Download Effects of high-dose glucose–insulin–potassium

Clinical Science (2001) 101, 37–43 (Printed in Great Britain)
Effects of high-dose glucose–insulin–potassium
on myocardial metabolism after coronary
surgery in patients with Type II diabetes
Zolta! n SZABO! *, Hans ARNQVIST†, Erik HA/ KANSON*, Lennart JORFELDT‡
and Rolf SVEDJEHOLM§
*Department of Cardiothoracic Anaesthesia, Linko$ ping Heart Centre, University Hospital, S-581 85 Linko$ ping, Sweden,
†Department of Medical Endocrinology, Linko$ ping Heart Centre, University Hospital, S-581 85 Linko$ ping, Sweden, ‡Department
of Thoracic Physiology, Karolinska Hospital, 104 01 Stockholm, Sweden, and §Department of Cardiothoracic Surgery, Linko$ ping
Heart Centre, University Hospital, S-581 85 Linko$ ping, Sweden
A
B
S
T
R
A
C
T
The effects of glucose–insulin–potassium (GIK) on cardiac metabolism have been studied
previously in non-diabetic patients after cardiac surgery. Although patients with diabetes
mellitus can be expected to benefit most from such treatment, the impact of GIK in diabetic
patients undergoing cardiac surgery remains unexplored. Therefore the present study
investigates the effects of high-dose GIK on myocardial substrate utilization after coronary
surgery in patients with Type II diabetes. A total of 20 patients with Type II diabetes undergoing
elective coronary surgery were randomly allocated to either post-operative high-dose GIK or
standard post-operative care, including insulin infusion if necessary to keep blood glucose below
10 mmol/l. Myocardial substrate utilization was studied using the coronary sinus catheter
technique. Haemodynamic state was assessed with the aid of Swan–Ganz catheters. High-dose
GIK caused a shift towards carbohydrate utilization, with significant lactate uptake throughout
the study period and significant uptake of glucose after 4 h. Arterial levels of non-esterified fatty
acids and β-hydroxybutyric acid decreased, and after 1 h no significant uptake of these substrates
was found. Increases in the cardiac index and stroke volume index were found in patients treated
with high-dose GIK. A decrease in systemic vascular resistance was found both in the control
group and in the high-dose GIK group. We conclude that high-dose GIK can be used in diabetic
patients after cardiac surgery to promote carbohydrate uptake at the expense of non-esterified
fatty acids and β-hydroxybutyric acid. This could have implications for treatment of the diabetic
heart in association with surgery and ischaemia.
INTRODUCTION
Glucose–insulin–potassium (GIK) has been re-appraised
in recent years for the treatment of the heart in association
with myocardial infarction and cardiac surgery [1–7].
The optimum doses of insulin in treatment of myocardial
infarction and of cardiac surgery appear to differ markedly. Due to neuroendocrine stress, high doses of insulin
are required to achieve maximal metabolic effects after
cardiac surgery in non-diabetic patients [8–10]. Although
patients with diabetes mellitus can be expected to benefit
most from such treatment, the impact of GIK on cardiac
Key words : coronary surgery, diabetes, glucose, β-hydroxybutyric acid, insulin, lactate, myocardial metabolism, non-esterified fatty
acids, potassium.
Abbreviations : BW, body weight ; GIK, glucose–insulin–potassium ; NEFA, non-esterified fatty acids.
Correspondence : Dr Zolta! n Szabo! (e-mail Zoltan.Szabo!lio.se).
# 2001 The Biochemical Society and the Medical Research Society
37
38
Z. Szabo! and others
metabolism in diabetic subjects remains unexplored.
Therefore the effects of high-dose GIK on myocardial
substrate utilization after elective coronary artery bypass
graft surgery in patients with Type II diabetes have been
investigated.
METHODS
Patients
A total of 20 patients with Type II diabetes undergoing
elective coronary surgery for stable angina pectoris were
studied. Exclusion criteria were a left ventricular ejection
fraction of 0.40, age 80 years, serious late complications of diabetes, liver disease, poorly controlled
diabetes or metabolic disturbance other than diabetes.
