Risks for clinical events throughout the GI tract were lower with a cyclooxygenase-2 inhibitor than with a nonselective nonsteroidal anti-inflammatory drug plus a proton-pump inhibitor.
The association between nonsteroidal anti-inflammatory drugs (NSAIDs) and upper gastrointestinal (GI) bleeding and ulceration is well documented. Such adverse events can be reduced by use of a cyclooxygenase (COX)-2 selective NSAID or a proton-pump inhibitor (PPI) with a nonselective NSAID.
However, prospective trials that evaluate these strategies to prevent NSAID-related events throughout the GI tract are lacking.
To address this issue, investigators conducted an industry-sponsored trial (CONDOR), involving 4484 patients with osteoarthritis or rheumatoid arthritis from 196 centers in 32 countries. Patients were randomized to receive celecoxib (Celebrex; 200 mg twice daily) or diclofenac (75 mg twice daily) plus omeprazole (20 mg once daily).
All patients had elevated risk for GI complications (age ≥60 or history of gastroduodenal ulcers). Patients with Helicobacter pylori infection and those taking anticoagulant or antiplatelet agents were excluded.
The study endpoints included overt upper or lower GI bleeding, perforation, gastric outlet obstruction, and anemia (decrease in hemoglobin level of ≥20 g/L) from a GI source or from presumed occult GI blood loss.
During 6-month follow-up, fewer patients receiving celecoxib experienced a primary endpoint event than did those receiving diclofenac plus omeprazole (0.9% vs. 3.8%; hazard ratio favoring celecoxib, 4.3; P<0.0001). p="0.03)" p="0.0006).">The authors concluded that the risk for an adverse event anywhere in the GI tract is lower for patients taking a COX-2 selective NSAID than for those taking a nonselective NSAID plus a PPI.
Comment: This paper provides important information about the prevention of lower GI complications from NSAID use. Of note, no between-group differences were seen in upper GI complications (3 in each group) or in overt lower GI bleeding (1 in each group). The difference in outcomes was explained by the large disparity in the number of patients with new-onset anemia (15 in the celecoxib group vs. 77 in the diclofenac plus omeprazole group).
Occult blood loss in the absence of overt bleeding is well documented in patients taking NSAIDs and is independent of gastric acid. In this study, the great majority of patients with anemia were assumed to have experienced occult blood loss in the absence of a documented source.
The number of patients with GI blood loss might be overstated, but it should not differ between groups.
These data suggest that a PPI reduces upper GI events in patients taking nonselective NSAIDs but does not reduce anemia from lower GI blood loss as much as the use of a selective NSAID.
David J. Bjorkman, MD, MSPH (HSA), SM (Epid.)
Published in Journal Watch Gastroenterology June 25, 2010