1 Department of Anaesthesia, Söder Hospital, Sweden. 2 Karolinska Institute, Stockholm, Sweden
* Corresponding author. E-mail: robert.hahn{at}sos.sll.se
Accepted for publication August 19, 2004.
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Methods. In 15 women, mean age 50 yr (range 3763), 2 days after an abdominal hysterectomy, a volume kinetic analysis was performed on an i.v. infusion of 12.5 ml kg1 (900 ml) of glucose 2.5% given over 45 min. The insulin resistance was measured by a glucose clamp, and it was compared with daily bioimpedance analyses, which indicated the hydration of the intra/extracellular body fluid spaces.
Results. The clearance of glucose was 0.42 litre min1 (0.60 litre min1 is normal) while the other five parameters in the kinetic model were similar to those obtained in healthy volunteers. Computer simulations indicated that in a 70-kg female, at steady state, the rate of infusion (ml min1) should be three times the allowed increase in plasma glucose (mmol litre1). To maintain a predetermined plasma dilution the corresponding rate factor was 160. The glucose uptake during clamping was 3.9 mg kg1 min1 (7.0 is normal), which, during the second day after hysterectomy, correlated with the dehydration of the intracellular space (r=0.77; P<0.002) and with the protein catabolism as indicated by the urinary excretion of 3-methylhistidine (r=0.76, P<0.002).
Conclusion. The anaesthetist can prescribe postoperative administration of glucose 2.5% to reach any desired plasma glucose level and dilution by using the two presented nomograms. Insulin resistance correlated with intracellular dehydration and protein catabolism.
Keywords: blood, haemodilution ; fluids, i.v ; metabolism, glucose ; pharmacokinetics
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In the present study, volume kinetics were applied to describe the distribution and elimination of the glucose and volume components of glucose 2.5% in the postoperative period after abdominal hysterectomy. The purpose was to use the results from these kinetic analyses, if consistent enough, to create nomograms, which would make it possible for the anaesthetist to predict how fast a glucose solution needs to be infused to reach any predetermined target with respect to plasma glucose and volume expansion. For explorative analyses, the results of volume kinetics were compared with those of a euglycaemic hyperinsulinaemic glucose clamp,2 3 6 7 fluid balance calculations and changes in cellular hydration as studied by bioimpedance.8 9
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The operating time (from incision to the last suture) and the surgical blood loss were recorded. After surgery, and during the follow-up period, each patient was equipped with a patient-controlled analgesic pump containing ketobemidone 1 mg ml1.
To quantify the degree of surgical stress, serum concentrations of two cytokines, interleukin-6 and vascular endothelial growth factor (VEGF), were measured using Quantikine kits (R&D Systems, Minneapolis, MN, USA) in the morning on the first day after surgery. The coefficient of variation of these measurements was less than 5%.
Glucose clamp
On the first morning after surgery, a euglycaemic hyperinsulinaemic clamp was performed. The glucose infusion rate required to maintain a constant level of plasma glucose (clamp) during infusion of a standardized amount of insulin is a measure of insulin sensitivity.6 7 The patients fasted overnight and, at 08:00 the i.v. infusion of insulin (Actrapid Novo, Copenhagen, Denmark) was started at a constant rate of 40 mU m2 min1 using an infusion pump. The insulin was added to 500 ml saline into which 20 ml of the patient's own plasma had been added to prevent adsorption by the plastic material of the infusion set. Glucose 20% was administered i.v. at a rate that was adjusted every 5 min depending on the immediately measured blood glucose level (YSI Inc., Yellow Springs, OH, USA), with the aim of maintaining euglycaemia (mean 6 mmol litre1). Serum insulin and serum C-peptide concentrations were measured every 15 min, using ELISA kits (Mercodia AB, Uppsala, Sweden).
