Intrathecal Dexmedetomidine Attenuates Hypercapnic but Not Hypoxic Cerebral Vasodilation in Anesthetized Rabbits

The procedures used in this study conformed to the Guiding Principles in the Care and Use of Animals approved by the Council of the American Physiologic Society, and the experimental protocols were approved by the Institutional Committee for Animal Care at Gifu University School of Medicine. The experiments were performed on 55 anesthetized rabbits weighing 2.3–2.6 kg. Each animal was anesthetized with pentobarbital sodium (20 mg/kg intravenously) and maintained with a continuous infusion of the same agent (4 mg · kg⁻¹· h⁻¹), and each was mechanically ventilated through a tracheotomy tube with oxygen-enriched room air (arterial oxygen content, 14–15 vol%). The tidal volume and respiratory rate were adjusted to maintain arterial carbon dioxide partial pressure between 35 and 40 mmHg. Catheters were placed in the femoral vein and artery for administration of fluid and drugs and for the continuous monitoring of the mean arterial blood pressure (MAP) and heart rate (HR) and for blood sampling, respectively. Rectal temperature was maintained between 37.5 and 38.5°C by means of a heating blanket.

A closed cranial window was used to observe the cerebral pial microcirculation. Each animal was placed in the sphinx posture, the scalp was retracted, and a 5-mm-diameter hole was made in the parietal bone. The dura and arachnoid membranes were opened carefully, and a ring with a cover glass was placed over the hole and secured with dental acrylic. The space under the window was filled with artificial cerebrospinal fluid (aCSF). The composition of aCSF was Na⁺151 mEq/l, K⁺4 mEq/l, Ca²⁺3 mEq/l, Cl⁻110 mEq/l, and glucose 100 mg/dl. This fluid was maintained at 37.0°C, pH 7.48, and continuously bubbled with a gas mixture of 5% CO²and air. Two polyethylene catheters were inserted through the ring; one was attached to a reservoir bottle containing aCSF to maintain the desired level of intrawindow pressure (5 mmHg), and the other was used for continuous monitoring of intrawindow pressure. The volume below the window was between 0.5 and 1 ml. The lumbar paraspinal muscles were exposed by a longitudinal midline skin incision from the lower lumbar level to the upper sacral level. After the spinous process had been removed using a rongeur, a laminectomy was performed with an electric drill at the sixth lumbar vertebra. A polyethylene catheter was inserted into the intrathecal space for a distance of 1 cm rostrally.

The diameters of three pial arterioles were measured in each cranial window using a videomicrometer (Olympus Flovel videomicrometer, Model VM-20; Flovel, Tokyo, Japan) on a television monitor, which received pictures from a microscope (Model SZH-10; Olympus, Tokyo, Japan). The data from the pial views were stored on videotape for later playback and analysis. We averaged the pial arteriolar diameters from each animal to yield an average diameter per animal and handled that single value statistically. Pial arterioles ranged from 50 to 120 μm in diameter.

The study was divided into three parts. In the first, we evaluated the effects of intrathecal DXM (Abbott Laboratories, North Chicago, IL) on the cerebral vasodilator response to hypercapnia. In the second, we evaluated the effects of DXM on the cerebral vasodilator response to hypercapnia in the presence of intrathecal yohimbine (Wako, Osaka, Japan), an α²-adrenoceptor antagonist. In the third, we evaluated the effect of DXM on the cerebral vasodilator response to hypoxia. All experiments were conducted after at least 30-min recovery from the surgical preparation.

In the first set of experiments, the changes in pial arteriolar diameter induced by hypercapnia and various physiologic variables (including MAP, HR, rectal temperature, arterial blood gas tensions, and p  H) were evaluated (STAT Profile-5; NOVA Biomedicals, Waltham, MA) whenever we measured the diameters of pial arterioles before and 30, 60, and 90 min after the intrathecal administration of DXM (2 μg/kg in 0.1 ml/kg normal saline; n = 7) or normal saline (0.1 ml/kg; n = 8) being given as a control. Hypercapnic challenges were induced before and 30, 60, and 90 min after the intrathecal administration of the aforementioned drugs by carbon dioxide gas addition to inspiratory gases. After 5 min at a stable level (arterial carbon dioxide partial pressure, approximately 60 mmHg), pial arteriolar diameters were measured.

