Effect of COX-1/COX-2 Inhibition versus Selective COX-2 Inhibition on Coronary Vasodilator Responses to Arachidonic Acid and Acetylcholine

The effect of a nonselective COX-1/COX-2 inhibitor, naproxen, was compared with a COX-2-selective inhibitor (SC-58236) on coronary vasodilatory responses in the anesthetized dog. Coronary vasodilation was induced by direct intracoronary injection of acetylcholine (ACH) and arachidonic acid (AA) in control animals and in those treated with either naproxen (1, 3, or 10 mg/kg p.o. 24 h prior to the experiment) or SC-58236 (1, 5, or 15 mg/kg p.o. 24 h prior to the experiment). Naproxen, at 10 mg/kg, significantly attenuated the AA-induced vasodilation (prostacyclin dependent) with no effect on ACH-induced vasodilation (nitric oxide dependent). SC-58236 failed to attenuate either AA- or ACH-induced vasodilation. Ex vivo assays were utilized to establish inhibition of COX-2 (lipopolysaccharide-stimulated prostaglandin E₂ formation) and COX-1 (serum thromboxane B₂) in blood taken from dogs administered 1, 3, or 10 mg/kg naproxen or 15 mg/kg SC-58236. Naproxen (3 and10 mg/kg) and SC-58236 (15 mg/kg) markedly reduced the lipopolysaccharide-induced prostaglandin E₂ formation, whereas SC- 58236 (15 mg/kg) had no effect on serum thromboxane B₂. Naproxen significantly reduced thromboxane B₂ at all three doses studied. Furthermore, naproxen (10 mg/kg p.o.) significantly inhibited the AA-induced platelet aggregation (thromboxane B₂ dependent), whereas SC-58236 had no effect. Collectively, these results demonstrate that SC-58236 is selective for COX-2, while naproxen is a nonselective inhibitor. These data also suggest that vasodilatory responses to AA in the dog are primarily COX-1 dependent. Selective COX-2 inhibition does not affect either prostacyclin or nitric oxide mediated vasodilation in the canine coronary circulation.