Background & aims: Our aim was to establish the incidence of symptomatic upper gastrointestinal ulcers, ulcer perforation, ulcer obstruction, or bleeding episodes (PUBs) associated with the use of selective cyclooxygenase-2 inhibitors at standard clinical doses compared with placebo. We report here on the PUB outcomes associated with the use of rofecoxib 25 mg in a 3-year, multicenter, double-blind, placebo-controlled trial designed to determine the effect of rofecoxib on the risk of recurrent neoplastic polyps of the colon.
Methods: A total of 2587 patients with a history of colorectal adenomas underwent randomization to 25 mg/day of rofecoxib or to placebo. Investigator-reported PUBs were adjudicated by an external blinded committee. Kaplan-Meier and Cox proportional hazards techniques were used to estimate incidence and relative risks of PUBs in an intention-to-treat analysis.
Results: Patients assigned to rofecoxib had a higher incidence of confirmed PUBs than those randomized to placebo (.88 vs .18 events per 100 patient-years; relative risk, 4.9; 95% confidence interval, 1.98-14.54). The incidence of confirmed complicated PUBs (ulcer perforation, obstruction, or bleeds) was low, but was numerically higher in the rofecoxib than in the placebo group (.23 vs .06 events per 100 patient-years; relative risk, 3.8; 95% confidence interval, .72-37.46; P = .14). Rofecoxib increased the incidence of confirmed PUBs vs placebo in both low-dose aspirin users and nonusers.
Conclusions: Among patients with a history of colorectal adenomas, the long-term use of 25 mg/day of rofecoxib was associated with an increased risk of clinically relevant upper gastrointestinal events when compared with placebo.