The glucose uptake (M value) was calculated during 1 h of steady state, usually between the 60 and 120 min of clamping. The normal M is about 7 mg kg1 min1.2 6 7
Volume kinetics
On the second postoperative day, at 08:00, after a light breakfast (consisting of one sandwich and one cup of tea or coffee), the patients were taken from the ward to the laboratory. An i.v. infusion of 12.5 ml kg1 (mean volume 888 ml, glucose load 22 g) of iso-osmotic glucose 2.5% with electrolytes (Na 70, Cl 45 and acetate 25 mmol litre1; Rehydrex, Pharmacia, Uppsala, Sweden) was given over 45 min via an infusion pump. Venous blood was drawn every 5 min for a 75-min period, and thereafter every 1015 min for up to 185 min for measuring the concentration of plasma glucose (GLU Gluco-quant reagent, Roche, Mannheim, Germany), blood haemoglobin, the red blood cells count, and the mean corpuscular volume (Hitachi 917, Hitachi Co., Naka, Japan). All indices of the erythrocytes were measured in duplicate and the mean value was used in the calculations. Serum insulin was measured every 45 min.
The data were used to calculate the kinetics of glucose and of the infused volume load. In brief, plasma glucose over time was analysed according to a one-compartment turnover model, which yielded the volume of distribution (Vd), glucose clearance (CL) and endogenous glucose production (ki en). In the absence of glucosuria, each mmol of glucose eliminated from the system implied that it was taken up by the cells and then, by virtue of osmosis, drew in 3.6 ml of fluid to a peripheral (intracellular) space, V3.3 4 The fate of the remaining excess fluid was studied according to volume kinetic principles where it expands a body fluid space of size V1, from which elimination occurs in the form of insensible fluid losses kb, which was fixed at 0.5 ml min1, and a dilution-dependent mechanism for which the rate is proportional by a parameter kr to the expansion of V1. Moreover, fluid is returned from the cells, mainly as a result of glucose oxidation, at a rate that is proportional to the fluid uptake in V3 by a parameter k31 (see Appendix). An analysis was also made to test for the possible expansion of an intermediate space, V2, but this did not yield statistical significance.3
The model parameters for the glucose and volume components of glucose 2.5% were calculated in each individual patient using Matlab version 6.5 (Math Works Inc., Natick, MA, USA). The kinetic models used a non-linear least-squares regression routine based on a modified GaussNewton method. Corrections were made for iatrogenicdilution caused by the blood sampling, which averaged 7 ml for each sampling procedure. The output parameters consisted of Vd, CL, ki en, V1, kr and k31 which, along with the uncertainty of the estimates, were expressed as their standard deviations (SD).
Glucosuria was not assessed as our previous studies show that such losses are very small, if at all present, with a short glucose peak of about 1112 mmol litre1.35
Bioimpedance and fluid balance
On four consecutive mornings, starting on the day of surgery, the extracellular fluid (ECF) and intracellular fluid (ICF) volumes were estimated by bioelectrical impedance using a Xitron 4000B Spectrum Analyzer (Xitron Technologies, San Diego, CA, USA).8 9 Each reported value is the mean of three successive recordings. The preoperative body weight was used as an input variable in all estimations of these physiological body fluid spaces, which were performed by the software delivered with the apparatus.
The total amount of fluid and electrolytes administered during surgery and until 08:00 the next morning (=day of surgery) was measured. Urine was collected from after induction of anaesthesia until 3 successive days for analyses of electrolytes, urea and creatinine. The urinary excretion of 3-methylhistidine (3-MH) on the second postoperative day was measured, using a LC3000 (Viotronic, Mainthal, Germany). This amino acid is found in actin and myosin in muscle and is used as an index of protein catabolism as 3-MH cannot be re-utilized for protein synthesis.10 11 Increased precision in quantifying muscular breakdown is obtained on reporting the ratio between the urinary 3-MH and the urinary creatinine concentration.12
Statistics
Results are expressed as the mean (SD). Because of skewed distributions, the median (interquartile range) was used where appropriate. Changes were evaluated by repeated-measures ANOVA and linear correlations by simple linear regression. P<0.05 was considered significant.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Glucose clamp
Glucose uptake during steady state with the clamp (M value) was 3.9 (1.8) mg kg1 min1 and the corresponding serum insulin level was 365 (69) pmol litre1. The C-peptide concentration, which is an index of endogenous insulin production, decreased gradually from 0.65 (0.28) to 0.42 (0.22) nmol litre1 during clamping (P<0.001).
