Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

JOURNAL CLUB - Gastroenterology

October 2011

Editorial Review:

Alan Barkun, MD, FRCPC, FACP, FACG

Director, Division of Gastroenterology, MUHC-Montreal General Hospital, Professor of Medicine and Surgery, McGill University, Montréal, Quebec

The concern that proton pump inhibitors (PPIs) reduce or blunt the ability of clopidogrel, a thienopyridine inhibitor of platelet function, to prevent cardiovascular (CV) events was initially generated by case-control data. Despite the inconsistency of this association across retrospective analyses, caution was once advised on the basis of biologic plausibility pending prospective studies to provide definitive data. A prospective randomized trial, in which patients on clopidogrel randomized to a PPI did not experience any increased risk in CV events, allayed this concern, providing the foundation for an evidence-based consensus recommendation. The use of PPIs in patients on clopidogrel with or without ASA is now recommended in the presence of multiple risk factors for gastrointestinal (GI) bleeding. This is important information from the perspective of both the GI and CV specialists attempting to prevent life-threatening events.

CV and GI Risks in Aging Patients

CV events, typically the product of a disease process that progresses over decades, occur most frequently in older individuals. GI bleeding events, which may stem from increased vulnerability of the gastric mucosa to injury as well as increased exposure to agents that damage mucosal integrity, particularly NSAIDs, are also most common in older individuals. As a result, many candidates for antiplatelet therapy, to prevent new or recurrent CV events, need also be considered as candidates for strategies to control GI risks.

As a potent inhibitor of platelet activity, clopidogrel is indicated for patients with an acute coronary syndrome (ACS) or previous myocardial infarction (MI), more particularly following coronary stent insertion for varying durations of time. Often employed in combination with ASA, it is associated with a substantial reduction in major CV events, including CV death (Bell et al. Can J Cardiol 2011;27(suppl A):S1-S59). By reducing gastric acid secretion, PPIs reduce the risk of mucosal injury from a variety of sources, including NSAIDs (Lazzaroni et al. Drugs 2009;69:51-69). In patients with risk factors for GI bleeding, including older age, history of peptic ulcers, especially bleeding ulcers, or treatment with anticoagulants, PPIs are indicated for gastroprotection to prevent upper GI bleeding, including potentially life-threatening events (Rostom et al. Aliment Pharmacol Ther 2009;29:481-96).

Revisiting Antiplatelet and PPI Interaction

The biologic plausibility for an interaction between clopidogrel and PPIs is based on their common CYP450 isoenzyme 2C19 metabolic pathway. Variability in clopidogrel metabolism has already been established in relationship to genetic CYP2C19 polymorphisms that can lead to both substantial loss of antiplatelet activity and an increased risk of CV events relative to those without loss-of-function alleles (Mega et al. N Engl J Med 2009;360:354-62). The importance of the CYP2C19 isoenzyme for clopidogrel metabolism generated interest in the potential for drugs metabolized by this hepatic pathway, including PPIs, to inhibit clopidogrel activity through competitive inhibition.

However, the pharmacodynamic studies attempting to demonstrate this interaction have been variable and conflicting. In a systematic review of 18 studies evaluating a drug interaction between PPIs and clopidogrel, 10 were judged to be of low scientific quality (de Aquino Lima JP, Brophy JM. BMC Medicine 2010;8:81). Of the 5 studies judged to be of moderate or high quality, none reported a statistically significant association. A review of the effects of PPIs on platelet reactivity in patients taking clopidogrel also demonstrated inconsistent results (Lettino M. Eur J Intern Med 2010;21:484-9).

Several studies appeared to corroborate the potential for PPIs to inhibit the activity of clopidogrel. In a nested case-control study conducted in Canada with more than 13,000 individuals (Juurlink et al. CMAJ 2009;180:713-8), current use of PPIs in patients receiving clopidogrel was associated with a hazard ratio (HR) of 1.27 (95% CI, 1.03-1.57) for a recurrent MI in multivariable analysis. In a retrospective cohort study of 8,205 patients taking clopidogrel after an ACS, those who also took a PPI had a significantly higher risk of hospitalizations for recurrent ACS (14.6% vs. 6.9%; HR 1.86) and revascularization procedures (15.5% vs. 11.9%; HR 1.49), but not for all-cause mortality (Ho et al. JAMA 2009;301(9):937-44).

