Metabolic interventions in acute myocardial infarction

Metabolic
interventions
in
acute
myocardial
infarction
Metabolic interventions in acute myocardial infarction
The study described in this thesis was supported by a grant of the Netherlands Heart
Foundation (99.028).
Financial support by the Netherlands Heart Foundation for the publication of this thesis is
gratefully acknowledged.
The Rijksuniversiteit Groningen is gratefully acknowledged for the support of this thesis.
Isala series nummer 52
CIP-GEGEVENS KONINKLIJKE BIOBLIOTHEEK, DEN HAAG
van der Horst, J.C.C.
Metabolic interventions in acute myocardial infarction.
Proefschrift Groningen. Met literatuur-opgave en samenvatting in het Nederlands.
ISBN: 90-74991-51-3
© Copyright 2005 J.C.C. van der Horst
All rights are reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted, in any form or by any means – electronic, mechanical, photocopy,
recording, or otherwise – without the prior written permission of the author.
Cover
Suikerhart, Anneke Bloema, The Factory II, Beilen, The Netherlands
Lay-out
Rémon Kortman & Iwan van der Horst
Printed by
Visible Edge BV
2
Metabolic interventions in acute myocardial infarction
Proefschrift
ter verkrijging van het doctoraat in de
Medische Wetenschappen
aan de Rijksuniversiteit Groningen
op gezag van de
Rector Magnificus, dr. F. Zwarts,
in het openbaar te verdedigen op
maandag 31 januari 2005
om 16.15 uur
door
Johannes Cornelis Clemens van der Horst
geboren op 9 oktober 1971
te Wateringen
3
Metabolic interventions in acute myocardial infarction
Promotores
Prof. Dr. F. Zijlstra
Prof. Dr. R.O.B. Gans
Copromotores
Dr. H.J.G. Bilo
Dr. M.J. de Boer
Beoordelingscommissie
Prof. Dr. G. van den Berghe
Prof. Dr. D.J. van Veldhuisen
Prof. Dr. B.W.O. Wolffenbuttel
4
5
Metabolic interventions in acute myocardial infarction
Contents
Chapter 1
General introduction and aims of the thesis
Chapter 2
Results of the Glucose-Insulin-Potassium Study (GIPS)
2.1
9
Glucose-insulin-potassium infusion in patients treated with primary
29
angioplasty for acute myocardial infarction (GIPS): a randomized trial
2.2
The impact of glucose-insulin-potassium infusion in acute myocardial
47
infarction on infarct size and left ventricular ejection fraction
2.3
Preservation of myocardial function by glucose-insulin-potassium
63
infusion as adjunct to primary angioplasty
2.4
Electrocardiographic evidence of reperfusion after glucose-insulin-
67
potassium infusion in patients treated with primary angioplasty for ST
segment elevation myocardial infarction
2.5
Glucose and potassium derangements by glucose-insulin-potassium
85
infusion related to 1-year mortality in acute myocardial infarction
2.6
Long-term survival after glucose-insulin-potassium infusion and primary
99
angioplasty in ST segment elevation myocardial infarction: the
randomized Glucose Insulin Potassium Study (GIPS)
Chapter 3
Glucose-insulin-potassium
solutions
as
adjunctive
therapy
in 113
myocardial infarction: current evidence and potential mechanisms
Chapter 4
4.1
Admission hyperglycemia and outcome in acute myocardial infarction
Prognostic value of admission glucose in non-diabetic patients with 125
myocardial infarction
4.2
Hyperglycemia
reduces
myocardial
reperfusion
after
primary 139
percutaneous coronary intervention for ST elevation myocardial
infarction
4.3
Hyperglycemia and no reflow phenomenon in acute myocardial 149
infarction
6
Contents
Chapter 5
5.1
Time-averaged hyperglycemia and outcome in critically ill patients
Glucose at admission or time-averaged glucose during hospital stay:
153
predict major adverse cardiac events in patients with acute myocardial
infarction?
5.2
The hyperglycemic index is a tool to assess glucose control: a
167
retrospective study
5.3
Relationship of baseline glucose and mortality during medical critical
179
illness?
Chapter 6
Glucose-insulin-potassium
and
reperfusion
in
acute
myocardial
183
infarction: rationale and design of the Glucose-Insulin-Potassium
Study-2 (GIPS-2)
Chapter 7
Summary and future perspectives
199
Chapter 8
Dutch summary
213
List of abbreviations
221
Acknowledgements
223
7
Metabolic interventions in acute myocardial infarction
8
Introduction and aims of thesis
Chapter 1
General introduction and aims of the thesis
Diagnosis and treatment of ST segment elevation myocardial infarction
9
Metabolic interventions in acute myocardial infarction
Introduction
Treatment strategies for acute ST segment elevation myocardial infarction (MI) have
evolved over the last 25 years. In the 1950s and 1960s, it was debated whether coronary
thrombosis was the cause or the consequence of ST segment elevation MI. In the 1960s
and 1970s, treatment of ST segment elevation MI patients consisted of bed rest for up to a
month. Mortality was reduced with the emergence of Coronary Care Units and treatment
of the arrhythmias. Landmark studies by DeWood in the early 1980s showed that
occlusion of the coronary artery was the critical event leading to ST segment elevation
MI.1 Reperfusion therapy became the cornerstone of acute treatment for ST segment
elevation MI. Preferentially, acute coronary reperfusion is nowadays accomplished (1)
mechanically by primary percutaneous coronary intervention (PCI), previously called
primary transluminal coronary angioplasty (PTCA) with or without stenting or (2)
pharmacologically with intravenous fibrinolytic therapy. Recent evidence suggests that
apart from improved PCI techniques, adjunctive use of platelet glycoprotein IIb/IIIa
receptor blockers and metabolic interventions, such as glucose-insulin-potassium (GIK)
infusion may enhance procedural success and improve clinical outcome. Together these
developments have stimulated renewed efforts to determine the optimal therapeutic
strategy for patients with ST segment elevation MI.
Pathophysiology
The first papers on the clinical diagnosis of MI date from the early 20th century. Obrastzow
and Stracheschenko in 1910 and Herrick in 1912 described the features of a sudden
obstruction of a coronary artery.2;3
Myocardial infarction is the consequence of disruption, fissuring or hemorrhage of a
vulnerable coronary artery plaque, complicated by various degrees of intraluminal
thrombosis, embolization, and subtotal or total obstruction to perfusion. The residual
antegrade or collateral flow, and the volume and location of affected myocardium
determine the characteristics of the clinical presentation. Patients with complete occlusion
may manifest ST segment elevation MI, if the lesion occludes an artery supplying a
substantial volume of the myocardium. A similar occlusion in the presence of extensive
collaterals may present as MI without ST segment elevation. ST segment elevation MI is
diagnosed by the presence of a clinical syndrome of new-onset ischemia with either rest
pain or a crescendo pattern of ischemic pain on minimal exertion, and elevated enzymatic
markers together with electrocardiographic evidence of acute ischemic injury. The
predictive accuracy of ST segment elevation for a final diagnosis of MI is very high. The
definition of MI proposed by the European Society of Cardiology (ESC)4, the American
10
Introduction and aims of thesis
College of Cardiology (ACC)5, and the American Heart Association (AHA)5 requires a
typical clinical syndrome plus a rise and fall in creatine kinase-MB (CK-MB) or tropinin.
Predictors of death in patients with myocardial ischemia and
infarction
A number of prognostic models have been developed in populations of patients with ST
segment elevation MI to determine the predictive value of several characteristics to
predict outcome.6-9 In the multinational, observational Global Registry of Acute Coronary
Events (GRACE) the value of baseline clinical and demographic characteristics on hospital
mortality was predicted in an unselected population of patients with acute coronary
syndrome.10 Killip class, age, blood pressure, cardiac arrest, positive enzymatic markers,
serum creatinine level, ST segment deviation, and heart rate contained most of the
prognostic information. Although acute coronary syndromes are usually categorized
according to the presence or the absence of ST segment elevation at the time of
presentation, this variable did not appear to be important for determining the risk of death
after accounting for the presence of ST segment deviation. The risk of major
cardiovascular complications and death is dependent on acute and pre-existing risk
factors (table 1).
Table 1. High risk factors and markers of outcome
•
Age
•
Previous cardiovascular disease
•
ST segment deviation
•
Rhythm disturbances (bundle branch block, ventricular fibrillation, cardiac arrest)
•
Signs of heart failure (Killip class ≥2)
•
Glucose derangement (elevated glycosylated hemoglobin or diabetes mellitus)
•
Renal dysfunction (raised serum creatinine, raised blood urea nitrogen, reduced
•
Elevated inflammatory markers (C reactive protein, interleukin-6)
•
Extent of coronary artery disease on angiography (multi-vessel disease)
•
Large enzymatic infarct size
creatinine clearance, micro-albuminuria)
General measures
The underlying principles of the treatment of ST segment elevation MI patients are to
provide relief of ischemia and pain. In the presence of heart failure or shock, assisted
11
Metabolic interventions in acute myocardial infarction
ventilation with positive end expiratory pressures may be required.4;5 Reperfusion of
critically ischemic myocardium is crucial in those with acute ST elevation or (new-onset)
left-bundle branch block or posterior MI. Hemodynamic support may be necessary in
patients with hypotension or cardiogenic shock, i.e., patients with Killip class 4 at
admission. If so indicated, this supports intra-aortic balloon pumping to stabilize the
patient for PCI. Specific measures may be required to control hypertension so as to
reduce myocardial wall stress, and to treat acute heart failure.
Reperfusion therapy
Early and effective reperfusion therapy is the cornerstone of treatment for acute ST
segment elevation MI. Restoration of antegrade flow in the occluded artery can be
achieved by PCI and/or fibrinolytic therapy. To evaluate the coronary blood flow in
patients with the Thrombolysis in Myocardial Infarction (TIMI) flow is determined. The
restoration of TIMI grade 3 flow, i.e., optimal flow, is achieved in approximately 9 out of
10 patients treated with PCI as compared to 5-7 out of 10 patients treated with fibrinolytic
therapy.11 Early restoration of antegrade flow is related to diminished enzymatic infarct
size, preserved left ventricular function, prevention of recurrent infarction, and short-term
as well as long-term survival benefit.
Fibrinolytic therapy
The first two large-scale, placebo-controlled, randomized trials that compared fibrinolytic
therapy with placebo demonstrated dramatic benefits for streptokinase. These two trials
showed that streptokinase reduced 30-day mortality rates from 13% to 10.7% (P<0.001)12
and 12% to 9.2% (P<0.001)13. These results also showed the synergistic benefits of
antiplatelet agents and fibrinolytic therapy, since 30-day mortality was reduced to a
greater extent by the combination of aspirin and streptokinase (13.2% versus 8.0%,
P<0.001) than by aspirin alone (11.8% versus 9.4%, P<0.001).12;13 Subsequent long-term
data from both trials confirmed that the mortality benefit with streptokinase persisted for
at least 10 years. Similar trials with other fibrinolytic agents have shown complementary
findings, and a systematic overview of all trials randomizing more than 1000 patients to
fibrinolytic therapy or placebo (total N=58600) reported a significant reduction in 30-day
mortality with fibrinolytic therapy compared to placebo (11.5% versus 9.6%).14 Prehospital
administration of fibrinolytic therapy has been proposed as a means of further reducing
time to reperfusion. Several studies have analyzed the potential advantages of prehospital
fibrinolytics. A recent meta-analysis (N=6434) that combined data from six randomized
trials showed a 17% reduction in mortality with prehospital fibrinolytics versus hospitaladministered fibrinolytic therapy (P=0.03).15 Fibrinolytic therapy is limited by various
12
Introduction and aims of thesis
safety and efficacy issues, such as contraindications and intracranial hemorrhage.
However, until now the combination of glycoprotein IIb/IIIa receptor blockers or specific
anti-thrombin (bivalirudin) and fibrinolytic therapy have not shown to improve survival,
and may be associated with increased bleeding.
Primary percutaneous coronary intervention
Primary percutaneous coronary intervention (PCI) achieves reperfusion through
mechanical recanalization of the infarct-related artery rather than through lysis of the
coronary thrombus with fibrinolytic therapy. Dotter and Judkins were the first to propose
the concept of reperfusion of the coronary artery by a catheter technique.16 In 1977,
Grüntzig performed the first percutaneous balloon angioplasty.17 One year later, he and
his colleagues reported that over a period of 18 months angioplasty had been used in 50
patients.18 The technique was successful in 32 patients, reducing the stenosis from a
mean of 84% to 34%. Percutaneous balloon angioplasty without the use of fibrinolytic
therapy for acute MI was first described by Hartzler and colleagues in 1983.19 The first
stents were implanted in 1985.20
The number of trials comparing primary PCI with fibrinolytic therapy has been relatively
small; however, these trials found an advantage for primary PCI. The first three
randomized clinical trials comparing primary PCI with various fibrinolytic regimens were
published in 1993. Zijlstra and colleagues found in 142 patients that compared to
streptokinase primary PCI was associated with a lower incidence of the combined endpoint of recurrent infarction or angina, death, stroke, reocclusion, and heart failure (19%
versus 47% P=0.001).21 Grines and colleagues found that primary PCI resulted in a lower
rate of nonfatal reinfarction and death compared to tissue-type plasminogen activator
(5.1% versus 12%, P=0.02) with a trend towards reduced overall mortality (2.6% versus
6.5%, P=0.06).22 Gibbons and colleagues investigated in 108 patients the effect on
myocardial salvage by technetium-99m-sestamibi and could not detect any improvement
with primary PCI when compared to tissue-type plasminogen activator.23 These
pioneering trials were too small to determine the magnitude of the impact of mechanical
reperfusion on mortality. Subsequently, Weaver and colleagues performed a metaanalysis incorporating data from the 10 available early trials that compared primary PCI
with fibrinolytic therapy. Primary PCI was associated with a significantly lower rate of 30day mortality (4.4% versus 6.5%, P=0.02), as well as with a significantly lower rate of the
combined end-point of death or nonfatal reinfarction (7.2% versus 11.9%, P<0.001).
Furthermore, Zijlstra and colleagues evaluated the 5-year results in patients randomly
assigned to primary PCI versus streptokinase and showed a significant reduction in
mortality in the primary PCI group (13.4% versus 23.9%, P=0.01).24
13
Metabolic interventions in acute myocardial infarction
Current data are available from 23 published randomized controlled trials with 7739
patients.25 Eight trials compared primary PCI to streptokinase (N=1837), and 15 primary
PCI with fibrin-specific agents (N=5902). Of the 3867 patients randomly assigned to
fibrinolytic therapy, most (76%, N=2939) received a fibrin-specific agent (tissue-type
plasminogen activator). Stents were used in twelve and platelet glycoprotein IIb/IIIa
receptor blockers in eight trials. The included trials differ in many respects, including
patient sample size, type of fibrinolytic therapy, and whether the stents were used with or
without platelet glycoprotein IIb/IIIa receptor blockers. Primary PCI was found to be more
effective than fibrinolytic therapy in reducing short-term and long-term major adverse
clinical events, including death. It was also associated with better clinical outcomes,
regardless of the type of fibrinolytic agent used or whether the patient required emergent
transfer to another hospital for primary PCI. Thus, primary PCI reduces mortality for
patients with ST segment elevation MI even in high-risk patients. In the ‘Should we
emergenly revascularize Occluded Coronaries for cardiogenic shocK’ (SHOCK) trial it was
observed that at 1 year patients treated with PCI or coronary artery bypass grafting had a
lower mortality than patients receiving fibrinolytic therapy (34% versus 47%, P=0.03).
Postprocedural TIMI grade 3 flow rates in primary PCI trials have been as high as 73% to
97%, surpassing the 50% to 60% early TIMI grade 3 flow rates demonstrated with
fibrinolytic therapy. Infarct-related artery reocclusion is also much less frequent after
mechanical recanalization. Complications of PCI include the need for vascular repair and
development of acute renal failure (approximately 3%).26;27
Additional treatment
To salvage viable myocardium by re-establishing coronary blood flow with the rapid use
of reperfusion strategies is an essential first step. However, reperfusion of ischemic
myocardium carries with it an inherent risk. Paradoxically, the process of reperfusion itself
can result in myocyte death. This phenomenon is termed reperfusion injury. Mitochondria
play a key role in determining cell fate during exposure to stress. Their role during
ischemia/reperfusion is particularly critical due to the conditions that promote both
apoptosis by the mitochondrial pathway and necrosis by irreversible damage to
mitochondria in association with mitochondrial permeability transition. Mitochondrial
permeability transition is caused by the opening of permeability transition pores in the
inner mitochondrial membrane, leading to matrix swelling, outer membrane rupture,
release of apoptotic signaling molecules such as cytochrome c from the intermembrane
space, and irreversible injury to the mitochondria. During ischemia, factors such as
intracellular calcium accumulation, fatty acid accumulation, and reactive oxygen species
progressively
14
increase
mitochondrial
susceptibility
to
mitochondrial
permeability
Introduction and aims of thesis
transition, increasing the likelihood that mitochondrial permeability transition will occur on
reperfusion. Since functional cardiac recovery ultimately depends on mitochondrial
recovery, cardioprotection by ischemic and pharmacological preconditioning needs to
involve the prevention of mitochondrial permeability transition. Experimental studies in
animals suggest that it is possible to limit the amount of myocardial damage during
ischemia and the early reperfusion periods. A variety of pharmacological approaches to
prevention of injury (including vasodilators, adhesion molecule blockers, and receptor
blockers of complement fractions) has been investigated. One of these potential additional
treatments
is
metabolic
intervention.
28
dichloroacetate, L-carnitine
Drugs
such
as
ranolazine,
trimetazidine,
and glucose-insulin-potassium (GIK) infusion and glucagon-
like peptide29 have mechanisms of action distinct from traditional anti-ischemic drugs.30
These agents work by shifting myocardial energy metabolism away from free fatty acids
(FFA) toward glucose as a source of fuel. Since these agents are well tolerated and do not
affect heart rate or blood pressure, they conceivably could supplement traditional antiischemic drug therapy with little risk.
Glucose-insulin-potassium infusion in myocardial ischemia
In the timespan of almost a century, a large amount of experimental evidence has been
accumulated
that
underlines
the
importance
of
glucose
metabolism
during
ischemia/reperfusion of the heart. As early as 1912, Goulston suggested that treatment
with glucose could be beneficial in several heart diseases.31 The first experimental results
on the mechanical effects of insulin and glucose in the isolated heart were made by
Visscher and Muller in 1926.32 In 1935, Evans and colleagues showed that in the ischemic
myocardium the uptake of glucose is increased.33 Almost 30 years later, Sodi-Pallares and
colleagues suggested that metabolic interference during myocardial ischemia with GIK
infusion decreased electrocardiographic signs of ischemia.34 They also showed that GIK
infusion resulted in a lower occurrence of arrhythmias.34 They attributed this effect mainly
to the influx of potassium in ischemic cardiomyocytes.35 In order to further stimulate
potassium transport into the cell, insulin was administered.36 Consequently, the rise of
intercellular calcium is curtailed by the influx of potassium and so the incidence of
arrhythmias is reduced.37-40 However, systemic infusion of insulin stimulates the uptake of
glucose in many celltypes41, which may result in hypoglycemic episodes.42 Consequently,
it is not possible to administer potassium and insulin in high concentrations without
adding glucose. Interventions in the glucose metabolism in the clinical arena, whether or
not used to correct acute hyperglycemia, encompass three potentially effective elements:
glucose, insulin and potassium.
15
Metabolic interventions in acute myocardial infarction
Basic mechanism of GIK protection
Ischemia induces many changes in the heart’s metabolism, including shifts from aerobic
fatty acid metabolism to anaerobic glycolysis, which provides energy for critical
myocardial cellular function (figure 1).43-45
Figure 1. The main changes that occur in peripheral and myocardial metabolism during
the development of acute myocardial ischemia. [adapted from Oliver MF. Am J Med
2002;112:305-311]
CoA = coenzyme A; FFA = free fatty acid; TG = triglyceride.
During most clinical ischemic syndromes, including acute MI, residual or collateral blood
flow usually provides at least 10% of the normal level of perfusion to a significant portion
of the ischemic myocardium. This small amount of perfusion provides such a level of
oxygen delivery that oxidative ATP synthesis from both glucose and free fatty acids
greatly exceeds ATP synthesis from anaerobic glycolysis.46;47 Thus, a mixture of aerobic
and anaerobic metabolism occurs. With progressively severe ischemia, anaerobic
glycolysis becomes a progressively more important source of energy for a limited amount
of ATP, which may or may not suffice to support the most essential cellular functions.