In the second set of experiments, intrathecal yohimbine (20 μg/kg) was given as pretreatment, and 30 min later, DXM was administered intrathecally (2 μg/kg in 0.1 ml/kg normal saline; n = 7). Measurements of physiologic variables were made as in the first set of experiments.

In the third set of experiments, we examined the effects of intrathecal DXM (2 μg/kg in 0.1 ml/kg normal saline; n = 7) or normal saline (0.1 ml/kg; n = 7) as a control, on the cerebral vasodilation induced by mild (oxygen content, 8–10 vol%) or moderate (oxygen content, 5–7 vol%) hypoxia (OSM3, Radiometer, Copenhagen, Denmark). We performed a hypoxic challenge test before and 30 min after the intrathecal administration of DXM. Fifteen minutes after the arteriolar diameter returned to baseline value after the hypoxic challenge, we performed another hypoxic challenge test. We completed the hypoxic experiments within 60 min after the administration of DXM. We randomized the induction of mild or moderate hypoxia. These oxygen levels were produced by adding supplementary nitrogen to the inspired gas. Changes in physiologic variables were measured as in the first set of experiments.

In an additional set of experiments, we examined the effects of systemic DXM (2 μg/kg and 10 μg/kg; n = 6), topical DXM (10⁻⁸m and 10⁻⁶m; n = 6), and lumbar intrathecal clonidine (2 μg/kg; n = 7; Boehringer Ingelheim, Ridgefield, IL) on pial arteriolar diameter and on the change induced by hypercapnia. At 15 min after intravenous injection of 2 μg/kg DXM, we performed hypercapnic challenge. Fifteen minutes after arteriolar diameter returned to baseline value, we administered 10 μg/kg DXM intravenously. The second hypercapnic challenge was induced 15 min after intravenous injection of 10 μg/kg DXM. In topical administration studies, hypercapnia was induced 5 min after topical application of 10⁻⁸m DXM. We flushed the window continuously with aCSF for 30 min to establish baseline arteriolar diameter value and then administered 10⁻⁶m DXM. A second hypercapnic challenge was induced 5 min after topical administration of 10⁻⁶m DXM. For the intrathecal clonidine study, rabbits were treated similarly to those given DXM as described in the first set of experiments. Measurements of physiologic variables were made similarly as in the first set of experiments.

All physiologic variables within groups were examined by a one-way analysis of variance for repeated measurements followed by the Scheffé F test for post hoc  comparisons. The differences between groups were examined by a two-way analysis of variance. An unpaired t  test with Bonferroni’s correction was used to assess differences found between groups. A paired t  test was used to assess differences found between variables before and after hypercapnia or hypoxia. Significance was considered to be demonstrated at P < 0.05. All results are expressed as mean ± SD.

In the presence of DXM, the pial arteriolar dilator response to hypercapnia was significantly less than in control (normal saline) at 30, 60, and 90 min after the intrathecal DXM (P < 0.05;fig. 1). Although intrathecal yohimbine alone did not change arteriolar diameters, pretreatment with yohimbine completely blocked the attenuation of the arteriolar dilator response to hypercapnia induced by DXM (fig. 1). Intrathecal DXM into the lumbar region, per se , did not affect the pial arteriolar diameter (table 1). MAP after the intrathecal administration of DXM was significantly lower than that of control at 30, 60, and 90 min (P < 0.05), but it was not changed by hypercapnia in any of the groups (except at 30 min after the administration of DXM in the yohimbine-pretreatment group;table 2). HR did not change after the intrathecal administration of DXM in any of the groups, but it decreased significantly in response to hypercapnia (P < 0.05;table 2). Arterial p  H decreased significantly in response to hypercapnia without changes in arterial oxygen tension (table 2).

Intrathecal DXM did not alter the dilation of pial arterioles to mild or moderate hypoxia compared with control (fig. 2). Pial arteriolar diameter was not changed by the intrathecal DXM (table 3). MAP decreased after the intrathecal DXM and increased significantly during mild and moderate hypoxia (P < 0.05) but did not change in the response to mild or moderate hypoxia in the control group (table 4). HR did not change after the intrathecal administration of DXM but was decreased significantly by both mild and moderate hypoxia in both control and DXM groups (P < 0.05;table 4). Arterial p  H, carbon dioxide tension, and total hemoglobin were unchanged by either level of hypoxia in either group (table 4).

