Effects of dexmedetomidine on adrenocortical function, and the cardiovascular, endocrine and inflammatory responses in post-operative patients needing sedation in the intensive care unit

R. M. Venn1, A. Bryant2, G. M. Hall2 and R. M. Grounds3

1Department of Anaesthesia and Intensive Care, Worthing Hospital, Lyndhurst Road, Worthing, West Sussex BN11 2DH, UK. 2Department of Anaesthesia and Intensive Care Medicine, St George’s Hospital Medical School, Cranmer Terrace, London SW17 ORE, UK. 3Department of Anaesthesia and Intensive Care Medicine, St George’s Hospital, Blackshaw Road, London SW17 OQT, UK*Corresponding author

Accepted for publication: January 16, 2001


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
We have compared the effects of dexmedetomidine and propofol on endocrine, metabolic, inflammatory and cardiovascular responses in patients in the intensive care unit (ICU) after major surgery. Twenty patients who were expected to require 8 h of post-operative sedation and ventilation were allocated randomly to receive either an infusion of dexmedetomidine 0.2–2.5 µg kg–1 h–1 or propofol 1–3 mg kg–1 h–1. Arterial pressure, heart rate and sequential concentrations of circulating cortisol, adrenocorticotrophic hormone (ACTH), growth hormone, prolactin, insulin, glucose and interleukin 6 were measured. An ACTH stimulation test was performed in all patients who received dexmedetomidine. Heart rate was significantly lower in the dexmedetomidine patients. There were no differences in arterial pressure, cortisol, ACTH, prolactin and glucose concentrations between the two groups. A positive response to the ACTH stimulation test varied depending on the diagnostic criteria used. The insulin concentration was significantly lower in the dexmedetomidine group at 2 h (P=0.021), although this did not affect blood glucose concentrations. Growth hormone concentrations were significantly higher in dexmedetomidine-treated patients overall (P=0.036), but circulating concentrations remained in the physiological range. Interleukin 6 decreased in the dexmedetomidine group. We conclude that dexmedetomidine infusion does not inhibit adrenal steroidogenesis when used for short-term sedation after surgery.

Br J Anaesth 2001: 86: 650–6

Keywords: hormones, glucocorticoid, cortisol; hormones, growth; hormones, prolactin; hormones, adrenal; hormones, adrenocorticotrophic; intensive care; sedation, post-operative


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The majority of patients admitted for ventilation to the intensive care unit (ICU) after surgery require sedation and analgesia to reduce anxiety, tolerate mechanical ventilation and provide pain relief. Inappropriate sedation techniques can, however, cause morbidity and even mortality in the ICU, as occurred with the use of long-term infusions of the imidazole agent etomidate.1 Etomidate suppresses adrenocortical function by dose-dependent, reversible inhibition of two mitochondrial cytochrome P450 enzymes (P450c11 and P450scc), and this effect may persist for 24 h in critically ill patients after a single dose of etomidate.2

The highly selective and potent {alpha}2 agonist dexmedetomidine is an effective agent for post-operative sedation and analgesia in the ICU.3 It is an imidazole compound and therefore has the potential to exert similar inhibitory effects to etomidate on cortisol synthesis. In vitro and in vivo animal studies have shown that dexmedetomidine inhibits cortisol synthesis at concentrations above 10–6 M, whereas therapeutic plasma concentrations of dexmedetomidine in patients are less than 10–9 M.4 Consequently, inhibition of steroidogenesis is unlikely to occur clinically, and this supposition is supported by the absence of hypocortisolism in human volunteers and surgical patients after i.v. or i.m. use.5 6 However, inhibition may become important clinically when dexmedetomidine is given as an infusion in the ICU patient, especially if changes in circulating inflammatory mediators [tumour necrosis factor {alpha}, immunoreactive corticostatin, interleukin 6 (IL-6)] have already impaired the adrenocortical response to adrenocorticotrophic hormone (ACTH).7

Attenuation of the cardiovascular and neuroendocrine responses to surgery may improve outcome by beneficial effects on organ function,8 although modulating the inflammatory response is probably more important in improving recovery.9 The modification of cardiovascular responses after surgery by dexmedetomidine has been described recently.3 The neuroendocrine effects of the {alpha}2-adrenoceptor agonist clonidine are well known10 11 but have not been observed during surgery.12 13 However, dexmedetomidine modified the pituitary hormonal response to gynaecological laparoscopy,6 possibly because of its higher selectivity for the {alpha}2-adrenoceptor. There has been little work on the effect of {alpha}2-adrenoceptor agonists on the inflammatory response. Although IL-6 production is suppressed in vitro by p-aminoclonidine,14 oral premedication with clonidine had no effect on the IL-6 response after pancreatobiliary surgery.15

We compared the effects of dexmedetomidine and propofol on adrenocortical function and the cardiovascular, endocrine and inflammatory responses in post-operative patients requiring sedation and ventilation in the ICU.