Glycogen is rapidly mobilized during ischemia, and reduced glycogen concentrations
impair force development, calcium release, and contractile function.48 Key intermediates
of the Krebs cycle are also depleted, which may impair energy transfer.49
The ischemia-mediated increase in glucose utilization is characterized by enhanced rates
of exogenous glucose uptake in vivo, which requires greater rates of transport across the
16
Introduction and aims of thesis
plasma membrane.50;51 Of the seven reported members of the facilitative glucose
transporter family, GLUT-4 and GLUT-1 are the primary forms expressed in adult
mammalian heart muscle.52 During low-flow ischemia the expression of GLUT-4 is
doubled.53 Insulin increases the translocation of GLUT-4 via a pathway mediated by
phosphatidylinositol 3-kinase (PI3-K). During ischemia and hypoxia GLUT-4 translocation is
stimulated through a PI3-K-independent pathway. AMP-activated protein kinase plays a
role in the translocation during ischemia.54
Table 2. Mechanisms of GIK infusion during myocardial ischemia62-64
•
The yield of moles of ATP per mole of oxygen consumed is 11 percent higher for
•
Anti FFA effects
glucose than for FFA oxidation
o
Decrease of circulating FFA levels and myocardial FFA uptake
o
Increased esterification of intracellular FFA by increasing the supply of
alpha-glycerophosphate
•
Increased rate of ATP synthesis via anaerobic glycolysis with consequent
beneficial effects
o
Increased concentrations of phosphocreatine and ATP
o
Blunting of an increase in inorganic phosphate and ADP concentrations
o
Increased free energy yield from ATP hydrolysis
•
Increased myocardial glycogen
•
Improved sodium and calcium homeostasis
•
Increased tolerance to rises in intracellular calcium
•
Replenishment of citric acid cycle intermediates by anaplerosis
•
Increase of glucose and decrease of FFA oxidation during reperfusion
•
Activation of cell survival signalling pathways such as Akt
Opie proposed the glucose hypothesis: the enhanced uptake and metabolism of glucose
delays cellular damage.55;56 Glucose utilization during ischemia prevents the breakdown
of glycogen stores and leads to increased net intramyocardial glycogen synthesis, thereby
limiting enzymatic infarct size and contracture.47;57 Two studies showed that infusion of
GIK in isolated hearts with regional ischemia resulted in decreased infarct size, increased
high-energy phosphate levels, and improved ventricular function.58;59 Acute MI patients
treated with GIK also showed better stress tolerance and ischemic threshold
improvement, analyzed with technetium-99m-tetrofosmin-gated SPECT.60 The improved
energetic profile results in improved systolic and diastolic function during ischemia and
reperfusion, as well as coronary vasodilatation.47 Also, glucose uptake has been shown to
reduce hypoxia-induced apoptosis in cultured neonatal rat cardiac myocytes.61 The
17
Metabolic interventions in acute myocardial infarction
observed benefits of GIK infusion have been attributed to a number of mechanisms,
which are summarized in table 2 and in part discussed.
Glucose
The potential positive effects of glucose are based on the fact that glucose is a source of
energy for cells.65 The uptake of glucose into the cell is influenced by insulin, although
there is also an insulin-independent transport of glucose.66 It has been observed that
AMP-activated protein kinase is responsible for activation of glucose uptake and
glycolysis during low-flow ischemia.67 During MI, low-flow perfusion of the ischemic area
is often present, making the administration glucose useful.48 In an experimental study it
was observed that a high glucose concentration stimulated translocation of GLUT-4.68 It
was already know that the combination of glucose and insulin is more effective than either
one alone in stimulating glycolysis under ischemic conditions.47
Administering
glucose
can
prevent
insulin-induced
hypoglycemia.
When
the
hypoglycemic episodes persist or when they are severe (<2.7 mmol/L), convulsions, brain
damage and even brain death may occur.69 Hypoglycemia is also related to myocardial
ischemia.70 It has been shown that the prevention of hypoglycemia can prevent an
increase in enzymatic infarct size.71 When hypoglycemia occurs, the contraregulating
hormones are activated and result in an increased release of glucose.72 Increased glucose
release requires, in particular, an increase in glucagon and adrenaline. During MI the
levels of glucagon, adrenaline and aldosteron among others are already elevated.
Insulin
The potential positive effects of insulin during stress situations are multifarious.73;74 First,
insulin is involved in the uptake of glucose in tissues, including the myocardium, mainly
through GLUT-4 and partly through GLUT-1.75 However, the exact amount of uptake
during ischemia is disputable.76 Besides the stimulation of glucose uptake and the
stimulation of glycogen synthesis, insulin is also involved in gene transcription,
expression of various metabolic enzymes, the activation of various pathways with
mitogenic activity, and even fatty acid uptake. The insulin receptor substrate 1 has an
important role in realizing this pleiotrope.77 Both insulin-like growth factor (IGF) 1 and
insulin inhibit postischemic apoptosis, energetic failure and damage to cardiac tissue, in
vitro and in animals, possibly through reduced oxidative stress.78;79 Insulin increases the
bio-availability of IGF1 and suppresses hepatic synthesis of IGF1-binding protein, which
binds and limits free-circulating IGF1.80;81
Insulin stimulates protein synthesis in skeletal muscles and inhibits intracellular protein
breakdown in cardiac tissue.82-84 The preservation of myocardial cells by inhibiting
apoptosis and reduced destruction of proteins could be the reason that contractibility is
18
Introduction and aims of thesis
preserved.85 An additional factor is that insulin potentiates ischemic preconditioning;
however, this has not been proven irrefutably in clinical trials.86;87
Insulin has an anti-inflammatory effect that is caused by a reduction in oxidative stress.88
First, insulin reduces the pro-inflammatory effects of hyperglycemia. Insulin suppresses
the production of tumor necrosis factor α in macrophages, leucocytes and endothelium.89
Furthermore, insulin blocks the upregulation of the endothelial cell adhesion molecule
induced by hyperglycemia.90;91 Also, insulin inhibits macrophage-inhibitory factor, and
potentiates endothelial nitric oxide synthase and endothelin release.92 In a clinical study it
appeared that the oxidative stress that occurs in myocardial ischemia and during
reperfusion by primary PCI could not be suppressed by insulin.93
Insulin is shown to influence the adhesion of leucocytes and blood platelets during an
acute MI.94 A study with 48 patients with type 2 diabetes mellitus showed that intensive
treatment with insulin during an acute ischemic event (i.e., acute MI or unstable angina
pectoris) improves the fibrinolytic profile.95 The administration of insulin with the aid of an
algorithm led to lower mean blood-glucose values (6.9 mmol/L versus 11.4 mmol/L) and
to lower concentrations of tissue plasminogen activator, plasminogen activator inhibitor-1
and fibrinogen.
Insulin has vasodilating capacities in blood vessels of muscle tissue.96;97 Vasodilatation
during myocardial ischemia has been observed both in patients with and without diabetes
mellitus.97-99 This is advantageous, since it opens up the blood vessels in the myocardium,
enabling more glucose to reach the cells and preventing the accumulation of metabolites
that are toxic and cause mitochondrial damage and alterations in membrane ion
channels.98;100 Vasodilatation occurs, among others, by stimulating the production of nitric
oxide in the endothelium and by antagonizing endothelin, a potent vasoconstrictor.101;102
Even a small increase in myocardial blood flow can significantly reduce myocardial
ischemia.103
Conversely, administering insulin can potentially lead to an adverse reaction.104 It might
lead to enhanced polarisability of the cell membrane, stimulation of the sympathetic
nervous system, and inhibition of the parasympathic nervous system. In contrast to the
above-mentioned results, it was also found that insulin induced oxidative stress.105 The
effect of insulin on nuclear factor-κB (NF-κB) remains unclear.88;106 In an experimental
setting insulin in high amounts has inhibitory effects on intermediates involved in the
activation by platelet-derived growth factor.107
Potassium
It
appears
that
hypokaliemia
during
stress
situations
is
disadvantageous.108;109
Hypokaliemia frequently occurs in trauma patients and has been associated with a worse
score on the Glasgow Coma Score (GCS).110 Moreover, hypokaliemia has been associated
19
Metabolic interventions in acute myocardial infarction
with muscle necrosis and paralysis. In patients with myocardial ischemia, hypokaliemia
increases the risk of ventricular arrhythmias and acute cardiac arrest.111 Restoring the
potassium concentration in the cell and in serum may be accompanied by a decrease in
the incidence and severity of arrhythmias.112 However, when the effect of GIK infusion on
QT-time was analyzed, no effect was found.113 Consequently, administration of potassium
in stress situations is mandatory to patients who present with hypokaliemia as well as to
patients treated with insulin in order to prevent hypokaliemia. Moreover, hypokaliemia
appears to suppress the secretion of insulin, and in this way stimulates a (continued) state
of hyperglycemia.114
This thesis
This thesis is a new branch on the large tree of studies on optimal therapeutic strategy for
and understanding of ST segment elevation MI. In 1989, the Zwolle Myocardial Infarction
Study Group performed its first study of comparing PCI with streptokinase. Thereafter
studies on the effect of primary PCI, stenting, intra-aortic balloon counterpulsation,
prehospital treatment with heparin and glycoprotein IIb/IIIa receptor blockers have been
reported. Over the last years more research has been done to investigate the causative
mechanisms behind favorable and unfavorable outcome after treatment with primary PCI.
Currently, special emphasis is given to the effect of glucose derangements and patients
with diabetes mellitus. The concept to add a metabolic intervention to the treatment
strategy of primary PCI was formulated in 1997.
The purpose of this thesis is to investigate the effect of metabolic interventions and
disturbances on the clinical outcome and myocardial function in ST segment elevation MI
patients. Therefore, we have investigated whether GIK infusion in adjunction to primary
PCI reduces 30-day (Chapter 2.1) and 3-year mortality (Chapter 2.6) in ST segment
elevation MI patients. We also investigated the effects of GIK infusion on myocardial
infarct size (Chapter 2.2), left ventricular function (Chapters 2.2 and 2.3) and ST segment
elevation resolution (Chapter 2.4). Furthermore, we investigated the metabolic
derangements induced by GIK infusion, and the impact of metabolic derangements on
clinical outcome (Chapter 2.5). With the results of the GIPS we performed a new analysis
on all available results of GIK infusion on 30-day mortality (Chapter 3).
In the second part of this thesis we report our studies on the relation between
hyperglycemia and outcome. Based on a large body of evidence it is known that
hyperglycemia at admission is related to mortality. We were able to analyze the effect of
admission hyperglycemia on myocardial function (Chapter 4.1). The predictive value of
20
Introduction and aims of thesis
admission glucose is not strong and we hypothesized that persistent hyperglycemia in
both critically ill patients admitted to a Coronary Care Unit (Chapter 5.1) and an Intensive
Care Unit (Chapter 5.2 and 5.3) could be a better determinant for unfavorable outcome.
Based on the results found in the above-mentioned studies, we wrote the protocol of the
GIPS-2 a multi-center trial on the effect of GIK infusion in ST segment elevation MI
patients without signs of heart-failure and eligible for reperfusion therapy (Chapter 6).
Finally, this thesis purposes to give future directions for the implementation of metabolic
interventions in critically ill patients (Chapter 7).
References
1.
DeWood MA, Spores J, Notske R et al. Prevalence of total coronary occlusion during the
early hours of transmural myocardial infarction. N Engl J Med 1980;303:897-902.
2.
Obrastzow WP, Straschesko ND. [Zur kenntnis der thrombose der koranarterien des
herzens]. Z Klin Med 1912;71:116-32.
3.
Herrick JB. Clinical features of sudden obstruction of the coronary arteries. J Am Med
Assoc 1912;59:2015-20.
4.
Van de Werf F, Ardissino D, Betriu A et al. Management of acute myocardial infarction in
patients presenting with ST segment elevation. The Task Force on the Management of
Acute Myocardial Infarction of the European Society of Cardiology. Eur Heart J 2003;24:2866.
5.
Ryan TJ, Antman EM, Brooks NH et al. 1999 update: ACC/AHA Guidelines for the
Management of Patients With Acute Myocardial Infarction: Executive Summary and
Recommendations: A report of the American College of Cardiology/American Heart
Association Task Force on Practice Guidelines (Committee on Management of Acute
Myocardial Infarction). Circulation 1999;100:1016-30.
6.
Lee KL, Woodlief LH, Topol EJ et al. Predictors of 30-day mortality in the era of reperfusion
for acute myocardial infarction. Results from an international trial of 41,021 patients.
GUSTO-I Investigators. Circulation 1995;91:1659-68.
7.
Morrow DA, Antman EM, Charlesworth A et al. TIMI risk score for ST-elevation myocardial
infarction: A convenient, bedside, clinical score for risk assessment at presentation: An
intravenous nPA for treatment of infarcting myocardium early II trial substudy. Circulation
2000;102:2031-37.
8.
Morrow DA, Antman EM, Giugliano RP et al. A simple risk index for rapid initial triage of
patients with ST-elevation myocardial infarction: an InTIME II substudy. Lancet
2001;358:1571-75.
9.
Krumholz HM, Chen J, Wang Y, Radford MJ, Chen YT, Marciniak TA. Comparing AMI
mortality among hospitals in patients 65 years of age and older: evaluating methods of risk
adjustment. Circulation 1999;99:2986-92.
10.
Eagle KA, Lim MJ, Dabbous OH et al. A validated prediction model for all forms of acute
coronary syndrome: estimating the risk of 6-month postdischarge death in an international
registry. JAMA 2004;291:2727-33.
21
Metabolic interventions in acute myocardial infarction
11.
de Boer MJ, Reiber JH, Suryapranata H, van den Brand MJ, Hoorntje JC, Zijlstra F.
Angiographic findings and catheterization laboratory events in patients with primary
coronary angioplasty or streptokinase therapy for acute myocardial infarction. Eur Heart J
1995;16:1347-55.
12.
Effectiveness of intravenous thrombolytic treatment in acute myocardial infarction. Gruppo
Italiano per lo Studio della Streptochinasi nell'Infarto Miocardico (GISSI) Lancet 1986;1:397402.
13.
Randomised trial of intravenous streptokinase, oral aspirin, both, or neither among 17,187
cases of suspected acute myocardial infarction: ISIS-2. ISIS-2 (Second International Study
of Infarct Survival) Collaborative Group Lancet 1988;2:349-60.
14.
Indications for fibrinolytic therapy in suspected acute myocardial infarction: collaborative
overview of early mortality and major morbidity results from all randomised trials of more
than 1000 patients. Fibrinolytic Therapy Trialists' (FTT) Collaborative Group Lancet
1994;343:311-22.
15.
Morrison LJ, Verbeek PR, McDonald AC, Sawadsky BV, Cook DJ. Mortality and prehospital
16.
Dotter CT, Judkins MP. Transluminal treatment of arteriosclerotic obstruction. Description
17.
Gruntzig A. Transluminal dilatation of coronary-artery stenosis. Lancet 1978;1:263.
18.
Gruntzig AR, Senning A, Siegenthaler WE. Nonoperative dilatation of coronary-artery
thrombolysis for acute myocardial infarction: A meta-analysis. JAMA 2000;283:2686-92.
of a new technic and a preliminary report of its application. Circulation 1964;30:654-70.
stenosis: percutaneous transluminal coronary angioplasty. N Engl J Med 1979;301:61-68.
19.
Hartzler GO, Rutherford BD, McConahay DR et al. Percutaneous transluminal coronary
angioplasty with and without thrombolytic therapy for treatment of acute myocardial
infarction. Am Heart J 1983;106:965-73.
20.
Sigwart U, Puel J, Mirkovitch V, Joffre F, Kappenberger L. Intravascular stents to prevent
21.
Zijlstra F, de Boer MJ, Hoorntje JC, Reiffers S, Reiber JH, Suryapranata H. A comparison of
occlusion and restenosis after transluminal angioplasty. N Engl J Med 1987;316:701-6.
immediate coronary angioplasty with intravenous streptokinase in acute myocardial
infarction. N Engl J Med 1993;328:680-684.
22.
Grines CL, Browne KF, Marco J et al. A comparison of immediate angioplasty with
thrombolytic therapy for acute myocardial infarction. The Primary Angioplasty in
Myocardial Infarction Study Group. N Engl J Med 1993;328:673-79.
23.
Gibbons RJ, Holmes DR, Reeder GS, Bailey KR, Hopfenspirger MR, Gersh BJ. Immediate
angioplasty compared with the administration of a thrombolytic agent followed by
conservative treatment for myocardial infarction. The Mayo Coronary Care Unit and
Catheterization Laboratory Groups. N Engl J Med 1993;328:685-91.
24.
Zijlstra F, Hoorntje JC, de Boer MJ et al. Long-term benefit of primary angioplasty as
compared with thrombolytic therapy for acute myocardial infarction. N Engl J Med
1999;341:1413-19.
25.
Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenous thrombolytic
therapy for acute myocardial infarction: a quantitative review of 23 randomised trials.
Lancet 2003;361:13-20.
22
Introduction and aims of thesis
26.
Lindsay J, Apple S, Pinnow EE et al. Percutaneous coronary intervention-associated
nephropathy foreshadows increased risk of late adverse events in patients with normal
baseline serum creatinine. Catheter Cardiovasc Interv 2003;59:338-43.
27.
Rihal CS, Textor SC, Grill DE et al. Incidence and prognostic importance of acute renal
28.
Wolff AA, Rotmensch HH, Stanley WC, Ferrari R. Metabolic approaches to the treatment of
failure after percutaneous coronary intervention. Circulation 2002;105:2259-64.
ischemic heart disease: the clinicians' perspective. Heart Fail Rev 2002;7:187-203.
29.
Huisamen B, Genade S, Lochner A. Glucagon-like peptide-1 (GLP-1) protects the heart
30.
Opie LH. The metabolic vicious cycle in heart failure. Lancet 2004;364:1733-34.
31.
Goulston A. West Indian cane sugar in treatment of certain forms of heart diseases. British
32.
Visscher MB, Muller EA. The influence of insulin upon the mammalian heart. J Physiol
against ischaemia by activating glycoclysis. Cardiovasc J S Afr 2004;15:S15.
Medical Journal 1912;2:693-95.
1926;62:341-48.
33.
Evans CL, Grande F, Musu FJ. The glucose and lactate consumption of the dog's heart.
34.
Sodi-Pallares D, Testelli MR, Fishleder BL et al. Effects of an intravenous infusion of a
Quart J Exptl Physiol 1935;24:347-52.
potassium-glucose-insulin solution on the electrocardiographic signs of myocardial
infarction. A preliminary clinical report. Am J Cardiol 1962;9:166-81.
35.
Sodi-Pallares D, Bisteni A, Medrano GA, Testelli MR, de Micheli A. The polarizing treatment
of acute myocardial infarction. Possibility of its use in other cardiovascular conditions. Dis
Chest 1963;43:424-32.
36.
Zierler K, Rogus EM, Scherer RW, Wu FS. Insulin action on membrane potential and
37.
Aulbach F, Simm A, Maier S et al. Insulin stimulates the L-type Ca2+ current in rat cardiac
38.
Danielson KS, DeWeese JA, Mahoney EB. Evaluation of dextrose, insulin, and potassium
glucose uptake: effects of high potassium. Am J Physiol 1985;249:E17-E25.
myocytes. Cardiovasc Res 1999;42:113-20.
on ventricular irritability in acute myocardial infarction. J Thorac Cardiovasc Surg
1970;60:653-60.
39.
Harris AS, Toth LA, Tan EH. Arrhythmic and antiarrhythmic effects of sodium, potassium,
and calcium salts and of glucose injected into coronary arteries of infarcted and normal
hearts. Circ Res 1958;6:570-579.
40.
Schonekess BO, Brindley PG, Lopaschuk GD. Calcium regulation of glycolysis, glucose
oxidation, and fatty acid oxidation in the aerobic and ischemic heart. Can J Physiol
Pharmacol 1995;73:1632-40.
41.
Ferrannini E, Santoro D, Bonadonna R, Natali A, Parodi O, Camici PG. Metabolic and
42.
Sands MJ, Jr., McDonough MT, Neubauer SJ, Lemole GM, Spann JF. Hypoglycemia
hemodynamic effects of insulin on human hearts. Am J Physiol 1993;264:E308-E315.
complicating the use of solution of glucose, insulin and potassium. Chest 1975;67:363-65.
43.
Apstein CS. Increased glycolytic substrate protection improves ischemic cardiac
44.
Opie LH. Metabolic response during impending myocardial infarction. I. Relevance of
dysfunction and reduces injury. Am Heart J 2000;139:S107-S114.
studies of glucose and fatty acid metabolism in animals. Circulation 1972;45:483-90.
23
Metabolic interventions in acute myocardial infarction
45.
Whitmer JT, Idell-Wenger JA, Rovetto MJ, Neely JR. Control of fatty acid metabolism in
46.
Cave AC, Ingwall JS, Friedrich J et al. ATP synthesis during low-flow ischemia: influence of
47.
Eberli FR, Weinberg EO, Grice WN, Horowitz GL, Apstein CS. Protective effect of increased
ischemic and hypoxic hearts. J Biol Chem 1978;253:4305-9.
increased glycolytic substrate. Circulation 2000;101:2090-2096.
glycolytic substrate against systolic and diastolic dysfunction and increased coronary
resistance from prolonged global underperfusion and reperfusion in isolated rabbit hearts
perfused with erythrocyte suspensions. Circ Res 1991;68:466-81.
48.
Chin ER, Allen DG. Effects of reduced muscle glycogen concentration on force, Ca2+
release and contractile protein function in intact mouse skeletal muscle. J Physiol 1997;498
( Pt 1):17-29.
49.
Taegtmeyer H, Goodwin GW, Doenst T, Frazier OH. Substrate metabolism as a determinant
50.
Schwaiger M, Neese RA, Araujo L et al. Sustained nonoxidative glucose utilization and
51.
Guth BD, Wisneski JA, Neese RA et al. Myocardial lactate release during ischemia in swine.
for postischemic functional recovery of the heart. Am J Cardiol 1997;80:3A-10A.
depletion of glycogen in reperfused canine myocardium. J Am Coll Cardiol 1989;13:745-54.
Relation to regional blood flow. Circulation 1990;81:1948-58.
52.
Stephens JM, Pilch PF. The metabolic regulation and vesicular transport of GLUT4, the
53.
Young LH, Renfu Y, Russell R et al. Low-flow ischemia leads to translocation of canine heart
major insulin-responsive glucose transporter. Endocr Rev 1995;16:529-46.
GLUT-4 and GLUT-1 glucose transporters to the sarcolemma in vivo. Circulation
1997;95:415-22.
54.
Russell RR, III, Bergeron R, Shulman GI, Young LH. Translocation of myocardial GLUT-4 and
increased glucose uptake through activation of AMPK by AICAR. Am J Physiol
1999;277:H643-H649.
55.
Opie LH. Hypothesis: Glycolytic rates control cell viability in ischemia. J Appl Cardiol
56.
Opie LH. The glucose hypothesis: its relation to acute myocardial ischemia. J Mol Cell
57.
Owen P, Dennis S, Opie LH. Glucose flux rate regulates onset of ischemic contracture in
58.
Heng MK, Norris RM, Singh BN, Barratt-Boyes C. Effects of glucose and glucose-insulin-
1988;3:407-14.
