Glucose, insulin and potassium applied as perioperative hyperinsulinaemic normoglycaemic clamp: effects on inflammatory response during coronary artery surgery

L. Visser1, C. J. Zuurbier1, F. J. Hoek2, B. C. Opmeer3, E. de Jonge4, B. A. J. M. de Mol5 and H. B. van Wezel1,*

Departments of 1 Anaesthesia, 2 Clinical Chemistry, 3 Epidemiology and Biostatistics, 4 Intensive Care and 5 Cardiothoracic Surgery, Academic Medical Center, University of Amsterdam, Meibergdreef 9, 1105 AZ, Amsterdam, The Netherlands

* Corresponding author. E-mail: H.B.vanWezel{at}amc.uva.nl

Accepted for publication July 11, 2005.


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Background. The clinical benefits of glucose–insulin–potassium (GIK) and tight glycaemic control in patients undergoing coronary artery bypass grafting (CABG) may be partly explained by an anti-inflammatory effect. We applied GIK as a hyperinsulinaemic normoglycaemic clamp for >25 h and quantified its effect on systemic inflammation in patients undergoing CABG.

Methods. Data obtained in 21 non-diabetic patients with normal left ventricular function scheduled for elective coronary artery surgery, who were randomly allocated to a control or GIK group, were analysed. In GIK patients, regular insulin was infused at a fixed rate of 0.1 IU kg–1 h–1. The infusion rate of glucose (30%) was adjusted to maintain blood glucose levels within a target range of 4.0–5.5 mmol litre–1. Plasma concentrations of interleukins 6, 8 and 10, C-reactive protein (CRP) and serum amyloid A (SAA) were measured on the day of surgery and on the first and second postoperative days (POD1 and POD2).

Results. In the GIK group hypoglycaemia (glucose <2.2 mmol litre–1) did not occur, whereas hyperglycemia (glucose >6.1 mmol litre–1) developed in 15% of all measurements. In control patients, hyperglycaemia developed in >80% of all measurements in the presence of low endogenous insulin levels. CRP and SAA levels increased in both groups, with maximum levels measured on POD2. GIK treatment significantly reduced CRP and SAA levels. Interleukin levels increased significantly in both groups following cardiopulmonary bypass, but no differences were found between the groups.

Conclusion. Hyperinsulinaemic normoglycaemic clamp is an effective method of maintaining tight glycaemic control in patients undergoing CABG and it attenuates the systemic inflammatory response in these patients. This effect may partly contribute to the reported beneficial effect of glycaemic control in patients undergoing CABG.

Keywords: complications, systemic inflammatory response syndrome ; metabolism, glucose ; metabolism, hyperglycaemia ; metabolism, insulin ; surgery, cardiovascular


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Following the studies by Taegtmeyer and colleagues,1 Van den Berghe and colleagues2 and Lazar and colleagues,3 a number of glucose–insulin–potassium (GIK) infusion strategies designed to maintain tight glycaemic control have been increasingly used to improve clinical outcome in patients undergoing coronary artery bypass grafting (CABG). However, the exact mode of action underlying this approach remains to be elucidated. GIK research initially focused on and demonstrated the ability of insulin to influence substrate flux through myocardial metabolic pathways and transmembrane signalling. Furthermore, GIK infusions were found to attenuate post-ischaemic disturbances in lipid and glucose homeostasis, both in the setting of CABG and in acute myocardial infarction (AMI).410

Recent evidence from animal studies suggests that GIK has the potential to reduce the inflammatory response.11 12 Chaudhuri and colleagues13 were the first to demonstrate the anti-inflammatory effects of insulin in patients with AMI, as reflected by a reduction in the absolute increase in post-infarct levels of C-reactive protein (CRP) and serum amyloid A (SAA). To date, anti-inflammatory effects of insulin have not been described in the setting of CABG, although these patients may theoretically benefit particularly from such an effect since CABG is known to be accompanied by a substantial systemic inflammatory response.14

Furthermore, acute hyperglycaemia frequently develops in patients undergoing CABG, usually following cardiopulmonary bypass (CPB),15 and it has been demonstrated in both rats and humans that pro-inflammatory cytokine concentrations are increased by acute hyperglycemia.16 17 We hypothesize that maintaining tight glycaemic control with insulin attenuates the systemic inflammatory response in patients undergoing CABG.

The first aim of the present study was to describe the use of a perioperative (>25 h) hyperinsulinaemic normoglycaemic clamp technique as proposed by Svedjeholm18 in 1995 and recently applied intraoperatively (<4 h) by Carvalho.19 The second goal was to quantify its effect on markers of systemic inflammation in patients with normal left ventricular function scheduled for elective CABG.


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
After approval by the local medical ethics committee and obtaining written informed consent, 34 patients with normal left ventricular function scheduled for elective CABG were enrolled in the study. Patients with diabetes mellitus, ejection fraction <45%, unstable angina pectoris or atrioventricular conduction defects were excluded. Patients taking corticosteroids or non-steroidal anti-inflammatory drugs or undergoing additional surgical procedures (e.g. valve replacement or aneurysmectomy) were also excluded.

A locally designed randomization programme was used to allocate patients to a control group or a GIK group. Patients in the control group received standard institutional perioperative care. Patients allocated to the GIK group received additional infusions of insulin and glucose.