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The local research ethics committee of St. George's Healthcare approved the study (reference number 98.06.8) and written informed consent was obtained from all patients. We studied 20 adult patients (age 18 yr or older) admitted to the ICU after complex major abdominal or pelvic surgery and who were expected to require 8 h post-operative sedation and ventilation. Exclusion criteria were allergy to any of the trial drugs, pregnancy, severe hepatic disease, requirement for haemodialysis or haemofiltration, spinal or epidural anaesthesia, use of etomidate in the preceding 24 h or a history of corticosteroid treatment within the last 3 months.

The anaesthetic technique was decided by the individual anaesthetist and the dose of intra-operative analgesic was recorded. On arrival in the ICU, patients were allocated randomly, using sealed envelopes, to receive either dexmedetomidine or propofol infusions together with an alfentanil infusion for analgesia if required. Patients received a loading dose of dexmedetomidine 2.5 µg kg–1 h–1 over 10 min followed by a maintenance infusion of 0.2–2.5 µg kg–1 h–1 into a peripheral or central vein. Propofol was given undiluted as an infusion of 1–3 mg kg–1 h–1, and a bolus dose of 1 mg kg–1 was given initially if required. Alfentanil was infused at 0.25–1.0 µg kg–1 min–1 if the patient complained of pain and analgesia was required. The level of sedation was measured and recorded hourly using the Ramsay sedation score16 and bispectral index17, and patients were maintained at a Ramsay sedation score >2 by adjustments to the sedative regimen. Bolus doses of atracurium were given to provide muscle relaxation and paracetamol was used as an antipyretic if necessary. No other sedative or analgesic agents were given.

Patients were ventilated mechanically with oxygen-enriched air to attain acceptable blood gases and extubation was undertaken when indicated clinically. The dexmedetomidine infusion could be continued up to a maximum of 24 h and patients were changed to propofol infusions if they still required sedation and ventilation after this time.

Heart rate, arterial pressure, central venous pressure and oxygen saturation measurements were monitored continuously. Blood samples were obtained from all patients immediately on arrival in the ICU and before administration of sedative or analgesic infusions, and 2, 4, 6, 8, 12 and 24 h after commencement of sedative infusions, provided the patient had not been extubated and the sedative infusion had not been discontinued. In addition, after cessation of the infusion, a short ACTH stimulation test was performed in those patients who had received dexmedetomidine; blood samples were collected 30 and 60 min after administration of synthetic ACTH 250 µg i.v. (Synacthen; CIBA-Geigy, Macclesfield, UK). Samples were centrifuged immediately and serum was stored at –70°C, before analysis in duplicate. Serum concentrations of cortisol, ACTH, insulin, growth hormone (GH), prolactin and IL-6 were measured with commercially available enzyme-linked immunoassay (ELISA) kits. Serial glucose measurements were performed immediately on the ICU using a blood gas analyser (ABL 625; Radiometer, Copenhagen, Denmark), which was calibrated and referenced daily. IL-6 measurements were only performed at baseline and 2, 6, 8 and 12 h after admission to the ICU. The lower limits of sensitivity of the assays were 8.3 nmol litre–1 for cortisol (Milenia Cortisol, Diagnostic Products, CA, USA), 0.27 pmol litre–1 for ACTH (Active ACTH, Diagnostic Systems Laboratories, London, UK), 0.11 mIU litre–1 for growth hormone (Medgenix–HGH–EASIA, Biosource, Nivelles, Belgium), 7.6 mIU litre–1 for prolactin (Fertigenix–PRL–EASIA, Biosource, Nivelles, Belgium), 0.15 mIU litre–1 for insulin (Medgenix–INS–EASIA, Biosource, Nivelles, Belgium) and 0.70 pg ml–1 for IL-6 (Quantikine IL-6; R&D systems, Abingdon, UK). Intra- and inter-assay coefficients of variation were 6.9 and 8.0% respectively for cortisol, 8.6 and 6.6% for ACTH, 3.6 and 7.1% for GH, 6.1 and 8.3% for prolactin, 5.3 and 9.8% for insulin, and 4.2 and 6.4% for IL-6. The morning reference range for cortisol was 138– 690 nmol litre–1.

