| ||||||||||||||
|
|
|||||||||||||



*Department of Anesthesia, McGill University, Montreal, Quebec, Canada;
Clinic of Anesthesiology, University of Ulm, Ulm, Germany; and
Department of Anesthesia, St. Marien-Krankenhaus, Ludwigshafen, Germany
Address correspondence and reprint requests to Dr. Thomas Schricker, Department of Anesthesia, McGill University, Royal Victoria Hospital, Room S5.05, 687 Pine Ave. W., Montreal, Quebec, Canada H3A 1A1.
| Abstract |
|---|
|
|
|---|
Implications: Epidural analgesia combined with general anesthesia prevented thehyperglycemic response to surgery by decreasing endogenous glucose production.The increased glucose plasma concentration in patients receivingfentanyl/midazolam anesthesia was caused by a decrease in whole-body glucoseclearance. The hyperglycemic response observed during inhaled anesthesia withisoflurane was a consequence of both impaired glucose clearance and increasedglucose production.
| Introduction |
|---|
|
|
|---|
The goal of this investigation was to elucidate the mechanisms by which the hyperglycemic response to surgery is modified by different types of anesthesia. Therefore, the kinetics of intraoperative glucose metabolism (i.e., whole-body glucose production and glucose uptake) have been studied by stable isotope tracers in patients undergoing cystoprostatectomy receiving either epidural analgesia combined with general anesthesia, fentanyl/midazolam anesthesia, or inhaled anesthesia with isoflurane.
| Methods |
|---|
|
|
|---|
Patients were randomly assigned to receive one of three different anesthetic techniques. One group received epidural analgesia combined with general anesthesia with fentanyl and midazolam (n = 8). The second group received IV fentanyl and midazolam alone (n = 8). Patients in the third group received inhaled anesthesia with isoflurane (n = 7).
All patients were given premedication (20 mg chlorazepate orally) the night before surgery and 3 h before the operation. In the Epidural group, an epidural catheter was inserted at a thoracic level between T10 and T12, and segmental block (T4 through S5) was established with 0.5% bupivacaine, as judged from perception of pinprick. Anesthesia was induced with 0.15 mg/kg midazolam and 510 µg/kg fentanyl and was continued by boluses of midazolam. Epidural analgesia was maintained by the continuous administration of 0.25% bupivacaine 812 mL/h. Anesthesia in the Fentanyl/Midazolam group was induced with 0.15 mg/kg midazolam and 510 µg/kg fentanyl and maintained by intermittent boluses of fentanyl and midazolam to keep heart rate and mean arterial blood pressure within 20% of preoperative values. In the Isoflurane group, general anesthesia was induced with 5 mg/kg thiopentone and 1.5 µg/kg fentanyl. Anesthesia was continued by adjusting end-tidal isoflurane concentrations to keep heart rate and arterial pressure within 20% of preinduction values. In all patients, the trachea was intubated after neuromuscular block with 1.5 mg/kg succinylcholine. The patients lungs were ventilated with 30% oxygen in nitrous oxide at a respiratory rate of 10 breaths/min to maintain normocapnia. Supplemental doses of vecuronium were applied as needed for complete surgical muscle relaxation. During surgery, patients received a balanced electrolyte solution at a rate of 10 mL · kg-1 · h-1. Colloids (3% gelatin; molecular weight, 35,000) were administered to keep the pulmonary artery occlusion pressure (PAOP) between 10 and 15 mm Hg.
Hemodynamic monitoring was performed with a three-lead electrocardiogram monitor and radial artery catheterization for continuous blood pressure measurement. In addition, a catheter was placed in the superior vena cava via the right jugular internal vein, and a thermistor-tipped 7F Swan-Ganz catheter was advanced into the pulmonary artery as judged from the pressure contour.
The rates of appearance of glucose (Ra glucose; endogenous glucose production) and glycerol (Ra glycerol; lipolysis) were determined before and during the operation by using [6,6-2H2]glucose and [1,1,2,3,3-2H5]glycerol (Masstrace, Woburn, MA). Sterile isotope solutions were prepared by the hospital pharmacy as previously described (6).
Preoperative measurements were taken after an overnight fast 3 days before surgery. A superficial vein in the dorsum of the hand was cannulated and kept patent with continuous saline infusion (2 mL · kg-1 · h-1). A second catheter was placed in a superficial vein of the contralateral arm for the infusion of [6,6-2H2]glucose and [1,1,2,3,3-2H5]glycerol. After warming the hand in a heated air box to achieve arterialized blood (SaO2 >95%), blood samples were obtained to determine basal isotopic enrichment. Thereafter, priming doses of [6,6-2H2]glucose (4 mg/kg) and [1,1,2,3,3-2H5]glycerol (0.11 mg/kg) were administered, followed by continuous infusions of 0.05 mg · kg-1 · min-1 [6,6-2H2]glucose and 0.007 mg · kg-1 · min-1 [1,1,2,3,3-2H5]glycerol, respectively. Three arterialized blood samples were drawn after 100, 110, and 120 min of isotope infusion, when the tracer was assumed to have reached an isotopic steady-state. On the day of surgery, the primed continuous infusion of stable isotopes was started at the beginning of surgery and, in analogy to the preoperative measurement, arterialized blood samples were taken 100, 110, and 120 min after skin incision.
