Lanas A et al. (2006) Risk of upper gastrointestinal ulcer bleeding associated with selective cyclo-oxygenase-2 inhibitors, traditional non-aspirin non-steroidal anti-inflammatory drugs, aspirin and combinations. Gut 55: 1731–1738

Long-term NSAID use causes upper gastrointestinal bleeding (UGIB), but selective cyclo-oxygenase 2 inhibitors (coxibs) are thought to reduce this risk. Lanas and colleagues, however, found that low-dose aspirin combined with either coxibs or NSAIDs synergistically exacerbated patients' risk of UGIB; they recommend caution with coadministration of coxibs and NSAIDs—particularly in patients with high cardiovascular risk—until further data are obtained.

Their hospital-based, case–control study compared the risk of UGIB associated with non-aspirin NSAIDs, coxibs, aspirin, and combinations thereof, in 2,777 consecutive patients (mean age 61 years) with endoscopy-proven UGIB from peptic ulcers, and 5,335 controls matched for age, hospital and month of admission.

The NSAIDs with the lowest relative risk of UGIB were diclofenac, aceclofenac and ibuprofen, whereas piroxicam and ketorolac had the highest relative risks (1.4–2.5 versus 7.2–8.0 times that of controls). Worldwide, the authors noted that meloxicam—a moderate-risk NSAID—was prescribed more frequently after rofecoxib was withdrawn by the FDA, whereas the low-risk NSAIDs diclofenac and aceclofenac were sparsely used, and unavailable in the USA.

Rofecoxib, but not celecoxib, modestly increased patients' UGIB risk. Overall, coxibs increased UGIB risk to a lesser extent than either NSAIDs or cardioprotective aspirin (100–300 mg daily). Cardioprotective aspirin use accounted for 15% of UGIB cases, and analgesic aspirin (≥500 mg daily) was associated with a large excess risk of UGIB; the authors suggest that free availability and clinical use of analgesic aspirin should be restricted.