PPIs Appear Largely Safe in Multiyear Trial

Roxanne Nelson, RN, BSN

July 22, 2019

Proton pump inhibitors (PPIs) have become one of the most commonly prescribed classes of drugs, although concerns have been raised regarding potential harms linked to long-term use.

A new randomized trial found that for the most part, use of PPIs was safe and generally was not associated with the adverse events observed in other studies. However, at least one expert thinks the trial was underpowered and shouldn't erase concerns about the drugs.

Among more than 17,000 participants, there was no statistically significant difference between patients who received the PPI pantoprazole (Protonix, Wyeth) for up to 3 years and the group that received placebo with respect to safety events. The only exception was an increase in enteric infections among patients in the PPI group (1.4% vs 1.0%; odds ratio [OR], 1.33).

Lawrence B. Cohen, MD, professor of medicine at the Icahn School of Medicine at Mount Sinai in New York City, noted that this study "provides the strongest evidence to date supporting the safety of PPIs when used for up to 3 years."

Cohen, who was not involved in the new trial, pointed out that the safety of PPIs has recently come under scrutiny in scientific and lay publications, owing to concern about a variety of potential adverse consequences of their use.

"The evidence supporting these harms is based upon observational studies that demonstrate an association between chronic PPI use and various adverse outcomes, and the reported associations have usually been quite modest," he said. "Such an association does not establish a causal relationship, since residual confounding is an inherent weakness of observational studies."

However, Cohen added that further study is needed to assess the safety of longer-term PPI use.

The study, by Paul Moayyedi, MB ChB, PhD, professor in the Division of Gastroenterology, McMaster University, Hamilton, Canada, and colleagues, was published in Gastroenterology.

Associated Harms

PPIs were introduced to the US market in 1990. Over the years, several observational studies have linked PPI use to uncommon but serious adverse events. These include osteoporosis-related fractures, Clostridium difficile infection, community-acquired pneumonia, cerebrovascular events, kidney disease, gastric cancer, and increased mortality.

Given the widespread use of PPIs, it is important to ensure that long-term use of these agents is safe, said lead author Moayyedi. To assess safety, he and his colleagues tested the findings from past observational studies in an adequately powered randomized trial.

They had previously conducted a randomized trial that evaluated the use of high-dose PPIs and aspirin for patients with Barrett esophagus and found that PPIs could reduce the rate of high-grade dysplasia and esophageal adenocarcinoma. Importantly, both agents were well tolerated, and there were few serious adverse events.

"We can't say that there is no harm whatsoever, because you can't say that about any drug," Moayyedi told Medscape Medical News.

Risk for Enteric Infections Only

The Cardiovascular Outcomes for People Using Anticoagulant Strategies (COMPASS) trial was a three-by-two multicenter, double-blind, randomized, placebo-controlled trial that evaluated use of anticoagulants and PPIs in 17,598 participants who had stable cardiovascular and peripheral artery disease.

Participants were randomly assigned to receive pantoprazole (40 mg daily, n = 8791) or placebo (n = 8807) or either rivaroxaban (Xarelto, Janssen) (2.5 mg twice daily) with aspirin (100 mg once daily), rivaroxaban (5 mg twice daily) alone, or aspirin (100 mg) alone.

The study was conducted in this way in this particular cohort, Moayyedi explained, because the investigators needed an ethical method of evaluating the safety of PPIs.

"It isn't ethical to randomize people to placebo if they really need the medication," he told Medscape Medical News. "In this study, we had an opportunity to see if we could prevent bleeding with the use of PPIs in these patients, and we could also look at the potential harms of PPIs in a randomized trial."

Although the study was conducted in persons with cardiovascular disease, the results are generalizable to other populations, Moayyedi emphasized. The presence of stable cardiovascular disease would not alter any negative impact that PPIs might have.