Volume kinetics
The plasma glucose concentration (baseline 6.1 mmol litre1) was doubled during the glucose infusion but had returned to baseline after 60 min (Fig. 1A). The infused fluid diluted the plasma by 10%. The dilution was more rapidly normalized than plasma glucose (Fig. 1B). The insulin concentration increased 3-fold (Fig. 1C).
|
|
|
|
|
Simulations
The plasma glucose concentration expected to result from infusing glucose 2.5% at various rates is given in Figure 4, which is based on computer simulation using the kinetic parameters in Table 1. A corresponding nomogram for the expected plasma dilution is shown in Figure 5.
|
|
A simulation was also done to illustrate differences in plasma glucose and plasma dilution during and after infusion of glucose 2.5% in the postoperative, intraoperative, and laboratory settings (Fig. 6). For this purpose, parameter sets from the present and two previous studies were used.4 5
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
There is a common belief that glucose solution expands the plasma volume only very slightly, but glucose 2.5% and 5% both compare well with Ringer's solution in this respect.3 For example, increasing plasma glucose by 5 mmol litre1 over 40 min using glucose 2.5% expands the plasma volume by 10%, which is not negligible.
The kinetics of glucose 2.5% solution in the postoperative period is quite similar to that found in laboratory experiments, while it is much more affected during surgery. How different clinical situations compare can be highlighted in simulation experiments designed to predict the response to a fluid infusion with respect to plasma glucose, plasma dilution, volume expansion, and cellular hydration (Fig. 6). Estimates of the kinetic parameters can also be compared directly. In general, altered physiological conditions change the parameters expressing rates but only to a lesser degree those denoting volume.
The CL was between 0.55 and 0.70 litre min1 in the laboratory3 4 and 0.20 litre min1 during laparoscopy;5 in the present study during the postoperative period, we found its value to be 0.42 litre min1 (Table 1). These rates correspond to half-lives for infused glucose of 12, 30, and 16 min, respectively. Therefore, suitable rates of glucose administration are in-between those proposed during surgery and for healthy individuals. However, the endogenous glucose production at baseline was identical to that obtained in healthy volunteers if expressed per kg body weight. The reduction of this rate by 50% during laparoscopy5 would probably result in hypoglycaemia if not counteracted by insulin resistance.
With regard to volume kinetics, the size of the expandable central body fluid space (V1) was slightly decreased, as compared with the surgical setting, but still within the limits of previous laboratory results.4 The clearance constant for infused fluid (kr) was normalized to about 130 ml min1 from being reduced by as much as two-thirds during laparoscopy.5 Finally, the slope k31/V3, which denotes the rate of removal of excess water from the cells, was normal (Table 1).
The volume of distribution for glucose 2.5% solution might be anticipated to be equal to the ECF space, which is often taken as 20% of the body weight. However, both Vd for glucose and V1 for the fluid volume have consistently reached much lower values in our volume kinetic studies. The Vd reported here is 3040% smaller than the expected size of the ECF volume, but this can be explained by the existence of an interstitial pool, which exchanges glucose very slowly with plasma.13 Previous authors have suggested that the ECF space should be multiplied by a pool fraction of 0.65 to obtain the Vd for glucose that is of clinical interest.14 This is important when extrapolating the plasma glucose response to an exogenous glucose bolus.
Why the water of an isotonic fluid has such a small extracellular distribution (2.7 litre) is more difficult to explain, but can probably be attributed to the long time (2025 min) required to distribute Ringer's solution between a central (V1) and a peripheral (V2) extracellular body fluid space.15 With a high kr and half-life for glucose of only 16 min, the infused volume is either rapidly excreted or brought into the cells (V3) along with the glucose. The distribution does not occur fast enough to expand a V2.
All these comparisons and predictions are based on kinetic analyses of individual curves showing the plasma glucose and the haemoglobin-derived plasma dilution over time. As can be seen from Figure 1B, infused glucose and fluid were both effectively cleared from plasma. One may note that the simulated curve in Figure 1A did not capture the decrease of the plasma glucose below baseline during the third hour, while this rebound hypoglycaemia is incorporated in further calculations of volume kinetics.
