Mild to moderate hypothermia has neuroprotective and cardioprotective properties in animals (1–3). Mild hypothermia improves neurologic outcome in human survivors of out-of-hospital cardiac arrest (4,5). Because of its potential neuroprotective effects, therapeutic use of hypothermia is increasing. Unfortunately, even mild hypothermia has unwanted side effects such as shivering and vasoconstriction, which are associated with hemodynamic instability, increased oxygen consumption, increased levels of circulating catecholamines, and discomfort (6). Thus, it would be desirable to have therapies to prevent postoperative shivering and its related comorbidities.
Clonidine, an α-2 agonist, reduces the incidence of shivering and has sympatholytic, sedative and opioid-sparing properties (7,8). Whereas the sympatholytic and antishivering effects of clonidine may improve hemodynamic stability and reduce postoperative shivering, its sedative effects may hinder immediate postoperative neurological evaluation of neurosurgical patients. Neither the dose of clonidine to prevent postoperative shivering after mild intraoperative hypothermia nor the effect of clonidine on recovery from anesthesia in neurosurgical patients has been studied previously. Thus, the goals of this two-part study were 1) to determine the dose of clonidine to prevent postoperative shivering after mild hypothermia and 2) to evaluate the effect of clonidine on recovery from anesthesia in normothermic patients undergoing surgery for intracranial lesions.
With approval from the IRB of the University of California, San Francisco and written informed consent, we enrolled 48 patients undergoing elective supratentorial neurosurgical procedures into one of two studies. We included patients who were over 18 yr of age and ASA physical status II–IV. We excluded patients who had clinical evidence of increased intracranial pressure, those scheduled for an emergency surgery, and those taking clonidine. In study 1 we determined the ED50 of clonidine to prevent postoperative shivering after mild hypothermia (35°C). Study 2 was a prospective, randomized, double-blind, placebo controlled study to evaluate the effect of 3 μg/kg clonidine on recovery from anesthesia.
At least 1 day before surgery, all patients had a routine preoperative history assessment and physical examination, laboratory studies, and a head computed tomographic and/or magnetic resonance imaging scan. On the morning of surgery, patients were premedicated with midazolam 1–2 mg IV. After entry into the operating room, patients breathed 100% oxygen while anesthesia was induced IV with 3 μg/kg fentanyl and propofol or thiopental. A muscle relaxant of the anesthesiologist’s choice was administered to achieve neuromuscular blockade before tracheal intubation, along with an additional 2 μg/kg of fentanyl. After intubation, an esophageal temperature probe and a Foley urinary catheter were placed. Anesthesia was then maintained with 70% nitrous oxide in oxygen, a 2 μg · kg−1 · h−1 fentanyl infusion, and isoflurane (study 1) or either propofol or an inhaled anesthetic of the anesthesiologist’s choice (study 2). Management of ventilation and end-tidal CO2 was left to the discretion of the anesthesiologist.
Preinduction monitoring included a noninvasive blood pressure cuff, electrocardiograph, and a pulse oximeter. Arterial blood pressure also was monitored by a radial artery cannula. Baseline heart rate (HR) and arterial blood pressure (noninvasive or invasive) were measured after premedication, before induction of anesthesia while patients lay supine on the operating room table. Intraoperative and postoperative hemodynamic management was left to the discretion of the physician caring for the patient.
Arterial blood pressure (systolic [SBP], diastolic, and mean) and HR were measured continuously (Propaq 106, Protocol Systems, Beaverton, OR) from before administration of study drug until the end of study. Arterial blood pressure was measured via a radial artery cannula connected to a Transpac II transducer (Abbott Laboratories, North Chicago, IL). Hemoglobin oxygen saturation (Spo2) was measured by a pulse oximeter incorporated into the Propaq 106. End-tidal concentrations of anesthetic vapor, nitrous oxide, oxygen, and carbon dioxide were measured continuously using an infrared anesthetic gas monitor (Datex Ultima; Datex Instrumentarium, Helsinki, Finland). Hemodynamic and Spo2 data were recorded at 10-s intervals from the monitors through an automated data acquisition system.
Throughout the study, we recorded all medications and perioperative event times. Esophageal temperature was recorded until the end of the operation.