Demographic data are given in Table 1.
Clinical management
All patients were operated on before 12.00 hours. After
an overnight fast, β-blockers and calcium antagonists
were administered orally, but ACE (angiotensin-converting enzyme) inhibitors, oral anti-diabetic treatment
and insulin were withheld. The patients were premedicated intramuscularly with 8–10 mg of oxicodone
and 0.4–0.5 mg of scopolamine. Anaesthesia was induced
with thiopentone at a dose of 2–3 mg\kg body weight
(BW) and fentanyl at a dose of 30 µg\kg BW. Pancuronium bromide was used for neuromuscular blockade. Anaesthesia was maintained with fentanyl and
isoflurane. After sternotomy, 3 mg of heparin\kg BW
was given.
Cardiopulmonary bypass was conducted with a membrane oxygenator and a roller pump generating pulsatile
flow. The extracorporeal circuit was primed with crystalloid fluid containing no glucose or lactate (Ringer’s
acetate ; Braun2) and mannitol. Moderate haemodilution
(haematocrit 20–25 %) and moderate hypothermia (32–
Table 1
34 mC) were employed. A combination of antegrade and
retrograde delivery of St Thomas’ cold crystalloid cardioplegic solution was used for myocardial protection. After
the distal anastomoses were completed, the ascending
aorta was unclamped and the proximal vein anastomoses
were performed during partial aortic occlusion. Weaning
from cardiopulmonary bypass was started at a rectal
temperature of 36 mC. Heparin was neutralized with
protamine sulphate (3 mg\kg BW), with additional doses
given if the activated clotting time (measured with an
Automated Coagulation Timer ACT II ; Medtronic
Hemo Tec, Parker, CO, U.S.A.) exceeded 125 s. Immediately before closure of the sternum, 4 mg of pancuronium was given intravenously, and this was repeated
during the study period to prevent shivering. To
minimize the influence of post-operative events such as
extubation, pain, shivering and emotional stress during
the study period, the patients were ventilated normally
and sedated with a continuous infusion of midazolam (2–
6 mg\h). Analgesia during the study period was achieved
by ketobemidone infusion (1–4 mg\h) supplemented
with intermittent doses of fentanyl. Nitroglycerine
or nitroprusside was added if necessary to prevent
post-operative hypertension to a pressure greater than
150 mmHg. Ringer’s acetate was used for volume substitution. Tachycardia exceeding 90\min unrelated to
hypovolaemia was treated with intravenous boluses of
1–5 mg of metoprolol (Seloken2). Post-operative rewarming in the intensive care unit was facilitated by
radiant heat provided by a thermal ceiling. Shed mediastinal blood was routinely re-transfused after surgery.
Study protocol
The study was performed in accordance with the Helsinki
Declaration of Human Rights, and approved by the
ethics committee for medical research at Linko$ ping
University. Informed consent was obtained from each
patient.
Patient characteristics and intra-operative data
Values are meanspS.E.M. BMI, body mass index ; HbA1c, glycosylated haemoglobin
Parameter
Control (n l 10)
GIK (n l 10)
Age (years)
Weight (kg)
Length (cm)
BMI (kg/m2)
Female gender (%)
Hypertension (%)
Pre-operative long-term insulin treatment (%)
Pre-operative HbA1c (%)
Number of distal anastomoses
Cardiopulmonary bypass time (min)
Aortic cross-clamp time (min)
56p3
92p6
172p2
30.3p1.07
30
30
60
6.7p0.3
3.5p0.2
84p7
46p5
58p2
87p4
174p3
28.7p1.0
10
60
60
7.2p0.2
3.7p0.3
77p8
45p7
# 2001 The Biochemical Society and the Medical Research Society
Glucose–insulin–potassium and the diabetic heart
The patients were randomly allocated to groups
receiving either post-operative high-dose GIK treatment
(n l 10) or standard post-operative glucose control (n l
10). The study was not blinded because of the insulin
doses used. Originally three arms of the study were
considered, with one arm investigating the effect of a
GIK regime employing an insulin dose of 0.08
i.u.:h−":kg−" BW [11]. However, this arm was abandoned after the first patient, because of unacceptable
blood glucose control.