There is also no evidence that there are any differences within the PPI class for relative risk of a clinically significant interaction. Although some studies have demonstrated differences, such as a nested control study in which pantoprazole was associated with an increased risk of rehospitalization in patients taking clopidogrel after a MI (Stocki et al. Arch Intern Med 2010;170:704-10), a large retrospective study of >20,000 patients found no HR differences for frequency of CV events in patients on clopidogrel for esomeprazole, lansoprazole, omeprazole or pantoprazole (the decreased interaction noted with rabeprazole is not interpretable due to very small numbers in this subgroup) (Ray et al. Ann Intern Med 2010;152:337-45) (Figure 1). This lack of specificity has been noted in other authoritative studies as well (O’Donoghue et al. Lancet 2009;374:989-97). There is some variability in PPI activity on CYP isoenzymes but no persuasive evidence that these translate into meaningful differences in clinical outcome.

Such findings were hypothesis-generating because of the known limitations of retrospective analyses on determining causation. For example, the possibility that patients receiving PPIs were more likely to have comorbidities that placed them at a higher risk of MI recurrence could not be ruled out with these non-randomized data. Concern about this and other potential biases inherent to non-randomized studies was intensified by the lack of consistency among the case-control studies. Several trials, such as a large registry in France evaluating survivors of acute MI (Simon et al. Circulation 2011;123:474-82), were unable to corroborate an increased risk from concomitant clopidogrel and PPI exposure.

In addition to retrospective studies, data generated from 2 large prospective studies comparing clopidogrel to alternative antiplatelet strategies were also unable to detect an adverse effect on outcome from concomitant PPI use. The PRINCIPLE TIMI 44 (Prasugrel in Comparison to Clopidogrel for Inhibition of Platelet Activation and Aggregation-Thrombolysis in Myocardial Infarction-44) and TRITON-TIMI 38 (Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition with Prasugrel-Thrombolysis in Myocardial Infarction-38) studies were not specifically designed to evaluate an interaction between clopidogrel and PPI and did not prospectively stratify patients for this variable. Nonetheless, a pooled analysis of the 2 studies with nearly 14,000 patients receiving a PPI found no interaction between the PPI and the composite primary end point of CV events (O’Donoghue et al. Lancet 2009;374:989-97).

Figure 1.

COGENT: Resolving the Controversy

As defined by the principles of evidence-based medicine, the controversy about the risk of an interaction between clopidogrel and PPIs was largely resolved by a large, multicentre double-blind trial in which patients on ASA but otherwise at no particular increased risk for a GI event were randomized to the combination of clopidogrel and a PPI or to clopidogrel alone (Bhatt et al. N Engl J Med 2010;363:1909-17). In this study, called COGENT (Clopidogrel and the Optimization of Gastrointestinal Events Trial), the group receiving a PPI had a significant reduction in GI events (1.1% vs. 2.9%, HR 0.34; 95% CI, 0.18-0.63; P<0.001), which included a composite end point of overt or occult bleeding, symptomatic ulcers or erosions, obstruction or perforation. It is notable that there was an 87% reduction (HR 0.13; P=0.001) in the risk of overt upper GI bleeding for those who received omeprazole relative to those who received clopidogrel alone. CV event rates were not statistically different (4.9% vs. 5.7%, HR 0.99; 95% CI, 0.68-1.44; P=0.96) in the arm that received clopidogrel with a PPI relative to the group that received clopidogrel alone.

Consensus Guidelines

The importance of alerting clinicians to the appropriate use of clopidogrel and PPIs in individuals who could benefit from each provided the basis for an expert consensus that followed publication of the COGENT study (Abraham et al. Circulation 2010;122:2619-33). Sponsored by the American College of Gastroenterology (ACG), the American College of Cardiology (ACC) and the American Heart Association (AHA), the document was considered a timely response to new information relevant to managing 2 high-risk conditions. The major message of the document is that patients with multiple risk factors for GI bleeding have a further exacerbation of risk by the use of antiplatelet therapies. The guidelines encouraged use of PPIs with clopidogrel in patients at risk for a GI event, for which the risk of an interaction appears to be absent or low.