Study 1: Clonidine and Postoperative Shivering
Intraoperatively, at the beginning of dural closure (approximately 60 min before the end of surgery), the fentanyl infusion was discontinued and the study drug infusion initiated. The study drug infusion contained 0–5 μg/kg of clonidine added to a total volume of 50 mL normal saline and was administered via zero order infusion over 15 min. The dose for each patient was determined using Dixon’s up-and-down method. Clonidine dose for the first study patient was 3 μg/kg. The dose was then adjusted in 1-μg/kg increments for the following patients. If a patient shivered, the next patient received a larger dose; if he/she did not shiver, the dose for the following patient was decreased. The study continued until five independent crossovers (shivering—no shivering) were observed.
Isoflurane administration was discontinued an estimated 10 min before end of surgery. Residual neuromuscular blockade was reversed using neostigmine and glycopyrrolate when appropriate. Nitrous oxide was discontinued at the end of surgery and this action was defined as the end of anesthesia time. No prophylactic antiemetics or antihypertensive medications were administered.
The target core temperature was 35.0°C at the end of the operation. Active cooling and warming devices (forced air blankets) were used to regulate the patients’ temperature intraoperatively. Core temperature was measured using esophageal or tympanic temperature probes.
Incidence of shivering was recorded for the first hour after end of anesthesia by an observer blinded to the treatment (dose). We defined shivering as visible shivering for at least 1 min.
The ED50 of clonidine to prevent shivering was calculated using the modified Dixon’s up-and-down method, which uses the average clonidine concentrations of all independent pairs of patients that involved a crossover (i.e., shivering to nonshivering or nonshivering to shivering) to calculate the mean and standard deviation (9).
Study 2: Clonidine and Recovery from Anesthesia
Thirty-four patients were randomly assigned to a study group (clonidine or placebo) by sealed, numbered envelopes that were opened after informed consent was obtained. Our sample size analysis indicated that 34 subjects (17 in each group) would provide a 80% change (power = 0.8) to identify a 5-min difference in time to follow commands between the 2 groups at an α level of 0.05.
Intraoperatively, at the beginning of dural closure (approximately 60 min before the end of surgery), fentanyl infusion and inhaled anesthetics were discontinued. Simultaneously a propofol infusion (50–200 μg · kg−1 · min−1) was started (unless continuous). Patients then randomly received either a 15-min infusion of clonidine (3 μg/kg in 50 mL of normal saline) or a 15-min infusion of normal saline (50 mL). Propofol infusion was discontinued an estimated 10 min before the end of surgery. Residual neuromuscular blockade was reversed using neostigmine and glycopyrrolate when appropriate. Nitrous oxide was discontinued at the end of surgery and this action was defined the end of anesthesia time. No prophylactic antiemetics or antihypertensive medications were administered. Intraoperative temperature control was left to the discretion of the anesthesiologist.
Patient recovery was studied for 2 h after the end of anesthesia, as defined by the discontinuation of nitrous oxide. Once nitrous oxide was discontinued, we recorded the time required for patients to: 1) open their eyes, 2) follow simple commands (“squeeze my hand”), 3) be tracheally extubated, and, 4) be oriented to time and place. Rate of recovery was evaluated using a modified Aldrete Score every 15 min from the end of anesthesia until the patient obtained at least 10 of the 12 potential points. Visual analog scale (VAS) scoring was used to determine sedation (0 = completely awake to 100 = cannot stay awake), pain (0 = no pain to 100 = most pain I have ever felt), and nausea (0 = no nausea to 100 = as nauseous as I have ever been) before anesthesia and at 30, 60, 90, and 120 min after end of anesthesia. Sedation also was monitored by one of the investigators using the Ramsay scoring system; the same investigator monitored nausea and vomiting by observation at the same postoperative intervals.
Duration of anesthesia was defined as the time from induction of anesthesia to discontinuation of nitrous oxide. Duration of surgery was defined as the time from skin incision to last suture. Before analysis, continuously monitored hemodynamic and anesthetic gas data were reduced to 1-min median values. The area under the curve was used to quantify the incidence and severity of hemodynamic aberrations and was calculated as the sum of the amount HR or SBP exceeded the predetermined limits for each 1-min epoch (tachycardia: HR >100 bpm, bradycardia: HR <50 bpm, hypertension: SBP >160 mm Hg, hypotension SBP <90 mm Hg). Continuous data were analyzed by repeated-measures analysis of variance. Nonparametric data were analyzed by Kruskal-Wallis statistic. Categorical data were analyzed by χ2 test. Data are expressed as mean ± sd. P < 0.05 indicated statistical significance.