Sample size was determined from previous work, which indicated that with 20 patients there was a power of 90% to detect a 20% decrease in heart rate at a significance level of 5%. Cardiovascular data are shown as mean (SEM) values and comparisons were made by analysis of variance for repeated measures, followed by post hoc Bonferroni testing. Other data are shown as median values (interquartile range), and non-parametric analysis was undertaken. Analysis of serial measurements for unequal time points was performed with a two-stage method that used summary measures for cortisol, ACTH, prolactin, insulin and glucose concentrations.18 The area under the curve was calculated for each patient and between-group comparisons were made using the Mann–Whitney U-test. Insulin concentrations at several time points were less than the sensitivity of the assay and so data were ranked and the groups compared using the Kruskal–Wallis test. The ratios of IL-6 concentrations at 2, 6, 8 and 12 h relative to baseline were calculated and then compared using the Mann–Whitney U-test. P<0.05 was accepted as significant. All analysis was carried out using the Statview for Windows software package (version 4.57; Abacus Concepts, Berkeley, CA, USA).


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The groups were similar in age, operation time, dose of intra-operative fentanyl, APACHE II score, weight, mortality and duration of sedative infusion in the ICU (Table 1). One patient (patient P2) underwent an emergency operation; the remaining operations were elective procedures. In the dexmedetomidine and propofol groups, the intra-operative fentanyl requirement was 725 (575–1000) and 800 (700–1000) µg respectively and the duration of sedative infusions 10 (8–15) h and 12 (9–15) h. Two patients in the dexmedetomidine group and three patients in the propofol group received sedation for only 6 h because extubation was indicated clinically. Four patients in the dexmedetomidine group and five in the propofol group received sedation for >=12 h. Two patients (D1, D2) died in the dexmedetomidine group and one (P7) died in the propofol group, on days 14, 17 and 35 respectively after initially recovering well.


View this table:
[in this window]
[in a new window]
 
Table 1 Patient and operative characteristics. Summary statistics are reported as median (interquartile range). Patients: D=dexmedetomidine; P=propofol
 
Patients receiving propofol infusions required significantly more alfentanil [2.5 (2.2–2.9) mg h–1] than patients receiving dexmedetomidine [0.8 (0.65–1.2) mg h–1] (P=0.004). The median (range) dexmedetomidine infusion rate was 0.86 (0.45–1.06) µg kg–1 h–1. Sedation was similar in the two groups, as measured by the Ramsay sedation score and the bispectral index (P=0.68 and P=0.32 respectively).

Haemodynamics
Patients receiving dexmedetomidine had significantly lower heart rates compared with the propofol group (P=0.034), but there were no differences in arterial pressure between the two groups (systolic, P=0.6, diastolic P=0.36) (Fig. 1).



View larger version (25K):
[in this window]
[in a new window]
 
Fig 1 Mean (SEM) heart rate and arterial pressure. n=10 in both groups except at 7 and 8 h, when n=8 and n=7 in the dexmedetomidine and propofol groups respectively.

 
Adrenocortical function
No patients receiving dexmedetomidine had serum cortisol concentrations <138 nmol litre–1 during the study, but two patients in the propofol group had concentrations <138 nmol litre–1 at more than one sampling time. Cortisol concentrations <400 nmol litre–1 or <500 nmol litre–1 respectively were found in six and seven patients receiving dexmedetomidine, and in seven and nine patients receiving propofol infusions. There were no differences in cortisol (P=0.22) and ACTH concentrations (P=0.74) between the dexmedetomidine and propofol groups. For clarity, median (interquartile range) values of cortisol and ACTH concentrations are shown only for the first 8 h of infusion in the ICU (Fig. 2). Analysis of the cortisol and ACTH concentrations of individual patients showed a normal and appropriate inverse relationship between the two hormones.



View larger version (21K):
[in this window]
[in a new window]
 
Fig 2 Median (interquartile range) serum cortisol and ACTH. ACTH concentrations are on a logarithmic scale. n=10 in both groups except at 8 h, when n=8 and n=7 in the dexmedetomidine and propofol groups respectively.

