agonist clonidine is a commonly used centrally acting antihypertensive drug. There is interest in the potential role of clonidine and other more selective alpha2
agonists in anesthesia because they reduce opioid and volatile anesthetic requirements during surgery, reduce plasma catecholamine concentrations, improve hemodynamic stability, and have potent analgesic properties. 
receptors have a role in the regulation of the cerebral circulation, 
agonists, including clonidine, decrease cerebral blood flow (CBF), [3-6]
and produce constriction of cerebral arteries. [7-9]
Data from animal studies give conflicting results on the effect of clonidine on cerebrovascular reactivity to CO2
. Clonidine has been reported to decrease cerebral blood flow and reduce cerebral CO2
and to enhance cerebrovascular reactivity to CO2
Despite considerable interest regarding the effect of clonidine on CBF, there are no human studies that examine the effect of alpha2
agonists on cerebral CO2
Transcranial Doppler ultrasonography (TCD) is a noninvasive technique that offers reliable measurement of CBF velocity, [11,12]
and autoregulation. [14,15]
We used TCD to measure the influence of clonidine on CBF velocity, CO2
vasoreactivity, and the response of the cerebral circulation to increased arterial pressure.
Nine healthy volunteers, aged from 24 to 33 yr, gave informed consent to the study that was approved by the Committee on Human Research, University of California, San Francisco. Inclusion criteria included normal middle cerebral artery (MCA) blood flow velocities, as determined by TCD evaluation, and a negative history for cardiac or cerebrovascular disease. Volunteers were instructed to abstain from the use of alcohol, caffeine, and tobacco for 24 h before the study.
On arrival in the study room, the subject was placed in the supine position. Monitoring with non-invasive blood pressure (Dinamap 1846 SX, Critikon, Tampa, FL), electrocardiography (ECG), and pulse oximetry was commenced. A 20-gauge intravenous catheter was placed for fluid and drug administration. Lactated Ringer's solution was administered at a rate of 2-4 ml [center dot] kg sup -1 [center dot] min sup -1. A 22-gauge catheter was inserted into the radial artery to measure mean arterial pressure (MAP) continuously and to draw blood samples for measurement of PaCO2. The volunteer was then placed in a semirecumbent position for the remainder of the study. The transducer that measured blood pressure was zeroed at the level of the right atrium.
Measurement of Mean Blood Flow Velocity
Transcranial Doppler ultrasonography of the right MCA was performed using a Neurogard[registered sign] monitor (Medasonics, Fremont, CA) with the 2-MHz pulsed Doppler monitoring probe placed at the right transtemporal window. Insonation of the MCA was initiated at a depth of 45 mm. Confirmation of MCA identity was achieved by increasing insonation depth to visualization of the bidirectional flow pattern typical of the bifurcation of the internal carotid artery into the middle cerebral and anterior cerebral arteries. Insonation depth was then decreased to the point of maximal signal intensity, and the monitoring probe was locked into position with a Medasonics probe clamp and head strap. Baseline readings were obtained and included mean blood flow velocity in the right MCA (Vm) and pulsatility index (PI). 
Pulsatility index was defined as [(peak systolic blood flow velocity-peak diastolic blood flow velocity)/Vm].
Ten to 20 min were allowed for the conditions to stabilize after transition to each new experimental state, with continuous TCD monitoring throughout this time. The measured value of Vm was recorded only after it was observed to be stable for at least 1 min at the new steady-state experimental conditions. Measurements of Vm were obtained in triplicate at end-expiration and averaged to obtain a single value.
Measurement of Carbon Dioxide Vasoreactivity
Volunteers were rested in a quiet and dimly illuminated room with their eyes closed until TCD and cardiovascular measurements had been stable for at least 15 min. Then each volunteer breathed normally through a face mask attached to an anesthesia breathing circuit (FiO2, 0.21) with end-tidal CO2 measured at the circuit Y-piece (Capnomac, Datex Instrumentarium Inc., Helsinki, Finland). Ten to 15 min were allowed for the volunteer to adapt to the breathing apparatus, before baseline values of Vm, PI, end-tidal CO2, MAP, heart rate (HR), and PaCO2 were obtained.
