For most patients taking antiplatelet and/or anticoagulant therapies, the proton pump inhibitor (PPI) pantoprazole is unnecessary, based on findings from the prospective COMPASS trial, which involved more than 17,000 participants.
Pantoprazole may reduce the risk of bleeding from gastroduodenal lesions, but it is unlikely to prevent upper-gastrointestinal events, reported lead author, of McMaster University in Hamilton, Canada, and colleagues.
The investigators wrote in Gastroenterology, “Guidelines suggest that patients receiving the combination of antiplatelet and anticoagulant therapy should receive PPIs to reduce the risk of upper-GI bleeding. However … there are no randomized data to support the use of PPI therapy in patients taking oral anticoagulants, and a paucity of data relating to aspirin.”
To fill this knowledge gap, the investigators recruited 17,598 participants from 33 countries who had stable peripheral artery disease and cardiovascular disease. Participants were randomized to one of three groups: 100-mg aspirin once daily, 5-mg rivaroxaban twice daily, or a combination of 2.5-mg rivaroxaban twice daily with 100-mg aspirin once daily. This part of the trial was discontinued before completion because of early cardiovascular advantages associated with combination therapy over aspirin alone, and related findings were reported previously. While combination therapy did reduce cardiovascular risks, it had less favorable effects on gut health, highlighted by an associated increase in major GI bleeding events. Despite early cessation of the cardiovascular portion of the trial, the pantoprazole regimen was continued, offering a look at the effect of long-term PPI use on gut health.
At baseline, about two-thirds of participants (64%) were not taking a PPI, requiring randomization to either 40-mg pantoprazole once daily or matching placebo. The primary efficacy outcome was time to first upper-GI clinical event, defined as a composite of the following: upper-GI obstruction, perforation, at least five gastroduodenal erosions with at least 3 days of GI pain, symptomatic gastroduodenal ulcer involving at least 3 days of GI pain, overt upper-GI bleeding of unknown origin, occult bleeding (drop in hemoglobin of at least 2 g/dL), overt bleeding with a gastroduodenal lesion (active bleeding during endoscopy), or a symptomatic gastroduodenal ulcer involving at least 3 days of GI pain. In addition to this measure, the investigators evaluated a post-hoc endpoint with a looser definition of peptic ulcer events, most notably eliminating the requirement that a lesion be actively bleeding during endoscopy.
Most patients in the trial (78%) were male, and 23% were current smokers. Smaller proportions of the population were taking a nonsteroidal anti-inflammatory drug (5%) and/or had a history of peptic ulcer disease (2.6%). The median follow-up was 3.01 years, ranging from 2.49 to 3.59 years. Permanent discontinuations occurred at approximately equal rates in the pantoprazole (21%) and placebo (22%) group, after a median of 11 months (338 days). In both groups, more than 96% of participants who continued treatment took their medications as prescribed at least 80% of the time.
Analysis showed that upper-GI events occurred marginally less often in the pantoprazole group than the placebo group, but without statistical significance (1.2% vs. 1.3%; P = .35). Of the outcomes measured, only overt bleeding of gastroduodenal origin detected by radiography or endoscopy was statistically less common in the pantoprazole group than the placebo group, with a 48% reduced rate (0.2% vs. 0.4%; P = .03). No statistical efficacy differences or statistical interactions were detected between population subgroups.
“The data suggest that routine use of PPI therapy is not warranted for patients receiving low-dose rivaroxaban with or without aspirin for the prevention of atherothrombotic events in patients with stable coronary artery disease or symptomatic peripheral artery disease, as there was no overall impact on clinical upper-GI events or upper-GI bleeding,” the investigators wrote. “This is in contrast to previous systematic reviews of randomized trials reporting that PPIs were associated with a 50%-70% reduction in bleeding and symptomatic peptic ulcers related to nonsteroidal anti-inﬂammatory drugs, including in the critical care setting.”
Post-hoc analysis, which allowed for a broader definition of upper-GI events related to gastroduodenal ulcers, revealed a slightly greater reduction in risk of bleeding lesions in patients taking pantoprazole, compared with placebo (hazard ratio, 0.45), and additional risk reductions for peptic ulcers (HR, 0.46) and erosions (HR, 0.33). Ultimately, pantoprazole reduced the combined rate of post-hoc events by 56%.
The investigators noted that these ulcer- and erosion-reducing effects of pantoprazole align with previous reports. “It is therefore possible that PPIs might be beneficial for patients at particularly high risk for peptic ulcer disease who are also taking aspirin and/or anticoagulants,” the investigators concluded.
The COMPASS trial was funded by Bayer AG. The investigators disclosed additional relationships with Allergan, Takeda, Janssen, and others.
SOURCE: Moayyedi P et al. Gastro. 2019 May 2.