The high-dose GIK treatment has been described
previously [7]. Briefly fast-acting insulin (Actrapid
Novo2) was infused at a rate of 1 i.u.:h−":kg−" BW for
6 h. A bolus of 25 i.u. was also injected after 5 min.
A 30 % (w\v) glucose solution supplemented with
10 mmol\l magnesium and 40 mmol\l phosphate was
also infused, with the aim of keeping blood glucose
between 7 and 10 mmol\l. The average infusion rate was
83 ml\h during the study period. After stopping insulin
infusion, the glucose infusion was decreased gradually.
Potassium was infused separately.
During the study period, three of the control patients
required insulin infusion ranging from 1 to 10 i.u.\h to
keep blood glucose levels at 10 mmol\l.
Post-operatively, a coronary sinus catheter (Wilton
Webster Labs Inc., Altadena, CA, U.S.A.) was inserted
through the right internal jugular vein. The final midcoronary sinus position was confirmed by fluoroscopy
and measurement of oxygen saturation. Coronary sinus
blood flow (CF 300A Flowmeter ; Webster Labs Inc.)
was measured using the retrograde thermodilution technique. The mean of three measurements was used.
The study was started on average 3 h after release of the
aortic cross-clamp. Blood sampling from the coronary
sinus and radial artery was done in the basal state (before
starting GIK) and after 30 min and 1, 2 and 4 h.
Haemodynamic state and coronary sinus blood flow
were measured at the same time points. Samples for
glucose, lactate, glycerol, β-hydroxybutyric acid, glutamate and alanine were analysed in whole blood. Samples
for non-esterified fatty acids (NEFA) were analysed in
plasma. Details of biochemical analyses have been presented previously [12].
Myocardial fluxes of substrates were calculated as the
product of arterial–coronary-sinus blood or plasma
concentration differences and coronary sinus blood or
plasma flow, as appropriate [12]. A release of substrates
was defined as a myocardial flux value significantly less
than zero (P 0.05), whereas uptake of substrates was
defined as a myocardial flux significantly greater than
zero (P 0.05).
The oxygen consumption of the heart was estimated as
the product of the arterial–coronary-sinus blood oxygen
content difference and coronary sinus blood flow. Oxygen content was given by [B-HbiSO i(6.2i10−%)]
#
j(PO i0.01), where B-Hb represents blood level of
#
haemoglobin (in g\l), SO is oxygen saturation expressed
#
as a percentage, and PO is oxygen tension (in kPa).
#
Statistical methods
Statistical analyses were performed with a computerized
statistical package (Statistica 5.1 ; StatSoft, Inc., Tulsa,
OK, U.S.A.). ANOVA for repeated measures employing
the Tukey honest significant difference test was used to
analyse inter-group differences (between the GIK and
control groups) after the basal state and to analyse
changes occurring over time. The Mann–Whitney U-test
adjusted for repeated measures with the Bonferroni
correction was used to determine statistical differences
from zero. Statistical significance was defined as P
0.05. Data are presented as meanspS.E.M.
RESULTS
Clinical outcome
Pre-operative and intra-operative data are presented in
Table 1. There was no mortality. The median stay in the
intensive care unit was 1 day for both groups.
Haemodynamic results
Haemodynamic state was stable in both groups during
the study period (Table 2). In the high-dose GIK
group an increase in cardiac index occurred, from
2.1p0.1 litres:min−":m−# in the basal state to 2.9p0.2
litres:min−":m−# after 4 h (P 0.05). The stroke volume
index increased in the high-dose GIK group, and a
reduction in systemic vascular resistance during the study
period was observed in both groups. No significant
change in left ventricular stroke work index occurred in
either group. Inter-group differences (effect of GIK)
reached statistical significance only for the cardiac index,
but a borderline P value was found for the stroke volume
index.