This study also comprised a measurement of insulin resistance (M value) using a hyperinsulinaemic euglycaemic clamp. Our average M of 3.9 represents a decrease from the expected normal value of 7 mg kg1 min1,6 7 although an assessment of the degree of individual change would require that another glucose clamp was performed before surgery.2 We had expected that the CL for glucose would correlate with the M value, but this was not the case. The M value apparently cannot be used directly as a guide to how fast a glucose solution should be infused without leading to hyperglycaemia. On the other hand, the finding that CL divided by the peak insulin concentration correlated with the M value might be tested as a minimal model to estimate insulin resistance. Additional evaluation is certainly needed and also an account of the dose of administered glucose, as peak insulin, but not CL, is dose-dependent.4
Insulin resistance had several interesting implications for the postoperative period. First, bioimpedance indicated that surgery dehydrated the intracellular space from the second postoperative day onwards, and the degree of dehydration correlated with the insulin resistance (Fig. 3B). A similar dehydration by 10% was described by Stillström and colleagues,16 who also found that Ringer's solution causes more ICF loss than glucose 2.5%. This result is consistent with the fact that water accompanies the uptake of glucose into the cell, as modelled in the volume kinetic analysis. Furthermore, 3-MH excretion, which was used as an index of protein catabolism, was higher in patients with pronounced intracellular dehydration (Fig. 3D).
The succession in which these relationships develop may provide a clue to how they operate. The insulin clamp was applied on the first postoperative day when intracellular dehydration had not yet developed. This suggests that insulin resistance develops first and dehydration later. The next correlation that between dehydration and protein catabolism, is consistent with a theory by Häussinger, which holds that protein catabolism is triggered by cellular dehydration.17 18 A possible chain of events is that insulin resistance leads to intracellular dehydration, which leads to protein catabolism.
The interleukin-6 response suggests that abdominal hysterectomy inflicts a moderately severe trauma in the patients. The rise in this serum concentration is proportional to the degree of surgical trauma.19 In the first postoperative morning after abdominal hysterectomy, we obtained levels midway between those obtained after laparoscopic and open cholecystectomy.2 20 The cytokine VEGF was measured because it greatly increases vascular permeability21 and is up-regulated in response to factors involved in the inflammatory cascade, such as hypoxia, endothelin-1, and interleukin-1, while nitric oxide acts as an inhibitor.22 However, the serum concentration of VEGF was not elevated. It is plausible that the measured alterations in cellular hydration and insulin resistance reported by us would be more pronounced with a more severe trauma response to the surgery.23
Limitations of this study include the circumstance that the endogenous glucose production was estimated by a turnover kinetic model. This assumes that CL for glucose is unchanged down to a plasma glucose level of zero, which assumption has not been tested. Radioactive tracers show that endogenous glucose production 2 h after an abdominal hysterectomy is 40% lower than that reported by us for the second postoperative day,24 although this rate agrees quite well with our findings during surgery.5 Moreover, fluid intake and feeding could only be monitored on the day of surgery until the clamp was performed. The data on urinary output collected later can therefore only be used to indicate an expected normal recovery (Table 2) and we can only assume that the intake of meat, which would affect the index of protein catabolism, was small during the first postoperative day. The bioimpedance calculations could also have been slightly different using body weights measured daily. The patient with the smallest 3-MH excretion in the study was excluded from the linear regression analysis in Figure 3, where she appears as an outlier, as we suspect that the urinary collection was incomplete. This exclusion hardly changed the slopes of these relationships, but Figure 3 should still merely be regarded as explorative extensions of the present study, and further evaluation is clearly needed.
In conclusion, the kinetics of glucose 2.5% solution are essentially normal 2 days after hysterectomy with the exception of a reduced glucose clearance. Infusion therapy can be planned using nomograms, which outline the expected plasma glucose and dilution of this fluid. The study also reported intercorrelations between postoperative insulin resistance, cellular dehydration and an index of protein catabolism.
