Preferential venoconstriction by cyclooxygenase inhibition in vivo without attenuation of nitroglycerin venodilation.

Because prostacyclin is a rather potent venodilator in vivo, we analyzed the effect of cyclooxygenase inhibition on venous tone in 14 anesthetized dogs during ganglionic and beta-adrenergic blockade and atraumatic conditions. Effective vascular stiffness (a reciprocal of effective vascular compliance) as a variable of integrated venous tone was 0.30 +/- 0.01 mm Hg.kg/ml (n = 35) and was augmented up to twofold by diclofenac (1, 3, and 10 mg/kg i.v.), ibuprofen (6 and 60 mg/kg), or indomethacin (5 mg/kg) parallel to augmentations in central venous pressure, while the rise in arterial pressure was less than half of the increase induced by equivenoconstrictor dosages of norepinephrine. After preconstriction by indomethacin or diclofenac, nitroglycerin (1.5 micrograms/kg/min) lowered effective vascular stiffness (by 24 +/- 2% or 23 +/- 5%, respectively), similarly as during preconstriction by norepinephrine (by 24 +/- 4%). Long-term cyclooxygenase inhibition (diclofenac 2 x 1 mg/kg/day for 4 days) did not modify arterial pressure, heart rate, or hematocrit levels in conscious dogs at rest, but it lowered plasma volume to 52.5 +/- 1.9 ml/kg (sham treatment: 59.1 +/- 1.6 ml/kg, p less than 0.05, n = 4). In conclusion, venoconstriction by clinical dosages of cyclooxygenase inhibitors does not interfere with the venodilator action of nitroglycerin and is compensated chronically by adjustments of plasma volume.