Metabolic findings
Arterial levels of NEFA, β-hydroxybutyric acid and
glycerol were significantly lower in the group receiving
high-dose GIK treatment than in the control group
(Table 3). In the control group, the heart extracted NEFA
and β-hydroxybutyric acid, but no uptake of carbohydrate substrates was observed (Table 4). In the highdose GIK group, uptake of lactate occurred throughout
the study period, and at 4 h significant uptake of glucose
was also observed (Table 4). No uptake of NEFA was
found during high-dose GIK treatment, and uptake of βhydroxybutyric acid was observed only during the first
1 h of the study period. Inter-group differences caused
by GIK reached statistical significance only for the higher
uptake of lactate and the lower uptake of βhydroxybutyric acid (Table 4). The difference in lactate
# 2001 The Biochemical Society and the Medical Research Society
39
40
Z. Szabo! and others
Table 2
Haemodynamic results
Values are meanspS.E.M. HR, heart rate ; CI, cardiac index ; BSA, body surface area ; MAP, mean arterial pressure ; CVP, central venous pressure ; PCWP, pulmonary
capillary wedge pressure ; SVRI, systemic vascular resistance index ; LVSWI, left ventricular stroke work index ; SI, stroke index ; S VO2, mixed venous oxygen saturation.
Statistically significant inter-group differences due to GIK (ANOVA repeated-measures design) are indicated in the right-hand column ; ns, not significant. Post hoc
differences are indicated as follows : **P 0.01, ***P 0.001 denote significant differences compared with the basal state (0 h) ; †P 0.05, ††P 0.01,
†††P 0.001 denote significant inter-group differences.
Variable
Group
Basal
30 min
1h
2h
4h
HR (beats/min)
Control
GIK
Control
GIK
Control
GIK
Control
GIK
Control
GIK
Control
GIK
Control
GIK
Control
GIK
Control
GIK
88p5
77p4
2.1p0.1
2.1p0.1
79p4
88p5
6p1
6p1
10p2
10p1
2907p238
3086p210
27.0p2.6
33.4p2.1
24.9p1.4
28.4p2.0
63.8p1.8
68.5p2.2
87p5
83p5
2.1p0.1
2.3p0.1
81p4
80p4
6p1
6p1
10p2
10p1
2832p254
2621p149
27.6p2.1
30.5p2.3
25.6p1.3
28.2p2.1
64.5p1.6
67.3p1.6
89p5
83p5
2.1p0.1
2.5p0.2
75p4
74p4**
8p1
7p1
9p1
11p2
2521p229
2247p178***
25.5p2.0
31.3p3.4
25.3p1.4
31.0p3.0††
59.2p2.2
67.2p1.6††
89p4
83p5
2.1p0.1
2.4p0.1
73p3
71p3***
7p1
7p1
9p1
11p1
2512p141
2137p126***
25.3p2.3
29.2p2
25.0p1.7
30.2p1.8†
60.7p2.5
67.7p1.5†††
84p4
82p4
2.3p0.1
2.9p0.2***
72p3
68p3***
8p1
7p1
9p1
11p1
2260p174**
1667p112***
27.6p1.7
33.5p2.2
28.0p1.7
36.6p2.1***†††
61.1p2.2
70.0p1.5†††
CI (litres:min−1:m−2 BSA)
MAP (mmHg)
CVP (mmHg)
PCWP (mmHg)
SVRI (dyne:s:cm−5:m−2 BSA)
LVSWI (g:beat−1:m−2 BSA)
SI (ml:beat−1:m−2 BSA)
S VO2 (%)
Table 3
P (ANOVA)
ns
0.017
ns
ns
ns
ns
ns
ns (0.06)
ns
Arterial concentrations of substrates and insulin
Values are meanspS.E.M. Statistically significant inter-group differences due to GIK (ANOVA repeated-measures design) are indicated in the right-hand column ; ns, not
significant. Post hoc differences are indicated as follows : **P 0.01, ***P 0.001 denote significant differences compared with the basal state (0 h) ; ††P
0.01, †††P 0.001 denote significant inter-group differences.