Insulin and glucose infusions
After insertion of a central venous catheter and baseline blood sampling, hyperinsulinaemic normoglycaemic clamping was started and continued throughout the period of CPB until 24 h after release of the aortic cross-clamp. Soluble insulin (Actrapid, NovoNordisk, Copenhagen, Denmark) was infused continuously at a fixed rate of 0.1 IU·kg–1·h–1.

A separate mixture of glucose 30% (Baxter-Clintec Benelux SA, Brussels, Belgium), potassium chloride 80 mmol litre–1 and phosphate 60 mmol litre–1 was infused at a variable rate adjusted to maintain blood glucose levels within a target range of 4.0–5.5 mmol litre–1. This rather narrow range was used to create a ‘buffer’ in case of unexpected overshoot hypoglycaemia or hyperglycaemia during the clamping period. A wider range of blood glucose levels between 2.2 and 6.1 mmol litre–1 was considered normoglycaemic as described by Van den Berghe and colleagues.2

The infusion of glucose was started at a rate of 0.5 ml·kg–1·h–1. In all patients the first glucose samples were taken 15 and 30 min after the start of the glucose and insulin infusions in order to obtain the initial direction and degree of change of plasma glucose levels under hyperinsulinaemic conditions. Adjustments to the glucose 30% infusion rate were made as follows.

When plasma glucose was <3.0 mmol litre–1 the infusion rate of glucose 30% was increased by 10 ml h–1 and one or more additional 10 ml boluses of glucose 30% were administered, depending on the absolute value of plasma glucose. Plasma glucose levels were checked 5–15 min later. If required, the infusion rate of glucose 30% was increased again by 5–10 ml h–1.

When plasma glucose was between 3.0 and 4.0 mmol litre–1, the infusion rate of glucose 30% was increased by 5–10 ml h–1. The plasma glucose concentration was checked 15–30 min later.

When plasma glucose was within the target range of 4.0–5.5 mmol litre–1 the infusion rate of glucose 30% remained unchanged, unless there was a difference of >0.5 mmol litre–1 in the plasma glucose level compared with the previous measurement. In such a situation the glucose infusion rate was increased or decreased by 5–10 ml h–1. Plasma glucose levels were checked 60–120 min later.

When plasma glucose was >5.5 mmol litre–1 the infusion rate of glucose 30% was reduced by 5–10 ml h–1. Plasma glucose levels were checked 60–120 min later.

Anaesthetic management
Calcium-channel blockers and long-acting nitrates were given until the evening before surgery, and ß-adrenoceptor blocking agents were continued until the morning of surgery. Lorazepam 2–3 mg was given as premedication 2 h before surgery.

Anaesthesia was induced with sufentanil 3 µg kg–1 (Sufenta®, Janssen-Cilag, Tilburg, The Netherlands) and propofol 50–100 mg (Fresenius Kabi, Den Bosch, The Netherlands). Pancuronium bromide 0.1 mg kg–1 (Pavulon®, Organon, Oss, The Netherlands) was given for muscle relaxation. Morphine 20 mg was given as a slow bolus injection before start of surgery. Anaesthesia was maintained with a continuous infusion of propofol 2–5 mg·kg–1·h–1. The lungs were ventilated with air–oxygen (). Dexamethasone and {alpha}2-adrenoceptor agonists were not used in any of the participating patients. After induction of anaesthesia, a flow-directed pulmonary artery catheter (Edwards Lifesciences, Irvine, CA, USA) was inserted into the right internal jugular vein.

Cardiopulmonary bypass
The CPB system was primed with 1850 ml of priming liquid consisting of Haemaccel 1000 ml (Behring, Malburg, Germany), Ringer's lactate 500 ml, aprotinin 200 ml (Bayer, Leverkusen, Germany), mannitol 20% solution 100 ml and sodium bicarbonate 8.4%, 50 ml. After systemic heparinization (300 U kg–1) (Leo Pharmaceutical Products, Weesp, The Netherlands), CPB was initiated with cannulas placed in the ascending aorta and right atrium. Activated clotting time was kept >480 s. The non-pulsatile flow rate was maintained at 2.4 litre min–1 m–2 during the cooling and rewarming phases. At a temperature of 30–32°C flow rate was kept at 2.2 litre min–1 m–2. For myocardial protection, patients received high-potassium cold crystalloid cardioplegia (1000 ml) (St Thomas type II containing potassium 20 mmol, calcium 2 mmol, magnesium 16 mmol, chloride 203 mmol and procaine hydrochloride 273 mg, given at a temperature of 4°C) during aortic cross-clamping. The haematocrit during CPB was maintained >20%. After termination of CPB, heparin was antagonized with protamine hydrochloride (ICN Pharmaceuticals Holland B.V.) at a ratio of ~1:1. Residual volume from the extracorporeal circuit was infused into the patient after a cell-saving process.

After surgery, patients were admitted to the intensive care unit (ICU) and treated according to a standardized clinical protocol. Fluid administration consisted of NaCl 0.9% and hydroxyethyl starch 6% of molecular weight 200 kDa (Haes-Steril, Fresenius Kabi, Den Bosch, The Netherlands). Throughout the ICU stay, a continuous infusion of glucose 5% was given at a rate of 30 ml h–1 (1.5 g h–1) to all patients through a central venous line. All patients in both groups stayed for a relatively long period in the ICU as a result of the study design, i.e. continuation of hyperinsulinaemic normoglycaemic clamping for 24 h in the reperfusion period.