After baseline measurements were attained (i.e., normocapnia), hypocapnia or hypercapnia was induced, the order being alternated with each volunteer. For hypocapnia, volunteers hyperventilated until the end-tidal CO2 tension decreased by 10 mmHg. When a stable plateau of end-tidal CO2 tension was achieved, Vm (hypocapnia) was measured. Volunteers then breathed normally for 10 to 15 min. To induce hypercapnia, a rebreathing technique was used. The soda lime was removed from the breathing circuit, and fresh gas flow (oxygen only) was reduced to only that required to maintain an FiO2 of 0.21-0.30. The end-tidal CO2 tension was monitored until it increased by 10 mmHg. Air was then added to the fresh gas flow, sufficient to maintain the end-tidal CO2 tension stable, and Vm (hypercapnia) was measured. Arterial blood samples were taken simultaneously for analysis of PaCO2 at each measurement of Vm.
Test of Response to Increased Arterial Pressure
After the CO2 reactivity test had been performed, the volunteer was allowed 10 to 15 min to recover. Initial measurements of MAP (MAPi) and Vm (Vmi) were made, and then a 30-mmHg increase in MAP was induced slowly with an infusion of phenylephrine (0.5-4.0 micro gram [center dot] kg sup -1 [center dot] min sup -1). The final Vm (Vmf) was measured when MAP had stabilized at its final increased level (MAPf). All measurements during this phase were made with the subject normocapnic and breathing room air. Because Vm is significantly influenced by PaCO2, arterial blood samples for analysis of Pa sub CO2 were drawn during all measurements of Vm to document if any change in PaCO2 occurred. All the measurements so far described were made before the administration of clonidine.
Assessment of Clonidine Effect
Each volunteer acted as his or her own control. In the first phase of the study, the measurements described previously were made before the administration of clonidine (preclonidine measurements). The second phase of the study began 90 min after an oral dose of clonidine, 5 micro gram/kg. In this phase of the study, the measurements described previously were repeated (postclonidine measurements). In addition, because clonidine decreased the volunteers' blood pressure, CO2 vasoreactivity was measured twice, first at the postclonidine MAP and again when MAP had been restored to the preclonidine level (+/- 5%) by a phenylephrine infusion (MAP-restored measurements).
To assess the effect of clonidine on the response to increased arterial pressure, Vm was measured at three different blood pressures; the original postclonidine MAP, MAP-restored to the preclonidine level (obtained by an infusion of phenylephrine), and finally when MAP was increased 30 mmHg above the MAP-restored level (obtained by increasing the infusion of phenylephrine).
Analysis of Data
CO2 Vasoreactivity. Linear regression was used to calculate the slope of the relationship of PaCO2 and Vm for each volunteer. The PaCO2 versus Vm relationship derived by linear regression for each volunteer was used to calculate the percent change in Vm/mmHg change in PaCO2 over the range of that subject's PaCO2 values.
Response to Increased Arterial Pressure. Because Vm values are significantly affected by PaCO2, we first compared values for PaCO2 before making comparisons of Vm, e.g. the initial Vm values before and after clonidine.
Comparisons between two treatment groups were by paired t test and between three treatment groups were by repeated measures ANOVA with Student-Newman-Keuls test. All results are presented as mean +/- SD, unless otherwise stated. A P value of < 0.05 was considered statistically significant.
General Variables. After clonidine administration, Vm decreased from 62 +/- 9 cm/s to 48 +/- 8 cm/s (P < 0.01); PI increased from 0.84 +/- 0.06 to 1.13 +/- 0.13 (P < 0.01); MAP decreased from 86 +/- 10 mmHg to 63 +/- 5 mmHg (P < 0.01); neither heart rate, nor PaCO2
changed (Table 1
). The decrease in Vm after clonidine was not reversed by increasing MAP to the preclonidine level. Arterial oxygen saturation as measured by pulse oximetry was > or = to 97% in all subjects.