Metabolite
Glucose (mmol/l)
Group
Control
GIK
Lactate (mmol/l)
Control
GIK
NEFA (mmol/l)
Control
GIK
Glycerol (µmol/l)
Control
GIK
β-Hydroxybutyric acid (µmol/l) Control
GIK
Glutamate (µmol/l)
Control
GIK
Alanine (µmol/l)
Control
GIK
Insulin (pmol/l)
Control
GIK
Basal
30 min
7.16p1.32 7.01p1.43
6.14p0.47 8.52p0.53***
1.15p0.40 1.01p0.38
1.22p0.22 1.33p0.18
0.67p0.17 0.81p0.25
0.63p0.84 0.49p0.67†††
91p44 136p58
63p10
42p6†††
150p58 229p63
100p57
79p52
168p9
174p7
154p11 160p15
243p21 224p22
306p41 304p50
94p53
32p13
84p28 12088p2559***†††
1h
2h
7.48p1.60
7.54p1.36
8.16p0.70**
7.37p0.70
1.15p0.47
1.08p0.44
1.65p0.11**
1.42p0.10
0.77p0.25
–
0.34p0.59**†††
–
134p75
121p77
40p4†††
40p3†††
321p79
450p84
28p14
11p3††
165p6
159p7
141p7
133p9
234p21
215p14
285p35
265p21
34p15
44p12
10301p782***††† 11125p1464***†††
uptake was explained by a markedly higher rate of
extraction during high-dose GIK treatment (P l 0.008).
In the high-dose GIK group, the average fractional
extraction of lactate increased from 15.9 % in the basal
# 2001 The Biochemical Society and the Medical Research Society
4h
7.95p1.21
6.97p0.43
1.12p0.51
1.36p0.65
0.69p0.24
0.27p0.70***†††
79p32
37p3
395p117
14p3††
154p7
135p10
234p25
251p20
62p15
14007p1509***†††
P (ANOVA)
ns
ns
0.0007
0.007
0.005
ns
ns
0.0001
state to 33.8 % at 1 h (P 0.05), and thereafter ranged
between 25 % and 32 %. In the control group, the average
fractional extraction rate of lactate ranged from a peak of
4.9 % in the basal state to 0.6 % after 4 h. Myocardial
Glucose–insulin–potassium and the diabetic heart
Table 4
Myocardial flux of substrates and oxygen consumption
Values are meanspS.E.M. CS, coronary sinus ; MV O2, myocardial oxygen consumption. Statistically significant inter-group differences due to GIK (ANOVA repeatedmeasures design) are indicated in the right-hand column ; ns, not significant. Post hoc differences are indicated as follows : *P 0.05, **P 0.01 denote significant
differences compared with the basal state (0 h) ; ††P 0.01 denotes significant inter-group differences. Statistically significant uptake or release of substrates is
indicated by : ‡P 0.05.
Parameter
Flux ( µmol/min)
Glucose
Lactate
NEFA
Glycerol
β-Hydroxybutyric acid
Glutamate
Alanine
CS blood flow (ml/min)
MV O2 ( µmol/min)
Group
Basal state
Control
GIK
Control
GIK
Control
GIK
Control
GIK
Control
GIK
Control
GIK
Control
GIK
35p23
23p11
16p13
k6p16
k10p5
20p36
23.9p13.5
11.4p8.6
9.2p6.7
23.4p7.2‡ 31.9p12‡
74.0p12**††‡
11.8p3.6‡
6.7p1.4‡
4.3p1.3‡
8.5p1.7‡
3.9p1.8
3.4p2.5
2.4p2.4
0.8p0.6
0.4p0.3
k2.0p2.6
0.8p0.3‡ k1.3p0.5‡
2.4p5.8
5.9p3.7‡ 11.1p4.0‡
3.7p1.5‡
3.9p3.1
2.1p1.7‡
1.9p2.4‡
2.3p1.2‡
2.5p0.9‡
3.3p1.3
4.3p2.4‡
2.8p0.4‡
1.2p5.8
0.4p1.5 k4.1p1.1
k2.0p3.0 k1.2p3.0 k6.1p3.1
Control
GIK
Control
GIK
141p38
122p19
485p147‡
519p95‡
30 min
92p11
126p27
337p47‡
449p82‡
uptake of glutamate was observed in both groups, but
significant release of alanine was only recorded once in
the control group.