 
Three methods were used for interpreting the short ACTH stimulation test after dexmedetomidine infusion. In the first method, the response to ACTH was considered adequate if the post-ACTH cortisol concentration at 30 or 60 min was >400 nmol litre–1, and in the second method a concentration >550 nmol litre–1 was considered adequate. In the third method, if the difference between the pre-ACTH and highest post-ACTH cortisol concentration was >200 nmol litre–1, this was considered an adequate response. Five patients failed the last method, although four of these patients had either high or high normal basal cortisol concentrations. Two patients failed the second method, and one patient failed all three methods for determining an adequate response to ACTH (Table 2).


View this table:
[in this window]
[in a new window]
 
Table 2 Assessment of adrenocortical function using three different criteria to interpret the short ACTH stimulation test in patients who had received dexmedetomidine infusions. +=pass, –=fail
 
Endocrine and inflammatory changes (Table 3)
Cortisol and ACTH data are reported above. There were no significant between- or within-group differences in prolactin and glucose concentrations (P=0.25 and P=0.22 respectively).

Two patients in the propofol group and one receiving dexmedetomidine required an insulin infusion to control blood glucose concentration and so were excluded from the analysis of the insulin results. Of the remainder, three patients in the dexmedetomidine group and one patient in the propofol group had insulin concentrations below the sensitivity of the ELISA assay for the whole study period. Insulin concentrations were consistently lower in the dexmedetomidine group, but this was statistically significant only at 2 h (P=0.021).

Serum concentrations of GH increased from baseline in patients receiving dexmedetomidine in the ICU, although this rise was not statistically significant. There was no significant change in the propofol group. There was a statistically significant difference between the groups overall for GH secretion (P=0.036) (Table 3).


View this table:
[in this window]
[in a new window]
 
Table 3 Median (interquartile range) blood glucose, serum insulin, prolactin, GH and IL-6 concentrations in patients infused with dexmedetomidine (D) or propofol (P). D vs P refers to between-group differences determined by estimation of the area under the curve. There was a significant difference in insulin concentrations between the groups at 2 h (P=0.021). n=10 in both groups except at 8 h, when n=8 and n=7 in the dexmedetomidine and propofol groups respectively for all measurements, apart from insulin, when n=7 and n=6 respectively
 
Interleukin-6 response
Serum IL-6 concentrations were similar at baseline in the two groups, and patients receiving dexmedetomidine showed a decrease in IL-6 concentrations during the study, in contrast to patients receiving propofol (Table 3). However, there were no statistically significant differences between the dexmedetomidine and propofol groups throughout the study.


    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The results show clearly that there were no differences in serum cortisol or ACTH concentrations between patients receiving post-operative propofol or dexmedetomidine infusions in the ICU (Fig. 2). Propofol infusions have been shown previously to have no significant effects on adrenal steroidogenesis in critically ill patients,19 and the findings from our study imply that dexmedetomidine acts similarly. There were no differences in the doses of fentanyl given intra-operatively (Table 1), and although fentanyl may attenuate the endocrine responses to surgery,20 21 this only occurs at doses much greater than those used in this study. Patients receiving propofol for sedation required significantly more alfentanil in the ICU, but this is unlikely to have influenced the endocrine changes because opioids are unable to attenuate an established cortisol response.22

Imidazole compounds are potent, non-selective inhibitors of cytochrome P450 enzymes, through the binding of the unhindered nitrogen in the imidazole ring to the catalytic haem iron atom of the protoporphyrin in the resting cytochrome P450 enzyme.23 Dexmedetomidine does not inhibit cytochrome P450 enzymes, including those involved in steroidogenesis, at therapeutic concentrations.4 6 It is difficult to show that a new drug does not affect adrenocortical function in the critically ill because many other factors may be influencing steroidogenesis, including the pathology and other medications. A single test of adrenal function in the ICU remains elusive, and so we assessed steroidogenesis by measuring sequential cortisol and ACTH concentrations and by performing an adrenal cortex stimulation test at the end of the dexmedetomidine infusion.