Blood samples and measurement points
Samples in the operating room and ICU were taken from the radial artery catheter. Blood samples taken on the ward were collected by venipuncture. Insulin and glucose samples were collected at the following predetermined time points: (1) before induction of anaesthesia (‘awake’); (2) after insertion of the pulmonary artery catheter, but before the start of clamping (‘baseline’); (3) 15 min before CPB, before heparinization (‘before CPB’); (4) immediately after release of the aortic cross-clamp ( ‘reperfusion’); (5) 2 h after release of the aortic cross-clamp (‘2 h reperfusion’); (6) on the day of surgery (DOS) between 6 and 8 pm; (7–10) on the first and second postoperative days (POD1 and POD2) in the morning between 6 and 8 a.m. and in the evening between 6 and 8 p.m. (POD1 6–8 a.m., POD1 6–8 p.m., POD2 6–8 a.m. and POD2 6–8 p.m., respectively). Blood for CRP and SAA measurements was collected at baseline and at time points 4–10. Cytokines were measured at baseline and at 2 h after reperfusion, on the DOS between 6 and 8 p.m. and on POD1 between 6 and 8 a.m.

Cardiac enzymes (CK-MB) were sampled on arrival in the ICU and every 3 h thereafter until it was clear that levels were past the peak level.

CRP levels were measured using an immunoturbidimetric assay. The detection limit of the assay was 0.1 mg litre–1. SAA levels were measured using an N-latex-SAA testkit (DADE Behring, Leusden, The Netherlands) and a ProSpec nephelometer (DADE Behring, Leusden, The Netherlands). The detection level of the assay was 0.695 mg litre–1. Blood samples for interleukin (IL) measurements were centrifuged at 1500 rpm for 15 min and then stored in aliquots at –80°C until required for further analysis. IL-6, IL-8 and IL-10 were measured by sandwich ELISA using a commercially available kit (BD Biosciences, The Netherlands).

Insulin was determined by a luminescence enzyme immunoassay (Immulite, Diagnostic Products Corporation, Los Angeles, CA). The intra- and interassay coefficients of variation were 3–5% and 6–9%, respectively. The detection limit was 2 mU litre–1.

The following postoperative data were obtained to estimate the presence and degree of a systemic inflammatory response syndrome (SIRS): leucocyte count measured once daily, starting preoperatively until POD5; heart rate measured from the start of reperfusion until POD5; body temperature measured from the start of reperfusion until POD5; ventilatory frequency measured hourly starting before extubation until 5 h after extubation.

Statistical analysis
Demographic data were reported as mean (SD) for continuous variables in the case of an approximately normal distribution, median (IQR) for non-normal distributed variables and as counts (proportions) for categorical variables. Two-way analyses of variance (ANOVA), by treatment group and time, for repeated measurements were performed on the outcome parameters. The main model assumption (homogeneity of error variances) was tested by Levene's test, and the same analysis was performed on rank-order-transformed variables if necessary. When a significant main effect between treatment groups was found (P<0.05), means at subsequent time points (with Bonferroni correction) were compared between treatment groups. A t-test was performed to analyse differences in areas under the curve (AUCs) for CRP and SAA. Data analyses were performed with SPSS 11.5. P-values <0.05 were considered statistically significant.


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Patients
Thirteen patients were excluded from detailed analysis for the following reasons: diabetes mellitus (n=2), unstable angina pectoris (n=3), chronic corticosteroid therapy (n=1) and insufficient insulin therapy during cardiopulmonary bypass (n=5). In addition, two more patients were excluded: one patient in the control group who developed acute ischaemic heart failure following bypass and died 2 h after arrival in the ICU, and one patient in the GIK group who developed an AMI after bypass, requiring successful coronary angioplasty shortly after arrival in the ICU.

The characteristics of the remaining 21 patients are summarized in Table 1. Patients were comparable with respect to age, weight and BMI, extent of coronary artery disease and preoperative medication. An overview of perioperative characteristics is given in Table 2. Table 3 gives an overview of fluid administration, output and balance in the operating room and in the ICU on the DOS and in the ICU on POD1.


View this table:
[in this window]
[in a new window]
 
Table 1 Patient characteristics. Data are presented as mean (range), mean (SD) or number of patients. PTCA, percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass grafting

 

View this table:
[in this window]
[in a new window]
 
Table 2 Surgical details and clinical characteristics.

 

View this table:
[in this window]
[in a new window]
 
Table 3 Fluid administration and fluid output. Data are presented as mean (SD). For the GIK group, glucose infusion is not included in fluid administration but reported separately as glucose 30% administration. It is included in the calculation of fluid balance. Fluid loss includes urinary output and blood loss. No correction was made for evaporation during open chest conditions. DOS, day of surgery; POD1, first postoperative day

 
Insulin
Plasma insulin levels are presented in Figure 1A. At baseline, there were no differences in insulin levels between the control group (47 (SD 30) pmol litre–1) and the GIK group (46 (19) pmol litre–1). In the control group, insulin levels were significantly elevated when compared with baseline on the morning of POD1 and on the evening of POD2. In the GIK group, hyperinsulinaemic normoglycaemic clamping resulted in insulin levels ranging from 600 to 800 pmol litre–1. After insulin infusion was discontinued, plasma levels quickly decreased to similar concentrations to those observed in the control group. On the evening of POD2 the insulin concentration in the GIK group was significantly higher than at baseline.