At t = 110 min of pre- and intraoperative isotope infusion, plasma concentrations of metabolic substrates (glucose, glycerol, lactate, and nonesterified fatty acids) and hormones (insulin, glucagon, cortisol, epinephrine, and norepinephrine) were determined. Preoperative hemodynamic monitoring included heart rate (HR) and mean arterial blood pressure by noninvasive oscillometric measurements. During the operation, PAOP, cardiac output (CO), and whole-body oxygen uptake (
O2) were recorded. CO was determined by using the Swan-Ganz catheter thermodilution technique. PAOP was quantified by means of the balloon inflation technique.
O2 was calculated with the Fick principle.
The isotopic plasma enrichments (atom percent excess; APE) of [6,6-2H2]glucose and [1,1,2,3,3-2H5]glycerol were determined by gas chromatography-mass spectrometry as previously described (7). The APE values used in the calculation of the Ra glucose and glycerol were the means of the three APE measurements obtained at the two study periods. The Ra glucose and the Ra glycerol were derived from the equation Ra = I x (APEinf/APEpl - 1), where APEinf is the tracer enrichment in the infusate, APEpl is the tracer enrichment in plasma at steady state, and I is the tracer infusion rate. Steady-state conditions were assumed when the coefficient of variation of the three enrichment values was <5%.
In the physiologic steady-state, whole-body glucose uptake equals the rate of endogenous glucose production. Because glucose uptake increases proportionally as blood glucose concentrations increase, changes in whole-body glucose uptake do not necessarily reflect corresponding changes in the tissues ability to take up glucose. This may be because most glucose uptake occurs in noninsulin-sensitive tissues, and the rate of uptake is to a large extent determined by the diffusion gradient for glucose. Thus, the rate of glucose uptake has to be corrected for the prevailing plasma glucose concentration. The resulting value, the glucose clearance rate, calculated as the Ra glucose divided by the corresponding plasma glucose concentration, represents an index of the ability of tissues to take up glucose.
The fractional plasma clearance rate of glycerol was calculated as the Ra of glycerol divided by the glycerol plasma concentration.
Plasma concentrations of glucose, glycerol, lactate, and nonesterified fatty acids (NEFA) were measured by using enzymatic assays (Boehringer Mannheim GmbH, Mannheim, Germany). The mean intraassay and interassay coefficients of variance were 3.0% and 4.5% for glucose, 2.5% and 3.5% for glycerol, 2.5% and 4.0% for lactate, and 3.0% and 4.2% for NEFA. Insulin, glucagon, and cortisol were analyzed with radioimmunoassays (INS-RIA-100, Medgenix Diagnostics, Brussels, Belgium, for insulin; Double Antibody Glucagon RIA, Diagnostic Products Corporation, Los Angeles, CA, for glucagon; and DSL-200 SP Aktive Cortisol, Diagnostic System Laboratories, Sinsheim, Germany, for cortisol). The mean intraassay and interassay coefficients of variance were 5.6% and 8.3% for insulin, 4.8% and 8.0% for glucagon, and 3.0% and 7.0% for cortisol. Catecholamine concentrations were quantified by means of reversed-phase high-performance liquid chromatography (Chromakon 500; Kotron, Eiching, Germany) with electrochemical detection. The mean intraassay and interassay coefficients of variance were 5.0% and 4.5% for epinephrine and 3.2% and 2.5% for norepinephrine, respectively.
Differences between the groups were tested by using the analysis of variance with post hoc analysis by the Kruskal-Wallis test. Within-group comparison of variables was made by Students t-test. A probability of P < 0.05 was considered to be significant. Values are given as mean ± SEM.
| Results |
|---|
|
|
|---|
At baseline, the hemodynamic conditions were similar in the three groups (Table 1). During surgery, HR significantly decreased in both the Fentanyl/Midazolam group and the Epidural group to a value that was significantly less than in patients during isoflurane anesthesia (P < 0.05). HR and mean arterial blood pressure in the Epidural group were significantly less when compared with patients receiving fentanyl/midazolam anesthesia (P < 0.05). There were no significant differences among the three groups with respect to CO, PAOP, and
O2 during the operation. The hematocrit significantly decreased during surgery to a comparable extent in all groups (P < 0.05).
|
|
|
| Discussion |
|---|
|
|
|---|
In contrast to patients in the Isoflurane group, who showed an increase in the plasma glucose concentration by 40%, epidural analgesia suppressed the hyperglycemic response to surgery. This finding is in line with the results of numerous studies demonstrating that epidural analgesia with local anesthetic prevents the increase in the plasma glucose concentration during abdominal surgery (13). In agreement with previous study protocols that used similar amounts of fentanyl, fentanyl/midazolam anesthesia in this investigation did not completely suppress, but blunted, the intraoperative hyperglycemia when compared with patients receiving inhaled anesthesia (4,8).