Cardiol 1970;1:107-15.
globally underperfused rat hearts. Circ Res 1990;66:344-54.
potassium on haemodynamics and enzyme release after acute myocardial infarction. Br
Heart J 1977;39:748-57.
59.
Maroko PR, Libby P, Sobel BE et al. Effect of glucose-insulin-potassium infusion on
myocardial infarction following experimental coronary artery occlusion. Circulation
1972;45:1160-1175.
60.
Marano L, Bestetti A, Lomuscio A et al. Effects of infusion of glucose-insulin-potassium on
myocardial function after a recent myocardial infarction. Acta Cardiol 2000;55:9-15.
61. Malhotra R, Brosius FC, III. Glucose uptake and glycolysis reduce hypoxia-induced
apoptosis in cultured neonatal rat cardiac myocytes. J Biol Chem 1999;274:12567-75.
62.
Apstein CS, Gravino FN, Haudenschild CC. Determinants of a protective effect of glucose
and insulin on the ischemic myocardium. Effects on contractile function, diastolic
24
Introduction and aims of thesis
compliance, metabolism, and ultrastructure during ischemia and reperfusion. Circ Res
1983;52:515-26.
63.
Apstein CS, Taegtmeyer H. Glucose-insulin-potassium in acute myocardial infarction: the
64.
Apstein CS. The benefits of glucose-insulin-potassium for acute myocardial infarction (and
65.
Gerich JE. Control of glycaemia. Baillieres Clin Endocrinol Metab 1993;7:551-86.
66.
Mizock BA. Alterations in fuel metabolism in critical illness: hyperglycaemia. Best Pract Res
67.
Russell RR, III, Li J, Coven DL et al. AMP-activated protein kinase mediates ischemic
time has come for a large, prospective trial. Circulation 1997;96:1074-77.
some concerns). J Am Coll Cardiol 2003;42:792-95.
Clin Endocrinol Metab 2001;15:533-51.
glucose uptake and prevents postischemic cardiac dysfunction, apoptosis, and injury. J Clin
Invest 2004;114:495-503.
68.
Ramasamy R, Payne JA, Whang J, Bergmann SR, Schaefer S. Protection of ischemic
myocardium in diabetics by inhibition of electroneutral Na+-K+-2Cl- cotransporter. Am J
Physiol Heart Circ Physiol 2001;281:H515-H522.
69.
Mitrakou A, Ryan C, Veneman T et al. Hierarchy of glycemic thresholds for
counterregulatory hormone secretion, symptoms, and cerebral dysfunction. Am J Physiol
1991;260:E67-E74.
70.
Desouza CV, Murthy SN, Diez J et al. Differential effects of peroxisome proliferator activator
receptor-alpha and gamma ligands on intimal hyperplasia after balloon catheter-induced
vascular injury in Zucker rats. J Cardiovasc Pharmacol Ther 2003;8:297-305.
71.
Libby P, Maroko PR, Braunwald E. The effect of hypoglycemia on myocardial ischemic
72.
Rizza RA, Cryer PE, Gerich JE. Role of glucagon, catecholamines, and growth hormone in
injury during acute experimental coronary artery occlusion. Circulation 1975;51:621-26.
human glucose counterregulation. Effects of somatostatin and combined alpha- and betaadrenergic blockade on plasma glucose recovery and glucose flux rates after insulininduced hypoglycemia. J Clin Invest 1979;64:62-71.
73.
Das UN. Insulin: an endogenous cardioprotector. Curr Opin Crit Care 2003;9:375-83.
74.
Groeneveld AB, Beishuizen A, Visser FC. Insulin: a wonder drug in the critically ill? Crit Care
75.
McNulty PH, Jacob R, Deckelbaum LI, Young LH. Effect of hyperinsulinemia on myocardial
2002;6:102-5.
amino acid uptake in patients with coronary artery disease. Metabolism 2000;49:1365-69.
76.
Liedtke AJ, Hughes HC, Neely JR. Effects of excess glucose and insulin on glycolytic
77.
Sun XJ, Wang LM, Zhang Y et al. Role of IRS-2 in insulin and cytokine signalling. Nature
78.
Aikawa R, Nawano M, Gu Y et al. Insulin prevents cardiomyocytes from oxidative stress-
79.
Jonassen AK, Brar BK, Mjos OD, Sack MN, Latchman DS, Yellon DM. Insulin administered
metabolism during experimental myocardial ischemia. Am J Cardiol 1976;38:17-27.
1995;377:173-77.
induced apoptosis through activation of PI3 kinase/Akt. Circulation 2000;102:2873-79.
at reoxygenation exerts a cardioprotective effect in myocytes by a possible anti-apoptotic
mechanism. J Mol Cell Cardiol 2000;32:757-64.
25
Metabolic interventions in acute myocardial infarction
80.
Nygren J, Carlsson-Skwirut C, Brismar K, Thorell A, Ljungqvist O, Bang P. Insulin infusion
increases levels of free IGF-I and IGFBP-3 proteolytic activity in patients after surgery. Am J
Physiol Endocrinol Metab 2001;281:E736-E741.
81.
Timmins AC, Cotterill AM, Hughes SC et al. Critical illness is associated with low circulating
concentrations of insulin-like growth factors-I and -II, alterations in insulin-like growth factor
binding proteins, and induction of an insulin-like growth factor binding protein 3 protease.
Crit Care Med 1996;24:1460-1466.
82.
Madibally SV, Solomon V, Mitchell RN, Van De WL, Yarmush ML, Toner M. Influence of
insulin therapy on burn wound healing in rats. J Surg Res 2003;109:92-100.
83.
Solomon V, Madihally S, Mitchell RN, Yarmush M, Toner M. Antiproteolytic action of insulin
84.
Young LH, Dahl DM, Rauner D, Barrett EJ. Physiological hyperinsulinemia inhibits
85.
Whitlow PL, Rogers WJ, Smith LR et al. Enhancement of left ventricular function by
in burn-injured rats. J Surg Res 2002;105:234-42.
myocardial protein degradation in vivo in the canine heart. Circ Res 1992;71:393-400.
glucose-insulin-potassium
infusion
in
acute
myocardial
infarction.
Am
J
Cardiol
1982;49:811-20.
86.
Kersten JR, Schmeling TJ, Orth KG, Pagel PS, Warltier DC. Acute hyperglycemia abolishes
87.
Kersten JR, Montgomery MW, Ghassemi T et al. Diabetes and hyperglycemia impair
ischemic preconditioning in vivo. Am J Physiol 1998;275:H721-H725.
activation of mitochondrial K(ATP) channels. Am J Physiol Heart Circ Physiol
2001;280:H1744-H1750.
88.
Dandona P, Aljada A, Mohanty P et al. Insulin inhibits intranuclear nuclear factor kappaB
and stimulates IkappaB in mononuclear cells in obese subjects: evidence for an antiinflammatory effect? J Clin Endocrinol Metab 2001;86:3257-65.
89.
Rosenzweig T, Braiman L, Bak A, Alt A, Kuroki T, Sampson SR. Differential effects of tumor
necrosis factor-alpha on protein kinase C isoforms alpha and delta mediate inhibition of
insulin receptor signaling. Diabetes 2002;51:1921-30.
90.
Booth G, Stalker TJ, Lefer AM, Scalia R. Elevated ambient glucose induces acute
inflammatory events in the microvasculature: effects of insulin. Am J Physiol Endocrinol
Metab 2001;280:E848-E856.
91.
Booth G, Stalker TJ, Lefer AM, Scalia R. Mechanisms of amelioration of glucose-induced
endothelial dysfunction following inhibition of protein kinase C in vivo. Diabetes
2002;51:1556-64.
92.
Kuboki K, Jiang ZY, Takahara N et al. Regulation of endothelial constitutive nitric oxide
synthase gene expression in endothelial cells and in vivo : a specific vascular action of
insulin. Circulation 2000;101:676-81.
93.
Diaz-Araya G, Nettle D, Castro P et al. Oxidative stress after reperfusion with primary
coronary angioplasty: lack of effect of glucose-insulin-potassium infusion. Crit Care Med
2002;30:417-21.
94.
Bertuglia S, Giusti A, Fedele S, Picano E. Glucose-insulin-potassium treatment in
combination
with
dipyridamole
Diabetologia 2001;44:2165-70.
26
inhibits
ischaemia-reperfusion-induced
damage.
Introduction and aims of thesis
95.
Melidonis A, Stefanidis A, Tournis S et al. The role of strict metabolic control by insulin
infusion on fibrinolytic profile during an acute coronary event in diabetic patients. Clin
Cardiol 2000;23:160-164.
96.
Legtenberg RJ, Houston RJ, Oeseburg B, Smits P. Physiological insulin concentrations
protect against ischemia-induced loss of cardiac function in rats. Comp Biochem Physiol A
Mol Integr Physiol 2002;132:161-67.
97.
McNulty PH, Pfau S, Deckelbaum LI. Effect of plasma insulin level on myocardial blood flow
98.
Rogers WJ, Russell RO, Jr., McDaniel HG, Rackley CE. Acute effects of glucose-insulin-
and its mechanism of action. Am J Cardiol 2000;85:161-65.
potassium infusion on myocardial substrates, coronary blood flow and oxygen
consumption in man. Am J Cardiol 1977;40:421-28.
99.
100.
Sundell J, Knuuti J. Insulin and myocardial blood flow. Cardiovasc Res 2003;57:312-19.
Nava P, Carbo R, Guarner V. Coronary and femoral arterial contraction with high glucose,
insulin, and glucose-insulin-potassium solution: effects of hypoxia. Heart Vessels
2002;16:57-63.
101.
Nava P, Collados MT, Masso F, Guarner V. Endothelin mediation of insulin and glucoseinduced changes in vascular contractility. Hypertension 1997;30:825-29.
102.
Verma S, Yao L, Stewart DJ, Dumont AS, Anderson TJ, McNeill JH. Endothelin antagonism
uncovers insulin-mediated vasorelaxation in vitro and in vivo. Hypertension 2001;37:32833.
103.
Apstein CS, Deckelbaum L, Mueller M, Hagopian L, Hood WB, Jr. Graded global ischemia
and reperfusion. Cardiac function and lactate metabolism. Circulation 1977;55:864-72.
104.
Quinones-Galvan A, Ferrannini E. Metabolic effects of glucose-insulin infusions:
myocardium and whole body. Curr Opin Clin Nutr Metab Care 2001;4:157-63.
105.
Kashiwagi A, Shinozaki K, Nishio Y et al. Endothelium-specific activation of NAD(P)H
106.
Golovchenko I, Goalstone ML, Watson P, Brownlee M, Draznin B. Hyperinsulinemia
oxidase in aortas of exogenously hyperinsulinemic rats. Am J Physiol 1999;277:E976-E983.
enhances transcriptional activity of nuclear factor-kappaB induced by angiotensin II,
hyperglycemia, and advanced glycosylation end products in vascular smooth muscle cells.
Circ Res 2000;87:746-52.
107.
Goalstone ML, Natarajan R, Standley PR et al. Insulin potentiates platelet-derived growth
factor action in vascular smooth muscle cells. Endocrinology 1998;139:4067-72.
108.
He FJ, MacGregor GA. Fortnightly review: Beneficial effects of potassium. BMJ
2001;323:497-501.
109.
Gennari FJ. Disorders of potassium homeostasis. Hypokalemia and hyperkalemia. Crit Care
110.
Beal AL, Scheltema KE, Beilman GJ, Deuser WE. Hypokalemia following trauma. Shock
Clin 2002;18:273-88, vi.
2002;18:107-10.
111.
Schulman M, Narins RG. Hypokalemia and cardiovascular disease. Am J Cardiol
112.
Schwartz AB. Potassium-related cardiac arrhythmias and their treatment. Angiology
1990;65:4E-9E.
1978;29:194-205.
27
Metabolic interventions in acute myocardial infarction
113.
Wolk R, Lusawa T, Ceremuzynski L. Effects of glucose-insulin-potassium infusion on QT
dispersion in patients with acute myocardial infarction. Ann Noninvasive Electrocardiol
2001;6:50-54.
114.
Rowe JW, Tobin JD, Rosa RM, Andres R. Effect of experimental potassium deficiency on
glucose and insulin metabolism. Metabolism 1980;29:498-502.
28
GIK and 30-day outcome
Chapter 2.1
Glucose-insulin-potassium infusion in patients treated with
primary angioplasty for acute myocardial infarction (GIPS): a
randomized trial
Iwan CC van der Horst, Felix Zijlstra, Arnoud WJ van ‘t Hof, Carine JM Doggen, MenkoJan de Boer, Harry Suryapranata, Jan CA Hoorntje, Jan-Henk E Dambrink, Rijk OB Gans,
Henk JG Bilo, on behalf of the Zwolle Myocardial Infarction Study Group
Journal of the American College of Cardiology 2003;42:784-91
29
Metabolic interventions in acute myocardial infarction
Abstract
Background
A combined treatment of early and sustained reperfusion of the infarct-related coronary
artery and the metabolic modulation with GIK infusion has been proposed to protect the
ischemic myocardium.
Methods
From April 1998 to September 2001, 940 patients with an acute MI and eligible for PCI
were randomly assigned, by open-label, to either a continuous GIK infusion for 8-12 hours
or no infusion.
Results
30-day mortality was 23 of 476 patients (4.8%) receiving GIK compared to 27 of 464
patients (5.8%) in the control group, relative risk 0.82 (95% CI 0.46–1.46). In 856 patients
(91.1%) without signs of heart failure (Killip class 1), 30-day mortality was 5 of 426 patients
(1.2%) in the GIK group versus 18 of 430 patients (4.2%) in the control group, relative risk
0.28 (0.1–0.75). In 84 patients (8.9%) with signs of heart failure (Killip class ≥2), 30-day
mortality was 18 of 50 patients (36%) in the GIK group versus 9 of 34 patients (26.5%) in
the control group, relative risk 1.44 (0.65–3.22).
Conclusion
GIK infusion as adjunctive therapy to PCI in acute MI did not result in a significant
mortality reduction in all patients. In the subgroup of 856 patients without signs of heart
failure a significant reduction was seen. The effect of GIK infusion in patients with signs of
heart failure (Killip class ≥2) at admission is uncertain.
30
GIK and 30-day outcome
Introduction
Since the early sixties, glucose-insulin-potassium (GIK) infusion has been advocated as
therapy in the early hours following acute myocardial infarction (MI).1 A meta-analysis
involving nine of these early studies with 1932 patients described that GIK might play an
important role in reducing in-hospital mortality after acute MI.2 In four studies in which a
high-dose was used a non-significant reduction in mortality from 12% in the control group
to 6.5% in the GIK group was observed.3-6 High-dose GIK denotes an infusion of 30 g of
glucose, 50 units of insulin, and 80 mmol potassium per liter at a rate of 1.5 mL/kg/hour.7
This dose of GIK suppresses arterial free fatty acid levels and myocardial free fatty acid
uptake, and induces a maximal increase in myocardial glucose uptake. Therefore the main
effect of the GIK infusion is considered to be the beneficial effect of administration of
glucose to the ischemic myocardium.8-10
At present, primary coronary intervention (PCI) is regarded to be the most effective
reperfusion treatment of acute MI.11-13 The complementary role of GIK to reperfusion
therapy has been proposed.14;15 Therefore, we performed a randomized trial to investigate
the value of GIK when combined with PCI for acute MI.
Methods
Patients
All consecutive patients with symptoms consistent with acute MI of >30 minutes duration,
presenting within 24 hour after the onset of symptoms and with a ST segment elevation of
more than 1 mm (0.1 mV) in two or more contiguous leads on the electrocardiogram, or
new onset left bundle branch block, were evaluated for inclusion in this single center
study. Both patients presented at our center and patients referred for treatment of highrisk MI, from nine referring hospitals without facilities to perform coronary interventions,
were included. Patients were excluded when pre-treated with thrombolysis or when an
illness associated with a marked restricted life expectancy was present. Before
randomization baseline characteristics were recorded. The research protocol was
reviewed and approved by the medical ethics committee at our hospital, and patients
were included after informed consent.
Protocol
After admission patients were randomly assigned, with the use of a computerized
randomization program, to either GIK infusion or no infusion. In the GIK group, a
continuous infusion of 80 mmol potassium chloride in 500 mL 20% glucose with a rate of
31
Metabolic interventions in acute myocardial infarction
3 mL/kg/hour over an 8-12 hour period in a peripheral venous line was given, as soon as
possible. A continuous infusion of short-acting insulin (50 units Actrapid HM [Novo
Nordisk, Copenhagen, Denmark] in 50 mL 0.9% sodium chloride was started with the use
of a pump (Perfusor-FM, B. Braun, Melsungen, Germany). Baseline insulin-infusion dose
and adjustments of the insulin dose, to obtain blood-glucose levels between 7.0 and 11.0
mmol/L, were based on a modified algorithm.16 Glucose levels were based on
measurements
of whole-blood
glucose
(Modular
System,
Roche/Hitachi,
Basel,
Switzerland). The infusion was continued for 8 to 12 hour period in order to administer at
least 1.5 liter and not over 2 liters. The infusion was stopped earlier only if the clinical
situation deteriorated due to volume overload.
After the infusion was started, all patients went to the catheterization laboratory where
both coronary arteries were visualized. PCI was performed with standard techniques if the
coronary anatomy was suitable for PCI. Successful reperfusion was defined by TIMI grade
3 blood flow in combination with myocardial blush grades 2 and 3.17;18 Additional standard
treatment consisted of nitroglycerin intravenously, and aspirin 500 mg intravenously or
300 mg orally. During PCI 10000 IU heparin was administered as bolus. Low molecular
weight heparin was given for 24 hours thereafter. Standard therapies after PCI included
aspirin 80 mg, clopidogrel 75 mg, beta-blockers, lipid lowering agents and angiotensin
converting enzyme inhibitors or angiotensin II receptor blockers since they are likely to be
protective in most patients, were used according to the current international
guidelines.19;20
Statistical analysis
The primary endpoint of the study was 30-day mortality. Secondary endpoints were
recurrent infarction, repeat PCI and the composite incidence of death, recurrent infarction,
or repeat PCI. The incidence of death was evaluated in predefined subgroups including
age <60 year versus ≥60 year, men versus women, anterior MI versus non-anterior MI,
diabetics versus non-diabetics, and Killip class 1 versus Killip class ≥2, successful
reperfusion versus no successful reperfusion. Analyses were performed according to the
intention-to-treat principle. Relative risks were calculated by dividing the cumulative
incidence rate in the GIK group by the cumulative incidence rate in the control group.
Differences between group means at baseline were assessed with the two-tailed Student’s
32
GIK and 30-day outcome
Table 1. Baseline clinical and demographic characteristics and initial therapies
Characteristics
Number of patients
GIK group
Control group
476
464
Age, years (mean±SD)
59.9±11.9
60.8±12.0
Men
351 (73.7)
368 (79.3)
Referred patients
201 (42.3)
180 (38.8)
Body mass index (mean±SD)
26.7±3.8
27.0±4.0
Previous MI
52 (10.9)
53 (11.4)
Previous PCI
22 (4.7)
24 (5.2)
Previous CABG
14 (2.9)
12 (2.6)
History of stroke
17 (3.6)
15 (3.2)
Diabetes mellitus
50 (10.5)
49 (10.6)
Hypertension
134 (28.2)
130 (28.0)
Dyslipidemia
94 (19.7)
95 (20.5)
Currently smoker
225 (47.3)
237 (51.1)
Positive family history
195 (41.0)
179 (38.6)
Time to admission, min (IQR)*
150 (100-215)
150 (105-234)
Door to balloon time, min (IQR)
45 (31-64)
48 (34-69)
Killip class 1
426 (89.5)
430 (92.7)
Killip class 2
24 (5.0)
14 (3.0)
Killip class 3
14 (2.9)
14 (3.0)
Killip class 4
12 (2.5)
6 (1.3)
Anterior MI
250 (52.9)
224 (49.1)
Multi-vessel disease
249 (52.3)
242 (52.2)
PCI
436 (91.6)
424 (91.4)
Stent**
255 (58.5)
236 (55.7)
GP IIb/IIIa receptor blocker**
96 (22.1)
109 (25.7)
In-hospital CABG
19 (4.0)
19 (4.1)
TIMI 3 flow after PCI
459 (96.4)
435 (93.8)
Successful reperfusion‡
394 (82.8)
384 (82.8)
Intra-aortic balloon pump
45 (9.5)
42 (9.1)
Mechanical ventilation
12 (2.5)
4 (0.9)
Data are number (%) unless otherwise indicated. MI = myocardial infarction. IQR = interquartile
range. PCI = percutaneous coronary intervention. CABG = coronary artery bypass grafting. * Time
to admission denotes time between onset of symptoms and until admission. † In-hospital CABG =
CABG defined as patients treated initially conservative, followed by elective CABG within 7 days. ‡
Successful reperfusion denotes TIMI 3 flow and blush grade 2 or 3. ** Percentage defined as the
percentage of the patients that underwent PCI.
33
Metabolic interventions in acute myocardial infarction
t-test. Chi-square analysis or Fisher’s exact test was used to test differences between
proportions. Survival was calculated by the Kaplan-Meier product-limit method. The Logrank test was used to evaluate differences in survival curves between the two treatment
groups. The Cox proportional-hazards regression model was used to calculate relative
risks adjusted for differences in baseline characteristics. Statistical significance was
considered a two-tailed P-value <0.05. The Statistical Package for the Social Sciences
(SPSS Inc., Chicago, IL, USA) version 10.1 was used for all statistical analysis.
Figure1. Kaplan-Meier curve of the overall population
100
Survival (%)
90
80
GIK
Control
70
60
0
0
5
10
15
20
25
30
Time after randomization (days)
Power analysis
We postulated that infusion with GIK would induce a risk reduction of 66 percent.14 With
an estimated absolute mortality rate of 6% in the control group, a power of 0.80 and an α
of 0.05, 820 patients needed to be included in the study. In order to have 820 patients in
each major predefined subgroup (patients in Killip class 1, patients without diabetes
mellitus, and patients with successful reperfusion), we planned to enrol 940 patients.