View larger version (23K):
[in this window]
[in a new window]
 
Fig 1 Time dependence of (A) plasma insulin and (B) plasma glucose concentrations in the control and GIK groups, respectively. The hyperinsulinaemic clamp period is indicated by the black horizontal bar. Data are presented as mean (SEM). *P<0.05 between groups; {dagger}P<0.05 within group; two-way ANOVA for repeated measurements with Bonferroni correction.

 
Glucose
Glucose levels are presented in Figure 1B. At baseline, plasma glucose levels were similar between the groups, at 5.6 (0.5) mmol litre–1 and 5.5 (0.7) mmol litre–1 in the control and GIK groups, respectively. In the control group, glucose levels increased significantly compared with baseline after 2 h of reperfusion until final measurements on the evening of POD2. At 2 h of reperfusion until POD1 6–8 a.m., glucose levels were significantly higher in control patients compared with the levels in the GIK group.

In the GIK group, normoglycaemia was achieved at all predetermined measurement points during the clamp period, except for levels taken before CPB (i.e. after the start of the clamp procedure). After discontinuation of the hyperinsulinaemic normoglycaemic clamp, glucose concentrations increased significantly above baseline to levels in the range of 8 mmol litre–1, which was similar to those found in the control group (Fig. 1B).

The scattergram shown in Figure 2 gives an overview of all glucose measurements in both groups. Hypoglycaemia (<2.2 mmol litre–1) did not occur at any time in either group. Hyperglycaemia (>6.1 mmol litre–1) occurred in 15% of all measurements in the GIK group and in >80% of all measurements in the control group.



View larger version (21K):
[in this window]
[in a new window]
 
Fig 2 Scattergram showing all glucose measurements in GIK (clamp) and control (non-clamp) patients. Baseline measurements were taken at t=0, after which the hyperinsulinaemic normoglycaemic clamp was started in the GIK group. Glucose measurements in both groups were continued until 24 h after release of aortic cross-clamping. The horizontal line drawn at a glucose level of 2.2 mmol litre–1 represents the lower limit of normoglycaemia used by van den Bergheand colleagues.2 The horizontal line at a glucose level of 3 mmol litre–1 represents the lower glucose limit used in the present study (see Methods). The horizontal line at a glucose level of 4 mmol litre–1 shows the lower limit of the narrow range used as a buffer to prevent overshoot hyperglycaemia and hypoglycaemia. The horizontal line at a glucose level of 5.5 mmol litre–1 shows the upper limit of this buffer range. The horizontal line at a glucose value of 6.1 mmol litre–1 shows the upper limit of normoglycaemia used in the present study.

 
Potassium
In the control group the mean serum potassium levels remained within the institutional normokalaemic range (3.5–4.5 mmol litre–1) at all times.

In the GIK group, mean serum potassium levels also remained within the normokalaemic range before and during GIK infusion. However, after discontinuation of the GIK infusions, hyperkalaemia developed in the GIK group on POD1 in the evening (mean serum potassium 4.6 mmol litre–1, range 4.1–5.9 mmol litre–1) and on POD2 in the morning (mean serum potassium 4.6 mmol litre–1, range 4.1–5.2 mmol litre–1). This was not associated with cardiac arrhythmias.

Cytokines
Table 4 summarizes the data obtained from the cytokine measurements. CABG was associated with significantly increased levels of both IL-6 and IL-8 after cardiopulmonary bypass compared with baseline. Although a trend towards lower IL-6 and IL-8 levels in the GIK compared with the control group was observed, this did not reach statistical significance.


View this table:
[in this window]
[in a new window]
 
Table 4 Perioperative plasma cytokine concentrations (pg ml–1). Data are presented as median (25th–75th percentile).

 
C-reactive protein
Data concerning CRP are presented in Figure 3. Preoperative levels averaged 3.3 (0.5) mg litre–1 for the control group and 3.9 (2.5) mg litre–1 for the GIK group. In both groups, CRP levels started to rise on POD1 and remained high until the evening of POD2. Two-way ANOVA revealed an overall lowering effect (P=0.044) on CRP levels of the hyperinsulinaemic normoglycaemic clamp compared with control conditions. In the morning of POD1 (during insulin infusion) the mean (SEM) CRP values were significantly reduced (P<0.05) in the insulin group compared with the control group (49 (8) mg litre–1 compared with 93 (3) mg litre–1 in the control and GIK groups, respectively).



View larger version (13K):
[in this window]
[in a new window]
 
Fig 3 Time dependence of CRP production. Data are presented as mean (SEM). *P<0.05 between groups; two-way ANOVA for repeated measurements with Bonferroni correction. {dagger}P<0.05 between groups at separate time points. Note that at some measurement points the error bars are too small to be visible on the scale used.

 
In the evening of POD1 (shortly after discontinuation of insulin infusion) the mean (SEM) CRP levels were also significantly reduced (P<0.05) in the insulin group compared with the control group (107 (12) mg litre–1 compared with 160 (8) mg litre–1 in the GIK and control groups, respectively).