Because measurements of glucose plasma concentrations alone do not provide insight into the underlying dynamic biochemical pathways, whole-body glucose production and glucose clearance in this study were assessed by a stable isotope tracer technique using [6,6-2H2]glucose.
Epidural analgesia significantly decreased endogenous glucose production and glucose clearance, resulting in an unaltered plasma glucose concentration during surgery. Although the effect of epidural analgesia on the intraoperative kinetics of glucose metabolism has not been investigated so far, a 20% reduction of glucose production by epidural blockade has been reported in patients studied between one and five days after various surgical interventions (9). Fentanyl/midazolam anesthesia did not affect endogenous glucose production, but it decreased glucose clearance, leading to a small but significant increase in the plasma glucose concentration. These findings are in agreement with the results of a previous study showing that neuroleptanesthesia does not influence splanchnic glucose release during cholecystectomy, as calculated from the measurement of splanchnic blood flow and the arteriohepatic venous difference of glucose (8). Isoflurane anesthesia in this study was associated with an increase in endogenous glucose production accompanied by a decreased glucose clearance. The effect of inhaled anesthesia on glucose kinetics during surgery has not yet been determined in humans. The results of animal studies, however, provide evidence that the hyperglycemic action of isoflurane, at least in part, is caused by the stimulation of glucose production (10). Our results support the contention that hyperglycemia during surgery performed under inhaled anesthesia is caused by both an increase in endogenous glucose production rate and a reduction in whole-body glucose uptake.
Under postabsorptive conditions, glycogenolysis constitutes approximately 75% of total endogenous glucose production, the remainder being derived from gluconeogenesis (11). In this study, the entire contribution of gluconeogenesis to total glucose production could not be quantified because the tracer used did not allow the identification of the metabolic pathway where glucose was produced. However, previous studies have shown an increased rate of gluconeogenesis during surgical trauma (12). Gluconeogenesis in the liver is a highly oxygen-consuming pathway, accounting for 50% of hepatic oxygen consumption in the postabsorptive state (13). Furthermore, gluconeogenesis has been proposed to occupy a central position in catabolic pathways, because muscle protein is broken down to supply amino acids serving as precursors for de novo glucose synthesis. It has been hypothesized that, by reducing accelerated gluconeogenesis, the amino acid release from the muscle can be reduced, resulting in better preservation of body protein (14). In a recent investigation, a significant correlation between protein breakdown and glucose production was observed after colorectal surgery, confirming this interdependence between protein and glucose metabolism (6). Thus, the decrease in hepatic glucose production as seen during epidural analgesia assumes clinical importance regarding the energy balance of the liver and protein catabolism.
Several factors may account for the effects of anesthesia on glucose homeostasis observed in our study. Hormone infusion studies in volunteers indicate that increased circulating concentrations of catecholamines and cortisol can mimic the perioperative changes in carbohydrate metabolism (15). It has been proposed that the endocrine mechanisms by which the metabolic effects of anesthesia are mediated are likely to be hormonal. In agreement with the results of previous studies that used isoflurane in similar doses, isoflurane anesthesia did not inhibit the intraoperative increase in the plasma concentrations of cortisol and catecholamines (16,17), which all stimulate hepatic gluconeogenesis and counteract the peripheral action of insulin, resulting in impaired glucose use and hyperglycemia (14). The plasma concentrations of cortisol and catecholamines tended to increase during fentanyl/midazolam anesthesia, but these changes did not reach statistical significance. Because total suppression of the endocrine response has been attained only with large-dose fentanyl anesthesia (50100 µg/kg) (4), the average dose of 15 µg/kg fentanyl applied in this protocol was not sufficient to completely block the endocrine and hyperglycemic response. In accordance with earlier observations, epidural analgesia inhibited the endocrine response to surgery, as reflected in unchanged cortisol, epinephrine, and norepinephrine plasma concentrations (2,3). This unaltered endocrine milieu in the Epidural group, however, does not necessarily explain the significant decrease in endogenous glucose production observed during surgery. Preoperative measurements in this study were performed three days before surgery after a 12-h fast. Because of bowel preparation on the day before the operation, patients at the time of surgery were fasting for approximately 26 hours. One study demonstrated that total glucose production significantly decreases with prolonged fasting, from 10.2 µmol · kg-1 · min-1 after 16 hours to 8.6 µmol · kg-1 · min-1 after 22 hours of fasting (18). These findings support the conclusion that the decrease in glucose production observed during surgery in the Epidural group was a consequence of fasting and that this physiologic effect was uninfluenced by epidural analgesia.