Interim analysis with regard to safety was planned and performed after inclusion of 240
patients, and the study was continued according to protocol. Enrolment of patients started
in April 1998 and ended September 2001.
34
GIK and 30-day outcome
Results
Study population
Of the 940 patients enrolled, 476 patients were randomly assigned to the GIK group and
464 patients to the control group. Table 1 shows clinical and demographic characteristics
of the groups. With the exception of gender there were no statistical significant
differences between both groups. GIK infusion was started 15 to 20 minutes after
admission. Average door to balloon time was 45 minutes in the GIK group and 48 minutes
in the control group. After coronary angiography, 860 patients (90.5%) underwent PCI, 38
patients (4.0%) were referred for CABG within 7 days after initial stabilization, and 42
patients (4.5%) were treated conservatively. There were no major differences in bloodglucose levels between the GIK group and the control group at admission (median bloodglucose level 8.5 mmol/L in both groups) and 16 hours after admission.
Figure 2. Kaplan-Meier curve of Killip class 1 and Killip class ≥2 patients
100
Survival (%)
90
80
70
60
0
0
10
20
30
Time after randomization (days)
GIK - Killip 1
GIK - Killip ≥2
Control - Killip 1
Control - Killip ≥2
30-day mortality
Fifty of 940 patients (5.3%) had died at 30 days. Twenty-three of 476 patients (4.8%) in the
GIK group versus 27 of 476 patients (5.8%) in the control group, relative risk 0.82 (95%
35
Metabolic interventions in acute myocardial infarction
confidence interval 0.46–1.46) (table 2, figure 1). Causes of death are represented in table
5. In 856 of the 940 patients (91.1%) without signs of heart failure (Killip class 1), the
mortality rate was 5 of 426 patients (1.2%) in the GIK group versus 18 of 430 patients
(4.2%) in the control group, relative risk 0.28 (0.1–0.75) (table 2, table 4, figure 2). In this
subgroup of patients a higher number of patients died of heart failure in the control group
(0.7% in the GIK group versus 2.8% in the control group) (table 5). In the 84 patients
(8.9%) with signs of heart failure (Killip class ≥2), mortality rate was 18 of 50 patients
(36%) in the GIK group versus 9 of 34 patients (26.5%) in the control group, relative risk
1.44 (0.65–3.22). In this subgroup of patients 34.0% in the GIK group versus 20.6% in the
control group died of heart failure (table 5, figure 2).
Table 2. Effect of GIK on 30-day mortality in the overall population and in subgroups
Characteristic
GIK group
Control group
RR (95% CI)
P-value
mortality (%)
mortality (%)
All patients
23/476 (4.8)
27/464 (5.8)
0.82 (0.46–1.46)
0.50
<60 year
7/235 (3.0)
6/204 (2.9)
1.01 (0.34–3.07)
0.98
≥60 year
16/241 (6.6)
21/260 (8.1)
0.81 (0.41–1.59)
0.54
Men
9/351 (2.6)
16/368 (4.3)
0.58 (0.25–1.33)
0.19
14/125 (11.2)
11/96 (11.5)
0.98 (0.42–2.26)
0.95
Women
Time ≤180 min
9/304 (3.0)
13/288 (4.5)
0.65 (0.27–1.53)
0.32
Time >180 min
14/172 (8.1)
14/176 (8.0)
1.03 (0.47–2.20)
0.94
2/50 (4.0)
6/49 (12.2)
0.30 (0.06–1.56)
0.16
No diabetes mellitus
21/426 (4.9)
21/415 (5.1)
0.97 (0.52–1.81)
1.00
Killip class 1
5/426 (1.2)
18/430 (4.2)
0.27 (0.10–0.74)
0.01
Killip class 2
3/24 (12.5)
2/14 (14.3)
0.86 (0.13–5.87)
0.88
Killip class 3
7/14 (50.0)
4/14 (28.6)
2.50 (0.53–11.93)
0.25
Killip class 4
8/12 (66.7)
3/6 (50.0)
2.00 (0.27–14.78)
0.49
Anterior MI
18/250 (7.2)
19/224 (8.5)
0.84 (0.43–1.64)
0.60
Diabetes mellitus
Non-anterior MI
5/226 (2.2)
8/240 (3.3)
0.66 (0.21–2.04)
0.46
Multi-vessel disease
19/249 (7.6)
24/242 (9.9)
0.75 (0.40–1.41)
0.37
Single-vessel disease
4/227 (1.8)
3/222 (1.4)
1.31 (0.29–5.91)
0.73
TIMI 3 flow after PCI
17/459 (3.7)
20/435 (4.6)
0.80 (0.41–1.55)
0.51
No TIMI 3 flow after PCI
6/17 (35.3)
7/29 (24.1)
1.71 (0.46–6.35)
0.51
Successful*
9/394 (2.3)
16/384 (4.2)
0.54 (0.24–1.23)
0.14
No successful
14/82 (17.1)
11/80 (13.8)
1.29 (0.55–3.05)
0.56
Relative risks were determined by dividing the cumulative incidence rates in the GIK group and the
control group. RR = relative risk; CI = confidence interval; MI = myocardial infarction; TIMI =
Thrombolysis in Myocardial Infarction. * Successful reperfusion denotes TIMI 3 flow and blush
grade 2 or 3.
36
GIK and 30-day outcome
In 99 patients (10.5%) with a history of diabetes mellitus, 2 of the 50 patients (4.0%) in the
GIK group died versus 6 of 49 patients (12.2%) in the control group, relative risk 0.30
(0.06–1.56). Mortality rates in the other predefined subgroups are shown in table 2. Figure
3 shows relative risks and 95% confidence intervals in each subgroup. Recurrent
infarction occurred in 11 patients: 4 patients (0.8%) in the GIK group and 7 patients (1.5%)
in the control group, relative risk 0.55 (0.16–1.90) (table 4). The composite endpoint was
comparable in the GIK group and the control group, relative risk 0.79 (0.50–1.24). In
patients without heart failure (Killip class 1) the composite endpoint showed a significant
advantage of GIK, relative risk 0.48 (0.27–0.87). Age, gender, previous MI, smoking status,
Killip class, anterior MI, and multi-vessel disease were associated with 30-day mortality in
the overall population.
In multivariate analysis in the overall study age, gender, previous MI, anterior MI, Killip
class, and multi-vessel disease remained independent prognostic factors for 30-day
mortality. After adjustment for these factors in the overall population, relative risk of
mortality with GIK became 0.61 (0.34–1.10, P=0.09). In the group of Killip class 1 patients
the beneficial effect of GIK remained significant with an adjusted relative risk of 0.28 (0.10–
0.77, P=0.01).
Discussion
The 30-day mortality rate was lower in the GIK group (4.8%) compared to the control
group (5.8%), but was not statistically significant. In the large predefined subgroup of 856
patients without signs of heart failure (Killip class 1), a significant reduction of mortality
was seen (1.2% in the GIK compared to 4.2% in the control group). In the 84 patients with
signs of heart failure (Killip class ≥2) we did not observe a significant difference in
mortality between the two groups. In the other predefined subgroups, such as patients
with successful reperfusion, GIK did not result in a significant beneficial effect (P=0.14,
table 2). The definition of several subgroups, particularly with small number of patients,
carries the risk of finding significance due to chance alone, i.e. either beneficial or
detrimental. Nevertheless, the principal finding of benefit of GIK in patients without signs
of heart failure and in particular the relative risk and confidence interval (figure 3) are
unlikely caused by chance.
Previous studies with GIK
The beneficial effect of GIK in acute MI has been suggested by several clinical
studies.3;4;6;21-25 The first small study that combined GIK with thrombolysis reported an
37
Metabolic interventions in acute myocardial infarction
improved ejection fraction and wall motion.5 In the Estudios Cardiologicos Latinoamerica
(ECLA) pilot trial a beneficial effect on in-hospital mortality was observed primarily in the
subgroup of patients in which GIK was combined with thrombolysis.14
Figure 3. Relative risk of GIK on 30-day mortality in overall population and in predefined
subgroups
All patients
<60 year
≥ 60 year
Men
Women
Diabetes mellitus
No diabetes mellitus
Killip class 1
Killip class ≥2
Anterior MI
Non-anterior MI
Multi-vessel disease
Single-vessel disease
TIMI 3 flow
No TIMI 3 flow
Successful
No successful
0
1
GIK better
2
Control better
3
However, the Polish-Glucose-Insulin-Potassium (Pol-GIK) trial with 954 patients using lowdose GIK, mortality in the GIK group was even significantly higher than in the control
group (8.9% versus 4.8%, P=0.01).26 When compared to the results of the ECLA pilot trial
of the subgroup of 252 patients treated with GIK and reperfusion the mortality rate
reduction in our study, i.e. 1% in the overall population, was much smaller.14
38
2/214 (0.9)
5/416 (1.2)
Single-vessel disease
TIMI 3 flow after PCI
2/64 (3.1)
No successful reperfusion
6/66 (9.1)
12/364 (3.3)*
5/26 (19)
13/404 (3.2)*
3/211 (1.4)
15/219 (6.8)†
5/224 (2.2)
13/206 (6.3)†
14/387 (3.6)*
4/43 (9.3)
10/166 (6.0)
2/5 (40)
1/12 (5.3)
2/3 (67)
1/21 (4.8)
0/8 (-)
3/16 (19)
1/12 (17)
2/12 (8.3)
1/21 (4.8)
2/3 (67)
1/8 (13)
2/16 (13)
1/6 (17)
2/18 (11)
3/14 (21)
0/10 (-)
GIK group
1/4 (25)
1/10 (10)
1/2 (50)
1/12 (8.3)
0/6 (-)
2/8 (25)
1/8 (13)
1/6 (17)
2/13 (15)
0/1 (-)
2/6 (33)
0/8 (-)
1/5 (20)
1/9 (11)
1/9 (11)
1/5 (20)
Control group
Killip class 2
4/6 (67)
3/8 (38)
1/1 (100)
6/13 (46)
1/4 (25)
6/10 (60)
0/2 (-)
7/12 (58)
7/12 (58)
0/2 (-)
5/9 (56)
2/5 (40)
6/7 (86)
1/7 (14)
4/10 (40)
3/4 (75)
GIK group
1/6 (17)
3/8 (38)
0/0 (-)
4/14 (29)
0/3 (-)
4/11 (36)
1/6 (17)
3/8 (38)
3/11 (27)
1/3 (33)
0/2 (-)
4/12 (33)
2/7 (29)
2/7 (29)
3/8 (38)
1/6 (17)
Control group
Killip class 3
6/7 (86)
2/5 (40)
3/3 (100)
5/9 (56)
1/1 (100)
7/11 (64)
2/4 (50)
6/8 (75)
8/11 (73)
0/1 (-)
5/5 (100)
3/7 (43)
4/5 (80)
4/7 (57.1)
5/8 (63)
3/4 (75)
GIK group
3/4 (75)
0/2 (-)
1/1 (100)
2/5 (40)
0/2 (-)
3/4 (75)
1/2 (50)
2/4 (100)
2/4 (50)
½ (50)
2/2 (100)
1/4 (25)
1/1 (100)
2/5 (40)
3/5 (60)
0/1 (-)
Control group
Killip class 4
and blush grade 2 or 3.
MI = myocardial infarction; TIMI = Thrombolysis in Myocardial Infarction. ** Time denotes time to admission. ‡ Successful reperfusion denotes TIMI 3 flow
Relative risks were determined by dividing the cumulative incidence rates in the GIK group and the control group. * P-value of RR <0.05; † P-value of RR <0.01.
3/336 (0/8)
Successful reperfusion‡
0/10 (-)
3/212 (1.4)
Multi-vessel disease
No TIMI 3 flow
2/208 (1.0)
0/44 (-)
Diabetes mellitus
3/218 (1.4)
3/150 (2.0)
Time >180 min**
Non-anterior MI
2/276 (0.7)
Time ≤180 min**
Anterior MI
8/264 (3.0)
3/107 (2.8)
Women
5/382 (1.3)
7/83 (8.4)
2/319 (0.6)
Men
No diabetes mellitus
14/238 (5.9)*
4/209 (1.9)
≥60 year
11/347 (3.2)*
4/192 (2.1)
<60 year
Control group
GIK group
1/217 (0.5)
Subgroup
Killip class 1
Table 3. Effect of GIK on 30-day mortality in patients with Killip class 1, 2, 3 and 4
Metabolic interventions in acute myocardial infarction
Several facts could explain the difference between both studies. First, all patients in both
groups of our study were treated with PCI compared to prior studies which used
preferentially thrombolysis or no reperfusion therapy. Secondly, the mortality rate is low,
due to the effect of our standard care for patients with acute MI, i.e. the short time from
onset of symptoms to treatment, the use of additional therapies, and a high rate of
stenting (57.1%). Thirdly, in our study the number of patients with signs of heart failure at
admission was much higher than in the ECLA pilot trial. In this subgroup of patients an
increased mortality rate was observed in our study. It is harder to reduce the mortality
even more in a population with such a low mortality rate. The mortality rate in the control
group in the ECLA pilot trial was 15.2%, a percentage that was even higher than the
mortality rate of patients in the control group of patients that did not receive reperfusion
therapy.
GIK in patients with signs of heart failure
The effect of GIK infusion in patients with signs of heart failure (Killip class ≥2) is uncertain.
In spite of liberal use of intra-aortic balloon pumping (9.3%), more patients in the GIK
group needed mechanical ventilation (2.5% in the GIK group versus 0.9% in the control
group). Heart failure was an important cause of death in patients admitted with Killip class
≥2, see table 5. Therefore it is a possibility that in some patients the potential positive
effects of GIK on metabolism were overruled by a negative effect of the volume load on
the hemodynamic state.27-29 For a patient of 80 kg in our study, our infusion rate resulted
in the infusion of 1920 ml in 8 hours. Administering such a volume load to patients with an
impaired left ventricular function may induce hemodynamic instability, even though the
metabolic effect may be beneficial. However, we did not expect this infusion rate to be
detrimental since in early studies hemodynamic disturbances, such as pulmonary
congestion, dyspnoea, and oedema, occurred in only 1-3% of patients treated with GIK.3
In experimental and clinical studies, the infusion of insulin and glucose did increase the
contractile force30-32, although others found that GIK made no differences on
hemodynamic parameters.6
Studies with strict glucose regulation
In the Diabetes Insulin-Glucose in Acute Myocardial Infarction (DIGAMI) study with 620
patients with an acute MI and diabetes mellitus or glucose at admission >11.0 mmol/L,
the combination of insulin-glucose infusion followed by intensive insulin treatment
resulted in better glucose regulation in the insulin-glucose group.16 At 1-year follow-up,
mortality was lower in the insulin-glucose group, with 18.6% versus 26.1% in the control
group (P=0.03).33;34
40
GIK and 30-day outcome
It is suggested that GIK is mainly beneficial by the effects of insulin on the myocardium
and the whole body.35-37 Results of a study of intensive insulin therapy in critically ill
patients to maintain blood glucose between 4.4 mmol/L and 6.1 mmol/L gives direction
for future research in which metabolic interventions in acute MI are directed towards strict
glucose regulation.38;39
Table 4. Clinical endpoints at 30 days in overall population and Killip class 1 patients
GIK
Control
Overall population
N=476
N=464
RR*
Adjusted RR*
P-value
Death
23 (4.8)
27 (5.8)
0.82 (0.46–1.46)
0.61 (0.34–1.10)
0.09
Recurrent MI
4 (0.8)
7 (1.5)
0.56 (0.16–1.90)
0.42 (0.12–1.51)
0.19
Repeat PCI
16 (3.4)
20 (4.3)
0.77 (0.40–1.49)
0.74 (0.38–1.44)
0.37
Composite endpoint
38 (8.0)
46 (9.9)
0.79 (0.50–1.24)
0.68 (0.44–1.05)
0.08
Killip class 1
N=426
N=430
Death
5 (1.2)
18 (4.2)
0.28 (0.10–0.75)
0.28 (0.10–0.77)
0.01
Recurrent MI
3 (0.7)
6 (1.4)
0.50 (0.13–2.02)
0.39 (0.09–1.63)
0.20
Repeat PCI
12 (2.8)
18 (4.2)
0.66 (0.32–1.40)
0.64 (0.30–1.33)
0.23
Composite endpoint
18 (4.2)
36 (8.4)
0.48 (0.27–0.87)
0.47 (0.27–0.83)
0.01
Data are RR and 95% confidence interval. Relative risk adjusted for age, gender, history, Killip class,
infarct location, and multi-vessel disease by using Cox proportional hazards regression. In subgroup
of Killip class 1 patients RR adjusted for age, gender, history of MI, infarct location and multi-vessel
disease. † P-value for adjusted relative risk. ‡ Repeat PCI denotes coronary angioplasty after initially
conservative or coronary angioplasty within 30 days. § Composite endpoint denotes the occurrence
of death or recurrent infarction or repeat PCI. PCI = percutaneous coronary intervention, RR =
relative risk, CI = confidence interval.
GIK and time to reperfusion
It has been postulated that through reduction in the extent of ischemic myocardial
damage and suppression of free fatty acid levels, GIK therapy prevents reperfusion
injuries that may occur after successful revascularization.3 Protection of the cell membrane
of ischemic cardiac cells may also improve reflow after reperfusion and protect against no
reflow phenomenon by reducing cell swelling and microvascular compression. It has been
estimated that GIK therapy has the potential to protect ischemic myocardium before
reperfusion for 10 hours or even longer.40 Our study offered the opportunity to analyse the
effect of GIK in patients with different ischemic time, i.e. time to admission and door to
balloon time. We could not find a clear relation between the time delays and the effect of
GIK (table 2, table 3).
41
Metabolic interventions in acute myocardial infarction
Adverse effects
With the possible exception of volume overload in patients with Killip class ≥2 there were
no major adverse effects of GIK. Phlebitis did not occur in our study, even though the
majority of patients received the infusion via a peripheral venous line. It is most likely that
this is due to the short infusion period, maximally 12 hours. In the ECLA pilot trial, 17%
did developed mild phlebitis and only 2% developed severe phlebitis after 24 hour of GIK
infusion.14
Table 5. Causes of death in GIK group and control group
Killip class 1
Number of patients
Killip class ≥2
GIK group
Control group
GIK group
Control group
426
430
50
34
Cardiac
Myocardial rupture
1 (0.2)
1 (0.2)
-
2 (5.9)
Recurrent infarction
-
1 (0.2)
1 (2.0)
-
3 (0.7)
12 (2.8)
17 (34.0)
7 (20.6)
-
3 (0.7)
-
-
Non-cardiac
1 (0.2)
1 (0.2)
-
-
All causes
5 (1.2)
18 (4.2)
18 (36.0)
9 (26.5)
Heart failure
Sudden death
Data are number (%). Percentage defined as the percentage of all patients in the subgroup.
Study limitations
With the observed low mortality rates in the GIK group and control group in the overall
population our study could not detect a significant difference in mortality. A more realistic
risk reduction for future research may be a risk reduction of 30% in patients with Killip
class 1. Therefore, in future studies over 2000 patients have to be included. By including
940 patients in our present study, the number of patients in some of the subgroups, such
as in patients with a history of diabetes mellitus, were too small to draw definite
conclusions. The subgroups of patients in our study were predefined, but we did not use
a method to correct for multiple comparisons to analyze them. The role of GIK in patients
with signs of heart failure is unclear and the volume load may be detrimental. A more
tailored approach in patients with heart failure at admission seems warranted, such as the
admission of glucose 30% or an infusion of no more than 500 ml or a period of infusion of
12 to 24 hours.3 Alternatively, in these patients the GIK infusion should be combined with
other measures such as continuous monitoring of hemodynamic parameters.30 Our study
had an open-label design, since the need to administer a large volume of fluid, frequent
blood-glucose monitoring and adjustments of the insulin dose in the study group is both
unethical and unpractical to organize in a placebo-controlled study.
42
GIK and 30-day outcome
Conclusion
GIK infusion as adjunctive therapy to PCI in acute MI seems promising. In the large
subgroup of patients without signs of heart failure (over 90% of all patients) we did find a
significant beneficial effect. The favourable results of GIK in patients without signs of heart
failure did encourage us to start a large, multi-center, randomized trial to determine the
optimal metabolic management in patients with a heart rate <90 beats/min, a systolic
blood pressure >100 mmHg, and no third heart sound or pulmonary rales. Alternative
regimens for metabolic intervention may be necessary for patients with signs of heart
failure.
Contributors
This study was supported by a generous grant from the Netherlands Heart Foundation
(99.028). No other conflicts of interest or financial disclosure.
References
1.
Sodi-Pallares D, Testelli MR, Fishleder BL et al. Effects of an intravenous infusion of a
potassium-glucose-insulin solution on the electrocardiographic signs of myocardial infarction.
A preliminary clinical report. Am J Cardiol 1962;9:166-81.
2.
Fath-Ordoubadi F, Beatt KJ. Glucose-insulin-potassium therapy for treatment of acute
myocardial infarction: an overview of randomized placebo-controlled trials. Circulation
1997;96:1152-56.
3.
Heng MK, Norris RM, Singh BN, Barratt-Boyes C. Effects of glucose and glucose-insulinpotassium on haemodynamics and enzyme release after acute myocardial infarction. Br Heart
J 1977;39:748-57.
4.
Rogers WJ, Stanley AW, Jr., Breinig JB et al. Reduction of hospital mortality rate of acute
myocardial infarction with glucose-insulin-potassium infusion. Am Heart J 1976;92:441-54.
5.
Satler LF, Green CE, Kent KM, Pallas RS, Pearle DL, Rackley CE. Metabolic support during
coronary reperfusion. Am Heart J 1987;114:54-58.
6.
Stanley AW, Jr., Prather JW. Glucose-insulin-potassium, patient mortality and the acute
myocardial infarction: results from a prospective randomised study. Circulation 1978;57:61.