Patients (n = 14) were 52 ± 14 yr old, 170 ± 10 cm tall, and 79 ± 12 kg in weight. Patients’ core temperature was 35.1 ± 0.2°C at the end of anesthesia. Five of the 14 patients shivered postoperatively. Onset of shivering ranged from 7 to 32 min after the end of anesthesia. Figure 1 illustrates the series of clonidine doses and crossover pairs. The ED50 of clonidine to prevent shivering was 1.1 ± 1.5 μg/kg.
There were no significant differences in age, weight, gender, ASA class, duration of surgery, duration of anesthesia or time from beginning of dural closure to end of anesthesia between the two groups. The ages in the clonidine and saline groups were 51 ± 16 and 47 ± 14 yrs, respectively; weight was 85 ± 21 and 76 ± 13 kg, duration of surgery was 367 ± 136 and 372 ± 126 min, duration from beginning of dural closure to end of anesthesia was 81 ± 23 and 84 ± 23 min, and duration from beginning of study drug administration to end of anesthesia was 66 ± 22 and 71 ± 22 min. Of the patients, 14 were male and 20 female. Of the 34 operations, 30 were for resection of an intracranial tumor (n = 15 in each group), one for resection of an intracranial arteriovenous malformation (clonidine group), two for trigeminal neuralgia (one in each group), and one for resection of an intracranial cyst (saline group).
Arterial blood pressure and HR values at baseline and at beginning of study drug infusion did not differ between the groups (Table 1). At the end of anesthesia SBP was lower in the clonidine group compared with the saline group (P < 0.05). Clonidine infusion was associated with an initial increase in arterial blood pressure, followed by a gradual decline. On the average SBP increased by 7 mm Hg and peaked approximately 3 min after the beginning of the clonidine infusion.
Before study drug infusion, of the 34 patients, 21 received isoflurane (n = 9, 12 respectively for the clonidine and saline groups), seven received desflurane (n = 4, 3) and 6 received propofol (n = 4, 2). Reflecting the switch from inhaled anesthetics to propofol at dural closure, at the end of anesthesia the end-tidal isoflurane and desflurane concentrations were small, at 0 vol% and 0.26 ± 0.21 vol% for the clonidine and 0.06 ± 0.07 vol% and 0.22 ± 0.15 vol% for the saline groups, respectively. Total amount of propofol infused from beginning of study drug infusion to end of anesthesia was 314 ± 172 mg and 409 ± 111 mg for the clonidine and saline groups, respectively (P = not significant).
Vasoactive medications were required intraoperatively after the study drug infusion by 12 and 8 patients in the clonidine and saline groups, respectively. Labetalol was administered to 2 and 4 patients, phenylephrine to 8 and 3 patients, ephedrine to 4 and 2 patients, and esmolol to 0 and 1 patients in the clonidine and saline groups, respectively. Esophageal temperatures were 36.1 ± 0.8°C and 36.0 ± 0.8°C at the beginning of study drug infusion, 36.2 ± 0.8°C and 35.9 ± 0.8°C at the end of study drug infusion, 36.1 ± 0.8°C and 35.9 ± 0.9°C at the end of surgery, and 36.4 ± 0.6°C and 36.3 ± 0.8°C 2 hours after end of anesthesia in the clonidine and saline groups, respectively, (P = not significant for all).
HR and SBP values at the end of the 1-h recovery period (Table 1) and the amount of hypertension, hypotension, tachycardia, and bradycardia (as measured by area under the curve) from beginning of study drug infusion to end of study did not differ between the groups (Table 1). To control postoperative arterial blood pressure, one patient in the clonidine group and three patients in the saline group received labetalol.
The median time to reach an Aldrete Score of 10 was 15 min (first measurement time point) in both groups. Two patients in the clonidine group did not reach an Aldrete Score of 10 within 30 min. Postoperative recovery event time data are shown in Table 2 and illustrated in Figure 2. None of the recovery event times (open eyes, extubation, follow commands, oriented, Aldrete score) differed significantly.
Postoperative nausea was observed in 8 and 5 patients in the clonidine and saline groups, respectively (P = not significant); 6 and 4 patients in the clonidine and saline groups, respectively, received medication for nausea (P = not significant).