Hypocortisolism, defined as a serum cortisol concentration below the lower limit of the normal range (<138 nmol litre–1), is probably an inappropriate definition for patients in the ICU. Because surgery, intercurrent illness and ICU interventions will stimulate the hypothalamic–pituitary– adrenal axis, various investigators have suggested lower cortisol limits of 400 or 500 nmol litre–1 to be appropriate measures of adrenal insufficiency in these patients.24 The majority of patients studied would fall into this category at some time in the ICU, a finding reported recently in other ICU patients.25 These authors concluded that relative dysfunction of the adrenal glands is a frequent occurrence in the ICU patient, although it is unclear what an appropriate circulating cortisol value is in these patients. The exaggerated ACTH response at low concentrations of cortisol seen with etomidate,26 implying maximal ACTH secretion in order to overcome complete inhibition of cytochrome P450 enzymes, was not seen in this study.

Our results differ from those of Aho and colleagues, who noted that the cortisol response was attenuated during gynaecological laparoscopy.6 Pre-operative samples provided baseline cortisol values in this study and were lower than those found in our post-operative patients. It is possible that there were small differences in the endocrine response in the dexmedetomidine group that we were unable to detect with the power of the study; our sample size would be expected to show a 50% change in cortisol values based on Aho’s work ({alpha} value 0.05, ß value 0.8).

An ACTH stimulation test is used clinically to diagnose relative adrenal insufficiency, because isolated normal or high cortisol measurements may not reflect the patient’s ability to respond to further physiological stress and low cortisol levels do not necessarily imply functional adrenal insufficiency. However, the usefulness and interpretation of this test in both post-operative27 and intensive care7 patients remains controversial. There is no agreement in intensive care medicine on the criteria that should be used in evaluating the changes in cortisol after ACTH stimulation. Therefore, it is not surprising that our patients sedated with dexmedetomidine showed differing pass rates depending on the method used to interpret the test. Five patients failed to show an incremental rise in cortisol >200 nmol litre–1 after ACTH stimulation (Table 1), reflecting the inadequacy of this criterion in diagnosing adrenocortical dysfunction. The majority of patients had high normal or high basal cortisol values and the magnitude of the incremental rise was inversely proportional to the basal cortisol concentration.27 It is notable that over 30% of healthy volunteers failed to meet this criterion in one study.28 There are several criteria in the literature for diagnosing adrenal hypofunction based on the peak cortisol concentration after ACTH stimulation.27 We chose the lowest value (400 nmol litre–1) and a commonly used value (550 nmol litre–1). Two patients failed to meet two of our selected criteria, but both these patients had an uneventful post-operative ICU and hospital recovery with no clinical suspicion of adrenocortical insufficiency.

Dexmedetomidine failed to modulate the remaining hormonal response to surgery, with the exception of a transient decrease in insulin and augmentation of GH secretion. Alpha-2 adrenoceptor agonists can cause hyperglycaemia in humans12 29 and even the wild boar.30 The mechanism is thought to involve postsynaptic {alpha}2-adrenoceptor stimulation of pancreatic beta cells, which inhibits insulin release, although activation of hepatic glycogenolysis and an increase in GH may be involved.31 Hyperglycaemia was not found after surgery when oral clonidine was given,32 and this may reflect attenuation of the sympathoadrenal response. We found no differences in glucose concentrations between the groups, and there were no differences in the requirements for exogenous insulin infusion.

Agonists of the {alpha}2-adrenoceptor, including dexmedetomidine, are known to stimulate GH secretion.5 Increased protein turnover and a negative nitrogen balance are characteristic features of patients in the ICU, and this may prolong ventilation as a result of respiratory muscle weakness and delay mobility. This may be attributable partly to resistance to GH and the decreased production and action of insulin-like growth factor.33 It was thought that GH therapy may be beneficial in the critically ill, but a recent large multicentre study showed that mortality was increased at least two-fold by the use of large doses of GH in these patients.34 We found an increase in GH in the dexmedetomidine group that was within the physiological range, and this may be advantageous for patients in the ICU.

Dexmedetomidine may influence the inflammatory response and there is some evidence that this has a beneficial effect on outcome.9 IL-6 is the principal cytokine released after surgery, and the response is modulated partly by endogenous corticosteroids.35 In this study, any interaction between IL-6 and cortisol would have been similar in the two groups. Circulating IL-6 concentrations reflect the inflammatory response to surgical trauma, and the similarity of baseline concentrations indicates that the severity of surgery was equivalent in the groups. IL-6 usually peaks about 12 h after surgery,8 and it is notable that patients receiving dexmedetomidine showed a continuing decline in IL-6 concentrations (Table 3). The lack of a statistically significant difference between the groups may be the result of the small sample size, and further investigation is warranted. According to our results, a sample size of 40 would have an 80% power of detecting a 50% difference in IL-6 ({alpha} value 0.05).

