Preclinical pharmacology of celecoxib and demonstration of superior GI safety compared with NSAIDs in dogs

Specific cyclooxygenase (COX)-2 inhibitors are expected to offer improved GI safety over current NSAIDs in the treatment of inflammation and pain. Celecoxib, a specific COX-2 inhibitor, exhibits a 375-fold selectivity for COX-2 based on the ratio of IC50 's for inhibition of recombinant human COX isoforms. Marked selectivity for COX-2 is also apparent when comparing the respective ED50's for decreases in gastric and inflammatory prostaglandins in vivo. These data suggest that anti-inflammatory doses of celecoxib will not be ulcerogenic. The relationship of selectivity to GI safety was studied with these compounds in dogs because their sensitivity to NSAID-induced GI injury is useful for predicting improved safety within this class. Meloxicam or nabumetone were given orally to dogs for 2 weeks at near-therapeutic dosages; celecoxib was given at supertherapeutic dosages for 13 weeks. Evidence of GI and renal injury was observed in dogs receiving meloxicam at plasma concentrations as low as three-times therapeutic; and in all dogs receiving nabumetone, even at therapeutic plasma levels, similar to NSAIDs. No evidence of GI injury was seen with celecoxib after a longer duration of dosing at plasma concentrations 16-fold greater than therapeutic. We conclude that: 1) celecoxib is at least an order of magnitude safer than meloxicam and nabumetone, which are poorly differentiated from standard NSAIDs, 2) a high degree of selectivity may be needed for improved GI safety in man, and 3) in vitro data alone are inadequate for predicting improved GI safety.