. Hypo-responsiveness to prasugrel occurred in no patients after the loading dose and in 10.0% on a PPI vs. 0% not on a PPI (p=0.025) after 15 days of maintenance therapy.

TRITON-TIMI 38: At randomization, 4529/13,608 (33.3%) subjects were on a PPI (1844 on pantoprazole, 1675 on omeprazole, 613 on esomeprazole, 441 on lansoprazole, and 66 on rabeprazole).³ Baseline characteristics of the randomized treatment groups were similar; however, subjects on a PPI were significantly more likely to be older, female, and white; have an index diagnosis of unstable angina/non-ST-elevation MI; live in North America or Western Europe; have a history of peptic ulcer disease; and have lower baseline hemoglobin levels vs. subjects not on a PPI. As shown in tables 1 and 2, the primary composite endpoint of CV death, MI, or stroke occurred in a similar portion of subjects on a PPI vs. not on a PPI in the clopidogrel and prasugrel groups. Adjustment of the hazard ratio (HR) for potential confounders and the propensity to treat with a PPI revealed no significant association between the primary endpoint and PPI use with clopidogrel (p=0.46) and prasugrel (p=0.97). Additionally, PPI use in either group was not associated with an increased risk of MI, stent thrombosis, or reduced bleeding risk.

Table 1. Kaplan-Meier event rates (% (n/N)) for efficacy and safety endpoints in TRITON-TIMI 38 (clopidogrel).³

Sensitivity analyses examined the consistency of data based on PPI use at 3 days, 3 and 6 months after randomization, and after short- or long-term follow-up, since they could be discontinued or initiated at any time.³ Use of PPIs at randomization was not associated with an increased risk of CV death, MI, or stroke at 3 days of follow-up in clopidogrel (HR 1.00, 95% confidence interval [CI] 0.80 to 1.27) and prasugrel (HR 1.14, 95% CI 0.88 to 1.46) subjects or at 30 days of follow-up in clopidogrel (HR 0.98, 95% CI 0.80 to 1.21) and prasugrel (HR 1.09, 95% CI 0.87 to 1.37) subjects. Similarly, PPI use was not associated with a risk of early stent thrombosis at <30 days in clopidogrel (HR 1.17, 95% CI 0.76 to 1.81) and prasugrel (HR 0.76, 95% CI 0.36 to 1.61) subjects. Subjects who were taking a PPI for the entire duration of follow-up (n=2814) did not have an increased risk of CV death, MI, or stroke vs. subjects who never took a PPI (n=6912) with clopidogrel (HR 1.05, 95% CI 0.85 to 1.30) or prasugrel (HR 1.10, 95% CI 0.88 to 1.39). Among patients who had not discontinued the study drug, there was no significant association between PPI use and risk of the primary endpoint with clopidogrel (HR 0.94, 95% CI 0.80 to 1.11) or prasugrel (HR 0.99, 95% CI 0.83 to 1.19).

There was no independent association between individual PPIs and H₂-receptor antagonists and risk of MI or composite endpoint.³ Baseline PPI or H₂-receptor antagonist use was not associated with the risk of CV death, MI, or stroke for clopidogrel (HR 0.80, 95% CI 0.51 to 1.26) or prasugrel (HR 0.91, 95% CI 0.55 to 1.51). There was no significant interaction between PPI use, clopidogrel or prasugrel use, and risk of CV death, MI, or stroke. Patients on prasugrel had a lower incidence of this primary endpoint vs. clopidogrel regardless of whether they were on a PPI (HR 0.85, 95% CI 0.71 to 1.01) or not (HR 0.80, 95% CI 0.70 to 0.90, p-interaction 0.57) at baseline.

One thought was that individuals with reduced CYP2C19 activity would be more vulnerable to further suppression of this enzyme activity by PPIs.³ Among patients in whom this information was available, 357/1477 on clopidogrel and 372/1466 on prasugrel had a single reduced-function CYP2C19 allele in TRITON-TIMI 38; 34% on clopidogrel and 33% on prasugrel were also on a PPI at randomization. In these patients with a single reduced-function CYP2C19 allele, there was no significant difference in the incidence of the primary endpoint in patients on a PPI vs. not on a PPI with either clopidogrel or prasugrel. Similarly, subjects without a reduced-function CYP2C19 allele (wild-type carriers) had no difference in the incidence of the primary endpoint between those on a PPI vs. not on a PPI with either clopidogrel or prasugrel. Repeated analyses with stronger CYP2C19 inhibitor PPIs (esomeprazole, lansoprazole, omeprazole, and rabeprazole) revealed similar results. It should be noted, however, that this analysis was underpowered to detect a difference among genetic variants.

Discussion
Previous observational studies found an increased risk of adverse events when clopidogrel and a PPI were used together.^3,7 Omeprazole in particular was shown to reduce the anti-platelet effects of clopidogrel, prompting the Food and Drug Administration to issue a warning regarding concomitant use of the drug with PPIs outside of a strong indication. Indeed, this analysis demonstrated some in vitro antiplatelet attenuation of clopidogrel and prasugrel, loading and maintenance doses, with a PPI vs. no PPI. However, the higher CV event rates previously found in population-based studies could have been due to differences in risk factors. For example, increased age and number of comorbidities may signal the need for PPI therapy as well as worse outcomes. Therefore, the more homogenous TRITON-TIMI 38 study population may have reduced confounding by indication and led to more comparable event rates between patients on or not on a PPI, even prior to adjustment.

Additional theories are that PPIs reduce absorption of other medications by suppressing gastric acid secretion and increasing gastric pH, or their inhibition of CYP2C19 could reduce the bioactivation of thienopyridines.^3,7 In TRITON-TIMI 38, however, a reduced-function CYP2C19 allele did not lead to an increased risk of CV outcomes in patients on prasugrel. The authors stated that the higher potency of prasugrel led to almost no subjects qualifying as hypo-responsive with a PPI, compared to even high loading doses of clopidogrel. They also suggested that any modest anti-platelet attenuation may have been too small to demonstrate a clinical adverse effect. In any case, surrogate endpoints are no substitute for clinical events. This analysis also demonstrated the superior efficacy of prasugrel vs. clopidogrel, regardless of PPI use, in reducing the risk of CV death, MI, or stroke in TRITON-TIMI 38.

Although several covariates were adjusted for and sensitivity analyses were performed, additional limitations should be considered.^3,7 For instance, PPI prescriptions were at the discretion of the physician in both trials analyzed. This was a post-hoc analysis and TRITON-TIMI 38 was not intended to evaluate the use of PPIs. Compliance with PPIs could not be guaranteed throughout the follow-up period. In addition, subgroups such as different types of PPIs or genetic polymorphisms may have been underrepresented, rendering the analysis underpowered to demonstrate an adverse pharmacodynamic or clinical outcome.

In conclusion, this analysis found no association between the use of PPIs with the thienopyridines, clopidogrel or prasugrel, and an increase in the risk of CV events.³ Clinical data showing a definitive interaction are still scarce; a randomized controlled trial is necessary to prove causality between adverse clinical outcomes and PPI use with thienopyridines. In patients at an increased risk for GI bleeding, these findings imply that there is no need to avoid PPIs while on thienopyridines at this time.

References: