| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anesthesia, Toronto Western Hospital, University Health Network, University of Toronto, Toronto, Ontario, Canada
Address correspondence and reprint requests to Pirjo H. Manninen, MD, FRCPC, Department of Anesthesia, Toronto Western Hospital, 399 Bathurst St., Toronto, ON, M5T 2S8, Canada. Address e-mail to Pirjo.Manninen{at}uhn.on.ca.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Dexamethasone is often used during craniotomy to reduce cerebral edema and injury. In rat models of focal cerebral ischemia, dexamethasone has been associated with increased blood glucose concentrations and increased brain injury. When insulin was added to restore blood glucose concentrations to normal (6.8 ± 0.9 mmol/L) in dexamethasone-treated rats, neurological outcome was similar to that in nondexamethasone-treated rats (3).
It has not been routine practice at our institution, nor in many others, to measure blood glucose concentrations in nondiabetic patients during craniotomy, because it is often believed that these patients do not develop harmful blood glucose increases. A nonrandomized study comparing patients undergoing craniotomy who received 10 mg of dexamethasone or placebo showed that at 4 h from the induction of anesthesia, the dexamethasone group had a larger blood glucose concentration (8.3 versus 5.7 mmol/L; P < 0.05) (4). No further measurements were taken; thus, the peak of the blood glucose effect may have been missed. Also, patients who had been taking preoperative dexamethasone were not studied, and these patients form a large part of our clinical practice.
The purpose of this study was to document for 12 h the blood glucose concentrations of nondiabetic patients undergoing craniotomy and to compare patients who received intraoperative dexamethasone, with or without preoperative dexamethasone, with patients who did not receive any dexamethasone. The information obtained was used to identify which patients are at risk of hyperglycemia during craniotomy.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Glucose-containing fluids were withheld from the commencement of the preoperative fast to the end of the study period. There were no other restrictions on anesthetic or surgical management. The decision to administer dexamethasone or not was made solely by the surgeon. All patients who were taking preoperative dexamethasone received 10 mg of IV dexamethasone sodium phosphate injection USP (Sabex, Boucherville, Canada) at the induction of anesthesia and another 4 mg IV 6 h later; this is the standard dose used at our institution. The remaining patients were administered intraoperative dexamethasone at the same time and dose if requested by the surgeon. Anesthesiologists and surgeons were free to treat hypoglycemia or hyperglycemia with glucose or insulin, respectively, if they thought that this was clinically indicated. Any such treatment was recorded, and any blood glucose measurements after such treatment were excluded from analysis.
Blood glucose analysis was performed with an Accucheck Advantage blood glucose meter (Roche Diagnostics Corporation) that was calibrated daily. The blood glucose concentration was measured just before the induction of anesthesia and every 2 h thereafter for 12 h. Blood was drawn from the patient's arterial line when possible, but if an arterial line was not present, a fingerprick capillary blood glucose was measured.
For analysis, patients were divided into three groups:
The peak blood glucose concentration recorded for each patient was used for analyzing the two primary outcomes of this study: 1) differences among groups for peak blood glucose concentration and 2) differences among groups for the change from preinduction blood glucose concentration to peak blood glucose concentration. Power analysis based on data from a previous study (4) showed that 10 patients per group would have 97% power at 
= 0.05 to detect a 2.3 mmol/L difference between groups. Therefore, we decided to recruit patients for this study until there were at least 30 patients in total and at least 8 patients in each group. Secondary outcomes were the difference in preinduction blood glucose concentration between patients who were taking preoperative dexamethasone and those who were not, the time course of blood glucose concentration changes, and the correlation between the duration of surgery and peak blood glucose concentrations.