Plasma insulin levels in the high-dose GIK and control
groups are given in Table 3. The average glucose infusion
rate in the high-dose GIK group ranged between 4.7p0.2
and 5.1p0.6 mg:min−":kg−" BW.
DISCUSSION
To our knowledge, this is the first study to investigate the
effects of GIK on cardiac metabolism in patients with
diabetes. The main finding was that high-dose GIK
promoted the myocardial uptake of carbohydrate substrates at the expense of NEFA and β-hydroxybutyric
acid early after coronary surgery in patients with Type II
diabetes.
Myocardial substrate utilization after cardiac surgery
has been studied previously in non-diabetic patients
[10,12,13]. The metabolic state in these patients was
characterized by elevated blood glucose and plasma
NEFA, reliance on NEFA for myocardial energy uptake
and restricted uptake of carbohydrates. It was also
demonstrated that GIK could enhance the myocardial
uptake of carbohydrate substrates ; however, due to
neuroendocrine stress, insulin doses of up to 1
i.u.:h−":kg−" BW were required to achieve maximal
1h
111p21
149p27
381p69‡
518p61‡
2h
4h
k19p28
27p14
0.2p6.0
38.9p19‡
–
–
0.5p0.6
k0.0p0.2
13.9p5.7‡
0.5p0.3
1.9p0.8‡
1.4p1.0‡
k5.7p2.6‡
k0.4p1.3
5p20
62p18‡
3.7p6
50.5p10.5‡
7.9p2.7
2.0p1.2*
0.6p0.6
0.3p0.3
9.0p3.4‡
0.6p0.3
2.8p0.5‡
2.7p0.7‡
k6.7p1.7
k1.5p2.7
100p12
112p25
388p60‡
424p82‡
122p15
115p22
386p46‡
415p72‡
P (ANOVA)
ns
0.003
ns
ns
0.004
ns
ns
ns
ns
metabolic effects [8–10,13]. This dosage was used in the
present study, leading to an approximate 150-fold increase in plasma insulin. Insulin levels of this magnitude
are known to be associated with vasodilatation [14].
Our haemodynamic results are in keeping with this, and
consequently an increased cardiac index was found in the
high-dose GIK group.
In the basal state, satisfactory haemodynamic and
metabolic recovery had occurred in both groups. The
control patients were treated with insulin, if necessary, to
keep blood glucose below 10 mmol\l. In spite of this, no
uptake of carbohydrate substrates was observed, and
NEFA and β-hydroxybutyric acid were the major
substrates taken up by the heart. In contrast, high-dose
GIK enhanced the myocardial uptake of carbohydrate
substrates, in particular lactate, at the expense of NEFA
and β-hydroxybutyric acid. The effect of high-dose GIK
on myocardial glucose uptake was not as evident as
that on myocardial lactate uptake. Insufficient statistical
power due to study size and the relationship between
analytical precision and fractional extraction rate could
partly explain this discrepancy. Also, the possibilities of
an attenuated effect of GIK on myocardial glucose uptake
in diabetic patients, or a predominant direct or indirect
activation of pyruvate dehydrogenase underlying the
action of insulin (discussed below), under these circumstances have to be considered. However, despite
conservative statistical assessment, significant uptake of
# 2001 The Biochemical Society and the Medical Research Society
41
42
Z. Szabo! and others
glucose was observed at the end of the study period in the
high-dose GIK group, and the uptake of glucose and
lactate would have sufficed to explain the entire oxygen
consumption, assuming that all substrates taken up by
the heart were metabolized. These metabolic findings
could have clinical implications, as energy derived from
carbohydrates has been claimed to be important for the
preservation of mechanical function, structure and ionic
balance in association with myocardial ischaemia [8,15].
