A major international trial has found that celecoxib is safer than conventional nonsteroidal anti-inflammatory drugs when it comes to avoiding gastrointestinal events in patients with arthritis.
The data come from the GI portion of the multinational PRECISION trial, which directly compared a celecoxib, a cyclooxygenase-2 (COX-2) inhibitor, with two nonselective NSAIDs. The cardiovascular portion of the study, which found the COX-2 inhibitor noninferior for safety, was published previously (N Engl J Med 2016;375:2519-2529).
“Clinically significant events and iron-deficiency anemia occur half as often with the COX-2 inhibitor celecoxib when compared to ibuprofen or naproxen,” said Neville Yeomans, MD, professor emeritus at Western Sydney University, in Penrith, Australia. “The addition of low-dose aspirin did not annul the GI safety advantage of the COX-2 inhibitor.” Dr. Yeomans presented the GI data at the 2017 Digestive Disease Week (abstract 766).
The PRECISION trial randomly assigned 24,018 patients with rheumatoid arthritis or osteoarthritis into one of three treatment arms at 923 centers in 13 countries.
With such a large patient population and a mean follow-up of 34 months, the PRECISION trial is the most ambitious effort yet undertaken to evaluate the differences between COX-2-selective and nonselective NSAIDs. There was no placebo arm, so the relative cardiovascular and GI safety is relevant only to the comparison of these agents, but the study was designed to reflect real-world practice.
The three drugs were compared on the basis of prespecified clinically significant GI events that were adjudicated by a blinded panel of GI specialists. These events included gastroduodenal hemorrhage, gastric outlet obstruction, intestinal or colonic perforation, intestinal or colonic hemorrhage, acute GI hemorrhage of unknown origin, symptomatic gastric or duodenal ulcer, and iron-deficiency anemia of GI origin. Comparisons were made on a modified intention-to-treat basis to ensure that patients were being compared while taking their assigned medication, or up to one month afterward.
While GI event rates were relatively low, the protection afforded by celecoxib was large when calculated by hazard ratio (HR). Relative to ibuprofen, celecoxib cut the risk for events by 57% (HR, 0.43; P<0.001). For naproxen, the reduction in risk was 49% (HR, 0.51; P<0.001).
“The addition of aspirin to the COX-2-selective inhibitor did result in more GI events. In those who received celecoxib plus aspirin, there were still about half as many events as seen in those who received a nonselective NSAID plus aspirin,” said Dr. Yeomans, noting that these data conflict with some previous reports.
At 100 mg twice daily, the dosage of celecoxib was lower than in initial studies with the drug, but consistent with current indications in many countries, Dr. Yeomans said. The study dosage of naproxen was 375 mg twice daily, and the study dosage of ibuprofen was 600 mg three times per day. All patients were prescribed the proton pump inhibitor esomeprazole, which was taken by 99% of participants in the study, he said.
“The rate of most of the individual events was lower in patients randomized to celecoxib relative to the nonselective NSAIDs, but the only difference that reached statistical significance by itself was the lower rate of gastric and duodenal ulcers,” Dr. Yeomans reported. He noted that in patients with rheumatoid arthritis, steroids increased the risk for GI events in those receiving either ibuprofen or naproxen, but not in those receiving celecoxib.
The effect of Helicobacter pylori on risk for GI events in the context of assigned therapies also was evaluated in this study. Surprisingly, and in contrast to many guidelines that recommend a test-and-treat strategy for H. pylori infection before initiating NSAIDs, a positive H. pylori serology was not associated with an increased risk for GI events. Although Dr. Yeomans reiterated that essentially all patients also were receiving a PPI, he said the results argue that a test-and-treat strategy may not be needed.
The low rate of GI events in PRECISION was unexpected. According to Dr. Yeomans, these occurred at a rate of about 0.5% per year. Rates of events and iron-deficiency anemia were higher for patients with rheumatoid arthritis than for those with osteoarthritis, but the overall rate of small-bowel bleeding occurred at a lower rate than predicted. The relative cardiovascular and GI safety of the agents tested in the trial may not be representative of unselected patients with osteoarthritis and rheumatoid arthritis because established cardiovascular disease or high risk for cardiovascular disease were inclusion criteria, Dr. Yeomans cautioned.
Increased risk for cardiovascular disease is important to consider when evaluating the new data, said James Scheiman, MD, chief of the Division of Gastroenterology and Hepatology at the University of Virginia School of Medicine, in Charlottesville.
“We learn that in addition to PPI cotherapy, there are GI benefits of a COX-2-selective agent over a nonselective agent without adverse cardiovascular outcomes among these aspirin-using patients,” Dr. Scheiman said. “The low GI event rate and low absolute risk reduction suggests that the cost-effectiveness of the COX-2-selective agent is modest in this setting.
“However, a selective NSAID may be quite valuable for some patient populations, certainly those at increased GI risk. The study also reemphasizes the value of PPI cotherapy in reducing upper GI events due to NSAIDs with or without aspirin,” Dr. Scheiman added.
Dr. Scheiman reported financial relationships with Aralez and Stryker. Dr. Yeomans reported no relevant financial conflicts of interest.