Possible Cardiovascular Risk of PPIs Lingers in 2 New Studies

Despite substantial limitations, 2 recent studies appear to keep alive concern over potential adverse cardiovascular effects of proton pump inhibitors (PPIs), not only in ACS patients but also in the general population. Both findings point away from a drug-drug interaction with clopidogrel and toward other speculative mechanisms.

PPIs in Medically Managed ACS Patients

In the first study, a prespecified analysis of the TRILOGY ACS trial published online June 10, 2015, ahead of print in the American Heart Journal, use of PPIs was linked to a lower incidence of MI with prasugrel than clopidogrel despite similar levels of platelet reactivity.  

In the main trial, 7,243 patients under 75 years with unstable angina or NSTEMI treated medically without revascularization were randomized to clopidogrel or prasugrel. Those not previously on an antiplatelet regimen received a loading dose of 300 mg of clopidogrel or 30 mg of prasugrel followed by a maintenance dose of 75 mg of clopidogrel or 10 mg of prasugrel. At a median follow-up of 17 months, rates of bleeding and of the primary composite endpoint (cardiovascular death, MI, or stroke) were similar between groups.

In the current analysis, Jose Carlos Nicolau, MD, PhD, of the University of São Paulo Medical School’s Hospital das Clinicas (São Paulo, Brazil), and colleagues focused on the influence of PPI use, with additional analysis of 2,049 patients who underwent serial platelet reactivity assessment with the VerifyNow P2Y12 assay (Accumetrics).

At baseline, 23% of the overall cohort was on a PPI. These patients were older and more often initially diagnosed with NSTEMI and more commonly had previous CABG and peptic ulcer disease. They were also more likely to be treated with ACE inhibitors/angiotensin receptor blockers, statins, and low-dose aspirin and to have undergone angiography prior to randomization.

In unadjusted outcomes, patients treated with clopidogrel plus a PPI had higher rates of the primary endpoint and MI than those who received clopidogrel without a PPI or prasugrel with or without a PPI (log rank P < .001 for both). No differences were seen for cardiovascular death or stroke (log rank P = .3 and log rank P = .39). There was an interaction between PPI use and randomized treatment for the primary endpoint (P for interaction = .02), which was driven by the advantage of the prasugrel groups over the clopidogrel groups over follow-up (P for interaction = .004).

After adjustment, the influence of baseline PPI use was seen only in a lowered risk for MI (HR 0.61; 95% CI 0.42-0.88) with prasugrel vs clopidogrel but not the other endpoints (P = .01 for interaction). This pattern was seen only for omeprazole and pantoprazole.

Over 30 months, there were 81 adjudicated TIMI major or minor GI bleeding events (1.54%) that originated in the GI tract, with no difference in incidence between patients who did or did not use PPIs. Additionally, no treatment group had a higher rate of such bleeding (log-rank P = .08).

GI bleeding did predict subsequent PPI use (P < .001) but not MI (P = .61). TIMI major/minor GI bleeding did not modify the adjusted interaction between PPI use and prasugrel vs clopidogrel treatment with regard to MI risk (P for interaction = .04).

In several models, P2Y12 reaction unit (PRU) values were consistently lower with prasugrel vs clopidogrel, but PPI use did not modify the values for either drug. Taking into account the different PPI types (or no PPI), there was no difference in PRU values between treatment groups. In addition, patients’ weight had no influence on the impact of PPIs.

Disconnect Between Platelet, Clinical Findings

“This analysis doesn’t support any effect of PPIs on platelet function per se, so that is useful information and supports earlier studies,” coauthor Deepak L. Bhatt, MD, MPH, of Brigham and Women’s Hospital (Boston, MA), told TCTMD in a telephone interview. “But it is more challenging to know what to make of the lower frequency of MI with prasugrel vs clopidogrel in patients on PPIs without a drug-drug interaction blunting the effect of clopidogrel,” he said.

Because these signals are discordant—and the findings come from a substudy of a substudy—“I would be very cautious in interpreting them,” Dr. Bhatt observed. “There is nothing here that should change practice; it’s just something that future researchers should be aware of.”

Risk From PPIs May Not Involve Effect on Platelets

In the second study, investigators led by Nigam H. Shah, MBBS, PhD, of Stanford University (Stanford, CA), mined data on 2.9 million individuals from 3 databases, uncovering a link between PPI exposure and MI risk. Derived from a novel strategy, the findings for what the authors term “pharmacovigilance” were reported online June 10, 2015, in PLoS ONE.

Patients with gastroesophageal reflux disease (GERD) who were taking PPIs appeared to have an increased risk of MI (HR 1.16; 95% CI 1.09-1.24). In addition, adjusted analysis of a prospective cohort found a 2-fold higher risk of cardiovascular mortality regardless of clopidogrel use (HR 2.00; 95% CI 1.07-3.78). H₂ blockers, an alternate therapy for GERD, were not linked to increased cardiovascular risk.

“Our observations that PPI usage is associated with harm in the general population—including the young and those taking no antiplatelet agent—suggest that PPIs may promote risk via an unknown mechanism that does not directly involve platelet aggregation,” the authors say.

One plausible explanation, they observe, may be provided by their recent preclinical research showing that PPIs inhibit an enzyme that regulates nitric oxide production, resulting in impaired vasodilation.

Still, Dr. Shah and colleagues acknowledge, “[a]lthough our data-mining pipeline has high specificity and was validated to have high accuracy (89%), there is still a possibility that the association detected is a false positive.”  They concede that the findings are only “hypothesis generating” and require confirmation in a prospective randomized study.

Confounding Likely

Asked for his evaluation of the study, Dr. Bhatt responded: “We have to keep looking at the data because we wouldn’t want to miss a possible drug interaction for a commonly used therapy. But I think there’s a good chance that the findings are due to confounding.”

A likely confounder, he suggested, is patients presenting with “heartburn” who are given a PPI but really have unstable angina. Such patients may go on to have an MI, he added, and these instances of misdiagnosis can be difficult to tease out.

He noted that the randomized COGENT study showed no adverse effect of PPIs. “Putting aside the clopidogrel story [of potentially impaired metabolism], I think if PPIs were really causing MIs, we would have seen it in these high-risk, stented ACS patients,” he observed.

“I don’t think the current observational study—despite its size—would trump the randomized COGENT data,” Dr. Bhatt concluded, adding that the large upcoming COMPASS trial, which includes an arm randomized to pantoprazole, should provide further insight into the question of the safety of PPIs.

Meanwhile, he said, clinicians should continue to prescribe PPIs—but only when they are indicated and for the appropriate length of time.

 

---

Sources:

 

1. Nicolau JC, Bhatt DL, Roe MT, et al. Concomitant proton-pump inhibitor use, platelet activity, and clinical outcomes in patients with acute coronary syndromes treated with prasugrel vs. clopidogrel and managed without revascularization: insights from the TRILOGY ACS trial. Am Heart J. 2015;Epub ahead of print.
2. Shah NH, LePendu P, Bauer-Mehren A, et al. Proton pump inhibitor usage and the risk of myocardial infarction in the general population. PLoS ONE. 2015;10:e0124653.

 

Related Stories:

 

• Comparable Long-Term Safety, Efficacy with Prasugrel, Clopidogrel in ACS
• TRILOGY ACS Substudy: No Relationship Between Platelet Reactivity, Outcomes
• Novel Mechanism for Potential Cardiovascular Harm from PPIs Identified