7. Rackley CE, Russell RO, Jr., Rogers WJ, Mantle JA, McDaniel HG, Papapietro SE. Clinical
experience with glucose-insulin-potassium therapy in acute myocardial infarction. Am Heart J
1981;102:1038-49.
8.
Apstein CS, Gravino FN, Haudenschild CC. Determinants of a protective effect of glucose and
insulin on the ischemic myocardium. Effects on contractile function, diastolic compliance,
metabolism, and ultrastructure during ischemia and reperfusion. Circ Res 1983;52:515-26.
9.
Cave AC, Ingwall JS, Friedrich J et al. ATP synthesis during low-flow ischemia: influence of
increased glycolytic substrate. Circulation 2000;101:2090-2096.
43
Metabolic interventions in acute myocardial infarction
10.
Oliver MF, Opie LH. Effects of glucose and fatty acids on myocardial ischaemia and
11.
Grines CL, Browne KF, Marco J et al. A comparison of immediate angioplasty with
arrhythmias. Lancet 1994;343:155-58.
thrombolytic therapy for acute myocardial infarction. The Primary Angioplasty in Myocardial
Infarction Study Group. N Engl J Med 1993;328:673-79.
12.
Zijlstra F, de Boer MJ, Hoorntje JC, Reiffers S, Reiber JH, Suryapranata H. A comparison of
immediate coronary angioplasty with intravenous streptokinase in acute myocardial
infarction. N Engl J Med 1993;328:680-684.
13.
Zijlstra F, Hoorntje JC, de Boer MJ et al. Long-term benefit of primary angioplasty as
compared with thrombolytic therapy for acute myocardial infarction. N Engl J Med
1999;341:1413-19.
14.
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.
15.
Vanoverschelde JL, Janier MF, Bakke JE, Marshall DR, Bergmann SR. Rate of glycolysis
during ischemia determines extent of ischemic injury and functional recovery after
reperfusion. Am J Physiol 1994;267:H1785-H1794.
16.
Malmberg KA, Efendic S, Ryden LE. Feasibility of insulin-glucose infusion in diabetic patients
with acute myocardial infarction. A report from the multicenter trial: DIGAMI. Diabetes Care
1994;17:1007-14.
17.
The Thrombolysis in Myocardial Infarction (TIMI) trial. Phase I findings. TIMI Study Group N
18.
van 't Hof AW, Liem A, Suryapranata H, Hoorntje JC, de Boer MJ, Zijlstra F. Angiographic
Engl J Med 1985;312:932-36.
assessment of myocardial reperfusion in patients treated with primary angioplasty for acute
myocardial infarction: myocardial blush grade. Zwolle Myocardial Infarction Study Group.
Circulation 1998;97:2302-6.
19.
Ryan TJ, Antman EM, Brooks NH et al. 1999 update: ACC/AHA Guidelines for the
Management of Patients With Acute Myocardial Infarction: Executive Summary and
Recommendations: A report of the American College of Cardiology/American Heart
Association Task Force on Practice Guidelines (Committee on Management of Acute
Myocardial Infarction). Circulation 1999;100:1016-30.
20.
Van de Werf F, Ardissino D, Betriu A et al. Management of acute myocardial infarction in
patients presenting with ST segment elevation. The Task Force on the Management of Acute
Myocardial Infarction of the European Society of Cardiology. Eur Heart J 2003;24:28-66.
21.
Hjermann I. A controlled study of peroral glucose, insulin and potassium treatment in
22.
Medical Research Council Working Party on the Treatment of Myocardial Infarction.
myocardial infarction. Acta Med Scand 1971;190:213-18.
Potassium, glucose, and insulin treatment for acute myocardial infarction. Lancet
1968;2:1355-60.
23.
Mittra B. Potassium, glucose, and insulin in treatment of myocardial infarction. Lancet
24.
Pentecost BL, Mayne NM, Lamb P. Controlled trial of intravenous glucose, potassium, and
1965;2:607-9.
insulin in acute myocardial infarction. Lancet 1968;1:946-48.
44
GIK and 30-day outcome
25.
Pilcher J, Etishamudin M, Exon P, Moore J. Potassium, glucose and insulin in myocardial
26.
Ceremuzynski L, Budaj A, Czepiel A et al. Low-dose glucose-insulin-potassium is ineffective in
infarction. Lancet 1967;1:1109.
acute myocardial infarction: results of a randomized multicenter Pol-GIK trial. Cardiovasc
Drugs Ther 1999;13:191-200.
27.
Crexells C, Chatterjee K, Forrester JS, Dikshit K, Swan HJ. Optimal level of filling pressure in
the left side of the heart in acute myocardial infarction. N Engl J Med 1973;289:1263-66.
28.
Sjostrand F, Edsberg L, Hahn RG. Volume kinetics of glucose solutions given by intravenous
29.
Wildenthal K, Mierzwiak DS, Mitchell JH. Acute effects of increased serum osmolality on left
30.
Mantle JA, Rogers WJ, Smith LR et al. Clinical effects of glucose-insulin-potassium on left
infusion. Br J Anaesth 2001;87:834-43.
ventricular performance. Am J Physiol 1969;216:898-904.
ventricular function in acute myocardial infarction: results from a randomized clinical trial. Am
Heart J 1981;102:313-24.
31.
Sasso FC, Carbonara O, Cozzolino D et al. Effects of insulin-glucose infusion on left ventricular
function at rest and during dynamic exercise in healthy subjects and noninsulin dependent
diabetic patients: a radionuclide ventriculographic study. J Am Coll Cardiol 2000;36:219-26.
32.
Whitlow PL, Rogers WJ, Smith LR et al. Enhancement of left ventricular function by glucose-
33.
Malmberg K, Ryden L, Efendic S et al. Randomized trial of insulin-glucose infusion followed
insulin-potassium infusion in acute myocardial infarction. Am J Cardiol 1982;49:811-20.
by subcutaneous insulin treatment in diabetic patients with acute myocardial infarction
(DIGAMI study): effects on mortality at 1 year. J Am Coll Cardiol 1995;26:57-65.
34.
Malmberg K. Prospective randomised study of intensive insulin treatment on long term
survival after acute myocardial infarction in patients with diabetes mellitus. DIGAMI (Diabetes
Mellitus, Insulin Glucose Infusion in Acute Myocardial Infarction) Study Group. BMJ
1997;314:1512-15.
35.
Das UN. Insulin: an endogenous cardioprotector. Curr Opin Crit Care 2003;9:375-83.
36.
Groeneveld AB, Beishuizen A, Visser FC. Insulin: a wonder drug in the critically ill? Crit Care
37.
McNulty PH. Glucose and insulin management in the post-MI setting. Curr Diab Rep
38.
van den Berghe G, Wouters P, Weekers F et al. Intensive insulin therapy in the critically ill
39.
van den Berghe G, Wouters PJ, Bouillon R et al. Outcome benefit of intensive insulin therapy
2002;6:102-5.
2002;2:37-44.
patients. N Engl J Med 2001;345:1359-67.
in the critically ill: Insulin dose versus glycemic control. Crit Care Med 2003;31:359-66.
40.
Eberli FR, Weinberg EO, Grice WN, Horowitz GL, Apstein CS. Protective effect of increased
glycolytic substrate against systolic and diastolic dysfunction and increased coronary
resistance from prolonged global underperfusion and reperfusion in isolated rabbit hearts
perfused with erythrocyte suspensions. Circ Res 1991;68:466-81.
45
Metabolic interventions in acute myocardial infarction
46
GIK and myocardial function
Chapter 2.2
The impact of glucose-insulin-potassium infusion in acute
myocardial infarction on infarct size and left ventricular
ejection fraction
Iwan CC van der Horst, Jan Paul Ottervanger, Arnoud WJ van ‘t Hof, Stoffer Reiffers, Kor
Miedema, Jan CA Hoorntje, Jan-Henk E Dambrink, AT Marcel Gosselink, Maarten WN
Nijsten, Harry Suryapranata, Menko-Jan de Boer, Felix Zijlstra
Submitted
47
Metabolic interventions in acute myocardial infarction
Abstract
Background
Favorable clinical outcomes have been observed with glucose-insulin-potassium infusion
(GIK) in acute myocardial infarction (MI). The mechanisms of this beneficial effect have not
been delineated clearly. GIK has metabolic, anti-inflammatory and profibrinolytic effects
and it may preserve the ischemic myocardium. We aimed to assess the effect of GIK
infusion on infarct size and left ventricular function, as part of a randomized controlled
trial.
Methods
940 patients treated with primary percutaneous coronary intervention (PCI) for acute MI
were randomized to GIK infusion or no infusion. Endpoints were the creatine kinase MBfraction (CK-MB) and left ventricular ejection fraction (LVEF). CK-MB levels were
determined at 0, 2, 4, 6, 24, 48, 72 and 96 hour after admission and the LVEF before
discharge.
Results
There was no difference in the time course or magnitude of CK-MB release between the 2
groups: peak CK-MB level was 249±228 IU/L in the GIK group versus 240±200 IU/L in the
control group (NS). The mean LVEF was 43.7±11.0 % in the GIK group versus 42.4±11.7%
in the control group (P=0.12). A LVEF ≤30% was observed in 18% in the controls and in
12% in the GIK group (P=0.01).
Conclusion
Treatment with GIK preserves myocardial function as determined by LVEF. No effect was
observed on the pattern or magnitude of enzyme release.
48
GIK and myocardial function
Introduction
Glucose-insulin-potassium (GIK) infusion in acute myocardial infarction (MI) has been
suggested to be of clinical benefit.1-4 Both animal studies and early studies in patients
investigating the influence of GIK on infarct size have shown conflicting results.5-14
Experimental studies on the effect of GIK on preservation of left ventricular function, as
determined
by
hemodynamic
parameters,
demonstrated
a
beneficial
effect.15;16
Furthermore, in a small study in patients with acute MI treated with thrombolysis there
was a significant improvement in left ventricular function over a 10 day period 17. Recently,
a small randomized trial has suggested a beneficial effect of the addition of GIK to primary
PCI in 37 patients.18 Unfortunately, two large studies of GIK in the era of reperfusion did
not report detailed information on the effects of GIK on myocardial function.1;17;19 The
Polish-Glucose-Insulin-Potassium (Pol-GIK) trial with 954 patients using low-dose GIK,
found no difference between median creatine kinase (CK) levels (920 IU/L in GIK patients
versus 925 IU/L in controls).19 Recently, in the Reevaluation of Intensified Venous
Metabolic Support for Acute Infarct size Limitation (REVIVAL) trial with 312 MI patients the
combination of GIK and primary PCI had no effect on myocardial salvage.20 In the
Glucose-Insulin-Potassium Study (GIPS) with GIK infusion as adjunctive therapy to
primary PCI in acute MI the 30-day mortality was not significantly reduced in the overall
population.4 In the predefined subgroup of patients without heart-failure at admission
mortality was 1.2% in GIK patients versus 4.2% in controls (P=0.01). In this study we were
able to determine the effect of GIK on cumulative enzyme release and left ventricular
ejection fraction.
Methods
Study population
Outline of the study and 30-day clinical follow-up have been reported 4. All consecutive
patients with symptoms consistent with acute MI of >30 minutes duration, presenting
within 24 hours after the onset of symptoms and with ST segment elevation of more than
1 mm (0.1 mV) in two or more contiguous leads on the electrocardiogram were evaluated
for inclusion in this study. Patients were excluded when pre-treated with thrombolysis or
when an illness associated with a marked restricted life expectancy was present. Before
randomization, age, gender, history of coronary artery bypass grafting (CABG), previous
PCI, stroke and MI, existence of diabetes mellitus, smoking status, heart rate, arterial
pressure, body mass index, Killip class, electrocardiographic site of infarction, time of
onset of symptoms, time of hospital admission in both referring hospital and in our
hospital were recorded. The research protocol was reviewed and approved by the
49
Metabolic interventions in acute myocardial infarction
medical ethics committee of our hospital, and patients were included after informed
consent.
Treatment protocol
Patients were randomly assigned to either GIK infusion or no infusion. Assignments to the
treatment groups were made with the use of a computerized randomization program. In
the GIK group, a continuous infusion of 80 mmol potassium chloride in 500 mL 20%
glucose with a rate of 3 mL/kg/hour in a peripheral venous line was given. A continuous
infusion of short-acting insulin (50 units Actrapid HM-Novo Nordisk, Copenhagen,
Denmark) in 50 mL 0.9% sodium chloride was also started with the use of a pump
(Perfusor-FM, B. Braun, Melsungen, Germany). Baseline insulin-infusion dose and hourly
adjustments of the insulin dose were based on an algorithm, to obtain blood-glucose
levels between 7.0 and 11.0 mmol/L, based on measurements of whole-blood glucose
concentration. The infusion was started as soon as possible after randomization.
All patients went to the catheterization laboratory as soon as possible after admission and
randomization. At the laboratory both coronary arteries were visualized. PCI was
performed with standard techniques if the coronary anatomy was suitable for angioplasty.
Additional treatment consisted of unfractionated intravenous heparin, nitroglycerin and
aspirin. After sheath removal, low-molecular-weight heparin was given for 24 hours.
Before discharge additional pharmacological treatment consisting of aspirin, clopidogrel,
coumarins, beta-blockers, angiotensin-converting enzyme inhibitors, and statins were
initiated according to the guidelines of the American College of Cardiology/American
Heart Association.20 Enrollment of patients started in April 1998 and ended September
2001.
Measurements
Enzymatic infarct size was estimated by serial measurements of CK-MB fraction. The first
measurement was taken as soon as possible after admission. Thereafter frequent CKdeterminations were performed according to a schedule that called for 4 to 8
measurements in the first 96 hours. To accommodate the problem in practice that exact
predefined times for blood sampling were not always followed, the actual times of
sampling, expressed as minutes after the moment of randomization were recorded. To
allow optimal comparison of the time courses of CK-MB levels and to best approximate
the area under the CK-MB curves we developed an algorithm. The algorithm estimated
the time course of CK-MB for each patient before calculating means for patient groups.
According to the known kinetics of CK-MB that with a rapid rise and a much slower
decrease the following characteristic time points with accompanying time intervals with
defined: 0 hr (interval from 60 min before to 45 min after randomization); 2 hr (45min to 3
50
GIK and myocardial function
hr); 4 hr (3hr to 5 hr); 6 hr (5hr to 12hr); 24hr (12hr to 36hr); 48hr (36hr to 60hr); 72hr
(60hr to 84hr) and 96hr (84hr to 108hr). Per patient the algorithm interpolated the CK-MB
levels based on the precise times the measurements were performed. Next it determined
for each of the predefined intervals if an actual measurement had been performed in that
interval. If no measurement was available for a particular interval a ‘not available’ value
was generated for that interval. Accordingly inappropriately interpolated values were
avoided. If one or more measurements were available for an interval, the CK-MB value for
the time point with the corresponding interval was calculated from the area under the
interpolated curve was divided by the duration of that interval. CK-MB was determined
enzymatically on a Hitachi 717 automatic analyzer according to the International
Federation of Clinical Chemistry (IFCC) recommendation at 30 degrees Celsius. A peak CKMB above the 75% percentile (i.e. the highest quartile) was defined as high enzyme
release. Time to peak CK-MB was determined.
Left ventricular ejection fraction (LVEF) was measured before discharge by radionuclide
ventriculography or by echocardiography. Radionuclide ventriculography was performed
with the multiple gated equilibrium method following the labeling of red blood cells of the
patient with technetium-99m-pertechnate. A General Electric 300 gamma camera with a
low-energy all-purpose parallel-hole collimator was used. Global ejection fraction was
calculated by a General Electric Star View computer using the fully automatic PAGE
program. Two-dimensional echocardiography was performed by observers who were
unaware of the randomization outcome and clinical data. A LVEF ≤30% was defined as a
poor left ventricular function.
Statistical analysis
All analyses were by intention to treat. Endpoints were pattern and magnitude of CK-MB
release, peak CK-MB, high enzymatic enzyme release, defined as a peak enzyme release in
the highest quartile; mean LVEF, and a LVEF lower than or equal to 30%. To allow
comparison of the patterns of CK-MB release, the CK-MB values at 0, 2, 4, 6, 24, 48, 72
and 96 hour after admission were calculated for each individual patient with linear
interpolation. Differences between group means were assessed with the two-tailed
Student’s t-test. Chi-square analysis was used to test differences between proportions.
The Cox proportional-hazards regression model was used to calculate relative risks
adjusted for differences in baseline characteristics. To predict the independent association
between treatment with GIK and a peak CK-MB release above the highest quartile or LVEF
≤30%, a multivariate logistic regression analysis was performed. Statistical significance
was considered a two-tailed P-value <0.05. The Statistical Package for the Social Sciences
(SPSS Inc., Chicago, IL, USA) version 11.5 was used for all statistical analysis.
51
Metabolic interventions in acute myocardial infarction
Table 1. Baseline clinical and demographic characteristics and initial therapies
Characteristics
Number of patients
GIK group
Control group
476
464
Age, years (mean±SD)
59.9±11.9
60.8±12.0
Men
351 (73.7)
368 (79.3)
Referred patients
201 (42.3)
180 (38.8)
Body mass index
26.7±3.8
27.0±4.0
Previous MI
52 (10.9)
53 (11.4)
Previous PCI
22 (4.7)
24 (5.2)
Previous CABG
14 (2.9)
12 (2.6)
History of stroke
17 (3.6)
15 (3.2)
Diabetes mellitus
50 (10.5)
49 (10.6)
Hypertension
134 (28.2)
130 (28.0)
Dyslipidemia
94 (19.7)
95 (20.5)
Currently smoker
225 (47.3)
237 (51.1)
Positive family history
195 (41.0)
179 (38.6)
Time to admission, min (IQR)*
150 (100-215)
150 (105-234)
Door to balloon time, min (IQR)
45 (31-64)
48 (34-69)
Killip class 1
426 (89.5)
430 (92.7)
Killip class 2
24 (5.0)
14 (3.0)
Killip class 3
14 (2.9)
14 (3.0)
Killip class 4
12 (2.5)
6 (1.3)
Anterior MI
250 (52.9)
224 (49.1)
Multi-vessel disease
249 (52.3)
242 (52.2)
PCI
436 (91.6)
424 (91.4)
Stent**
255 (58.5)
236 (55.7)
GP IIb/IIIa receptor blocker**
96 (22.1)
109 (25.7)
In-hospital CABG
19 (4.0)
19 (4.1)
TIMI 3 flow after PCI
459 (96.4)
435 (93.8)
Successful reperfusion‡
394 (82.8)
384 (82.8)
Intra-aortic balloon pump
45 (9.5)
42 (9.1)
Mechanical ventilation
12 (2.5)
4 (0.9)
Data are number (%) unless otherwise indicated. MI = myocardial infarction. IQR = interquartile
range. PCI = percutaneous coronary intervention. CABG = coronary artery bypass grafting. * Time
to admission denotes time between onset of symptoms and until admission. † In-hospital CABG =
CABG defined as patients treated initially conservative, followed by elective CABG within 7 days. ‡
Successful reperfusion denotes TIMI 3 flow and blush grade 2 or 3. ** Percentage defined as the
percentage of the patients that underwent PCI.
52
GIK and myocardial function
Results
Study population
Of the 940 patients who underwent randomization, 476 were assigned to receive GIK
infusion. There were no differences in the baseline characteristics of the two treatment
groups (table 1). The mean time between onset of symptoms and admission was 150
minutes, and 75% of the patients were admitted within 225 min. The interquartile range
(IQR) in the GIK group was 100 to 215 minutes and in the control group 105 to 234
minutes. After coronary angiography, 860 patients (90.5%) underwent PCI, 38 patients
(4%) were referred for CABG within 7 days, and 42 patients (4.5%) were treated
conservatively. 30-day mortality was 4.8% in the GIK patients and 5.8% in the control
patients.4
Table 2. Comparison of patients with a peak CK-MB release beneath the highest quartile
and patients with a CK-MB release above the highest quartile
Characteristic
High enzyme release
Low enzyme release
232
708
59±12
61±12
0.17
Men
179 (77)
540 (76)
0.86
Hypertension
64 (28)
200 (28)
0.87
Previous MI
24 (10)
81 (11)
0.72
Diabetes mellitus
28 (12)
71 (10)
0.39
Killip class 1
192 (83)
664 (94)
<0.001
Anterior MI
161 (69)
313 (44)
<0.001
TIMI flow 0 before PCI
189 (82)
377 (53)
<0.001
TIMI flow 3 after PCI
219 (94)
675 (95)
0.6
GIK treatment
121 (52)
355 (50)
0.6
Number of patients
Age, years (mean±SD)
P-value
Data are number (%) unless otherwise indicated. MI = myocardial infarction. TIMI = Thrombolysis
in Myocardial Infarction. PCI = percutaneous coronary intervention.
Enzyme release
Data on the effect of GIK on enzymatic infarct size (CK-MB) were completely available in
923 patients (98%), 470 patients treated with GIK and 462 control patients. Of each patient
at least a single measurement was available. A mean of 6.9 CK-MB determinations were
available per patient. As indicated in figure 1, there was no difference in pattern or
magnitude of the CK-MB release between the two groups.
53
Metabolic interventions in acute myocardial infarction
Mean peak CK-MB±SD was 249±228 IU/L in the GIK group versus 240±200 IU/L in the
control group (NS). In GIK patients the median time to peak CK-MB was 450 minutes (IQR
300-733) compared to 468 minutes (IQR 285-738) in the control patients (NS). High
enzyme release (i.e. peak CK-MB >349 IU/L) was found in 111 patients (24%) in the GIK
group versus 121 patients (25%) in the control group (P=0.6). Differences between
patients with a peak CK-MB release beneath 349 IU/L and above 349 IU/L are represented
in table 2. Characteristic related with high enzyme release were Killip class ≥2, anterior MI,
and TIMI 0 flow before PCI.