Serum amyloid A
Figure 4 shows data for SAA. Baseline levels of SAA were 1.8 (1.0) mg litre–1 and 1.4 (1.2) mg litre–1 for the control and GIK groups, respectively. In both groups, SAA levels started to rise following reperfusion and maximum levels were measured on the evening of the POD2. Compared with control patients, overall SAA levels were lower in GIK patients (P=0.028). In the morning of POD 1 the mean (SEM) SAA values were significantly (P<0.05) reduced in the insulin group compared with the control group (179 (7) mg litre–1 compared with 295 (18) mg litre–1, respectively). In the evening of POD1 the mean (SEM) SAA levels were also significantly (P<0.05) reduced in the insulin group compared with the control group (431 (39) mg litre–1 versus 675 (36) mg litre–1, respectively).



View larger version (15K):
[in this window]
[in a new window]
 
Fig 4 Time dependence of SAA levels. Data are presented as mean (SEM). *P<0.05 between groups; two-way ANOVA for repeated measurements with Bonferroni correction. {dagger}P<0.05 between groups at separate time points. Note that at some measurement points the error bars are too small to be visible on the scale used.

 
CK-MB
CK-MB levels measured during the first 24 h postoperatively were not indicative of myocardial infarction (cut-off value 80 µg litre–1) in any of the reported patients. There were no significant differences in postoperative CK-MB levels between the control and GIK groups. All measured mean group levels were <30 µg litre–1.

Clinical parameters reflecting the presence and degree of SIRS
Of the parameters analysed to estimate the potential presence and degree of SIRS, only the leucocyte count was significantly lower (P=0.01) in the GIK group on the evening of the day of surgery (5.2 (0.3) leucocytes ml–1 compared with 12.1 (4.4) leucocytes ml–1 in the GIK group and the control group, respectively).

There were no additional significant differences in leucocyte count or in any of the other analysed indicators of the presence and degree of SIRS (heart rate, body temperature or ventilatory frequency) between the groups at any time during the study.


    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
This study confirms that hyperinsulinaemic normoglycaemic clamping is an effective procedure for maintaining tight glycaemic control perioperatively. The second major finding is that hyperinsulinaemic normoglycaemic clamping significantly reduces parameters of the systemic inflammatory response associated with CABG.

Although GIK infusions have been applied in the setting of cardiac surgery and AMI for several decades, there is still no standardized protocol available. In contrast with the widely used GIK cocktail technique, where GIK is administered as one mixture with a fixed composition and infused at a set rate, hyperinsulinaemic normoglycaemic clamping allows for adjustments in the infusion rate of glucose in response to changes in perioperative systemic insulin sensitivity and endogenous insulin and glucose production. Studies where GIK is infused as a cocktail have been hampered by hyperglycaemic and/or hypoglycaemic episodes.20 21 In the recently published CREATE-ECLA trial it was observed that GIK administered as a cocktail resulted in higher blood glucose levels than in control patients, which may have blunted the potential benefits of insulin.22

Using a different approach, i.e. intensive insulin therapy in combination with fixed high infusion rates of glucose as partial nutritional support, Van den Berghe and colleagues2 23 reported 5.2% hypoglycemia, defined as glucose levels <2.2 mmol litre–1. This seems a very low glucose value to use as the lower limit of normoglycemia (especially during routine clinical use of GIK), but it was apparently not associated with complications in the controlled study situation described.2 23 Using the criteria of van den Berghe's study we detected no hypoglycaemia in our patients receiving GIK, while hyperglycaemia (>6.1 mmol litre–1) occurred in 15% of all measurements taken during the clamp period (Fig. 2). In our control group, hypoglycaemic events were not seen either, but hyperglycaemia occurred in >80% of all measurements. Possible explanations for the high incidence of hyperglycaemia (mean peak value 8.5 mmol litre–1; see Fig. 1) in our control patients include the surprisingly low endogenous insulin concentrations in that group (in the presence of hyperglycaemia) and intolerance to the 5% glucose infusion (1.5 g of glucose per hour) which was started after arrival in the intensive care unit in all our patients. The latter suggestion is supported by the higher glucose levels (~12 mmol litre–1) found in Van den Berghe's study in ICU patients, where glucose was infused at a rate of 8 g h–1, i.e. five times higher than in our study. As indicated by Mazuski and colleagues,24 this emphasizes the potential risks of parenteral nutrition with glucose without concomitant control of blood glucose levels. The importance and potential effect on outcome of this ‘iatrogenic’ hyperglycaemia should be further investigated.

Several other factors may contribute to the occurrence of hyperglycaemia, such as increased hepatic glucose production due to elevated concentrations of cortisol, hepatic insulin resistance, catecholaminergic stimulation of glucose production, inhibition of insulin release by catecholamine stimulation of pancreatic {alpha} receptors and inhibition of insulin release through morphine-induced opening of pancreatic K+-ATP channels, analogous to inhibition of insulin secretion by diazoxide.2527 Finally, hypothalamic insulin resistance may also play a role.2830

With regard to the systemic inflammatory responses, our data demonstrate an attenuation of inflammation with hyperinsulinaemic normoglycaemic clamping in patients undergoing CABG. The release pattern of both CRP and SAA in our control patients has been observed by others.3136 The increase in CRP and SAA over time was significantly reduced in patients receiving GIK.

