The incidence of upper gastrointestinal adverse events in clinical trials of etoricoxib vs. non-selective NSAIDs: an updated combined analysis

Objective: In spite of numerous studies demonstrating the serious gastrointestinal (GI) toxicity associated with non-selective non-steroidal anti-inflammatory drugs (NSAIDs), many patients at high GI risk continue to receive prescriptions for these drugs, often without gastroprotective agents. Etoricoxib, a COX-2 specific inhibitor, was developed to provide similar efficacy and less GI toxicity than non-selective NSAIDs. We compared the incidence of upper GI Perforations, symptomatic gastroduodenal Ulcers, and upper GI Bleeding (PUBs) in a combined analysis of all randomized, double-blind, clinical trials of chronic treatment with etoricoxib versus NSAIDs completed by June 2003. Research design and methods: Data for 5441 individual subjects with osteoarthritis, rheumatoid arthritis, or ankylosing spondylitis were pooled from all 10 multinational etoricoxib trials completed by June 2003. Information on suspected PUBs was prospectively collected in all protocols, and all investigator-reported PUBs were judged by a blinded, external adjudication committee using pre-specified criteria. PUBs were analyzed using Cox proportional hazards models using terms for treatment and known PUB risk factors. Main outcome measure: The incidence of confirmed PUBs among patients treated with etoricoxib 60 mg, 90 mg, or 120 mg (combined N = 3226) was compared to that among patients treated with ibuprofen, diclofenac, or naproxen (combined N = 2215). Results: The incidence of PUBs over 44.3 months was significantly lower with etoricoxib vs. NSAIDs [cumulative incidence 1.24% vs. 2.48%, p < 0.001; rate/100 patient-years 1.00 vs. 2.47; relative risk 0.48, 95% Confidence Interval (CI) 0.32, 0.73]. Results of analysis of events occurring during the first year of treatment and subgroup analyses were consistent with the primary result. Conclusions: Treatment with etoricoxib was associated with a significantly lower incidence of PUBs than was treatment with non-selective NSAIDs. The difference was consistent in subgroups of patients defined by a variety of known risk factors.