Table 3. Comparison of patients with a low left ventricular ejection fraction (LVEF ≤30%)
with those with preserved left ventricular ejection fraction (LVEF >30%)
Characteristic
Low LVEF
Preserved LVEF
125
698
Age, years (mean±SD)
63±12
59±12
0.008
Men
98 (78)
537 (77)
0.7
Hypertension
43 (34)
182 (26)
0.05
Previous MI
18 (14)
63 (9)
0.06
Diabetes mellitus
29 (23)
57 (8)
0.008
Killip class 1
109 (87)
660 (95)
0.002
Anterior MI
115 (92)
291 (42)
<0.001
TIMI 0 before PCI
40 (32)
291 (42)
0.04
TIMI 3 after PCI
114 (91)
677 (97)
0.002
GIK treatment
51 (41)
368 (53)
0.01
Number of patients
P-value
Data are number (%) unless otherwise indicated. MI = myocardial infarction. TIMI = Thrombolysis
in Myocardial Infarction. PCI = percutaneous coronary intervention.
Left ventricular ejection fraction
LVEF measurements were available in 419 patients (88%) treated with GIK and in 404
patients (87%) of the controls. Mean time between admission and LVEF measurement
was 2.65 days in the GIK group and 2.5 days in the control group (IQR 2 days in both
groups). The mean LVEF was 43.7±11.0% in the GIK group versus 42.4±11.7% in the
control group (P=0.12). Of the 823 patients with a measured LVEF before discharge a total
of 125 (15%) had an ejection fraction lower than or equal to 30%. Differences between
patients with LVEF ≤30% and those with a higher LVEF are summarized in table 3.
Treatment with GIK was associated with a lower frequency of a LVEF ≤30% (51 patients
(12%) compared to 74 patients (18%), P=0.01). Other important characteristics associated
with LVEF ≤30% were age, diabetes mellitus, Killip class, infarct location and TIMI flow
54
GIK and myocardial function
after PCI. To identify independent predictors of a low left ventricular ejection fraction
(LVEF ≤30%) all variables associated with LVEF ≤30% in univariate analysis were used in
multivariate analysis, see table 3. After adjustment, patients treated with GIK had a RR of
0.53 (95% confidence interval 0.35–0.81, P=0.01) for a LVEF ≤30% compared to controls.
In patients with Killip class 1 at admission the mean LVEF was 44.1±10.9 % in the GIK
group versus 42.8±11.5% in the control group (P=0.12). In patients randomized to GIK a
LVEF ≤30% was less often observed (45 patients (12%) versus 65 patients (17%), P=0.03).
The mean peak CK-MB ± SD was 463±318 IU/L in patients with LVEF ≤30% versus
215±166 IU/L in patients with LVEF >30% (P<0.001).
Figure 1. CK-MB release after admission in the GIK group and control group
250
CK-MB (IU/L)
200
GIK group
Control group
150
100
50
0
0
20
40
60
80
100
Time after admission (hours)
CK-MB release after admission in the GIK group (unbroken line) and control group (dashed line).
Mean are indicated for 0, 2, 4, 6, 24, 48, 72 and 96 hours. At no time point a significant difference
was observed.
Discussion
In this study, additional treatment with GIK in patients treated with primary PCI for acute
myocardial infarction did not decrease enzymatic infarct size but preserved left ventricular
function. Since these results are based on an intention to treat analysis of a randomized
controlled trial of the effects of GIK, we may have underestimated the effect on
preservation of myocardial function. In the control group more patients died, and these
55
Metabolic interventions in acute myocardial infarction
patients had a high enzyme release and probably as well a severely impaired LVEF.
Nevertheless, our data make a clinically relevant impact of GIK on enzyme release
unlikely, and at the same time show an important reduction in the number of patients with
a LVEF ≤30%. A LVEF ≤30% exposes these patients to a high risk for the development of
heart failure during follow-up, even after successful PCI.21 We also found a relation
between high enzyme release and poor left ventricular function. Possibly the relation
between GIK and poor LVEF is related to a selection bias of this sub-analysis. On the other
hand it can be hypothesized that GIK mediates effects on LVEF, as measured after 2 to 4
days after admission unrelated to enzymatic infarct size.
In our previous study on the effect of GIK on 30-day mortality we found a beneficial effect
of GIK on outcome in patients without signs of heart-failure at admission. Patients with
Killip class 1 had no smaller enzymatic infarct size in this analysis. However, the curve of
CK-MB over the first 96 hours after admission in GIK patients lies below the curve of
controls. Mean LVEF is larger in patients treated with GIK compared to control patients,
albeit not significant (P=0.12). Even as in the overall population the number of patients
with LVEF ≤30% is smaller in Killip class 1 patients.
Figure 2. CK-MB release after admission in the GIK group and control group of Killip class
1 patients
250
CK-MB (IU/L)
200
GIK group
Control group
150
100
50
0
0
20
40
60
80
100
Time after ademission (hours)
CK-MB release after admission in the GIK group (unbroken line) and control group (dashed line).
Mean are indicated for 0, 2, 4, 6, 24, 48, 72 and 96 hours. At no time point a significant difference
was observed.
56
GIK and myocardial function
Previous clinical studies with GIK
Several studies have reported enzymatic infarct size in patients treated with GIK, mainly in
patients without reperfusion therapy.12;14 Clinical studies in the era of reperfusion did
provide sparse information.1;19 In line with our results the Pol-GIK trial showed that the
median and maximal CK and maximal glutamic oxaloacetic transaminase did not differ
significantly between GIK and control patients.19 However, the Pol-GIK trial used low-dose
GIK and included patients with a lower risk, i.e. no patients with signs of congestion were
included. One study of 32 MI patients treated with thrombolysis found that infusion of GIK
to obtain glucose regulation with insulin resulted in a shorter time to peak CK and a
smaller total enzyme release.6 In line with our results on enzyme release the REVIVAL trial
found no effect of GIK treatment on the myocardial salvage index measured with
technetium-99m-sestamibi scintigraphy 6 to 8 hours after admission and after 7 to 14 days
(0.50 versus 0.48, P=0.96). GIK consisted of 1000 mL glucose 20% with 40 IU of insulin
and 64 mmol potassium chloride at a rate of 1.8 mL/hour for 24 hours. Only in the
predefined subgroup of patients with diabetes mellitus a significant difference in salvage
index was found (mean difference 0.19, 95% confidence interval 0.01-0.37).
In a trial of 37 patients treated with primary PCI a beneficial effect of GIK on myocardial
function was observed.18 The increase of the LVEF from baseline to 3 months was 12.5%
in the GIK group (P=0.002) versus 6.1% in the placebo group (NS). However, this may
also have been the result of a combination of small numbers of patients together with
profound changes in the LVEF in the 3-6 months after PCI for acute myocardial
infarction.22 In patients with diabetes mellitus and without myocardial ischemia infusion of
insulin and glucose did increase the contractile force, i.e. rest LVEF and LVEF during
dynamic exercise measured with radionuclide ventriculography.23
Animal studies
Several studies with animals have assessed the effects of GIK on infarct size in myocardial
infarction, but show conflicting results.5;9;11 A study with pigs showed that hearts treated
with GIK had significantly less tissue acidosis, higher wall motion scores, and the less
tissue necrosis.16 Similar results were obtained in a study with rats, in which not only the
administration of GIK before induction of ischemia but also during acute ischemia seemed
beneficial for myocardial function.10 A recent study in diabetic sheep found that GIK
improved left ventricular contractility as determined by stroke work efficiency.24
Mechanisms of action
We found a relationship between higher enzymatic infarct size and low LVEF. Since we
observed only an effect of GIK on LVEF the preservation of the left ventricular function
could not be declared by a reduction of myocardial necrosis. The potential mechanisms of
57
Metabolic interventions in acute myocardial infarction
the beneficial effects of GIK include metabolic effects, direct hemodynamic effects,
improvement of coronary flow and catecholamine mediated effects. The main effect of the
GIK infusion is supposed to be the beneficial effect of delivery of glucose to the ischemic
myocardium resulting in a suppression of free fatty acids.25-28 Treatment with glucose and
insulin inhibits lipolysis, thereby decreasing circulating free fatty acid levels, and
suppresses fatty acid oxidation. Moreover, in the postischemic myocardium contractile
dysfunction may also be caused by impairment of glucose oxidation and cytosolic proton
accumulation, whereas agents that enhance glucose oxidation in the postischemic heart
may also improve contractile function.29 After prolonged periods of low-flow ischemia
functional recovery is mainly determined by the extent of irreversible ischemic damage.
After brief episodes of ischemic damage, oxidative metabolism rapidly returns, well
before contractile activity is restored. Therefore GIK treatment is likely to be beneficial
when started soon after onset of ischemia. Protection of the cell membrane of ischemic
cardiac cells may also improve reflow after reperfusion and protect against the no reflow
phenomenon by reducing cell swelling and microvascular compression.30 These
microvascular protective effects on the coronary circulation are probably similar in
patients with and without diabetes mellitus.31;32 GIK infusion showed an improvement in
regional myocardial perfusion and function in segments adjacent to the infarct area in
patients recovering from an acute MI.33 Another beneficial mechanism may be the
protective effect of GIK on the cell membrane.34-36 Insulin promotes tolerance against
ischemic cell death via the activation of innate cell-survival pathways in the heart.37
However, in a small study of patients with MI treated with PCI, administration of GIK did
not improve the enzymatic antioxidant reserve.38
Study limitations
Left ventricular function was measured in 88% of patients. For patients transferred back to
the referring hospitals within 48 hours the measurement was often missing. The same
accounts for patients who died within 48 hours. The exact relation between LVEF and 30day mortality can not be determined. It is unknown whether these missing data have
influenced the results on the effect of GIK on left ventricular function. We determined the
left ventricular function before discharge and it is suggested that measurements after
long-term follow-up are more predictive.18
Conclusion
High-dose GIK infusion in patients with acute MI treated with primary PCI, results in a
lower rate of poor left ventricular function, defined as a left ventricular ejection fraction
≤30%. There is no effect of GIK on the pattern and magnitude of enzyme release.
58
GIK and myocardial function
Contributors
This study was supported by a generous grant from the Netherlands Heart Foundation
(99.028). No other conflicts of interest or financial disclosure.
References
1.
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.
2.
Fath-Ordoubadi F, Beatt KJ. Glucose-insulin-potassium therapy for treatment of acute
myocardial infarction: an overview of randomized placebo-controlled trials. Circulation
1997;96:1152-56.
3.
Malmberg K, Ryden L, Efendic S et al. 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 1995;26:57-65.
4.
van der Horst IC, Zijlstra F, van't Hof AW et al. Glucose-insulin-potassium infusion in patients
treated with primary angioplasty for acute myocardial infarction: the glucose-insulinpotassium study: a randomized trial. J Am Coll Cardiol 2003;42:784-91.
5.
Bellows SD, Kloner RA. Glucose-Insulin-Potassium Does Not Reduce Myocardial Infarct Size
in an Ischemic/Reperfusion Rabbit Model. J Thromb Thrombolysis 1998;5:25-27.
6.
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.
7.
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.
8.
Heng MK, Norris RM, Singh BN, Barratt-Boyes C. Effects of glucose and glucose-insulinpotassium on haemodynamics and enzyme release after acute myocardial infarction. Br Heart
J 1977;39:748-57.
9.
Heng MK, Norris RM, Peter T, Nisbet HD, Singh BN. The effect of glucose-insulin-potassium
on experimental myocardial infarction in the dog. Cardiovasc Res 1978;12:429-35.
10.
Jonassen AK, Aasum E, Riemersma RA, Mjos OD, Larsen TS. Glucose-insulin-potassium
reduces infarct size when administered during reperfusion. Cardiovasc Drugs Ther
2000;14:615-23.
11.
Maroko PR, Libby P, Sobel BE et al. Effect of glucose-insulin-potassium infusion on
myocardial
infarction
following
experimental
coronary
artery
occlusion.
Circulation
1972;45:1160-1175.
12.
Reduto L, Gulotta S, Morisson M. Effects of glucose-insulin-potassium infusion on ischemic
13.
Rogers WJ, Segall PH, McDaniel HG, Mantle JA, Russell RO, Jr., Rackley CE. Prospective
myocardium. Clinical Research 1974;22:685A.
randomized trial of glucose-insulin-potassium in acute myocardial infarction. Effects on
myocardial hemodynamics, substrates and rhythm. Am J Cardiol 1979;43:801-9.
59
Metabolic interventions in acute myocardial infarction
14.
Sundstedt CD, Sylven C, Mogensen L. Glucose-insulin-potassium-albumin infusion in the
early phase of acute myocardial infarction--a controlled study. Acta Med Scand 1981;210:6771.
15.
Whitlow PL, Rogers WJ, Smith LR et al. Enhancement of left ventricular function by glucose-
16.
Zhu P, Lu L, Xu Y, Greyson C, Schwartz GG. Glucose-insulin-potassium preserves systolic and
insulin-potassium infusion in acute myocardial infarction. Am J Cardiol 1982;49:811-20.
diastolic function in ischemia and reperfusion in pigs. Am J Physiol Heart Circ Physiol
2000;278:H595-H603.
17.
Satler LF, Green CE, Kent KM, Pallas RS, Pearle DL, Rackley CE. Metabolic support during
coronary reperfusion. Am Heart J 1987;114:54-58.
18.
Castro PF, Larrain G, Baeza R et al. Effects of glucose-insulin-potassium solution on
myocardial salvage and left ventricular function after primary angioplasty. Crit Care Med
2003;31:2152-55.
19.
Ceremuzynski L, Budaj A, Czepiel A et al. Low-dose glucose-insulin-potassium is ineffective in
acute myocardial infarction: results of a randomized multicenter Pol-GIK trial. Cardiovasc
Drugs Ther 1999;13:191-200.
20.
Pache J, Kastrati A, Mehilli J et al. A randomized evaluation of the effects of glucose-insulinpotassium infusion on myocardial salvage in patients with acute myocardial infarction treated
with reperfusion therapy. Am Heart J 2004;148:105.
21.
Henriques JP, Zijlstra F, van 't Hof AW et al. Angiographic assessment of reperfusion in acute
myocardial infarction by myocardial blush grade. Circulation 2003;107:2115-19.
22.
Ottervanger JP, van 't Hof AW, Reiffers S et al. Long-term recovery of left ventricular function
after primary angioplasty for acute myocardial infarction. Eur Heart J 2001;22:785-90.
23.
Sasso FC, Carbonara O, Cozzolino D et al. Effects of insulin-glucose infusion on left ventricular
function at rest and during dynamic exercise in healthy subjects and noninsulin dependent
diabetic patients: a radionuclide ventriculographic study. J Am Coll Cardiol 2000;36:219-26.
24.
Ramanathan T, Shirota K, Morita S, Nishimura T, Huang Y, Hunyor SN. Glucose-insulinpotassium solution improves left ventricular mechanics in diabetes. Ann Thorac Surg
2002;73:582-87.
25.
Apstein CS, Gravino FN, Haudenschild CC. Determinants of a protective effect of glucose and
insulin on the ischemic myocardium. Effects on contractile function, diastolic compliance,
metabolism, and ultrastructure during ischemia and reperfusion. Circ Res 1983;52:515-26.
26.
Depre C, Vanoverschelde JL, Taegtmeyer H. Glucose for the heart. Circulation 1999;99:578-
27.
Opie LH. The glucose hypothesis: its relation to acute myocardial ischemia. J Mol Cell Cardiol
28.
Rackley CE, Russell RO, Jr., Rogers WJ, Mantle JA, McDaniel HG, Papapietro SE. Clinical
88.
1970;1:107-15.
experience with glucose-insulin-potassium therapy in acute myocardial infarction. Am Heart J
1981;102:1038-49.
29.
Lopaschuk GD, Wambolt RB, Barr RL. An imbalance between glycolysis and glucose
oxidation is a possible explanation for the detrimental effects of high levels of fatty acids
during aerobic reperfusion of ischemic hearts. J Pharmacol Exp Ther 1993;264:135-44.
60
GIK and myocardial function
30.
Eberli FR, Weinberg EO, Grice WN, Horowitz GL, Apstein CS. Protective effect of increased
glycolytic substrate against systolic and diastolic dysfunction and increased coronary
resistance from prolonged global underperfusion and reperfusion in isolated rabbit hearts
perfused with erythrocyte suspensions. Circ Res 1991;68:466-81.
31.
McNulty PH, Pfau S, Deckelbaum LI. Effect of plasma insulin level on myocardial blood flow
32.
Sundell J, Laine H, Nuutila P et al. The effects of insulin and short-term hyperglycaemia on
and its mechanism of action. Am J Cardiol 2000;85:161-65.
myocardial blood flow in young men with uncomplicated Type I diabetes. Diabetologia
2002;45:775-82.
33.
Marano L, Bestetti A, Lomuscio A et al. Effects of infusion of glucose-insulin-potassium on
34.
Aikawa R, Nawano M, Gu Y et al. Insulin prevents cardiomyocytes from oxidative stress-
35.
Jonassen AK, Sack MN, Mjos OD, Yellon DM. Myocardial protection by insulin at reperfusion
myocardial function after a recent myocardial infarction. Acta Cardiol 2000;55:9-15.
induced apoptosis through activation of PI3 kinase/Akt. Circulation 2000;102:2873-79.
requires early administration and is mediated via Akt and p70s6 kinase cell-survival signaling.
Circ Res 2001;89:1191-98.
36.
Matsui T, Tao J, del Monte F et al. Akt activation preserves cardiac function and prevents
37.
Sack MN, Yellon DM. Insulin therapy as an adjunct to reperfusion after acute coronary
injury after transient cardiac ischemia in vivo. Circulation 2001;104:330-335.
ischemia: a proposed direct myocardial cell survival effect independent of metabolic
modulation. J Am Coll Cardiol 2003;41:1404-7.
38.
Diaz-Araya G, Nettle D, Castro P et al. Oxidative stress after reperfusion with primary coronary
angioplasty: lack of effect of glucose-insulin-potassium infusion. Crit Care Med 2002;30:41721.
61
Metabolic interventions in acute myocardial infarction
62
GIK and left ventricular function
Chapter 2.3
Preservation
of
myocardial
function
by
glucose-insulin-
potassium infusion as adjunct to primary angioplasty
Iwan CC van der Horst, Jan Paul Ottervanger, Felix Zijlstra
Critical Care Medicine 2004;32:906
63
Metabolic interventions in acute myocardial infarction
To the editor:
In 37 patients with acute myocardial infarction Castro and colleagues could not observe
that high-dose glucose-insulin-potassium (GIK) infusion in adjunction to primary
angioplasty did improve left ventricular ejection fraction (LVEF) and did not decrease the
enzymatic infarct size.1
In the recently published Glucose-Insulin-Potassium-Study (GIPS) we randomized 940
patients to primary coronary angioplasty and GIK infusion or no infusion.2 As secondary
endpoints we determined a left ventricular ejection fraction (LVEF) lower than 30%. LVEF
was measured before discharge by radionuclide ventriculography using the multiple
gated equilibrium method following the labeling of red blood cells of the patient with
technetium-99m-pertechnate. When our results are compared with the Castro study (table
1) we suggest that GIK can induce a beneficial effect on myocardial function. We observed
a LVEF <30% in 18% in the controls and in 12% in the GIK group (P=0.01). Finally, we
also observed a relation between preservation of myocardial function and survival. By
including more than 37 patients in their study possibly beneficial effects could have been
observed.
Table 1. Comparison of the Castro results and GIPS
Characteristics
GIPS
Castro study
GIK group
Control group
GIK group
Control group
476
464
18
19
59.9±11.9
60.8±12.0
55±13
55.2±11
Men
351 (74)
368 (79)
13 (72)
15 (79)
Hypertension
134 (28)
130 (28)
8 (44)
6 (32)
Diabetes mellitus
50 (11)
49 (11)
3 (17)
1 (5)
3.3
3.3
3.6
4.3
Infarct location anterior
250 (53)
224 (49)
9 (50)
4 (21)
Killip class 1
426 (90)
430 (93)
14 (77)
15 (78)
Single-vessel disease
227 (48)
222 (48)
14 (77)
12 (63)
TIMI 3 flow after PCI
459 (96)
435 (94)
16 (89)
17 (89)
LVEF, % (mean±SD)*
43.7±11.0
42.4±11.7
38.9
43.7
23 (4.8)
27 (5.8)
Number of patients
Age, years (mean±SD)
Time to angioplasty (hour)
30-day mortality
Data are number (%) unless otherwise indicated. * LVEF denotes left ventricular ejection fraction.
LVEF in the GIPS was determined before discharge i.e. after GIK infusion; the presented LVEF in the
Castro study was determined before GIK infusion.
64
GIK and left ventricular function
References
1.
Castro PF, Larrain G, Baeza R et al. Effects of glucose-insulin-potassium solution on
myocardial salvage and left ventricular function after primary angioplasty. Crit Care Med
2003;31:2152-55.
2.
van der Horst IC, Zijlstra F, van't Hof AW et al. Glucose-insulin-potassium infusion in patients
treated with primary angioplasty for acute myocardial infarction: the glucose-insulinpotassium study: a randomized trial. J Am Coll Cardiol 2003;42:784-91.
65
Metabolic interventions in acute myocardial infarction
66
GIK and ST segment elevation resolution
Chapter 2.4
Electrocardiographic evidence of improved reperfusion after
glucose-insulin-potassium infusion in patients treated with
primary angioplasty for ST segment elevation myocardial
infarction
Iwan CC van der Horst, Giuseppe de Luca, Jan Paul Ottervanger, Menko-Jan de Boer, Jan
CA Hoorntje, Harry Suryapranata, Jan-Henk E Dambrink, AT Marcel Gosselink, Felix
Zijlstra, Arnoud WJ van ‘t Hof
Submitted
67
Metabolic interventions in acute myocardial infarction
Abstract
Background
To evaluate the impact of adjunctive high-dose glucose-insulin-potassium (GIK) on ST
segment elevation resolution in patients with ST segment elevation myocardial infarction
(MI).
Methods
As part of a randomized controlled trial about GIK versus no GIK in patients treated with
primary percutaneous coronary intervention for ST segment elevation MI in a tertiary
referral center, we analyzed ST segment elevation resolution. Paired electrocardiographic
recordings (baseline and 3 hours after primary percutaneous coronary intervention) were
available in 612 patients (65%) out of 940 patients.