Cerebral CO2 Vasoreactivity. In all subjects, there was a statistically significant relationship between PaCO2 and Vm with correlation coefficients all > 0.95. The CO2 vasoreactivity slope decreased from a control of 2.2 +/- 0.4 to 1.2 +/- 0.5 cm [center dot] s sup -1 [center dot] mmHg sup -1 after clonidine administration (P <0.05). Restoration of MAP to the preclonidine level increased the slope to 1.6 +/- 0.5 cm [center dot] s sup -1 [center dot] mmHg sup -1 (P <0.05), still less than the preclonidine slope (P < 0.05). CO2 vasoreactivity, expressed as a percentage change in Vm, decreased after clonidine administration, 3.5 +/- 0.8 versus 2.4 +/- 0.8%/mmHg (P <0.05); this difference disappeared after restoration of MAPs to the preclonidine level, 3.1 +/- 1.2 %/mmHg.
Hypocapnia and hypercapnia were associated with minor changes in MAP, all of which deviated < 10% from the normocapnic value (Table 1
). Hypocapnia and hypercapnia induced statistically significant increases in HR (P < 0.05); these changes were similar in the three phases of the study (Table 1
). The administration of phenylephrine to restore MAP to control levels produced a decrease in HR (P < 0.05; Table 1
Response to Arterial Pressure Increase. Except for the situations wherein PaCO2
was deliberately manipulated to produce hypo- or hypercapnia, the PaCO2
values were similar during all phases of the study (Table 2
). Before clonidine administration, the Vm of 61 +/- 9 cm/s increased to 69 +/- 9 cm/s after an increase of 30 mmHg in MAP, a median increase of 13% (P < 0.05). After clonidine administration, Vm was 51 +/- 7 cm/s, and this value did not change, 54 +/- 10 cm/s, with restoration of MAP to the preclonidine level, although it increased by a median of 13% to 62 +/- 8 cm/s when MAP was increased a further 30 mmHg (P < 0.05; Figure 1
In this study, we found that clonidine decreases Vm. The decrease in Vm was not reversed by restoring MAP to the control level, indicating an effect of clonidine independent of the decrease in MAP. The decrease in Vm was accompanied by an increase in PI, and this observation is consistent with data from animal studies, suggesting that the decrease in Vm is a result of cerebral vasoconstriction. [5-9]
-adrenergic receptors on the cerebral arteries of pigs and cats are stimulated, vasoconstriction occurs. [7,9]
In dogs, alpha2
agonists decrease cerebral blood flow, but they do not alter cerebral metabolic rate, suggesting a direct vasoconstrictor effect. [5,6]
Our assumption that the decreased Vm after clonidine is a result of cerebral vasoconstriction is supported by the results of Zornow et al., who found a decreased Vm and increased PI with the alpha2
agonist dexmedetomidine. 
Clonidine-induced vasoconstriction also could explain, at least in part, our observations on CO2
vasoreactivity because CO sub 2 reactivity is influenced by cerebrovascular resistance (CVR). 
The greater the CVR, the lower would be the PaCO2
-mediated changes in CBF. Because alpha2
agonists produce cerebral [5,6]
the CVR would be greater after clonidine administration, and the CO2
vasoreactivity slope would be decreased, which was the effect we found.
In addition, part of the decrease in slope was probably a result of the clonidine-induced decrease in arterial pressure. [19,20]
With decreasing MAP, CO2
reactivity of the cerebral resistance vessels decreases and eventually is abolished. [19,20]
We corrected for the influence of decreased arterial pressure by restoring MAP to the preclonidine level, and the CO2
reactivity slope increased somewhat. However, the slope did not return to the preclonidine level because Vm did not change with the increase in MAP.
Analyzing CO2 reactivity in terms of relative percent change rather than slope showed a broadly similar pattern, i.e., a reduction in vasoreactivity after clonidine administration, which was reversed when MAP was restored to preclonidine levels. Our conclusions from the previous results are that clonidine has no direct effect to decrease CO2 vasoreactivity, but rather it mediates its effects through decreased Vm and MAP and increased CVR.