In conclusion, dexmedetomidine infusions did not inhibit adrenal steroidogenesis when used for short-term sedation in intubated and ventilated post-operative patients in the ICU. Dexmedetomidine attenuated the heart rate response to surgical trauma compared with propofol and may decrease the inflammatory response to surgical trauma. However, dexmedetomidine had only minor effects on the other endocrine changes.


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Watt I, Ledingham I. Mortality amongst multiple trauma patients admitted to an intensive therapy unit. Anaesthesia 1984; 39: 973–81[ISI][Medline]

2 Absalom A, Pledger D, Kong A. Adrenocortical function in critically ill patients 24h after a single dose of etomidate. Anaesthesia 1999; 54: 861–67[ISI][Medline]

3 Venn R, Bradshaw C, Spencer R, Brealey D, Caudwell E, Naughton C, et al. Preliminary UK experience of dexmedetomidine, a novel agent for postoperative sedation in the intensive care unit. Anaesthesia 1999; 54: 1136–42[ISI][Medline]

4 Maze M, Virtanen R, Daunt D, Banks SJ, Stover EP, Feldman D. Effects of dexmedetomidine, a novel imidazole sedative–anesthetic agent, on adrenal steroidogenesis: in vivo and in vitro studies. Anesth Analg 1991; 73: 204–8[Abstract]

5 Kallio A, Scheinin M, Koulu M, Ponkilainen R, Ruskoaho H, Viinamaki O, et al. Effects of dexmedetomidine, a selective alpha 2-adrenoceptor agonist, on hemodynamic control mechanisms. Clin Pharmacol Ther 1989; 46: 33–42[ISI][Medline]

6 Aho M, Scheinin M, Lehtinen AM, Erkola O, Vuorinen J, Korttila K. Intramuscularly administered dexmedetomidine attenuates hemodynamic and stress hormone responses to gynecologic laparoscopy. Anesth Analg 1992; 75: 932–9[Abstract]

7 Masterson GR, Mostafa SM. Adrenocortical function in critical illness. Br J Anaesth 1998; 81: 308–10[ISI][Medline]

8 Desborough JP. The stress response to trauma and surgery. Br J Anaesth 2000; 85: 109–17[Free Full Text]

9 Kennedy BC, Hall GM. Neuroendocrine and inflammatory aspects of surgery: do they affect outcome? Acta Anaesthiol Belg 1999; 50: 205–9[Medline]

10 Lal S, Tolis G, Martin JB, Brown GM, Guyda H. Effects of clonidine on growth hormone, prolactin, luteinising-hormone, follicle-stimulating hormone and thyroid-stimulating hormone in the serum of normal men. J Clin Endocrinol Metab 1975; 41: 827–32[Abstract]

11 Helbo-Hansen S, Fletcher R, Lundberg D, Nordstrom I, Werner O, Stahl E, et al. Clonidine and the sympathoadrenal response to coronary artery by-pass surgery. Acta Anaesthesiol Scand 1986; 30: 235–42[ISI][Medline]

12 Lyons FM, Bew S, Sheeran P, Hall GM. Effects of clonidine on the pituitary hormonal response to pelvic surgery. Br J Anaesth 1997; 78: 134–7[Abstract/Free Full Text]

13 Novak-Jankovic V, Paver-Eren V, Bovill JG, Ihan A, Osredkar J. Effect of epidural and intravenous clonidine on the neuro-endocrine and immune stress response in patients undergoing lung surgery. Eur J Anaesthesiol 2000; 17: 50–6[ISI][Medline]

14 Straub RH, Herrmann M, Berkmiller G, Frauenholz T, Lang B, Scholmerich J, et al. Neuronal regulation of interleukin 6 secretion in murine spleen: adrenergic and opioidergic control. J Neurochem 1997; 68: 1633–9[ISI][Medline]

15 Dorman T, Clarkson K, Rosenfield BA, Shanholtz C, Lipsett PA, Breslow MJ. Effects of clonidine on prolonged postoperative sympathetic response. Crit Care Med 1997; 25: 1147–52[ISI][Medline]

16 Ramsay MA, Savege TM, Simpson BR, Goodwin R. Controlled sedation with alphaxalone–alphadolone. Br Med J 1974; 2: 656–9[ISI][Medline]