Neither the Ramsay sedation scores nor the patient-recorded VAS scores for sedation or pain differed significantly between the groups (Table 3). Medication for postoperative pain was required in 9 and 13 patients in the clonidine and saline groups, respectively (P = not significant). Postoperative shivering was more frequent in the saline group (n = 7) compared with the clonidine group (n = 1) (P = 0.03).
We tested the ability of clonidine, an α-2 agonist, to prevent postoperative shivering in neurosurgical patients with mild hypothermia (35°C) and clonidine’s effects on emergence from anesthesia. The ED50 of clonidine to prevent postoperative shivering was 1.1 ± 1.5 μg/kg in neurosurgical patients whose target core temperature was 35°C at the end of surgery. The second part of this study demonstrated that 3 μg/kg of clonidine, a dose more than one standard deviation above the ED50 for shivering, administered to neurosurgical patients toward the end of surgery, had negligent sedative and hemodynamic effects and did not delay emergence from anesthesia.
Several previous studies have demonstrated that clonidine decreases the incidence, intensity, and duration of postoperative shivering (7). These effects, however, have not been studied in neurosurgical patients. Thus, our study is the first to demonstrate the antishivering effects of clonidine in neurosurgical patients. We felt that it was important to investigate clonidine-induced antishivering and the sedative effect in these patients for two reasons. First, therapeutic use of hypothermia has been increasing in neurosurgical patients for its potential neuroprotective effects, making this a patient population likely to be exposed to perioperative hypothermia (10). Second, the clonidine-induced sedative effect and its potential to delay emergence from anesthesia may limit the use of clonidine in neurosurgical patients.
Several studies have shown that 3–5 μg/kg of clonidine prevents postoperative shivering (7). However, none of the previous studies controlled for the patient’s temperature at the end of surgery. Thus, these data are not sufficient to predict the dose of clonidine needed to prevent shivering during mild hypothermia. Accordingly, our study differs significantly from most previous studies by controlling for the core temperature at the end of anesthesia.
Compared with data from thermoregulatory studies in healthy unmedicated volunteers, our data suggests that the ED50 of clonidine to prevent shivering in patients with controlled, mild hypothermia is surprisingly small, 1.1 ± 1.5 μg/kg (11). One explanation for this small ED50 is the potential confounding shivering threshold decreasing effect of residual anesthetic concentrations during the early postoperative period (12,13). Therefore, it is possible that the ED50 of clonidine may differ in patients receiving medications other than those in our study.
Clonidine is used to prevent and to treat postoperative shivering (7). However, the optimal dosing of clonidine to prevent shivering has not yet been determined. Data from volunteer studies suggest that α-2 agonist-induced reduction in shivering threshold is dose-dependent (14). We chose to administer clonidine toward the end of the operation so that the clonidine plasma levels would be high in the early postoperative period, when the patient is most likely to shiver.
Several anesthetic adjuncts have been used during anesthesia for neurosurgical patients. The optimal technique should provide intraoperative and postoperative hemodynamic stability as well as rapid and safe awakening that will allow neurologic evaluation soon after surgery. Clonidine seems like an attractive choice for an anesthetic adjunct for neurosurgical patients because of its antishivering, hemodynamic-stabilizing, and anesthetic-sparing effects. However, clonidine’s side effects, dose-dependent sedation, hypotension, and bradycardia may limit its use in neurosurgical patients (8). Our data suggest that in this patient population 3 μg/kg clonidine does not have clinically significant sedative or hemodynamic effects and may be useful as an anesthetic adjunct in neurosurgical patients.
Alpha-2 agonists have peripherally-mediated vasoconstrictive, and centrally and peripherally-mediated sympatholytic effects (15). We chose clonidine (instead of meperidine) for this study in part for its potential hemodynamic stabilizing effects (reduction in tachycardia and hypertension). However, the only hemodynamic effect we observed was a statistically, but not clinically, significant clonidine-induced decrease in arterial blood pressure at the end of surgery. During the 15-min intraoperative zero order 3 μg/kg clonidine infusion we observed an initial small increase in arterial blood pressure followed by a gradual decline. The initial increase in arterial blood pressure was most likely mediated by vascular smooth muscle α-2 adrenoceptors (16).