Results
We analyzed paired electrocardiograms of 310 patients randomized to GIK and 302
control patients. Baseline characteristics of these groups were comparable. Combined
complete (>70%) and partial (30-70%) resolution was more commonly observed in the
GIK group (87%) when compared to the control group (78%), relative risk 1.72 (95%
confidence interval 1.2-2.46, P<0.001). 1-year mortality was lower in patients with
combined complete and partial resolution compared to patients without resolution (3.8%
versus 10.3%, P=0.011). There was no difference in 1-year mortality between GIK and
control patients (5.5% versus 4.3%, P=0.58).
Conclusion
In patients with ST segment elevation MI treated with primary percutaneous coronary
intervention addition of GIK is associated with improved ST segment elevation resolution.
ST segment elevation resolution is related to improved 1-year survival. No benefit of GIK
on 1-year outcome was observed. Future trials should investigate whether GIK induced
improvement of ST segment elevation resolution results in more favorable clinical
outcome.
68
GIK and ST segment elevation resolution
Introduction
In patients with ST segment elevation myocardial infarction (MI) early resolution of ST
segment elevation is a useful predictor of final infarct size, left ventricular function and
clinical outcome after both thrombolytic and coronary interventional approaches.1-5
Primary percutaneous coronary intervention (PCI) results in a high proportion of patients
with complete ST segment elevation resolution.6 Persistent ST segment elevation shortly
after recanalization correlates with impaired myocardial reperfusion at the microcirculatory
level and with further extension of myocardial damage.7 Impaired microvascular
reperfusion, as defined by non-complete ST segment elevation resolution, may occur in
up to 33% of patients with ST segment elevation MI, especially in older patients with low
systolic pressure.8 In patients with Thrombolysis In Myocardial Infarction (TIMI) grade 3
flow after primary PCI, failed microvascular reperfusion, defined as impaired ST segment
elevation resolution, peak creatine kinase levels after 12 hours, and T-wave inversion after
24 hours, is associated with poor outcome at 1 year.9 A beneficial effect of a glucoseinsulin-potassium (GIK) infusion on electrocardiographic signs of myocardial infarction
(MI) has been suggested in experimental studies.10-15
In the Estudios Cardiologicos Latinoamerica (ECLA) pilot trial an beneficial effect of GIK on
mortality was observed primarily in the subgroup of patients in which GIK was combined
with thrombolysis (5.2% in GIK patients versus 15.2% in control patients, P=0.01).16
Recently, in the Reevaluation of Intensified Venous Metabolic Support for Acute Infarct
Size Limitation (REVIVAL) trial with 312 MI patients treated with primary PCI, GIK did not
result in a significant difference in mortality after 6 months (5.8% in the GIK group versus
6.4% in the control group, P=0.85).17 In the Glucose-Insulin-Potassium Study (GIPS)
patients with ST segment elevation MI were randomly assigned to either a high-dose GIK
infusion or no infusion in adjunction to primary PCI.18 30-Day mortality did not differ
significantly between GIK and control patients (4.8% versus 5.8%, P=0.50). In an
additional analysis of this trial, we investigated the effect of GIK on ST segment elevation
resolution, and re-evaluated the relation between ST segment elevation resolution and 1year mortality in patients enrolled in a randomized controlled trial.18
Methods
Patients
Between April 1998 and September 2001, 940 patients with symptoms consistent with
acute MI of >30 min duration, presenting within 24 hours after the onset of symptoms
and with a ST segment elevation of more than 1 mm (0.1 mV) in two or more contiguous
leads on the electrocardiogram were included in the GIPS. The exact details of the trial
69
Metabolic interventions in acute myocardial infarction
protocol have been previously described.18 In short, a total of 940 patients were randomly
assigned to either GIK infusion (N=476) or no infusion (N=464). Patients presented at our
center as well as patients referred for treatment of high-risk MI, from nine referring
hospitals without facilities to perform coronary interventions, were included. Assignments
to the groups were made with a computerized randomization program. In the GIK group, a
continuous infusion of 80 mmol potassium chloride in 500 mL 20% glucose with a rate of
3 mL/kg body weight/hour over an 8-12 hour period in a peripheral venous line was given.
A continuous infusion of short-acting insulin (50 units Actrapid HM (Novo Nordisk,
Copenhagen, Denmark) in 50 mL 0.9% sodium chloride was also started with the use of a
pump (Perfusor-FM, B. Braun, Melsungen, Germany). Baseline insulin-infusion dose and
hourly adjustments of the insulin dose were based on an algorithm, to obtain blood
glucose levels between 7.0 and 11.0 mmol/L, based on measurements of whole-blood
glucose (Modular System, Roche/Hitachi, Basel, Switzerland). GIK infusion was started 15
to 20 minutes after admission. Average door to balloon time was 45 minutes in the GIK
group and 48 minutes in the control group.18
Protocol
Primary PCI was performed with standard techniques if the coronary anatomy was
suitable for PCI. Angiographic successful reperfusion was defined as TIMI flow grade 3, in
combination with a myocardial blush grade 2 or 3. Additional standard treatment
consisted of nitroglycerin intravenously, and aspirin 500 mg intravenously or 300 mg
orally. During PCI 10000 IU heparin was administered as bolus. Low molecular weight
heparin was given for 24 hours after sheath removal. Standard therapies after PCI
included aspirin 80 mg and/or clopidogrel 75 mg in all patients that received a stent. Betablockers, lipid lowering agents and angiotensin converting enzyme inhibitors or
angiotensin II receptor blockers were used according to current international
guidelines.19;20
Whenever possible, patients remained in our hospital at least 24 hours before they were
transferred back to a referring hospital. When no Coronary Care Unit bed was available,
immediate transfer after PCI was arranged. Follow-up data were collected up to 18 months
after randomization via the register office, the General Practitioner, or by direct contact
with the patients or their relatives by telephone. The research protocol was reviewed and
approved by the medical ethics committee of our hospital, and patients were included
after informed consent.
Measurements
Electrocardiographic (ECG) recordings were obtained on admission to the hospital (first
ECG), and in the coronary-care unit (second ECG) 3 hours after PCI. All ECGs were
70
GIK and ST segment elevation resolution
analyzed as pairs by an independent core lab (Diagram BV, Zwolle, The Netherlands) and
graded for ST segment elevation resolution by 2 investigators who were unaware of the
clinical data, randomization, angiographic findings, and outcome data. The investigators
analyzed the ECGs independently and when disagreements did exist, a third investigator
graded the ST segment elevation resolution. The investigators were not blinded to the
sequence. The sum of ST segment elevation was measured 20 ms after the end of the
QRS complex in leads I, aVL, and V1–V6 for anterior, and leads II, III, aVF, and V5–V6 for
non-anterior myocardial infarction. ST segment elevation resolution was defined as
complete (>70% resolution), partial (30-70% resolution) or no resolution (<30%
resolution). In this analysis only patients of whom paired ECG recordings were available
were included.
Statistical analysis
All analyses concerning potential effects of GIK were by intention to treat. Differences
between group means at baseline were assessed with the two-tailed Student’s t-test. Chisquare analysis or Fisher’s exact test was used to test differences between proportions.
Survival was calculated by the Kaplan-Meier product-limit method. The Log-rank test was
used to evaluate differences in survival curves between the two treatment groups. The
Cox proportional-hazards regression model was used to calculate relative risks adjusted
for differences in baseline characteristics. To predict the independent association between
treatment with GIK and ST segment elevation resolution, a multivariate logistic regression
analysis was performed. Statistical significance was considered a two-tailed P-value
<0.05. The Statistical Package for the Social Sciences (SPSS Inc., Chicago, IL, USA)
version 11.5 was used for all statistical analysis.
Results
Of 940 patients enrolled in the GIPS, data on ST segment elevation resolution were
available in 310 patients (65%) randomized to GIK and 302 patients (65%) in the control
group. For patients with (transient) bundle branch block (N=56), external pacemaker or
AIVR (N=63), technical inadequate recordings (N=47), patients who died during primary
PCI (N=2), patients who were transferred to a referring hospital immediately after PCI
(N=138), and miscellaneous reasons (N=22) data were lacking. The baseline
characteristics of patients included in this analysis and the excluded patients are
presented in table 1. The included patients were less often referred from other hospitals
(35% versus 51%), had less often signs of heart failure at admission (Killip class 1 patients
93% versus 88%), and were more often treated with PCI (96% versus 83%).
71
Metabolic interventions in acute myocardial infarction
Table 1. Baseline characteristics of the patients with paired ECG recordings (included
patients) and patients without paired ECG recordings (excluded patients)
Characteristics
Included patients
Excluded patients
612
328
60±12
61±12
Men
472 (77)
247 (75)
Referred patients#
216 (35)
166 (51)
Previous MI
61 (10)
44 (13)
Previous PCI
26 (4.2)
20 (6.1)
Previous CABG
17 (2.8)
9 (2.7)
History of stroke
23 (3.8)
9 (2.7)
Diabetes mellitus
63 (10)
36 (11)
Hypertension
162 (27)
102 (31)
Dyslipidemia
123 (20)
66 (20)
Currently smoker
304 (50)
158 (48)
Positive family history
248 (41)
126 (38)
Time to admission, min (IQR)*
145 (104-215)
161 (110-240)
Door to balloon time, min (IQR)
50 (35-69)
38 (19-58)
569 (93)
287 (88)
43 (7)
41 (12)
Anterior MI
293 (48)
181 (55)
Multi-vessel disease
312 (51)
179 (55)
PCI
588 (96)
273 (83)
Stent**
336 (55)
155 (47)
GP IIb/IIIa receptor blocker**
130 (21)
74 (23)
In-hospital CABG†
11 (1.8)
27 (8.2)
TIMI 3 flow after PCI
585 (96)
309 (94)
Successful reperfusion*
513 (84)
265 (81)
Intra-aortic balloon pump
41 (6.7)
46 (14)
Mechanical ventilation
6 (1.0)
10 (3.0)
Number of patients
Age, years (mean±SD)
Killip class 1
Killip class ≥2
Data are number (%) unless otherwise indicated. MI = myocardial infarction. IQR = interquartile
range. PCI = percutaneous coronary intervention. CABG = coronary artery bypass grafting.
#Referred patients denotes MI patients referred from hospitals without facilities to perform coronary
interventions. *Time to admission denotes time between onset of symptoms and until admission.
†In-hospital CABG = CABG defined as patients treated initially conservative, followed by elective
CABG within 7 days. ‡Successful reperfusion denotes TIMI 3 flow and blush grade 2 or 3.
**Percentage defined as the percentage of the patients that underwent PCI.
72
GIK and ST segment elevation resolution
ST segment elevation resolution and outcome
In 372 patients (60.8%) complete ST segment elevation resolution was observed, partial
resolution in 133 patients (22%) and no resolution in 107 patients (17.5%). Baseline
characteristics of the 3 groups are presented in table 2. No complete ST segment
elevation resolution was more often observed in patients with anterior infarction location,
previous MI or diabetes mellitus, and Killip class >1 at admission.
Figure 1. ST segment elevation resolution and 1-year mortality
100
Survival (%)
90
Complete resolution
80
Partial resolution
No resolution
0
0
100
200
300
Time after randomization (days)
Log-rank test P-value is 0.014.
ST segment elevation resolution was related to left ventricular function. Prevalence of left
ventricular ejection fraction (LVEF) <30% was 11.8% in patients with complete ST
segment elevation resolution, compared to 20.3% in those with partial or no resolution
(P<0.01). Mortality after 1-year was 3.5% with complete and 4.5% with partial ST segment
elevation resolution compared to 10.3% with no ST segment elevation resolution
(P=0.016). 1-year mortality was lower in patients with combined complete and partial
resolution compared to patients with no ST segment elevation resolution (3.8% versus
10.3%, P=0.011). Figure 1 represents the relation between ST segment elevation
resolution and 1-year mortality (P= 0.014 using the Log-rank test).
73
Metabolic interventions in acute myocardial infarction
ST segment elevation resolution and GIK
Baseline characteristics were comparable between the 2 groups (table 3). A total of 299
patients (96%) in the GIK group and 289 (96%) in the control group underwent primary
PCI (table 4). No difference was observed in procedural success and myocardial blush
grades. The prevalence of ST segment elevation resolution was higher in patients treated
with GIK. Complete ST segment elevation resolution was observed in 198 patients (64%)
treated with GIK versus 174 patients (58%) in the control group (table 5). Partial ST
segment elevation resolution was observed in 72 patients (23%) treated with GIK versus
61 patients (20%) in the control group. Combined complete and partial resolution was
more commonly observed in the GIK group (87%) when compared to the control group
(78%), relative risk 1.72 (1.2-2.46, P<0.01). No ST segment elevation resolution was
observed in 40 patients (13%) in the GIK group compared to 67 patients (22%) in the
control group, relative risk (RR) 0.58 (95% confidence interval 0.39-0.81, P<0.01).
A total of 30 patients (4.9%) out of 612 patients died after 1-year. There were 2 patients
lost to follow-up. No difference in mortality was observed between the two groups (5.5%
versus 4.3%, P=0.58).
Table 2. Comparison of patients with complete, partial and no ST segment elevation
resolution
Characteristics
Complete resolution
Partial resolution
No resolution
372
133
107
59±12
60±11
64±11
Men
290 (78)
107 (81)
75 (70)
Diabetes mellitus
27 (7.3)
22 (17)
14 (13)
Hypertension
97 (26)
35 (26)
30 (28)
Killip class 1
349 (94)
126 (95)
94 (88)
Anterior MI
159 (43)
76 (57)
58 (54)
Successful reperfusion*
326 (88)
107 (81)
80 (75)
30-day mortality
10 (2.7)
3 (2.3)
9 (8.4)
LVEF ≤30%
40 (11.8)
22 (18.3)
21 (22.8)
1-year mortality
13 (3.5)
6 (4.5)
11 (10.3)
Number of patients
Age, years (mean±SD)
Data are number (%) unless otherwise indicated. MI = myocardial infarction. LVEF = Left ventricular
ejection fraction. *Successful reperfusion denotes TIMI 3 flow and blush grade 2 or 3.
74
GIK and ST segment elevation resolution
Discussion
This is the first study showing that adjunction of GIK to primary PCI for ST segment
elevation MI results in improved resolution of ST segment elevation. We also observed
that partial or complete ST segment elevation resolution was related to a favorable 1-year
outcome compared to patients with no resolution. However, we could not demonstrate a
direct relation between GIK and improved outcome. This analysis did not have the power
to assess the effects on mortality. We therefore only can hypothesize that GIK may be
beneficial by inducing ST segment elevation resolution, as a reflection of improved
myocardial reperfusion. We could not detect a relation between infusion of GIK and
myocardial blush grade, an angiographic parameter that reflects myocardial reperfusion.7
This may be related to the very short interval between epicardial patency and assessment
of myocardial blush grade. Studies with continuous ST segment monitoring have shown
that it takes sometime before myocardial perfusion is reestablished.16;34 We defined
successful reperfusion as the combination of TIMI 3 flow after PCI and blush grade 2 or 3
and there was no difference in the rate of successful PCI between the GIK and control
patients.
Table 3. Baseline characteristics of the patients with paired ECG recordings
Characteristics
GIK group
Control group
310
302
59±12
61±12
Men
230 (74)
242 (80)
Referred patients#
118 (38)
98 (33)
Previous MI
30 (9.7)
31 (10.3)
Previous PCI
12 (3.9)
14 (4.6)
Previous CABG
9 (2.9)
8 (2.6)
History of stroke
14 (4.5)
9 (3.0)
Diabetes mellitus
32 (10)
31 (10)
Hypertension
78 (25)
84 (28)
Dyslipidemia
94 (20)
95 (21)
Currently smoker
152 (49)
152 (50)
Positive family history
195 (41)
179 (39)
Number of patients
Age, years (mean±SD)
Data are number (%) unless otherwise indicated. MI = myocardial infarction. IQR = interquartile
range. PCI = percutaneous coronary intervention. CABG = coronary artery bypass grafting.
#Referred patients denotes MI patients referred from hospitals without facilities to perform coronary
interventions.
75
Metabolic interventions in acute myocardial infarction
ST segment elevation resolution
Primary PCI results in optimal epicardial recanalization in a large majority of patients.
Nevertheless, it has been shown that myocardial perfusion may be suboptimal in a
relevant percentage of patients. ST elevation segment elevation resolution is a cheap and
user-friendly method to evaluate myocardial perfusion. In an animal study the relation
between ST segment elevation and regional myocardial blood flow both 15 minutes and
24 hour after coronary occlusion was demonstrated.21 Several clinical studies have
described the relation between electrocardiographic changes after the start of
thrombolytic therapy and patency of the infarct-related vessel at follow-up coronary
angiography.22;23
Table 4. Infarct characteristics, hemodynamic status and primary reperfusion treatment
Characteristics
GIK group
Control group
310
302
Anterior MI
155 (50)
138 (46)
Killip class 1
284 (92)
285 (94)
26 (8)
17 (6)
Systolic BP, mmHg (mean±SD)
131±21
129±25
Diastolic BP, mmHg (mean±SD)
78±15
76±17
Heart rate, beats/min (mean±SD)
74±16
79±20
Time to admission, min (IQR)*
135 (99-210)
150 (105-240)
Door to balloon time, min (IQR)
49 (34-68)
51 (35-70)
Multi-vessel disease
162 (52)
150 (48)
PCI
299 (97)
289 (96)
Stent **
178 (57)
158 (52)
GP IIb/IIIa receptor blocker **
62 (20)
68 (23)
In-hospital CABG†
5 (1.6)
6 (2.0)
TIMI 3 flow after PCI
301 (97)
284 (94)
Successful reperfusion*
260 (84)
253 (84)
Intra-aortic balloon pump
20 (6.5)
21 (7.0)
Number of patients
Killip class ≥2
Data are number (%) unless otherwise indicated. MI = myocardial infarction. BP = blood pressure.
IQR = interquartile range. PCI = percutaneous coronary intervention. CABG = coronary artery
bypass grafting. *Time to admission denotes time between onset of symptoms and until admission.
**Percentage defined as the percentage of the patients that underwent PCI. †In-hospital CABG =
CABG defined as patients treated initially conservative, followed by elective CABG within 7 days.
‡Successful reperfusion denotes TIMI 3 flow and blush grade 2 or 3.
76
GIK and ST segment elevation resolution
Although complete ST segment elevation resolution predicts an open infarct artery
reasonably well, the absence of ST segment elevation resolution gives no information
about patency.24 It has been shown that the ECG will not predict infarct-vessel patency in
many patients, as almost half of the patients with persistent ST segment elevation had an
open epicardial vessel with TIMI 3 flow.6 ST segment changes after reperfusion therapy
may, therefore, reflect myocardial flow rather than epicardial flow and predict clinical
outcome better than epicardial vessel patency alone in patients treated with thrombolytic
therapy5;25, and primary PCI.6
Table 5. Comparison of patients with complete, partial and no ST segment elevation
resolution, according to treatment with GIK
GIK group
P for trend
Control group
Number of patients
310
<0.001
302
Complete resolution
198 (64)
174 (58)
Partial resolution
72 (23)
61 (20)
No resolution
40 (13)
67 (22)
Data are number (%). Resolution denotes resolution of the ST segment elevation.
GIK and ST segment elevation resolution
In several animal studies the effect of GIK on ST segment elevation resolution was
investigated. In dogs developing myocardial ischemia, GIK decreased epicardial ST
segment elevation.12 Decreased epicardial ST segment elevation correlates with a
reduction in the degree of eventual myocardial creatine kinase depletion and the
reduction in the histologic severity of myocardial necrosis.26 One study found that GIK
induced metabolic changes, such as reduced myocardial sodium and water and increased
lactate, were associated with decreased ST segment elevation, especially in the periinfarct zone.10 Another study showed that the infusion of GIK was associated with a
decrease in arterial plasma concentration of free fatty acids, a decrease in uptake of free
fatty acids, and a decreased degree of ST segment elevation.11 An effect that was not
observed in contols treated with saline. In 32 ST segment elevation MI patients were
treated with reteplase and alternately assigned to either GIK or saline with
potassiumchloride that ST segment elevation resolution at 90 minutes was >50% in 82%
of the GIK patients versus 62% in control patients (NS).15 Our results confirm these
observations: GIK results in a higher number of patients with partial or complete ST
segment elevation resolution.
77
Metabolic interventions in acute myocardial infarction
Table 6. 30-day mortality in patients treated with high-dose GIK and low-dose GIK versus
control patients of published trials
Included patient
Trial (ref)
30-day mortality
GIK patients
Control patients
GIK patients
Control patients
N
N
N (%)
N (%)
Heng
12
15
1 (8.3)
0 (-)
Stanley35
55
55
4 (7.3)
4 (7.3)
Rogers34
61
73
4 (6.5)
9 (12.3)
Satler37
10
7
0 (-)
0 (-)
135
139
10 (7.4)
16 (11.5)*
28
16
ECLA *
18
476
464
23 (4.8)
27 (5.8)
REVIVAL17
155
157
7 (4.5)
5 (3.2)
Total (high-dose)
904
910
49 (5.4)
61 (6.7)
Mittra
85
85
10 (11.8)
24 (23.5)
Pilcher33
49
53
6 (6.7)
12 (22.7)
100
100
15 (15.0)
16 (16.0)
GIPS
31
32
Pentecost
30
MRC
480
488
103 (21.5)
115 (23.6)
Hjermann 29
104
100
11 (9.6)
20 (20)
16
133
139
10 (7.5)
16 (11.5)*
ECLA *
P
0.30
36
Pol-GIK
494
460
44 (8.9)
22 (4.8)
Total (low-dose)
1445
1425
199 (13.8)
225 (15.8)
0.14
Total (all)
2349
2196
248 (10.6)
270 (12.3)
0.07
Data are number or number (%). Ref denotes reference. N denotes number of patients. ECLA =
Estudios Cardiologicos Latinoamerica. GIPS = Glucose-Insulin-Potassium Study. REVIVAL =
Reevaluation of Intensified Venous Metabolic Support for Acute Infarct Size Limitation. Pol-GIK =
Polish Glucose Insulin Potassium. *In both comparisons the same control group is used; in the total
of all patients this groups is accounted for once.