The production of the acute phase proteins CRP and SAA in the liver may be partly induced by IL-6 and occurs in response to tissue injury, infection and trauma.3739 We observed no GIK effect on IL-6, whereas CRP and SAA levels were significantly depressed by GIK. Mechanistically, the observed reduction in CRP and SAA levels may result from a suppressed transcriptional induction of the IL-6 responsive acute phase plasma protein genes with insulin.40 41 Although we observed an overall reduction in inflammation parameters with GIK, this reduction was mostly demonstrated during the clamp period only. Shortly after discontinuation of the hyperinsulinaemic clamp, acute phase protein levels became indistinguishable between the two groups. This rebound pattern coincides with our reported data on glucose, and may suggest that insulin treatment should be prolonged beyond 24 h in the reperfusion period.

CRP and SAA levels are elevated following AMI.42 43 It has been found that following AMI, peak CRP levels predict infarct size and outcome44 45 and the magnitude of peak levels correlates with the occurrence of heart failure.42 However, the nature of the ischaemic challenge and the amount and extent of tissue injury are not comparable between AMI and CABG, and therefore results of studies in one patient group may not apply to the other. AMI is characterized by mild local inflammation (peak CRP levels in the range of 10 mg litre–1) with a high degree of necrosis (mean top CK-MB levels in the range of 100–300 µg litre–1).13 In contrast, CABG is characterized by more severe systemic inflammation (peak CRP levels in the range of 200 mg litre–1 (Fig. 3)) with little necrosis (mean maximum CK-MB levels in the range of 30 µg litre–1). Therefore it is anticipated that the anti-inflammatory action of insulin is especially beneficial in patients with a high grade of systemic inflammation, i.e. CABG patients and patients suffering large complicated AMIs.

Studies relating acute phase protein concentrations to outcome have not been performed in patients undergoing CABG. However, it has recently been demonstrated by Bisoendial and colleagues46 that infusion of recombinant human CRP in humans leads to acute activation of inflammation as reflected by subsequent increases in IL-6, IL-8 and SAA. They concluded that CRP is not ‘just’ a non-specific acute phase respondent, but rather a mediator in the inflammatory cascade and atherothrombotic disease. Therefore a reduction of CRP levels may have the potential to affect clinical outcome in AMI and CABG. In addition, the reported significant reduction in leucocyte count on the evening of surgery by GIK in the present study offers a potential indication that GIK affects leucocyte activation and the cascade of events ultimately leading to (some degree of) SIRS, which is a condition associated with severe effects on outcome.47 However, the present study was underpowered for detecting such significant effects on clinical outcome because of the small size of the groups and the selection of low-risk patients with good left ventricular function and no comorbidity who generally do not develop postoperative complications. This type of patient was deliberately selected to reduce the possible influence of confounding factors on glucose metabolism and inflammatory response, such as the prolonged postoperative use of inotropic agents and/or vasoactive compounds including norepinephrine and differences in fluid balance between the groups (dilution).

Future studies of high-risk cardiac surgical patients (who may suffer postoperative complications as a result of their inability to control the perioperative systemic inflammatory response14) should be performed to assess the effect of hyperinsulinaemic normoglycaemic clamping on perioperative inflammatory control and its hypothetical association with reduced perioperative morbidity and mortality. This is especially important as, with increasing numbers of elderly and diabetic patients, the cohort of high-risk patients will grow during the next decades. Other fields of interest that may be beneficially affected by perioperative GIK infusion include the coagulation cascade, complement activation, neurohumoral stress responses, parameters reflecting insulin resistance (ketone bodies and lactate), the lipid profile and leucocyte function.

In conclusion, this study demonstrates that in patients undergoing CABG, GIK applied as a perioperative hyperinsulinaemic normoglycaemic clamp is an effective method of maintaining tight glycaemic control, which is associated with a reduction in the increase in systemic inflammation. We hypothesize that these effects contribute to the reported beneficial influence on outcome of tight glycaemic control with insulin in the setting of CABG, AMI and critical illness.


    Acknowledgments
 
The study was supported by The Netherlands Heart Foundation (grant no. 2001/B107).


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Taegtmeyer H, Goodwin GW, Doenst T, Frazier OH. Substrate metabolism as a determinant for postischemic functional recovery of the heart. Am J Cardiol 1997; 80: 3A–10A[CrossRef][Medline]

2 Van den Berghe G, Wouters P, Weekers F et al. Intensive insulin therapy in critically ill patients. N Engl J Med 2001; 345: 1359–67[Abstract/Free Full Text]

3 Lazar HL, Chipkin SR, Fitzgerald CA, Bao Y, Cabral H, Apstein CS. Tight glycemic control in diabetic coronary artery bypass graft patients improves perioperative outcomes and decreases recurrent ischemic events. Circulation 2004; 104: 1497–1502[CrossRef]

4 Opie LH, Bruyneel K, Owen P. Effects of glucose, insulin and potassium infusion on tissue metabolic changes within first hour of myocardial infarction in the baboon. Circulation 1975; 52: 49–57[Abstract/Free Full Text]

