Abstract

Prostaglandins are generated through two isoforms of the enzyme cyclooxygenase, the constitutively expressed cyclooxygenase (Cox)-1 and Cox-2, which is induced at sites of inflammation. Selective inhibition of Cox-2 is desirable as this may avoid the gastropathy and platelet inhibition seen with nonselective agents. Moreover, these agents will allow us to examine the relative contribution of the two isoforms to prostaglandin formation in man. We examined the activity of nimesulide, a Cox-2 selective nonsteroidal antiinflammatory drug, in vitro against purified enzymes and in vivo in man. Nimesulide 100 mg twice daily or aspirin 300 mg three times daily were administered randomly for 14 days to 20 subjects complaining of musculoskeletal pain. Serum thromboxane B₂ was determined as an index of Cox-1 activity and endotoxin-induced prostaglandin E₂ formation in whole blood as an index of Cox-2 activity. Urinary excretion of prostaglandin metabolites was determined by GC/MS. Nimesulide was highly selective against ovine Cox-2, so that at concentrations attained in vivo, it had no effect on Cox-1 but completely suppressed Cox-2. Aspirin markedly inhibited serum thromboxane B₂ (181.92 ± 19.77 to 2.83 ± 0.96 ng/ml, P < .002), whereas nimesulide had very little effect (207.53 ± 47.30 to 181.15 ± 54.59 ng/ml). In contrast, nimesulide suppresses endotoxin-induced prostaglandin E₂formation (35.03 ± 8.73 to 2.62 ± 0.95 ng/ml, P = .002). As expected, aspirin reduced TX metabolite excretion, whereas nimesulide had no significant effect. In contrast, both compounds suppressed PGI₂ formation to the same extent. The findings suggest that TX is largely Cox-1 derived. Moreover, Cox-2 is expressed in man and generates prostaglandin I₂.