Every 6 months during the trial, the researchers collected data on the various conditions that have previously been associated with PPI use: pneumonia, Clostridium difficile infection, other enteric infections, fractures, gastric atrophy, chronic kidney disease, diabetes, chronic obstructive lung disease, dementia, cardiovascular disease, cancer, hospitalizations, and all-cause mortality.

The median follow-up was 3 years, although some patients were followed for up to 5 years.

For the primary efficacy endpoint, there was no significant difference in the composite outcome of myocardial infarction, stroke, or cardiovascular death with pantoprazole compared to placebo (HR, 1.04; 95% confidence interval [CI], 0.93 – 1.15).

There were no statistically significant differences between pantoprazole and placebo regarding myocardial infarction (HR, 0.94; 95% CI, 0.79 – 1.12), stroke (HR, 1.16; 95% CI, 0.94 – 1.44), and acute limb ischemia (HR, 1.13; 95% CI, 0.73 – 1.75).

Hospitalization rates (HR, 1.04; 95% CI, 0.99 – 1.09) and all-cause mortality (HR, 1.03; 95% CI, 0.92 – 1.15) were also similar between pantoprazole and placebo cohorts.

During the study period, 864 new cancers were diagnosed. Of those, 169 were related to the gastrointestinal tract (86 in the pantoprazole group and 83 in the placebo group). There were no statistically significant differences in overall cancer rates (HR, 0.99; 95% CI, 0.87 – 1.13) or in any of the primary cancer sites between both patient groups.

There were also no statistically significant differences between the two groups for the other noncardiovascular health conditions for which data were collected, with one exception.

Enteric infections occurred more frequently in patients taking pantoprazole (odd ratio [OR], 1.33; 95% CI, 1.01 – 1.75). "It was statistically significant but only marginally so," Moayyedi said.

C difficile infection was about twice as common among patients taking pantoprazole, but because there were only 13 events, the difference was not statistically significant (nine vs four; OR, 2.26, P = .18).

"Many studies showed negative effects, but not all studies are created equal," Moayyedi said. "Observational studies are subject to confounding and bias, and even though they try to adjust for this, you can't adjust for everything."

For example, patients who receive prescriptions for PPIs may be generally sicker, and statistical adjustments in observational analyses cannot account for differences in known and unknown confounders.

"That is why randomized controlled trials remain the gold standard," he added.

Repurposed Data

However, one expert has taken a more critical eye to the report by Moayyedi and colleagues.

Ziyad Al-Aly, MD, director of the Clinical Epidemiology Center and chief of the Research and Education Service at Veterans Affairs St. Louis Health Care System, Missouri, pointed out that the study was not originally designed to assess adverse events and that it was not sufficiently powered to detect adverse events, given the number of patients and the duration of follow-up.

"It was designed for a different purpose, and the authors repurposed some data for convenience and looked at adverse events," he told Medscape Medical News.

Al-Aly emphasized that the original trial was designed to investigate cardiovascular outcomes with rivaroxaban, with and without aspirin. "The primary objectives of this trial do not include testing PPI at all," he said. "And most of the results suggest that there is a signal for safety concerns about PPI, but the study power is too weak to allow meaningful and robust testing of the research question. The absence of evidence is not evidence of absence.

"Robustly designed studies which showed adverse events among PPI users included hundreds of thousands of people who were followed for much longer," Al-Aly continued, "and in some instances, for a decade."

In comparison, this was a relatively small post hoc study of only 17,000 persons who were followed for 3 years. With regard to adverse events, that is a small number of patients, and the follow-up was short. "This is way underpowered to tell us anything of consequence," he said.

Al-Aly added that PPIs are overprescribed and overused. "We advise people to take PPI only when needed, for the shortest duration of time and the minimum dose necessary," he said. "The risk of adverse events is significant, and especially so in patients who are on PPI but do not need them."

The study was funded by Bayer AG. Moayyedi has received funding for research from Allergan and Takeda. Several coauthors have disclosed multiple relationships with industry, as noted in the original article.

Gastroenterology. Published online May 29, 2019. Abstract

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.