5 Russell RO Jr, Rogers WJ, Mantle JA, McDaniel HG, Rackley CE. Glucose–insulin–potassium, free fatty acids and acute myocardial infarction in man. Circulation 1976; 53 (Suppl): I207–9[Medline]

6 Dalby AJ, Bricknell OL, Opie LH. Effect of glucose–insulin–potassium infusions on epicardial ECG changes and on myocardial metabolic changes after coronary artery ligation in dogs. Cardiovasc Res 1981; 15: 588–98[ISI][Medline]

7 Coleman GM, Gradinac S, Taegtmeyer H, Sweeney M, Frazier OH. Efficacy of metabolic support with glucose–insulin–potassium for left ventricular pump failure after aortocoronary bypass surgery. Circulation 1989; 80: I91–6[Medline]

8 Lazar HL, Zhang X, Rivers S, Bernard S, Shemin RJ. Limiting ischemic myocardial damage using glucose–insulin–potassium solutions. Ann Thorac Surg 1995; 60: 411–16[Abstract/Free Full Text]

9 Svedjeholm R, Hakanson E, Vanhanen I. Rationale for metabolic support with amino acids and glucose–insulin–potassium (GIK) in cardiac surgery. Ann Thorac Surg 1995; 59 (Suppl): S15–22[CrossRef][ISI][Medline]

10 Diaz R, Paolasso EA, Piegas LS et al. Metabolic modulation of acute myocardial infarction. The ECLA (Estudios Cardiologicos Latinoamerica) Collaborative Group. Circulation 1998; 98: 2227–34[Abstract/Free Full Text]

11 Jeschke MG, Einspanier R, Klein D, Jauch KW. Insulin attenuates the systemic inflammatory response to thermal trauma. Mol Med 2002; 8: 443–50[ISI][Medline]

12 Jeschke MG, Klein D, Bolder U, Einspanier R. Insulin attenuates the systemic inflammatory response in endotoxemic rats. Endocrinology 2004; 145: 4084–93[Abstract/Free Full Text]

13 Chaudhuri A, Janicke D, Wilson MF et al. Anti-inflammatory and profibrinolytic effect of insulin in acute ST-segment-elevation myocardial infarction. Circulation 2004; 109: 849–54[Abstract/Free Full Text]

14 Laffey JG, Boylan JF, Cheng DC. The systemic inflammatory response to cardiac surgery: implications for the anesthesiologist. Anesthesiology 2002; 97: 215–52[CrossRef][ISI][Medline]

15 Chaney MA, Nikolov MP, Blakeman BP, Bakhos M. Attempting to maintain normoglycemia during cardiopulmonary bypass with insulin may initiate postoperative hypoglycemia. Anesth Analg 1999; 89: 1091–5[Abstract/Free Full Text]

16 Marfella R, Siniscalchi M, Esposito K et al. Effects of stress hyperglycemia on acute myocardial infarction: role of inflammatory immune process in functional cardiac outcome. Diabetes Care 2003; 26: 3129–35[Abstract/Free Full Text]

17 Esposito K, Nappo F, Marfella R et al. Inflammatory cytokine concentrations are acutely increased by hyperglycemia in humans: role of oxidative stress. Circulation 2002; 106: 2067–72[Abstract/Free Full Text]

18 Svedjeholm R, Huljebrant I, Hakanson E, Vanhanen I. Glutamate and high-dose glucose–insulin–potassium (GIK) in the treatment of severe cardiac failure after cardiac operations. Ann Thorac Surg 1995; 59: S23–30[CrossRef][ISI][Medline]

19 Carvalho G, Moore A, Qizilbash B, Lachapelle K, Schricker T. Maintenance of normoglycemia during cardiac surgery. Anesth Analg 2004; 99: 319–24[Abstract/Free Full Text]

20 Lazar HL, Chipkin S, Philippides G, Bao Y, Apstein C. Glucose–insulin–potassium solutions improve outcomes in diabetics who have coronary artery operations. Ann Thorac Surg 2000; 70: 145–50[Abstract/Free Full Text]

21 Woods AA, Taegtmeyer H. Reversal of glucose–insulin–potassium-induced hyperglycemia by aggressive insulin treatment in postoperative heart failure. An observational study. Cardiology 2004; 102: 82–8[CrossRef][ISI][Medline]

22 Mehta SR, Yusuf S, Diaz R et al. Effect of glucose–insulin–potassium infusion on mortality in patients with acute ST-segment elevation myocardial infarction: the CREATE–ECLA randomized controlled trial. JAMA 2005 26; 293: 437–46

23 Van den Berghe G, Wouters PJ, Bouillon R et al. Outcome benefit of intensive insulin therapy in the critically ill: insulin dose versus glycemic control. Crit Care Med 2003; 31: 359–66[CrossRef][ISI][Medline]

24 Mazuski JE, Bailey JA, Shapiro MJ. Intensive insulin therapy in critically ill patients. N Engl J Med 2002; 346: 1586–8[Free Full Text]

25 Ostenson CG, Cattaneo AG, Doxey JC, Efendic S. Alpha-adrenoceptors and insulin release from pancreatic islets of normal and diabetic rats. Am J Physiol 1989; 257: E439–43[ISI][Medline]

26 McPherson BC, Yao Z. Morphine mimics preconditioning via free radical signals and mitochondrial K(ATP) channels in myocytes. Circulation 2001; 103: 290–5[Abstract/Free Full Text]