Although the consensus document cautioned that only one prospective trial randomized patients on clopidogrel to PPI or placebo, this provides the best current evidence that the risk of an interaction is low in an unselected population. The consensus document advises that neither clopidogrel nor a PPI should be administered to those without a specific indication for these agents but does not make any distinction between PPIs as a factor in drug selection. Importantly, both CV and GI events are potentially life-threatening so that both therapies may be indicated when risk factors support the potential for protection from events. The greatest single risk factor includes a past history of peptic ulcer (especially bleeding ulcer); other single risk factors that warrant considering bleeding prophylaxis with a PPI are anticoagulant therapy or dual antiplatelet therapy. More than one of the following risk factors should also prompt consideration for PPI prophylaxis: age 60 or over, corticosteroid use, dyspepsia or GERD symptoms, and the use of NSAIDs, or known Helicobacter pylori infection (H. pylori testing and treating should be carried out for all patients with past ulcers, with confirmation of eradication in the case of complicated ulcers).

The significance of a consensus document that included both gastroenterologists and cardiologists is that it allowed specialists to develop an approach to risk management with equal attention to potentially competing clinical issues. The consensus document involved a multidisciplinary team that agreed GI risk may exceed CV risk, or the reverse, and that neither should be considered in isolation of the other. It is notable that these guidelines differ from those issued earlier by the U.S. Food and Drug Administration and the Canadian Health Protection Branch which issued warnings in 2009 about the potential for an interaction based on the retrospective evidence.

Summary

The continuing controversy about the concomitant use of clopidogrel and PPIs has been resolved in practical terms. Although there are unanswered questions regarding safety in subpopulations, evidence-based recommendations support the use of PPIs in patients on clopidogrel when these are indicated for prevention of bleeding due to risk factors. The consensus recommendations, based on the preponderance of evidence, including a prospective, randomized study, were created by a collaborative group of gastroenterologists and cardiologists. These recommendations emphasize that both CV and GI risks should drive treatment decisions, and that in patients at risk for GI bleeding, PPI prophylaxis in patients on clopidogrel is indicated.

questions and answers

Panel

Marc Bradette, MD, FRCPC

Université Laval, Québec City, Quebec

Pierre Laflamme, MD

Affiliate of Université Laval, Quebec, Quebec

Charles Ménard, BPharm, MD

Université de Sherbrooke, Sherbrooke, Quebec

One of the major conclusions from the ACG/AHA/ACC consensus is that “PPIs are appropriate in patients with multiple risk factors for GI bleeding who require antiplatelet therapy.” Do you agree that the available data support this statement and that GI risk should be considered when considering antiplatelet therapy?

Dr. Bradette: This is a clear recommendation with which I absolutely agree. Although it is critical to consider the GI bleeding risk when prescribing antiplatelet drugs, there really is no appropriate substitute for antiplatelets when the CV risk is significant. In patients at high risk of GI bleeding, PPIs represent an important adjunctive therapy for gastroprotection. Patients with an indication for an anti-inflammatory agent to control joint disease as well as an antiplatelet agent to prevent CV events, most particularly the elderly to whom antiplatelets are most commonly prescribed, are good candidates for a PPI. While gastroenterologists are generally aware of the GI risk, other specialists also need to be aware and need to recognize that the complications of an uncontrolled GI bleed can be just as serious as those of a CV event.

