Background: The predicted gastrointestinal tolerability of specific cyclooxygenase-2 inhibitors could be due to either a lack of 'topical' irritation and/or lack of effect on cyclooxygenase-1.
Methods: Key pathophysiologic steps (in vitro and in vivo uncoupling, intestinal prostanoid levels (prostaglandin E, thromboxane B2, and 6-keto-prostaglandin F1alpha), intestinal permeability (51Cr-ethylenediaminetetraacetic acid), inflammation (faecal excretion of a granulocyte marker protein), and ulcer counts) in enteropathy induced by nonsteroidal anti-inflammatory drugs were assessed after administration of indomethacin, 10 mg/kg, and 15 (roughly equipotent), 30, and 60 mg/kg of the preferential cyclooxygenase-2 inhibitor nimesulide.
Results: Indomethacin uncoupled oxidative phosphorylation at lower concentrations than nimesulide in vitro. Indomethacin was associated with electron microscopy changes suggestive of uncoupling in 60%-70% of enterocytes examined, whereas nimesulide affected 10%-30% of enterocytes, depending on the dose. Indomethacin increased intestinal permeability and caused inflammation and ulcers with 71%-96% reductions in prostanoid levels. Nimesulide at 15 mg/kg caused no damage, whereas 30 and 60 mg/kg nimesulide were associated with significant decreases in mucosal prostanoids (46%-75%), but only the 60-mg/kg dose caused a transient increase in intestinal permeability. However, at none of the doses did nimesulide cause intestinal inflammation or ulcers.
Conclusions: These results endorse the idea that selective cyclooxygenase-2 inhibitors may be associated with some gastrointestinal tolerance due to their selectivity for cyclooxygenase-2, inhibiting cyclooxygenase-1 at only very high doses, and reduced topical irritation.