Continuous data (blood glucose concentration, age, weight, height, duration of surgery, and fasting duration) were initially analyzed with analysis of variance, with post hoc Student's t-tests when analysis of variance indicated a statistical difference. Sex was compared with Fisher's exact test. Bonferroni's correction was used for multiple comparisons within groups. A correlation analysis was performed to quantify the strength of the association between the maximum blood glucose level and the duration of surgery. Statistical significance was determined if P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
Demographic data are presented in Table 2. Patients in group P were significantly taller than those in group I, but there was no significant difference in weight or body mass index between groups. Group P had a shorter duration of surgery than the other two groups, which is probably because this group had mostly supratentorial tumors removed, whereas the other groups had more complicated surgery. Patients taking preoperative dexamethasone were taking a dose between 12 and 24 mg/d for a mean duration of 12 days (range, 2–37 days). All patients in groups P and I received 10 mg of dexamethasone IV during anesthetic induction and a further 4 mg IV 393 ± 100 min (mean ± sd) later. There was no statistical difference in anesthetics used among groups. Induction was performed with propofol and either fentanyl (31 patients) or remifentanil (3 patients). Maintenance was performed with desflurane (27 patients), sevoflurane (7 patients), or nitrous oxide (16 patients). Rocuronium was used for muscle relaxation in 27 patients and cisatracurium in 7. Fentanyl was given to all patients upon emergence. Mannitol was given to 16 patients. No patient received treatment for hyperglycemia or hypoglycemia, although treating anesthesiologists and surgeons were free to do so if they wished. No patient received glucose, insulin, or oral hypoglycemic drugs during the study period. No corticosteroid, apart from dexamethasone, was given to any patient.
|
A total of 231 blood glucose measurements were made. Seven measurements were scheduled but not done during the study. This was due to different reasons, including human error, patient taken to radiology, and measurement device temporarily lost. In no case were two successive measurements missed. Capillary blood was measured a total of 10 times in 4 patients; the remaining measurements were performed on arterial blood.
Preinduction and peak blood glucose data are presented in Table 3. The blood glucose concentration before anesthetic induction (initial blood glucose) was larger in group P than group N, but not group I. Patients taking preoperative dexamethasone had a larger preinduction blood glucose than those not receiving preoperative dexamethasone (groups I and N combined) (6.1 versus 5.2 mmol/L; P < 0.05). With regard to the primary outcomes, group I had a statistically significantly larger peak blood glucose concentration and a greater change from preinduction to peak blood glucose concentration than the other two groups. The mean peak blood glucose concentration in the 12-h period in group I was 11.0 mmol/L. In this group, 7 of 8 patients had peak blood glucose concentrations between 10.0 and 13.8 mmol/L, and 1 patient showed only a small increase in blood glucose to 7.5 mmol/L. The time course of blood glucose changes is shown in Figure 1. The blood glucose increased from the preinduction measurement in all groups, with the largest change in group I. Patients in this group reached the peak blood glucose level an average of 8–10 h after the commencement of surgery, with a mean blood glucose at these time periods of 10.2 mmol/L. There was a moderate positive correlation between peak blood glucose and duration of surgery, with a correlation coefficient of 0.34 (P = 0.05).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Patients taking preoperative dexamethasone had a 0.9 mmol/L larger blood glucose concentration at the commencement of surgery compared with those who were not. This is unlikely to be clinically significant. From then on, all groups showed an increase in blood glucose. The patients not receiving dexamethasone had a mean peak blood glucose concentration of 7.8 mmol/L (reference range, 4.0–7.0 mmol/L). It is common for blood glucose concentrations to increase in small amounts during surgery as part of a stress response (5). This was also demonstrated in another craniotomy study (4). We found a moderate positive correlation (r = 0.34) between the duration of surgery and peak blood glucose concentration. This suggests that the longer the surgery takes, the more the stress response effect on blood glucose.
Group I had a much larger increase in blood glucose concentration than the other groups. This group had a mean peak blood glucose of 11.0 mmol/L (range, 7.5–13.8 mmol/L), representing a mean 5.5 mmol/L increase from baseline. The peak blood glucose concentration occurred 8–10 hours after the induction of anesthesia. Our data are consistent with a previous study that measured blood glucose concentrations for 4 hours (4). After 4 hours, their dexamethasone group had a mean blood glucose concentration of 8.3 mmol/L (compared with 7.9 mmol/L in this study). It is interesting that group I had larger blood glucose concentrations than group P, even though both groups received the same dose of dexamethasone during the study. A possible explanation is that patients develop some tolerance to the hyperglycemic effects of dexamethasone with chronic use. However, patients taking long-term steroids can develop diabetes mellitus, which would suggest that such tolerance does not occur. Another possibility is that the adrenal response to stress is suppressed in patients already taking dexamethasone and that, therefore, the increase in endogenous steroids, as a result of the surgical stress response, is obtunded.