Dr. Laflamme: GI bleeding caused by H. pylori and anti-inflammatories has decreased considerably with H. pylori treatment and the advent of PPIs and COX-2 inhibitors. We now see more GI bleeding caused by low-dose antiplatelet therapies in the elderly with heart disease, but also in patients sometimes taking antiplatelet agents for off-label indications. We see elderly patients with comorbidities and multiple GI bleeding risk factors present with severe GI bleeding, particularly those on ASA/ clopidogrel/warfarin, which aggravate cardiac problems, sometimes causing death. The literature clearly illustrates the role of PPIs in reducing GI bleeding risk and physicians should take these into consideration when prescribing an antiplatelet therapy. Antiplatelets should be prescribed according to recognized norms and their expected benefits should outweigh risks. I agree with this recommendation and it should be widely disseminated. However, patients on antiplatelets sometimes present with lower GI bleeding and/or an iron-deficiency anemia without an identifiable cause (after investigation) other than their medication. Thus, for patients prescribed antiplatelet therapy, we have to consider the adverse effects that cannot be prevented by PPIs.

Dr. Ménard: The addition of a PPI to antiplatelet therapy in patients with recognized risk factors for upper GI bleeding has been shown to lower the risk of bleeding at least by half, so I do agree with the ACG/AHA/ACC recommendations. However, we also have to consider an almost equivalent risk of lower GI bleeding with antiplatelets and there is no magic bullet shown to provide any protection. So antiplatelets should only be prescribed when the demonstrated benefit clearly outweighs the risk of GI bleeding.

From the standpoint of the experts who wrote the consensus document, PPIs are appropriate when the bleeding risk is high but not for routine bleeding prophylaxis when risk is low. But what about PPIs for other reasons, particularly GERD? What is your practice for patients without an elevated bleeding risk who have been prescribed antiplatelet therapy? Would you still consider a PPI?

Dr. Bradette: If a patient on an antiplatelet agent has an appropriate indication for a PPI, I will not hesitate to prescribe one of these agents. Although most of the typical indications for a PPI, such as GERD, are not life-threatening, these diseases can significantly impair quality of life. I am now convinced by the available data that there is no clinically significant interaction between PPIs and antiplatelet agents, so there is no justification for denying patients a therapy that can offer substantial clinical benefits.

Dr. Laflamme: While an asymptomatic patient prescribed an antiplatelet needs to take a PPI due to GI bleeding risk factors, a symptomatic patient who is or will be prescribed an antiplatelet agent but does have GI bleeding risk factors can also take a PPI. When the early data on PPI/antiplatelet interactions and their deleterious cardiac or vascular effects emerged, risk of using PPIs was not favourable. Occasionally, a pharmacist would call and refuse to fill the prescription or suggest pantoprazole, the only PPI recognized by Canadian and American regulatory bodies at the time. Many other publications have refuted these clinical findings but inhibition of clopidogrel activation by CYP2C19, in the presence of PPIs in in vitro and/or in vivo studies, supports a plausible biological mechanism of action. Other literature data associate a possible link of PPI use with a risk of pneumonia, C. difficile infection or fracture. The overall literature data recommend appropriate use of PPIs so that that their benefits outweigh potential risks. Moreover, patients should be informed so that they can put their therapy, symptoms and lifestyle in perspective.

Dr. Ménard: PPIs are the most effective medications for symptomatic relief of GERD. I think that PPIs should be prescribed at the lowest effective dose in patients with symptomatic GERD, whether or not those patients are on antiplatelet therapy. In my practice, I rarely find that patients are effectively relieved with anti-H₂.

When there was concern about a potential interaction between clopidogrel and PPIs, competition on the CYP2C19 pathway was suspected, because all PPIs use CYP2C19 to some degree. Yet no clinically significant interaction was observed in COGENT, which was one of the key evidencebased studies cited in the consensus statement. Does the choice of PPI matter and are there any data to support that relative risks from PPIs may differ?

Dr. Bradette: The in vitro studies have demonstrated a pharmacological interaction between PPIs and clopidogrel, but we now have a large body of observational evidence plus a prospective controlled trial to demonstrate that there is no clinically meaningful interaction. There is also no evidence that PPIs differ for relative risk of an interaction. I think it is reassuring that the PPI most likely to cause an interaction, omeprazole, was the one used in a prospective study demonstrating the absence of a clinically meaningful change in antiplatelet effects. There is clearly no evidence basis for suggesting one PPI may be less likely than another to interact with clopidogrel or any other antiplatelet agent.