Finally, we studied the effect of clonidine on the response of Vm to increased arterial pressure. Autoregulation is the process that regulates CBF in the situation of changes in MAP, and we studied only a small component of that whole complex system, specifically the static response of Vm to a 30-mmHg increase in MAP. Several reports have shown that TCD is a valid and useful technique for evaluation of various aspects of cerebral autoregulation. [14,15,21-23]
Because previous studies found no change in Vm with increased MAP, [22,23]
we expected to find no significant change in Vm with an increase of 30 mmHg in MAP. However, before clonidine administration, we found an increase of approximately 13% in Vm with an increase of 30 mmHg in MAP. We propose two reasons for this observation. First, because we used awake volunteers, a very stable set of study conditions, and a paired study design wherein each subject was his or her own control, we had sufficient statistical power to detect even a small difference in Vm. Second, we increased MAP by 30%, whereas previous studies increased it by a lesser amount, and the effect may either not have occurred or not have been detectable with the smaller increase in MAP. [22,23]
A similar effect of MAP increase on Vm also was observed after clonidine administration. For direct comparison with the preclonidine situation, the initial postclonidine MAP was restored to the preclonidine level, and then the Vm change was measured during an additional 30 mmHg increase in MAP. A 13% increase in Vm occurred, i.e., the same magnitude of change as was observed before clonidine administration. When the MAP was increased from the initial postclonidine level to the MAP-restored level, no change in Vm was noted. However, in this phase of the study, the increase in MAP averaged only 21 mmHg, and so the results cannot be compared with those obtained before clonidine when the MAP increase was 30 mmHg. The conclusion these data suggest is that the effect of increasing MAP on Vm is essentially the same after clonidine as it was before.
Some aspects of our methodology merit discussion. We used the technique of transcranial Doppler insonation to measure mean blood flow velocity in the MCA. This technique does not allow for measurement of absolute cerebral blood flow, but in any person, there is a good correlation between relative changes in flow velocity and CBF. 
Because of this good correlation, the validity of using TCD to evaluate CO2
vasoreactivity and Vm response to increased arterial pressure is well established. [11-15,21,22,24]
The goodness of the relationship of Vm to CBF requires that the diameter of the MCA remains constant. Direct intraoperative observation of the MCA showed that its diameter changed little (2.5%) with moderately increased MAP, and this degree of change would have minimal influence on our results. 
Similarly, our use of phenylephrine to increase MAP would be expected to have negligible effect on MCA diameter. 
Because our subjects were in a semirecumbent position and because the arterial pressure transducer was zeroed at the level of the right atrium, we measured MAP and not cerebral perfusion pressure (CPP). Because the subjects remained in the same position throughout the study, the relationship of MAP and CPP would have been constant.
In summary, we conclude that a dose of clonidine, 5 micro gram/kg, such as might be used for premedication produces significant effects on the cerebral circulation. It decreases MCA blood flow velocity and slightly attenuates the cerebrovascular response to changes in CO2 via its effects on Vm and MAP and has no effect on the response to moderately increased arterial pressure. We speculate that these effects are consistent with a direct cerebral vasoconstrictive effect of clonidine. Therefore, decreased CBF and mildly attenuated CO sub 2 vasoreactivity should be anticipated when oral clonidine is administered as a premedication or as an adjunct to anesthesia or pain management.
1. Maze M, Tranquilli W: Alpha-2 adrenoceptor agonists: defining the role in clinical anesthesia. Anesthesiology 1991; 74:581-605.
2. Kanawati IS, Yaksh TL, Anderson RE, Marsh RW: Effects of clonidine on cerebral blood flow and the response to arterial CO2. J Cereb Blood Flow Metab 1986; 6:358-65.
3. James IM, Larbi E, Zaimis E: The effect of the acute intravenous administration clonidine (St 155) on cerebral blood flow in man. Br J Pharmacol 1970; 39:198P-9.
4. Bertel O, Conen D, Radu EW, Muller J, Lang C, Dubach UC: Nifedipine in hypertensive emergencies. BMJ (Clin Res Ed) 1983; 286:19-21.
5. Zornow MH, Fleischer JE, Scheller MS, Nakakimura K, Drummond JC: Dexmedetomidine, an alpha 2-adrenergic agonist, decreases cerebral blood flow in the isoflurane-anesthetized dog. Anesth Analg 1990; 70:624-30.
6. Karlsson BR, Forsman M, Roald OK, Heier MS, Steen PA: Effect of dexmedetomidine, a selective and potent alpha 2-agonist, on cerebral blood flow and oxygen consumption during halothane anesthesia in dogs. Anesth Analg 1990; 71:125-9.