27 Doyle ME, Egan JM. Pharmacological agents that directly modulate insulin secretion. Pharmacol Rev 2003; 55: 105–31[Abstract/Free Full Text]

28 Obici S, Zhang BB, Karkanias G, Rossetti L. Hypothalamic insulin signaling is required for inhibition of glucose production. Nat Med 2002; 8: 1376–82[CrossRef][ISI][Medline]

29 Fioramonti X, Lorsignol A, Taupignon A, Penicaud L. A new ATP-sensitive K+ channel-independent mechanism is involved in glucose-excited neurons of mouse arcuate nucleus. Diabetes 2004; 53: 2767–75[Abstract/Free Full Text]

30 Wang R, Liu X, Hentges ST et al. The regulation of glucose-excited neurons in the hypothalamic arcuate nucleus by glucose and feeding-relevant peptides. Diabetes 2004; 53: 1959–65[Abstract/Free Full Text]

31 Aouifi A, Piriou V, Blanc P et al. Effect of cardiopulmonary bypass on serum procalcitonin and C-reactive protein concentrations. Br J Anaesth 1999; 83: 602–7[Abstract/Free Full Text]

32 Vanek T, Brucek P, Straka Z, Klepetar J, Maly M. Development of C-reactive protein and fibronectin levels in coronary surgery patients: a comparison of on-pump, off-pump sternotomy and off-pump left anterior small thoracotomy groups. Jpn Heart J 2002; 43: 211–18[CrossRef][ISI][Medline]

33 Bruins P, te Velthuis H, Yazdanbakhsh AP et al. Activation of the complement system during and after cardiopulmonary bypass surgery: postsurgery activation involves C-reactive protein and is associated with postoperative arrhythmia. Circulation 1997; 96: 3542–8[Abstract/Free Full Text]

34 Roth-Isigkeit A, Schwarzenberger J, Borstel T et al. Perioperative cytokine release during coronary artery bypass grafting in patients of different ages. Clin Exp Immunol 1998; 114: 26–32[CrossRef][ISI][Medline]

35 Roth-Isigkeit A, Borstel TV, Seyfarth M, Schmucker P. Perioperative serum levels of tumour-necrosis-factor alpha (TNF-alpha),IL-1 beta, IL-6, IL-10 and soluble IL-2 receptor in patients undergoing cardiac surgery with cardiopulmonary bypass without and with correction for haemodilution. Clin Exp Immunol 1999; 118: 242–6[CrossRef][ISI][Medline]

36 Roth-Isigkeit A, Hasselbach L, Ocklitz E et al. Inter-individual differences in cytokine release in patients undergoing cardiac surgery with cardiopulmonary bypass. Clin Exp Immunol 2001; 125: 80–8[CrossRef][ISI][Medline]

37 Uhlar CM, Whitehead AS. Serum amyloid A, the major vertebrate acute-phase reactant. Eur J Biochem 1999; 265: 501–23[Abstract/Free Full Text]

38 Pepys MB, Hirschfield GM. C-reactive protein: a critical update. J Clin Invest 2003; 111: 1805–12[Free Full Text]

39 Castell JV, Gomez-Lechon MJ, David M, Fabra R, Trullenque R, Heinrich PC. Acute-phase response of human hepatocytes: regulation of acute-phase protein synthesis by interleukin-6. Hepatology 1990; 12: 1179–86[ISI][Medline]

40 Campos SP, Baumann H. Insulin is a prominent modulator of the cytokine-stimulated expression of acute-phase plasma protein genes. Mol Cell Biol 1992; 12: 1789–97[Abstract]

41 Seshadri V, Fox PL, Mukhopadhyay CK. Dual role of insulin in transcriptional regulation of the acute phase reactant ceruloplasmin. J Biol Chem 2002; 277: 27903–11[Abstract/Free Full Text]

42 Berton G, Cordiano R, Palmieri R, Pianca S, Pagliara V, Palatini P. C-reactive protein in acute myocardial infarction: association with heart failure. Am Heart J 2003; 145: 1094–1101[CrossRef][ISI][Medline]

43 Shainkin-Kestenbaum R, Winikoff Y, Cristal N. Serum amyloid A concentrations during the course of acute ischaemic heart disease. J Clin Pathol 1986; 39: 635–7[Abstract]

44 Ueda S, Ikeda U, Yamamoto K et al. C-reactive protein as a predictor of cardiac rupture after acute myocardial infarction. Am Heart J 1996; 131: 857–60[CrossRef][ISI][Medline]

45 Anzai T, Yoshikawa T, Shiraki H et al. C-reactive protein as a predictor of infarct expansion and cardiac rupture after a first Q-wave acute myocardial infarction. Circulation 1997; 96: 778–84[Abstract/Free Full Text]

46 Bisoendial RJ, Kastelein JP, Levels JHM et al. Activation of inflammation and coagulation after infusion of C-reactive protein in humans. Circ Res 2005; 96: 714–16[Abstract/Free Full Text]

47 Rinder CS, Fontes M, Mathew JP et al. Neutrophil CD11b upregulation during cardiopulmonary bypass is associated with postoperative renal injury. Ann Thorac Surg 2003; 75: 899–905[Abstract/Free Full Text]