Furthermore, during high-dose GIK treatment, plasma
levels of NEFA decreased and no myocardial uptake of
NEFA was observed. Although NEFA normally constitute the major source of energy for the heart, a state
with elevated NEFA levels and myocardial substrate
uptake restricted to predominantly NEFA represents an
unfavourable metabolic situation for the ischaemic and
post-ischaemic heart, because of increased oxygen expenditure and the accumulation of toxic metabolites
[8,15]. Thus the present study shows that myocardial
substrate utilization can be modified in the desired
direction even in diabetic patients after cardiac surgery.
Also, the results of the present study concerning systemic
glucose uptake could have implications for critically ill
diabetic patients in settings other than cardiac surgery.
By use of high-dose GIK, substantial amounts of energy
in the form of glucose could be provided coincident with
the maintenance of acceptable blood glucose control. The
need for large doses of insulin to achieve this effect was
illustrated by the fact that one arm of our study originally
designed to investigate the impact of a GIK regime
employing an insulin dose of 0.08 i.u.:h−":kg−" BW [11]
was abandoned because of unacceptable blood glucose
control.
The methods employed here do not elucidate the
precise mechanisms behind the action of high-dose GIK
on myocardial substrate uptake. Normally, myocardial
uptake of carbohydrate substrates is not insulindependent. In our study high-dose GIK caused marked
decreases in arterial levels of NEFA and β-hydroxybutyric acid, which may have indirectly enhanced the
uptake of carbohydrates and the activity of pyruvate
dehydrogenase [8,15]. Certainly, the observed impact on
myocardial lactate uptake in the absence of a concomitant
increase in alanine release is compatible with enhanced
pyruvate dehydrogenase activity. Furthermore, insulin
can enhance the uptake of carbohydrates directly by
stimulating glucose transporters, and possibly by intracellular stimulation of pyruvate dehydrogenase
[15,16].
Despite encouraging early results with GIK for treatment of acute myocardial infarction in the 1960s and
1970s, it was abandoned due to inconclusive trials. In
retrospect, this was done without sufficient statistical
power [3]. A meta-analysis including all properly
randomized placebo-controlled trials on GIK demonstrated a significant decrease in mortality in acute
# 2001 The Biochemical Society and the Medical Research Society
myocardial infarction [3], and this was later supported by
the ECLA study [2]. With regard to diabetic subjects,
two clinical studies have demonstrated encouraging
results with GIK for the treatment of myocardial
infarction and in association with cardiac surgery [5,6].
These studies employed substantially lower doses of
insulin than in the present study, and the metabolic
impact was not investigated. Although it is appreciated
that effects other than purely metabolic ones may play a
role, it is conceivable that insulin doses sufficient to
achieve maximal metabolic effects could enhance the
efficacy of this treatment. Further studies are warranted
in order to clarify these issues, and to investigate potential
clinical benefits or hazards associated with high-dose
GIK in clinical practice.
ACKNOWLEDGMENTS
We thank Mats Fredriksson (Department of Occupational and Environmental Medicine, Linko$ ping University) for expert statistical advice. This study was
supported by grants from The Swedish Heart Lung
Foundation, The Swedish Medical Research Council
(Project no 04139), Stina och Birger Johanssons stiftelse,
Svenska
Diabetes
Fo$ rbundets
Forskningsfond,
O$ stergo$ tlands La$ ns Landsting and the Linko$ ping Heart
Centre.
REFERENCES
1
Apstein, C. S. and Taegtmeyer, H. (1997) Glucoseinsulin-potassium in acute myocardial infarction : the time
has come for a large, prospective trial. Circulation 96,
1074–1077
2 Diaz, R., Paolasso, E. A., Piegas, L. S. et al. (1998)
Metabolic modulation of acute myocardial infarction. The
ECLA (Estudios Cardiologicos Latinoamerica)
Collaborative Group. Circulation 98, 2227–2234
3 Fath-Ordoubadi, F. and Beatt, K. J. (1997) Glucoseinsulin-potassium therapy for treatment of acute
myocardial infarction : an overview of randomized
placebo-controlled trials. Circulation 96, 1152–1156
4 Lazar, H. L., Philippides, G., Fitzgerald, C. et al. (1997)
Glucose-insulin-potassium solutions enhance recovery
after urgent coronary artery bypass grafting. J. Thorac.