7. Busija DW, Leffler CW: Postjunctional alpha 2-adrenoceptors in pial arteries of anesthetized newborn pigs. Dev Pharmacol Ther 1987; 10:36-46.
8. Coughlan MG, Lee JG, Bosnjak ZJ, Schmeling WT, Kampine JP, Warltier DC: Direct coronary and cerebral vascular responses to dexmedetomidine. Significance of endogenous nitric oxide synthesis. Anesthesiology 1992; 77:998-1006.
9. Skarby T: Pharmacological properties of prejunctional alpha-adrenoceptors in isolated feline middle cerebral arteries; comparison with the postjunctional alpha-adrenoceptors. Acta Physiol Scand 1984; 122:165-74.
10. Izumi Y, Fukuuchi Y, Imai A, Isozumi K: Effect of clonidine on the cerebrovascular system in cats. Keio J Med 1993;42:64-9.
11. Aaslid R, Markwalder TM, Nornes H: Noninvasive transcranial Doppler ultrasound recording of flow velocity in basal cerebral arteries. J Neurosurg 1982; 57:769-74.
12. Bishop CC, Powell S, Rutt D, Browse NL: Transcranial Doppler measurement of middle cerebral artery blood flow velocity: A validation study. Stroke 1986; 17:913-5.
13. Markwalder TM, Grolimund P, Seiler RW, Roth F, Aaslid R: Dependency of blood flow velocity in the middle cerebral artery on end-tidal carbon dioxide partial pressure-A transcranial ultrasound Doppler study. J Cereb Blood Flow Metab 1984; 4:368-72.
14. Newell DW, Aaslid R, Lam A, Mayberg TS, Winn HR: Comparison of flow and velocity during dynamic autoregulation testing in humans. Stroke 1994; 25:793-7.
15. Larsen FS, Olsen KS, Hansen BA, Paulson OB, Knudsen GM: Transcranial Doppler is valid for determination of the lower limit of cerebral blood flow autoregulation. Stroke 1994; 25:1985-8.
16. Aaslid R: Transcranial Doppler Sonography. New York, Springer-Verlag, 1986, pp. 39-59.
17. Zornow MH, Maze M, Dyck JB, Shafer SL: Dexmedetomidine decreases cerebral blood flow velocity in humans. J Cereb Blood Flow Metab 1993; 13:350-3.
18. Ackerman RH, Zilkha E, Bull JW, Du Boulay GH, Marshall J, Russell RW, Symon L: The relationship of the CO sub 2 reactivity of cerebral vessels to blood pressure and mean resting blood flow. Neurology 1973; 23:21-6.
19. Schmidt JF, Waldemar G, Paulson OB: Arterial CO2 tension and cerebral vascular reactivity during the induction of acute hypertension and hypotension in the awake human. J Neurosurg Anesthesiol 1990; 2:92-6.
20. Harper AM, Glass HI: Effect of alterations in the arterial carbon dioxide tension on the blood flow through the cerebral cortex at normal and low arterial blood pressures. J Neurol Neurosurg Psychiatry 1965; 28:449-52.
21. Strebel S, Lam AM, Matta B, Mayberg TS, Aaslid R, Newell DW: Dynamic and static cerebral autoregulation during isoflurane, desflurane, and propofol anesthesia. Anesthesiology 95; 83:66-76.
22. Tiecks FP, Lam AM, Aaslid R, Newell DW: Comparison of static and dynamic cerebral autoregulation measurements. Stroke 1995; 26:1014-9.
23. Matta BF, Lam AM, Strebel S, Mayberg TS: Cerebral pressure autoregulation and carbon dioxide reactivity during propofol-induced EEG suppression. Br J Anaesth 1995; 74:159-63.
24. Fale A, Kirsch JR, McPherson RW: Alpha 2-adrenergic agonist effects on normocapnic and hypercapnic cerebral blood flow in the dog are anesthetic dependent. Anesth Analg 1994; 79:892-8.
25. Giller CA, Bowman G, Dyer H, Mootz L, Krippner W: Cerebral arterial diameters during changes in blood pressure and carbon dioxide during craniotomy. Neurosurgery 1993; 32:737-41.
© 1997 American Society of Anesthesiologists, Inc.