Both of our experimental designs have advantages and disadvantages for the study of systemically administered clonidine. The advantage of using Dixon’s up-and-down method to determine the ED50 of clonidine to prevent shivering is the relatively small number of patients needed to conduct the study. However, the disadvantages include the inability to determine a clonidine dose that would prevent shivering in all patients. Because anesthetics and opioids produce central thermoregulatory inhibition reducing the shivering threshold, to minimize anesthetic effects as confounding factors, we standardized the anesthesia toward the end of surgery in both studies. The first study was limited by using only mild core hypothermia (35°C). However, we felt that before using moderate hypothermia (34°C) we first needed to demonstrate our ability to prevent shivering in neurosurgical patients with mild hypothermia. The study was also limited by not measuring clonidine plasma concentrations, and by using only one clonidine dose (3 μg/kg). However, the 3-μg/kg clonidine dose is more than one standard deviation above the calculated ED50. Our sample size calculation for the second study was based on our primary end-point (to follow commands). Thus, this study may have been underpowered for several other variables, such as hemodynamics, that did not reach statistical significance.
In conclusion, the ED50 of clonidine to prevent postoperative shivering was 1.1 ± 1.5 μg/kg in this study of neurosurgical patients whose target core temperature was 35°C at the end of surgery. In addition, 3 μg/kg of clonidine administered to neurosurgical patients approximately 1 hour before the end of anesthesia neither delayed emergence from anesthesia nor had clinically significant sedative or hemodynamic effects.
1. Sano T, Drummond JC, Patel PM, et al. A comparison of the cerebral protective effects of isoflurane and mild hypothermia in a model of incomplete forebrain ischemia in the rat. Anesthesiology 1992;76:221–8.
2. Ridenour TR, Warner DS, Todd MM, McAllister AC. Mild hypothermia reduces infarct size resulting from temporary but not permanent focal ischemia in rats. Stroke 1992;23:733–8.
3. Dae MW, Gao DW, Sessler DI, et al. Effect of endovascular cooling on myocardial temperature, infarct size, and cardiac output in human-sized pigs. Am J Physiol Heart Circ Physiol 2002;282:H1584–91.
4. Bernard SA, Gray TW, Buist MD, et al. Treatment of comatose survivors of out-of-hospital cardiac arrest with induced hypothermia. N Engl J Med 2002;346:557–63.
5. Hypothermia after Cardiac Arrest Study Group. Mild therapeutic hypothermia to improve the neurologic outcome after cardiac arrest. N Engl J Med 2002;346:549–56.
6. Sessler DI. Complications and treatment of mild hypothermia. Anesthesiology 2001;95:531–43.
7. Kranke P, Eberhart LH, Roewer N, Tramer MR. Pharmacological treatment of postoperative shivering: a quantitative systematic review of randomized controlled trials. Anesth Analg 2002;94:453–60.
8. Maze M, Segal IS, Bloor BC. Clonidine and other alpha2 adrenergic agonists: strategies for the rational use of these novel anesthetic agents. J Clin Anesth 1988;1:146–57.
9. Dixon WJ. Staircase bioassay: the up-and-down method. Neurosci Biobehav Rev 1991;15:47–50.
10. Hindman BJ, Todd MM, Gelb AW, et al. Mild hypothermia as a protective therapy during intracranial aneurysm surgery: a randomized prospective pilot trial. Neurosurgery 1999;44:23–32.
11. Nicolaou G, Chen AA, Johnston CE, et al. Clonidine decreases vasoconstriction and shivering thresholds, without affecting the sweating threshold. Can J Anaesth 1997;44:636–42.
12. Matsukawa T, Kurz A, Sessler DI, et al. Propofol linearly reduces the vasoconstriction and shivering thresholds. Anesthesiology 1995;82:1169–80.
13. Xiong J, Kurz A, Sessler DI, et al. Isoflurane produces marked and nonlinear decreases in the vasoconstriction and shivering thresholds. Anesthesiology 1996;85:240–5.
14. Talke P, Tayefeh F, Sessler DI, et al. Dexmedetomidine does not alter the sweating threshold, but comparably and linearly decreases the vasoconstriction and shivering thresholds. Anesthesiology 1997;87:835–41.
15. Maze M, Tranquilli W. Alpha-2 adrenoceptor agonists: defining the role in clinical anesthesia. Anesthesiology 1991;74:581–605.
© 2005 International Anesthesia Research Society
16. Talke PO, Caldwell JE, Richardson CA, Heier T. The effects of clonidine on human digital vasculature. Anesth Analg 2000;91:793–7.