Dr. Laflamme: All pharmaceutical companies [that produce PPIs] publish their pharmacokinetic studies showing a lesser clopidogrel inhibition with their PPI. These findings have no corresponding clinical value as clinical studies on cardiac events have not shown any difference between available PPIs. Only one prospective study corroborates these retrospective data, but we have no comparative prospective data. The ongoing SPICE study may provide insight. In the absence of prospective comparative data and with retrospective data showing no clinical differences between PPIs, the choice of a PPI depends on prescriber preference. Moreover, we still do not know the role of pharmacogenetics in the prescription of antiplatelets, far less so when used with a PPI.

Dr. Ménard: I do not think that at this time we have sufficient prospective clinical evidence to support the choice of a specific PPI in combination with clopidogrel or prasugrel. It is possible that a subset of patients may have specific cytochrome activity that may alter the pharmacoavailability of antiplatelet therapy but this issue needs to be studied prospectively in a controlled study. An answer to that question would lead to a genetically tailored, optimized antiplatelet therapy.

In the eyes of some internists, risk of a CV event trumps risk of a GI event even though both can be life-threatening. Any advice on how the primary care community can weigh concomitant risks and provide prophylaxis against both threats appropriately?

Dr. Bradette: The consensus document makes an important point, which is that clinicians need to change their paradigm to include both CV and GI risk when considering whether to prescribe a PPI with an antiplatelet agent. There is probably a tendency among many clinicians, particularly gastroenterologists, to see CV events as more serious, but this overlooks the morbidity of GI complications even when they do not result in a fatal bleed. Patients who make it to the ER in time for effective control of a GI bleed may still have a more complicated clinical course (that includes preventable death), and the physician’s analysis should not be based solely on a limited comparison between imminent risks of CV or GI death. I think the whole concern about an interaction between PPIs and antiplatelet drugs has brought attention to the important balance between CV and GI risks. Based on this concern, we really have no choice but to include both in the balance.

Dr. Laflamme: A cardiac event may have a greater emotional impact on patients and physicians, but the bleeding complications are still serious, even with endoscopic interventions decreasing transfusions, hospitalization and morbidity to a certain extent. In addition, we now have PPIs which reduce the risk by 50%. A patient with GI bleeding risk factors must necessarily have a PPI with the antiplatelet prescription. The prospective COGENT study (N=3761) on concomitant use of a PPI (omeprazole) and clopidogrel has shown a significant risk reduction of GI bleeding without increased cardiac events.

Dr. Ménard: I do think that the combination of antiplatelet and PPI is safe. If antiplatelet therapy is indicated, just follow the simple ACG/ AHA/ACC consensus algorithm and the patient will receive the safest treatment available.

Now that there is an expert consensus with evidence-based recommendations, are you more reassured about using PPIs and clopidogrel together than you were, say, 2 or more years ago?

Dr. Bradette: The consensus document was not necessary for reassurance, as I had been following these data carefully and had already arrived at the same conclusion. However, I think that it should be reassuring to others who were unsure about how to find a balance between the CV and GI risks. An evidence-based document is always helpful, especially for those who are outside of their field of expertise. Again, a document that highlights both the CV and GI risk is useful for understanding that these must be considered in relationship to each other when antiplatelets are being prescribed.

Dr. Laflamme: Yes, I am more comfortable in prescribing a PPI with an antiplatelet, although I have never stopped prescribing PPIs at the minimal effective dose in order to prevent a GI bleed, treat severe esophagitis, relieve dyspepsia and/or reflux. Literature data are solid, although other prospective data are desirable because of the possible biological interaction between a PPI and an antiplatelet. As we do with NSAIDs, I think we need to consider the GI risk as much as the cardiac risk when prescribing an antiplatelet.

Dr. Ménard: We should always analyze retrospective data carefully. The problem when comparing patients with and without a PPI is that the population taking a PPI is usually sicker and will generally experience worse outcomes. I think that the uncertainty about the safety of the combination of PPI and clopidogrel is slowly fading away as the most eloquent prospective data fail to show an increased risk, though we are still awaiting a well-designed clinical and pharmacogenetic study that will look at this important question.