The cerebral pial arteriolar dilator responses to hypercapnia were not significantly altered by systemic administration of DXM 2 μg/kg but were attenuated by systemic administration of DEX 10 μg/kg (P < 0.05;fig. 3). The larger dose of systemic DXM (10 μg/kg) constricted pial arterioles compared with DXM 2 μg/kg (−3.0 ± 3.3% and −0.6 ± 2.5%, respectively). MAP did not change in response to hypercapnia but HR decreased significantly in the control and DXM 2 μg/kg groups (table 5). Arterial p  H decreased significantly in response to hypercapnia in the systemic DXM study (table 5).

The cerebral pial arteriolar dilator responses to hypercapnia were not significantly altered by topical DXM 10⁻⁸m but were attenuated by topical administration of DXM 10⁻⁶m (P < 0.05;fig. 3). Topical DXM 10⁻⁶m, but not 10⁻⁸m, constricted pial arterioles (−11.4 ± 3.7% and −0.2 ± 1.0%, respectively). MAP did not change in response to hypercapnia but HR decreased significantly in the control and DXM 10⁻⁸m and 10⁻⁶m groups (table 6).

Pial arteriolar diameter was not affected by the intrathecal clonidine (table 1). Although the cerebral dilator response to hypercapnia was significantly attenuated in the presence of clonidine, the attenuation of arteriolar dilation at 30, 60, and 90 min after clonidine administration was significantly less than that after DXM at the same time point (P < 0.05;fig. 1).

The major findings of the present study were that intrathecal administration of DXM into the lumbar spinal regions attenuated the dilation of cerebral pial arterioles induced by hypercapnia, and that this effect of DXM was completely blocked by yohimbine, an α²-adrenergic antagonist. Intrathecal DXM did not itself induce any vasoconstriction of cerebral vessels. In the study of topical administration of DXM into the cranial window, cerebral vasoreactivity to hypercapnia was attenuated by higher concentration (10⁻⁶m) of DXM but not by its lower concentration (10⁻⁸m), although a significant vasoconstriction of cerebral vessels was observed with its only higher concentration of DXM. Thus, the direct effect of DXM on the vascular smooth muscle of the central nervous system vasculature could not totally account for the attenuation of cerebral vasoreactivity to hypercapnia during intrathecal DXM. In addition, intrathecal DXM did not affect the dilation of pial arterioles induced by mild or moderate hypoxia. Therefore, it seems that α²-adrenergic stimulation has different effects on cerebral vasodilation depending on whether it results from hypercapnia or hypoxia.

Systemic α²-adrenergic agonists have been found to decrease the CBF 4–6or constrict cerebral vessels. 8The mechanisms underlying the change in cerebral vascular tone caused by systemic α²-adrenergic agonists may involve both a direct effect on cerebral vessels and a remote central neuronal effect without direct effect on cerebral vascular smooth muscle. 7,9Using the cranial window technique, topical application of DXM under the window causes cerebral vasoconstriction via  local α²-adrenoceptor stimulation. 1,10In the present study, the spinal intrathecal α²-adrenergic agonists did not itself change the diameter of cerebral vessels, suggesting that intrathecal DXM, with doses given at the lumbar region, does not have any stimulatory effects on α²-adrenoceptors of the cerebral vasculatures. Systemic DXM (2 μg/kg) induced no effect on both pial vascular diameter and reactivity to hypercapnia. In the topical DXM study, any concentration of DXM did not blunt the vasoreactivity to hypercapnia without direct cerebral vasoconstriction. Thus, the attenuated response of pial arterioles to hypercapnia in the presence of intrathecal DXM (2 μg/kg) is unlikely to be induced by its systemic effect as a result of systemic absorption, and likely to be due to, at least in part, indirect or remote effect on cerebral vasculature. However, because a larger dose of systemic DXM (10 μg/kg) attenuated their reactivity to hypercapnia, without measuring the serum or CSF level of DXM, we may not exclude the possibility that the systemic effect of intrathecal α²-adrenergic agonists on hypercapnic cerebral vasodilation might be involved in the present results.