17 Venn R, Cusack RJ, Rhodes A, Grounds RM. Monitoring the depth of sedation in the intensive care unit. Clin Intensive Care 1999; 10: 81–90

18 Matthews JNS, Altman DG, Campbell MJ, Royston P. Analysis of serial measurements in medical research. Br Med J 1990; 300: 230–5[ISI][Medline]

19 Aitkenhead AR, Pepperman ML, Willatts SM, Coates PD, Park GP, Bodenham AR, et al. Comparison of propofol and midazolam for sedation in critically ill patients. Lancet 1989; 2: 704–9[ISI][Medline]

20 Klingstedt C, Giesecke K, Hamberger B, Jarnberg PO. High and low dose fentanyl anaesthesia; circulatory and plasma catecholamine responses during cholecystectomy. Br J Anaesth 1987; 59: 184–8[Abstract]

21 Hall GM, Young C, Holdcoft A, Alaghband-Zadeh J. Substrate mobilisation during surgery. A comparison between halothane and fentanyl anaesthesia. Anaesthesia 1978; 33: 924–30[ISI][Medline]

22 Bent JM, Paterson JL, Mashiter K, Hall GM. Effects of high-dose fentanyl anaesthesia on the established metabolic and endocrine response to surgery. Anaesthesia 1984; 39: 19–23[ISI][Medline]

23 Testa B, Jenner P. Inhibitors of cytochrome P-450s and their mechanism of action. Drug Metab Rev 1981; 12: 1–118[ISI][Medline]

24 Jacobs HS, Nabarro JDN. Plasma 11-hydroxycorticosteroid and growth hormone levels in acute medical illnesses. Br Med J 1969; 2: 595–8[ISI][Medline]

25 Rydvall A, Brandstrom A-K, Banga R, Asplund K, Backlund U, Stegmayr BG. Plasma cortisol is often decreased in patients treated in an intensive care unit. Intensive Care Med 2000; 26: 545–1[ISI][Medline]

26 Fragen RJ, Shanks CA, Molteni A, Avram MJ. Effects of etomidate on hormonal responses to surgical stress. Anesthesiology 1984; 61: 652–56[ISI][Medline]

27 Nicholson G, Burrin JM, Hall GM. Peri-operative steroid supplementation. Anaesthesia 1998; 53: 1091–4[ISI][Medline]

28 Speckart PF, Nicoloff JT, Bethune JE. Screening for adrenocortical insufficiency with cosyntropin (synthetic ACTH). Arch Intern Med 1971; 128: 761–3[ISI][Medline]

29 Swislocki AL, Vestal RE, Reaven GM, Hoffman BB. Acute metabolic effects of clonidine and adenosine in man. Horm Metab Res 1993; 25: 90–5[ISI][Medline]

30 Wolkers J, Wensing T, Groot Bruinderink GW. Sedation of wild boar (Sus scrofa) and red deer (Cervus elaphus) with medetomidine and the influence on some haematological and serum biochemical variables. Vet Q 1994; 16: 7–9[ISI][Medline]

31 Angel I, Bidet S, Langer SZ. Pharmacological characterization of the hyperglycemia induced by alpha-2 adrenoceptor agonists. J Pharmacol Exp Ther 1988; 246: 1098–103[Abstract]

32 Nishina K, Mikawa K, Maekawa N, Shiga M, Obara H. Effects of oral clonidine premedication on plasma glucose and lipid homeostasis associated with exogenous glucose infusion in children. Anesthesiology 1998; 88: 922–7[ISI][Medline]

33 Timmins AC, Cotterill AM, Hughes SC, et al. Critical illness is associated with low circulating concentrations of insulin-like growth factors-I and -II, alterations in insulin-like growth factor binding proteins, and an induction of an insulin-like growth factor binding protein 3 protease. Crit Care Med 1996; 24: 1460–6[ISI][Medline]

34 Takala J, Ruokonen E, Webster NR, Nielsen MS, Zandstra DF, Vundelinckx G, et al. Increased mortality associated with growth hormone treatment in critically ill adults. N Engl J Med 1999; 341: 785–92[Abstract/Free Full Text]

35 Jameson P, Desborough JP, Bryant AE, Hall GM. The effect of cortisol suppression on interleukin-6 and white blood cell responses to surgery. Acta Anaesthesiol Scand 1997; 41: 304–8[ISI][Medline]