The magnitude of the blood glucose increase during surgery found in our patients—particularly the increase seen in group I—was possibly enough to result in increased neuronal injury due to ischemia. The smallest change in blood glucose concentration that has been associated with a difference in brain injury was 2.3 mmol/L in a study of global ischemia in monkeys (6). All three groups had a more than 2.3 mmol/L increase from preinduction to peak blood glucose concentration. In addition, group I had peak glucose concentrations of 4 mmol/L more than the reference range for blood glucose and had peak glucose concentrations at least 2.5 mmol/L above the other groups. An expert reviewing glucose modulation of ischemic brain injury recommended the target range for blood glucose in nondiabetics at high risk for cerebral ischemia to be 80–150 mg/dL (4.4–8.3 mmol/L) (1). The mean peak blood glucose in the patients receiving dexamethasone in our study was larger than this target range. This study was not designed to try to detect an outcome difference between groups and thus did not have the power to do so. We cannot conclude that the administration of dexamethasone caused increased morbidity—only that those patients had a level of hyperglycemia that is associated with impaired outcome from neuronal ischemia in other studies. If the current recommendations to avoid hyperglycemia during neurosurgery (1,7,8) are to be followed, then these results are clinically relevant.
This study has its limitations. First, it was not a randomized study. Patients in each of the different groups had different types of surgery. The type of surgery, the underlying pathology, or another unknown associated factors may have contributed to the blood glucose effect. We chose a nonrandomized design for this study because our main aim was to document the blood glucose changes in different patient groups and not specifically to determine a causal relationship between dexamethasone and hyperglycemia. In addition, our institution has a reasonably standard prescribing practice for dexamethasone (it is given for tumor surgery but not arteriovenous malformation, aneurysm, or epilepsy surgery), and it would be unethical to randomize patients to dexamethasone or placebo just to investigate the side effects of increasing blood glucose.
Second, patient height and duration of surgery were potentially confounding factors unevenly distributed in our groups. The patients in group P were taller than those in group I, which may have led to pharmacokinetic differences between the groups having dexamethasone, but because the weight and body mass index were similar between groups, this is unlikely to have been a significant confounding factor. The longer duration of surgery in group I may have biased our results. As discussed previously, the duration of surgery likely affects blood glucose concentrations. These limitations mean that we are unable to definitely attribute the cause of the blood glucose changes we found to dexamethasone, despite the strong association. However, it seems likely that dexamethasone is the major cause of these blood glucose changes because we know that it can cause an increased blood glucose concentration in nondiabetic patients who are not having surgery. This appears to be due to an increase in gluconeogenesis, with some reduction in end-organ sensitivity to insulin at larger doses or chronic administration (9,10). Dexamethasone has also been associated with blood glucose increases during cardiac surgery (4,11).
We studied only nondiabetic patients. This was because it is already standard practice to intensively monitor and treat blood glucose in diabetic patients who are having neurosurgery and thus we would not be likely to change clinical practice for diabetic patients. Therefore, our results cannot be extrapolated to the diabetic population.
In summary, we found that nondiabetic patients undergoing craniotomy have increased blood glucose concentrations in the 12 hours after the commencement of surgery. Patients who receive dexamethasone at the induction of anesthesia are at higher risk of hyperglycemia than those already taking dexamethasone before surgery or those not receiving dexamethasone. The magnitude of the blood glucose increases is clinically relevant, and we recommend that these patients have their blood glucose concentrations monitored perioperatively for at least 12 hours and receive treatment if necessary.
We thank Karolinah Lukitto for her assistance with data collection.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Hans, A. Vanthuyne, P. Y. Dewandre, J. F. Brichant, and V. Bonhomme Blood glucose concentration profile after 10 mg dexamethasone in non-diabetic and type 2 diabetic patients undergoing abdominal surgery Br. J. Anaesth., August 1, 2006; 97(2): 164 - 170. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