We assessed whole-body lipolysis in this protocol because there is recent evidence of a significant interdependence between the rates of lipolysis and glucose production (19). In contrast to fentanyl/midazolam and isoflurane anesthesia, epidural analgesia decreased the Ra glycerol, indicating a decrease in whole-body lipolysis during surgery. These results are in line with previous investigations showing an inhibitory effect of epidural analgesia on intraoperative lipolysis (3). Because the inhibition of lipolysis is associated with a decrease in glucose production independent of hormonal changes (19), the suppressive effect of epidural analgesia on lipolysis also may have contributed to the decrease in glucose production observed in the Epidural group.
Hepatic glucose production depends on the supply of precursors to the liver, i.e., lactate, alanine, and glycerol (20). In addition, the hepatic uptake of gluconeogenic substrates is a function of blood flow through the liver (21). Consequently, the effect of anesthesia on hemodynamics and liver perfusion gains metabolic importance. Secondary to sympathetic blockade and reduction of peripheral vascular resistance, intraoperative HR and mean arterial pressure in the Epidural group were less than in the other groups. To render all patients comparable with respect to their intravascular volume status, perioperative fluid loading was adjusted to keep PAOP between 10 and 15 mm Hg. From global hemodynamic measurements, however, one cannot predict changes of the regional blood flow in the splanchnic region. Although no formal assessment of hepatic blood flow was obtained, glycerol clearance represents an indirect variable of liver perfusion. From animal studies it is known that at plasma levels <1 mmol/L, there is almost complete extraction of glycerol in one single transhepatic passage (22). The clearance of substrates with such high extraction is profoundly affected by changes in the hepatic blood flow. In contrast to fentanyl/midazolam and inhaled anesthesia, glycerol clearance during epidural analgesia decreased in our study, lending support to the hypothesis that hepatic perfusion declined, which also has been reported by other investigators (23,24). Therefore, it cannot be ruled out that reduced liver perfusion concomitant with decreased substrate supply contributed to the diminished glucose production rate during epidural analgesia.
In conclusion, a stable isotope tracer technique was used to characterize the effect of different types of anesthesia on the hyperglycemic response to surgery and to elucidate the mechanisms whereby these effects are mediated. Epidural analgesia abolished the hyperglycemic response to surgery via a decrease in the glucose production rate. The moderate, but significant, increase in glucose plasma concentration during fentanyl/midazolam anesthesia was caused by a decrease in whole-body glucose uptake. The hyperglycemic response observed during inhaled anesthesia with isoflurane was a consequence of both impaired glucose uptake and increased glucose production.
| Acknowledgments |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
C. J. Zuurbier, P. J. M. Keijzers, A. Koeman, H. B. Van Wezel, and M. W. Hollmann Anesthesia's Effects on Plasma Glucose and Insulin and Cardiac Hexokinase at Similar Hemodynamics and Without Major Surgical Stress in Fed Rats Anesth. Analg., January 1, 2008; 106(1): 135 - 142. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Lattermann, G. Belohlavek, S. Wittmann, B. Fuchtmeier, and M. Gruber The Anticatabolic Effect of Neuraxial Blockade After Hip Surgery Anesth. Analg., October 1, 2005; 101(4): 1202 - 1208. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. S. Anderson, M. Thamotharan, D. Kao, S. U. Devaskar, L. Qiao, J. E. Friedman, and W. W. Hay Jr Effects of acute hyperinsulinemia on insulin signal transduction and glucose transporters in ovine fetal skeletal muscle Am J Physiol Regulatory Integrative Comp Physiol, February 1, 2005; 288(2): R473 - R481. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. S Lewis, S. L Kane-Gill, M. B. Bobek, and J. F Dasta Intensive Insulin Therapy for Critically Ill Patients Ann. Pharmacother., July 1, 2004; 38(7): 1243 - 1251. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Lattermann, F. Carli, L. Wykes, and T. Schricker Perioperative Glucose Infusion and the Catabolic Response to Surgery: The Effect of Epidural Block Anesth. Analg., February 1, 2003; 96(2): 555 - 562. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Schricker, R. Lattermann, P. Fiset, L. Wykes, and F. Carli Integrated analysis of protein and glucose metabolism during surgery: effects of anesthesia J Appl Physiol, December 1, 2001; 91(6): 2523 - 2530. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|