Cardiovasc. Surg. 113, 354–362
5 Lazar, H. L., Chipkin, S., Philippides, G., Bao, Y. and
Apstein, C. (2000) Glucose-insulin-potassium solutions
improve outcomes in diabetics who have coronary artery
operations. Ann. Thorac. Surg. 70, 145–150
6 Malmberg, K., Ryden, L., Efendic, S. et al. (1995)
Randomized trial of insulin-glucose infusion followed by
subcutaneous insulin treatment in diabetic patients with
acute myocardial infarction (DIGAMI study) : effects on
mortality at 1 year. J. Am. Coll. Cardiol. 26, 57–65
7 Svedjeholm, R., Huljebrant, I., Ha/ kanson, E. and
Vanhanen, I. (1995) Glutamate and high-dose glucoseinsulin-potassium (GIK) in the treatment of severe
cardiac failure after cardiac operations. Ann. Thorac.
Surg. 59, S23–S30
8 Svedjeholm, R., Ha/ kanson, E. and Vanhanen, I. (1995)
Rationale for metabolic support with amino acids and
glucose-insulin-potassium (GIK) in cardiac surgery. Ann.
Thorac. Surg. 59, S15–S22
Glucose–insulin–potassium and the diabetic heart
9
10
11
12
Nilsson, F. (1987) Insulin and glucose infusion in the
early postoperative phase of cardiac surgery. Effects on
systemic carbohydrate and lipid metabolism. Medical
Dissertation, University of Gothenburg, pp. 7–49
Svensson, S., Svedjeholm, R., Ekroth, R. et al. (1990)
Trauma metabolism and the heart. Uptake of substrates
and effects of insulin early after cardiac operations.
J. Thorac. Cardiovasc. Surg. 99, 1063–1073
Gradinac, S., Coleman, G. M., Taegtmeyer, H., Sweeney,
M. S. and Frazier, O. H. (1989) Improved cardiac
function with glucose-insulin-potassium after
aortocoronary bypass grafting. Ann. Thorac. Surg. 48,
484–489
Svedjeholm, R., Vanhanen, I., Hakanson, E.,
Joachimsson, P. O., Jorfeldt, L. and Nilsson, L. (1996)
Metabolic and hemodynamic effects of intravenous
glutamate infusion early after coronary operations. J.
Thorac. Cardiovasc. Surg. 112, 1468–1477
13
Svedjeholm, R., Hallhagen, S., Ekroth, R., Joachimsson,
P. O. and Ronquist, G. (1991) Dopamine and high-dose
insulin infusion (glucose-insulin-potassium) after a
cardiac operation. Effects on myocardial metabolism.
Ann. Thorac. Surg. 51, 262–270
14 Svensson, S. E., Berglin, W.-O. E., Ekroth, R., Milocco,
I., Nilsson, F. and William-Olsson, G. (1984)
Haemodynamic effects of a single large dose of insulin in
open heart surgery. Cardiovasc. Res. 18, 697–701
15 Stanley, W. C., Lopaschuk, G. D., Hall, J. L. and
McCormack, J. G. (1997) Regulation of myocardial
carbohydrate metabolism under normal and ischaemic
conditions. Potential for pharmacological interventions.
Cardiovasc. Res. 33, 243–257
16 Rao, V., Merante, F., Weisel, R. D. et al. (1998) Insulin
stimulates pyruvate dehydrogenase and protects human
ventricular cardiomyocytes from simulated ischemia.
J. Thorac. Cardiovasc. Surg. 116, 485–494
Received 2 January 2001/12 February 2001; accepted 28 March 2001
# 2001 The Biochemical Society and the Medical Research Society
43