![]() |
Appendix |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() | (1) |
![]() | (2) |
Fluid volume. The expansion of the central (v1) and the peripheral (v3) body fluid spaces during infusion of 2.5% glucose solution is given by:
![]() |
Simulations. The best estimate of the model parameters (Table 1) was inserted into the mathematical solutions to the differential equations shown above, which had been programmed into the Matlab software. A time-stepping method was then used to predict the glucose concentration and the plasma dilution over time.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2 Thorell A, Nygren J, Gutniak M, et al. The metabolic response to cholecystectomy: insulin resistance after open compared with laparoscopic operation. Eur J Surg 1996; 162: 18791[ISI][Medline]
3 Sjöstrand F, Edsberg L, Hahn RG. Volume kinetics of glucose solutions given by intravenous infusion. Br J Anaesth 2001; 87: 83443
4 Sjöstrand F, Hahn RG. Validation of volume kinetic analysis of glucose 2.5% solution given by intravenous infusion. Br J Anaesth 2003; 90: 6007
5 Sjöstrand F, Hahn RG. Volume kinetics of 2.5% glucose solution during laparoscopic cholecystectomy. Br J Anaesth 2004; 92: 48592
6 deFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol 1979; 273: E21423
7 Soop M, Nygren J, Brismar K, Thorell A, Ljungqvist O. The hyperinsulinaemic-euglycaemic glucose clamp: reproducibility and metabolic effects of prolonged insulin infusion in healthy subjects. Clin Sci 2000; 98: 36774[CrossRef][ISI][Medline]
8 Johnson HL, Virk SPS, Mayclin P, Barbieri T. Predicting total body water and extracellular fluid volumes from bioelectrical measurements of the human body. J Am Coll Nutr 1992; 11: 53947[Abstract]
9 de Lorenzo A, Andreoli A, Matthie J, Withers P. Predicting body cell mass with bioimpedance by using theoretical methods. J Appl Physiol 1997; 82: 154258
10 Crowe PJ, Dennison A, Royle GT. The effect of preoperative glucose loading on postoperative nitrogen metabolism. Br J Surg 1984; 71: 6357[ISI][Medline]
11 Iapichino G, Radrizzani D, Cambisano M, et al. Kinetics of body nitrogen loss during a whole day infusion and withdrawal of glucose and insulin in injured patients. Int Care Med 1995; 21: 44751[ISI][Medline]
12 Sjölin J. Skeletal Muscle Protein Metabolism in Human Infection. Evaluation of 3-Methylhistidine as a Marker of Myofibrillar Protein Degradation. Thesis, Uppsala University, 1988
13 Ferrannini E, Smith JD, Cobelli C, et al. Effect of insulin on the distribution and disposition of glucose in man. J Clin Invest 1985; 76: 35764[ISI][Medline]
14 Radziuk J, Pye S. Quantitation of basal endogenous glucose production in Type II diabetes. Importance of the volume of distribution. Diabetologia 2002: 45: 105384[CrossRef][ISI][Medline]
15 Hahn RG. Physiological or functional fluid spaces (Letter to the Editor). Anesth Analg 2002; 95: 2512
16 Stillström A, Persson E, Vinnars E. Postoperative water and electrolyte changes in skeletal muscle: a clinical study with three different intravenous infusions. Acta Anaesthesiol Scand 198; 31: 2848
17 Häussinger D. Regulation of metabolism by changes in cellular hydration. Clin Nutr 1995; 14: 412[ISI]
18 Finn PJ, Plank LD, Clark MA, Connolly AB, Hill GL. Progressive cellular dehydration and proteolysis in critically ill patients. Lancet 1996: 347: 6546[CrossRef][ISI][Medline]
19 Cruickshank AM, Fraser WD, Burns HJG, van Damme J, Shenkin A. Response of serum interleukin-6 in patients undergoing elective surgery of varying severity. Clin Sci 1990; 79: 1615[ISI][Medline]
20 Kristiansson M, Saraste L, Soop M, Sundqvist KG, Thörne A. Diminished interleukin-6 and C-reactive protein responses to laparoscopic versus open cholecystectomy. Acta Anaesthesiol Scand 1999; 43: 14652[CrossRef][ISI][Medline]
21 Bates DO, Lodwick D, Williams B. Vascular endothelial growth factor and microvascular permeability. Microcirculation 1999; 6: 8396[CrossRef][ISI][Medline]
22 Josko J, Gwózdz B, Jedrzejowska-Szypulka H, Hendryk S. Vascular endothelial growth factor (VEGF) and its effect on angiogenesis. Med Sci Monit 2000; 6: 104752[Medline]
23 Thorell A, Efendic S, Gutniak M, Häggmark T, Ljungqvist O. Development of postoperative insulin resistance is associated with the magnitude of the operation. Eur J Surg 1993; 159: 5939[ISI][Medline]
24 Schricker T, Carli F, Schreiber M, et al. Propofol/sufentanil anesthesia suppresses the metabolic and endocrine response during, not after, lower abdominal surgery. Anesth Analg 2000; 90: 4505
|