It is well known that hypercapnia induced a dilation of pial vessels. 8,11With regard to the effects of α²-adrenergic agonists on hypercapnic vasodilation, it has been reported that: (1) systemic DXM did not affect hypercapnic CBF reactivity 2; (2) the intracarotid injection of clonidine caused an enhanced cerebrovascular reactivity to hypercapnia 12; and (3) systemic DXM and clonidine both attenuated the hypercapnia-induced dilation of pial arterioles. 4,8However, there is little information about the effect on the reactivity of cerebral vessels that might be induced by α²-adrenergic agonists administered into the spinal intrathecal space. α²-Adrenergic agonists may act on a supraspinal brainstem center, possibly the locus coeruleus or rostral ventrolateral medulla. 7,13,14Indeed, Elam et al.  15reported that systemic clonidine suppressed locus coeruleus firing induced by hypercapnia, whereas Fale et al.  2inferred from the findings of Cedarbaum and Aghajanian 16that systemic DXM altered the CBF response to hypercapnia by a mechanism involving an indirect inhibition of neurons within locus coeruleus. It is possible that even a very small amount of 2 μg/kg DXM given into the lumbar region can reach the brainstem neurons via  the CSF circulation. Thus, locus coeruleus can be implicated in some way in the attenuation of hypercapnic vasodilation induced by α²-adrenergic stimulation by spinal intrathecally administered DXM. Furthermore, complete inhibition of the DXM-induced attenuation by yohimbine should also support the involvement of α²-adrenoceptor action for this attenuation. The activation of the sympathetic nervous system would increase cerebrovascular resistance and decrease arterial dilation during hypercapnia. 17–19However, it is not clear whether the result obtained may be related to the attenuated vasodilation to hypercapnia by DXM, which can reduce sympathetic outflow.

In the hypoxic study, we used the arterial oxygen content as an index of hypoxia, because the arterial oxygen content is related to the amount of hemoglobin present and is a more reliable guide to the level of hypoxia than is the arterial oxygen tension. 20,21It is well known that hypoxia increases CBF, but the mechanism underlying hypoxic vasodilation is not clearly defined at present. In the present study, the stimulation of α²-adrenergic receptors caused by intrathecal DXM did not attenuate the dilation of pial vessels caused by mild or moderate hypoxia. McPherson et al.  7reported that intracerebroventricular DXM did not affect the percentage increase in CBF seen during hypoxia, an observation that is consistent with the present results.

The reactivity of cerebral vessels to hypercapnic condition, but not to hypoxic condition, is suppressed by the presence of an α²-agonist around the spinal cord. The reasons for this difference remain unclear. Underwood et al.  22suggested that rostral ventrolateral medulla, an area of the brainstem, was involved in mediating the cerebral vasodilator response to hypoxia but not that to hypercapnia. Such differences between the responses to hypercapnic and hypoxic condition might be in some way related to the different results observed in these two conditions of the present study. In addition, we should consider a potential effect of pentobarbital as a basal anesthetic, which seems to have the least effect on isolated cerebral vessels. 23However, Fale et al.  2reported that pentobarbital would inhibit the cerebrovascular response to DXM. Thus, we cannot exclude the participation of pentobarbital in the present results. Moreover, because we studied a single dose of intrathecal DXM for the present study, we cannot completely exclude the possibility that a lack of dose–response study may have an effect on the difference between the hypercapnia and hypoxia studies.

In conclusion, the present study shows that intrathecal DXM attenuates the dilation of cerebral pial arterioles that occurs in response to hypercapnia, but not the dilation that occurs during hypoxia, even though these drugs given into the lumbar spinal region did not produce a vasoconstrictor effect on cerebral vessels. Because pretreatment with yohimbine completely prevented the decreased vasoreactivity induced by DXM during hypercapnia, this effect of DXM is likely to be mediated via α²-adrenergic stimulation. The differential effects of α²stimulation on the cerebral vasodilator responses induced by hypercapnia and hypoxia may result from differences in the central mechanisms involved.