MECHANISMS OF DEXMEDETOMIDINE-INDUCED CEREBROVASCULAR EFFECTS IN CANINE IN-VIVO EXPERIMENTS

Dexmedetomidine decreases cerebral blood flow without significantly affecting cerebral oxygen consumption in anesthetized dogs. To assess the direct cerebrovascular effects of dexmedetomidine, we investigated the responses of vasomotor tone to topical application of dexmedetomidine to pial vessels in vivo, using a parietal cranial window. Forty-one dogs were anesthetized with pentobarbital. In 20 dogs, we topically applied six concentrations of dexmedetomidine solution (10(-8), 10(-7), 10(-6), 10(-5), 10(-4), 10(-3) M) and directly measured pial arterial and venous diameters. In 10 dogs, the inhibitory effects of pretreatment of pial vessels with 10(-5) M yohimbine were examined after the application of 10(-5) M dexmedetomidine. In the remaining 11 dogs, the effects of 10(-3) M dexmedetomidine were evaluated in the presence of N(o)mega-nitro-L-arginine methyl ester (L-NAME), glibenclamide, or propranolol. Dexmedetomidine significantly constricted pial arteries and veins in a concentration-dependent manner (10(-7) M to 10(-4) M; P < 0.05). Yohimbine blocked dexmedetomidine-induced constriction of pial vessels (both large and small arteries and large veins P < 0.0001; small veins P < 0.005). However, when the highest concentration of dexmedetomidine (10(-3) M) was administered under the window, pial vessel diameter was not significantly altered. In the presence of glibenclamide, 10(-7) and 10(-3) M dexmedetomidine induced a significant decrease in pial arterial diameter compared with 10(-7) and 10(-3) M dexmedetomidine solution alone, respectively (P < 0.05). L-NAME or propranolol did not affect the dexmedetomidine-induced constriction. Although yo-himbine, glibenclamide, or propranolol did not change pial vascular diameter, L-NAME significantly constricted both pial arteries and veins (P < 0.05). Our study demonstrates that topical application of dexmedetomidine constricts both pial arterial and venous vessels in a concentration-dependent manner. The vasoconstrictor effects of dexmedetomidine appear to be mediated via activation of alpha(2)-adrenoceptors, although this action is accompanied by activation of adenosine triphosphate sensitive K+-channels as a counterbalancing vasodilatory effect. The present results also suggest that the resting tone of pial arteries and veins does not depend on alpha(2)- and beta-adrenergic control, but is influenced by nitric oxide.