Pantoprazole did not reduce risk for a composite of upper gastrointestinal events in stable arterial disease

Given the limitations of this large industry-funded trial, the only definitive conclusion one can draw is that patients on aspirin and rivaroxaban with low risk for bleeding may get a very small benefit from long-term proton-pump inhibitor therapy, an ACP Journal Club commentary said.


For patients taking rivaroxaban and/or aspirin for stable cardiovascular disease, adding a proton-pump inhibitor (PPI) did not affect the risk of upper gastrointestinal events, an industry-funded trial found. However, patients who were randomized to pantoprazole, 40 mg/d, did have less bleeding specifically due to gastroduodenal lesions than those on placebo. The study patients were also all randomized to one of three anticoagulation regimens: rivaroxaban, 2.5 mg twice daily, with aspirin, 100 mg once daily; rivaroxaban, 5 mg twice daily; or aspirin, 100 mg/d.

The study was published online May 2 by Gastroenterology and summarized in the May ACP Gastroenterology Monthly. The following commentary by Pradeep Dayanand, MD, and Steven Borzak, MD, appeared in the ACP Journal Club section of the Oct. 15 Annals of Internal Medicine.

Moayyedi and colleagues conducted the largest randomized trial to date on PPI therapy to prevent GI bleeding in patients with coronary artery disease who are receiving antithrombotic therapy. This is an urgent question, particularly in light of the common use of PPIs and growing concerns about possible risks associated with long-term use. A recently published secondary analysis of the COMPASS trial concluded that use of pantoprazole for 3 years is reasonably safe except for a possible increased risk for enteric infections. However, the analysis of the pantoprazole component of the trial is limited by the exclusion of more than a third of patients who were randomized in the rivaroxaban/aspirin component due to the “need for continuous PPI” as defined by local investigators or treating physicians. As a result, only 3% of patients had previous peptic ulcer disease and only 5% were receiving nonsteroidal anti-inflammatory drugs—a sample that had a decidedly low risk for bleeding and was less likely to benefit from gastric protection. A further concern is that after the antithrombotic groups were terminated at 23 months, “all patients continued on at least taking aspirin for the remainder of the trial”, but their bleeding events were counted as randomized in the rivaroxaban groups. Thus, this trial cannot answer the question of whether pantoprazole may preferentially benefit use of aspirin over rivaroxaban or the combination.

Despite the large size and statistical power of the trial, the only conclusion is that in patients who have a low risk for bleeding and are receiving aspirin and rivaroxaban, long-term PPI offers a very small benefit in reducing “overt gastroduodenal bleeding” but not overall GI bleeding. How to manage patients with a higher risk for GI bleeding who are receiving more intense anticoagulation (e.g., for atrial fibrillation stroke prevention) remains unanswered.