GIK and clinical outcome
The effect of GIK on 30-day mortality in ST segment elevation MI has been investigated by
several clinical trials.16-18;27-37 An overview of the results of published trials (October 2004)
is presented in table 6. Taken the results of all these trials together no significant beneficial
effect on 30-day mortality is observed, mortality in GIK patients is 10.6% versus 12.3% in
control patients (P=0.07). Care has to be taken since the individual trials differed in
design. Trials used GIK in several doses and for different length of time. A few studies
used GIK in combination with appropriate reperfusion therapy.16-18;37 When the results of
these trials are analyzed separately, 30-day mortality is 4.8% in GIK patients compared to
6.2% in control patients (P=0.27). The study by Satler and colleagues and the ECLA used
thrombolysis as reperfusion therapy. In the first trial only 17 patients were included and
78
GIK and ST segment elevation resolution
no one had died after 30-days. In the ECLA the observed effect in the sub-group of
patients treated primarily with thrombolysis the mortality rate in GIK patients was 5.2%
versus 15.2% in control patients (P=0.01) GIK in these patients seems promising, albeit
that this observation comes from a sub-group of a pilot trial in which the mortality rate in
the control group was high. The GIPS and REVIVAL treated patients primarily with primary
PCI.17;18 After 30-days the mortality rates in GIK and control patients was almost similar.
New trials as the ECLA GIK 2/CREATE trial, the Organization to Assess Strategies for
Ischemic Syndromes (OASIS)-6 trial and GIPS-2 currently randomize thousands of
patients to GIK or no GIK in adjunction to either thrombolysis and primary PCI. These trials
will determine whether GIK is of clinical benefit in combination with different reperfusion
strategies.
GIK and time to reperfusion
Based on the results of an experimental study it has been estimated that treatment with
GIK has the potential to protect ischemic myocardium before reperfusion for 10 hours or
even longer.38 Two studies in Sprague-Dawley rats found that GIK can preserve the
ischemic myocardium when started at reperfusion or even after reperfusion.39;40 The first
study observed that GIK treatment started at the onset of reperfusion reduced infarct size
to the same extent as GIK started throughout ischemia and reperfusion.39The other
showed that after 20 minutes of global ischemia contractile function was improved.
40
In
our study the time between admission and reperfusion averaged 45 minutes. The infusion
of GIK was started up to 30 minutes before PCI. We did not record the exact time of start
of the infusion and therefore did not perform additional analysis on the relation of the time
of infusion and ST segment elevation resolution. It can be suggested that GIK is likely to
have the most effect when started as soon as possible. However based on experimental
results, GIK may still induce favourable effects when started during or after reperfusion.
Mechanisms of action of GIK
There are several potential mechanisms of the effects of GIK in the treatment of acute
myocardial infarction.41 An important mechanism is the effect on substrate metabolism.
GIK induces a decrease in myocardial uptake of free fatty acids, and reduces the amount
of circulating free fatty acids and promotes the use of glucose as primary myocardial
energy substrate.42;43 Insulin may be the principle component of this therapy. It has been
demonstrated that hyperinsulinemia can increase coronary blood flow in patients with
significant coronary artery disease.38 Even a small increase in myocardial blood flow can
reduce myocardial ischemia.44 The vasodilative effects of insulin on coronary arteries are
probably similar in patients with and without diabetes mellitus.45;46 The observed relation
between GIK infusion and ST segment elevation resolution in our trial, i.e. microvascular
79
Metabolic interventions in acute myocardial infarction
reperfusion, support this concept. GIK infusion also resulted in an improvement of
regional myocardial perfusion and function mainly in segments adjacent to the infarcted
area as determined with technetium-99m-tetrofosmin-gated SPECT.47 Finally, insulin
promotes tolerance against ischemic cell death via the activation of innate cell-survival
pathways in the heart.48
Study limitations
Paired ECGs were available in only 65% of patients enrolled in the GIPS, due to various
reasons. Nevertheless, the groups were well-matched, and our analysis has adequate
statistical power. We performed a post-hoc analysis and our findings should be confirmed
in future prospective studies. Quantitative ST segment evaluation may be more accurate,
but recent post-hoc data from the Controlled Abciximab and Device Investigation to
Lower Late Angioplasty Complications (CADILLAC) trial have reconfirmed the usefulness
of visual grading in 3 categories.49 Grading of ST segment elevation resolution at a fixed
period of time after reperfusion therapy, in our study 3 hours, may have limitations.
Myocardial reperfusion is a dynamic process during which alternating episodes of ST
segment elevation resolution may occur.50 Fluctuation of the ST segment has been
described during or shortly after thrombolysis and primary PCI, and continuous ST
segment monitoring has shown to be a valid alternative.50;51 Additional treatment with
medication could influence early reperfusion and ST segment elevation resolution. We did
not record the time concomitant medication was started. Aspirin, heparin and
nitroglycerin were started early after admission. Clopidogrel was given to patients who
received a stent. However, there were no differences in pharmacotherapy between GIK
patients and controls (table 7).
Table 7. Additional treatment at hospital discharge according to treatment group
Medication
GIK group
Control group
Aspirin
421 (88.4)
398 (85.8)
Clopidogrel
194 (40.8)
192 (41.4)
Coumarins
36 (7.6)
45 (9.7)
Beta-blockers
392 (82.3)
381 (82.1)
ACE-inhibitors
245 (51.5)
244 (52.6)
AII receptor antagonists
7 (15)
3 (0.6)
15 (3.2)
19 (4.1)
Diuretics
46 (9.7)
67 (14.4)
Nitrates
70 (14.7)
67 (14.4)
Statins
285 (51.3)
271 (48.7)
Calcium channel blockers
Data are number (%).
80
GIK and ST segment elevation resolution
Conclusion
In patients with ST segment elevation MI, addition of GIK to primary PCI is associated with
improved ST segment elevation resolution. ST segment elevation resolution is related to
improved 1-year survival. A significant benefit of GIK on 1-year outcome was not
observed.
Contributors
This study was supported by a generous grant from the Netherlands Heart Foundation
(99.028). No other conflicts of interest or financial disclosure.
References
1.
De Lemos JA, Braunwald E. ST segment resolution as a tool for assessing the efficacy of
reperfusion therapy. J Am Coll Cardiol 2001;38:1283-94.
2.
Pepine CJ. Prognostic markers in thrombolytic therapy: looking beyond mortality. Am J
Cardiol 1996;78:24-27.
3.
Cooper HA, De Lemos JA, Morrow DA et al. Minimal ST segment deviation: a simple,
noninvasive method for identifying patients with a patent infarction-related artery after
fibrinolytic administration. Am Heart J 2002;144:790-795.
4.
Saran RK, Been M, Furniss SS, Hawkins T, Reid DS. Reduction in ST segment elevation after
thrombolysis predicts either coronary reperfusion or preservation of left ventricular function.
Br Heart J 1990;64:113-17.
5.
Schroder R, Dissmann R, Bruggemann T et al. Extent of early ST segment elevation
resolution: a simple but strong predictor of outcome in patients with acute myocardial
infarction. J Am Coll Cardiol 1994;24:384-91.
6.
van 't Hof AW, Liem A, de Boer MJ, Zijlstra F. Clinical value of 12-lead electrocardiogram after
successful reperfusion therapy for acute myocardial infarction. Zwolle Myocardial infarction
Study Group. Lancet 1997;350:615-19.
7. van 't Hof AW, Liem A, Suryapranata H, Hoorntje JC, de Boer MJ, Zijlstra F. Angiographic
assessment of myocardial reperfusion in patients treated with primary angioplasty for acute
myocardial infarction: myocardial blush grade. Zwolle Myocardial Infarction Study Group.
Circulation 1998;97:2302-6.
8.
Claeys MJ, Bosmans J, Veenstra L, Jorens P, De R, Vrints CJ. Determinants and prognostic
implications of persistent ST segment elevation after primary angioplasty for acute
myocardial infarction: importance of microvascular reperfusion injury on clinical outcome.
Circulation 1999;99:1972-77.
9.
Corbalan R, Larrain G, Nazzal C et al. Association of noninvasive markers of coronary artery
reperfusion to assess microvascular obstruction in patients with acute myocardial infarction
treated with primary angioplasty. Am J Cardiol 2001;88:342-46.
81
Metabolic interventions in acute myocardial infarction
10.
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.
11.
Haneda T, Ganz W, Burnam MH, Katz J. Metabolic effects of glucose-insulin-potassium in the
12.
Opie LH, Owen P. Effect of glucose-insulin-potassium infusions on arteriovenous differences
ischemic myocardium. Jpn Heart J 1978;19:376-82.
of glucose of free fatty acids and on tissue metabolic changes in dogs with developing
myocardial infarction. Am J Cardiol 1976;38:310-321.
13.
Ueno T, Ishiyama T, Morita Y, Hatanaka Y, Azuma J. An experimental study of the effect of
glucose-insulin-potassium solution on energy metabolism of infarcted cardiac muscle. Recent
Adv Stud Cardiac Struct Metab 1976;11:549-54.
14.
Sodi-Pallares D, Testelli MR, Fishleder BL et al. Effects of an intravenous infusion of a
potassium-glucose-insulin solution on the electrocardiographic signs of myocardial infarction.
A preliminary clinical report. Am J Cardiol 1962;9:166-81.
15.
Chaudhuri A, Janicke D, Wilson MF et al. Anti-inflammatory and profibrinolytic effect of
16.
Diaz R, Paolasso EA, Piegas LS et al. Metabolic modulation of acute myocardial infarction.
insulin in acute ST segment-elevation myocardial infarction. Circulation 2004;109:849-54.
The
ECLA
(Estudios
Cardiologicos
Latinoamerica)
Collaborative
Group.
Circulation
1998;98:2227-34.
17.
Pache J, Kastrati A, Mehilli J et al. A randomized evaluation of the effects of glucose-insulinpotassium infusion on myocardial salvage in patients with acute myocardial infarction treated
with reperfusion therapy. Am Heart J 2004;148:105.
18.
van der Horst IC, Zijlstra F, van't Hof AW et al. Glucose-insulin-potassium infusion in patients
treated with primary angioplasty for acute myocardial infarction: the glucose-insulinpotassium study: a randomized trial. J Am Coll Cardiol 2003;42:784-91.
19.
Ryan TJ, Antman EM, Brooks NH et al. 1999 update: ACC/AHA Guidelines for the
Management of Patients With Acute Myocardial Infarction: Executive Summary and
Recommendations: A report of the American College of Cardiology/American Heart
Association Task Force on Practice Guidelines (Committee on Management of Acute
Myocardial Infarction). Circulation 1999;100:1016-30.
20.
Van de Werf F, Ardissino D, Betriu A et al. Management of acute myocardial infarction in
patients presenting with ST segment elevation. The Task Force on the Management of Acute
Myocardial Infarction of the European Society of Cardiology. Eur Heart J 2003;24:28-66.
21.
Heng MK, Singh BN, Norris RM, John MB, Elliot R. Relationship between epicardial ST
segment elevation and myocardial ischemic damage after experimental coronary artery
occlusion in dogs. J Clin Invest 1976;58:1317-26.
22.
Klootwijk P, Langer A, Meij S et al. Non-invasive prediction of reperfusion and coronary artery
patency by continuous ST segment monitoring in the GUSTO-I trial. Eur Heart J 1996;17:68998.
23.
Clemmensen P, Ohman EM, Sevilla DC et al. Changes in standard electrocardiographic ST
segment elevation predictive of successful reperfusion in acute myocardial infarction. Am J
Cardiol 1990;66:1407-11.
82
GIK and ST segment elevation resolution
24.
Califf RM, O'Neil W, Stack RS et al. Failure of simple clinical measurements to predict
25.
Barbash GI, Roth A, Hod H et al. Rapid resolution of ST elevation and prediction of clinical
perfusion status after intravenous thrombolysis. Ann Intern Med 1988;108:658-62.
outcome in patients undergoing thrombolysis with alteplase (recombinant tissue-type
plasminogen activator): results of the Israeli Study of Early Intervention in Myocardial
Infarction. Br Heart J 1990;64:241-47.
26.
Maroko PR, Kjekshus JK, Sobel BE et al. Factors influencing infarct size following
experimental coronary artery occlusions. Circulation 1971;43:67-82.
27. Fath-Ordoubadi F, Beatt KJ. Glucose-insulin-potassium therapy for treatment of acute
myocardial infarction: an overview of randomized placebo-controlled trials. Circulation
1997;96:1152-56.
28.
Heng MK, Norris RM, Singh BN, Barratt-Boyes C. Effects of glucose and glucose-insulinpotassium on haemodynamics and enzyme release after acute myocardial infarction. Br Heart
J 1977;39:748-57.
29.
Hjermann I. A controlled study of peroral glucose, insulin and potassium treatment in
30.
Medical Research Council Working Party on the Treatment of Myocardial Infarction.
myocardial infarction. Acta Med Scand 1971;190:213-18.
Potassium, glucose, and insulin treatment for acute myocardial infarction. Lancet
1968;2:1355-60.
31.
Mittra B. Potassium, glucose, and insulin in treatment of myocardial infarction. Lancet
1965;2:607-9.
32.
Pentecost BL, Mayne NM, Lamb P. Controlled trial of intravenous glucose, potassium, and
33.
Pilcher J, Etishamudin M, Exon P, Moore J. Potassium, glucose and insulin in myocardial
insulin in acute myocardial infarction. Lancet 1968;1:946-48.
infarction. Lancet 1967;1:1109.
34.
Rogers WJ, Stanley AW, Jr., Breinig JB et al. Reduction of hospital mortality rate of acute
35.
Stanley AW, Jr., Prather JW. Glucose-insulin-potassium, patient mortality and the acute
36.
Ceremuzynski L, Budaj A, Czepiel A et al. Low-dose glucose-insulin-potassium is ineffective in
myocardial infarction with glucose-insulin-potassium infusion. Am Heart J 1976;92:441-54.
myocardial infarction: results from a prospective randomised study. Circulation 1978;57:61.
acute myocardial infarction: results of a randomized multicenter Pol-GIK trial. Cardiovasc
Drugs Ther 1999;13:191-200.
37.
Satler LF, Green CE, Kent KM, Pallas RS, Pearle DL, Rackley CE. Metabolic support during
38.
Eberli FR, Weinberg EO, Grice WN, Horowitz GL, Apstein CS. Protective effect of increased
coronary reperfusion. Am Heart J 1987;114:54-58.
glycolytic substrate against systolic and diastolic dysfunction and increased coronary
resistance from prolonged global underperfusion and reperfusion in isolated rabbit hearts
perfused with erythrocyte suspensions. Circ Res 1991;68:466-81.
39.
Jonassen AK, Aasum E, Riemersma RA, Mjos OD, Larsen TS. Glucose-insulin-potassium
reduces infarct size when administered during reperfusion. Cardiovasc Drugs Ther
2000;14:615-23.
83
Metabolic interventions in acute myocardial infarction
40.
Angelos MG, Murray HN, Gorsline RT, Klawitter PF. Glucose, insulin and potassium (GIK)
during reperfusion mediates improved myocardial bioenergetics. Resuscitation 2002;55:32936.
41.
Apstein CS. Increased glycolytic substrate protection improves ischemic cardiac dysfunction
42.
McDaniel HG, Papapietro SE, Rogers WJ et al. Glucose-insulin-potassium induced alterations
and reduces injury. Am Heart J 2000;139:S107-S114.
in individual plasma free fatty acids in patients with acute myocardial infarction. Am Heart J
1981;102:10-15.
43.
Rackley CE, Russell RO, Jr., Rogers WJ, Mantle JA, McDaniel HG, Papapietro SE. Clinical
experience with glucose-insulin-potassium therapy in acute myocardial infarction. Am Heart J
1981;102:1038-49.
44.
Apstein CS, Deckelbaum L, Mueller M, Hagopian L, Hood WB, Jr. Graded global ischemia and
45.
McNulty PH, Pfau S, Deckelbaum LI. Effect of plasma insulin level on myocardial blood flow
46.
Sundell J, Laine H, Nuutila P et al. The effects of insulin and short-term hyperglycaemia on
reperfusion. Cardiac function and lactate metabolism. Circulation 1977;55:864-72.
and its mechanism of action. Am J Cardiol 2000;85:161-65.
myocardial blood flow in young men with uncomplicated Type I diabetes. Diabetologia
2002;45:775-82.
47.
Marano L, Bestetti A, Lomuscio A et al. Effects of infusion of glucose-insulin-potassium on
48.
Sack MN, Yellon DM. Insulin therapy as an adjunct to reperfusion after acute coronary
myocardial function after a recent myocardial infarction. Acta Cardiol 2000;55:9-15.
ischemia: a proposed direct myocardial cell survival effect independent of metabolic
modulation. J Am Coll Cardiol 2003;41:1404-7.
49.
McLaughlin MG, Stone GW, Aymong E et al. Prognostic utility of comparative methods for
assessment of ST segment resolution after primary angioplasty for acute myocardial
infarction: the Controlled Abciximab and Device Investigation to Lower Late Angioplasty
Complications (CADILLAC) trial. J Am Coll Cardiol 2004;44:1215-23.
50.
Shah PK, Cercek B, Lew AS, Ganz W. Angiographic validation of bedside markers of
51.
Krucoff MW, Croll MA, Pope JE et al. Continuous 12-lead ST segment recovery analysis in the
reperfusion. J Am Coll Cardiol 1993;21:55-61.
TAMI 7 study. Performance of a noninvasive method for real-time detection of failed
myocardial reperfusion. Circulation 1993;88:437-46.
84
GIK and metabolic derangements
Chapter 2.5
Glucose and potassium derangements by glucose-insulinpotassium infusion related to 1-year mortality in acute
myocardial infarction
Iwan CC van der Horst, Jorik R Timmer, Jan Paul Ottervanger, Henk JG Bilo, Kor
Miedema, Rijk OB Gans, Menko-Jan de Boer, Maarten WN Nijsten, Felix Zijlstra
Submitted
85
Metabolic interventions in acute myocardial infarction
Abstract
Background
Infusion of high-dose glucose-insulin-potassium infusion (GIK) appears to be beneficial in
acute myocardial infarction (MI). It has been shown that GIK induces derangement in
systemic glucose and potassium levels. We thought to investigate the effect of metabolic
derangements on outcome in MI patients.
Methods
Patients with signs and symptoms of ST segment elevation MI and treated with primary
percutaneous coronary intervention (PCI) were randomized to no infusion or high-dose
GIK, i.e. 80 mmol potassium chloride in 500 mL 20% glucose with a rate of 3 mL/kg/hour
and 50 units short-acting insulin in 50 mL 0.9% sodium chloride for 12 hours. We have
investigated the effect of GIK infusion on serum values of glucose and potassium and the
relation between derangement and 1-year mortality.
Results
In 940 patients a total of 6991 glucose values and 7198 potassium values were obtained.
Mean serum glucose was 9.3±4.5 mmol/L in the GIK group (N=476) compared to 8.4±2.9
mmol/L in the control group (N=464) (P<0.001). Hyperglycemia (glucose >11.0 mmol/L)
occurred in 337 patients (70.8%) treated with GIK and in 157 controls (33.8%) (P<0.001).
A total of 48 hyperglycemic patients (9.4%) died versus 21 patients (4.7%) without
hyperglycemia (P=0.004). In patients with hyperglycemia 1-year mortality was 8.3% in the
GIK group versus 12.7% in controls (P=0.14). Hypoglycemia (glucose ≤3.0 mmol/L)
occurred in 10 patients of whom 2 patients (20%) died, both treated with GIK.
Hyperkalemia (potassium >5.5 mmol/L) was present in 26 GIK patients (5.5%) and 15
controls (3.2%) (P=0.11), and of these patients 9 (34.6%) in the GIK group and 9 (60%) in
the control group died (P=0.19). Hypokalemia (potassium ≤3.5 mmol/L) occurred in 112
patients (23.5%) in the GIK group and 191 patients in the control group (41.2%) (P<0.001).
A total of 30 patients (9.9%) with hypokalemia died, 10 GIK patients (8.9%) and 20
controls (10.5%) (P=0.84).
Conclusion
In ST segment elevation MI patients treated with primary PCI high-dose GIK induced
hyperglycemia
and
hyperkalemia
and
prevented
hypokalemia.
These
metabolic
derangements were related to 1-year mortality. The beneficial effect of GIK on mortality
may be dimished by these derangements.
86
GIK and metabolic derangements
Introduction
In the past three decades an extensive body of evidence has accumulated that
underscores the importance of glucose metabolism during ischemia and reperfusion of
the heart. A meta-analysis involving nine early clinical studies with 1932 patients indicated
that GIK might play an important role in reducing in-hospital mortality after acute
myocardial infarction (MI).1 In the four studies in which a higher dose was used, a
reduction in mortality from 12% in the control group to 6.5% in the GIK group was
observed.2-5 In these studies high-dose GIK denoted an infusion of at least 30 grams of
glucose, 50 units of insulin, and 80 mmol potassium per liter at a rate of 1.5 mL/kg/hour
This dose maximally suppressed arterial free fatty acids (FFA) levels and myocardial FFA
uptake, and induced a maximal increase in myocardial glucose uptake.6 Thereafter in the
Estudios Cardiologicos Latinoamerica (ECLA) pilot trial the combination of reperfusion and
GIK was related to a 10% 30-day mortality reduction (5.2% versus 15.2%, P<0.001). In a
randomized trial with 940 patients we studied if primary percutaneous coronary
intervention (PCI) with high-dose GIK was beneficial.7 After 30 days the mortality rate in
the overall population was 4.8% in the GIK group compared to 5.8% in the control group
(P=0.50). In the large predefined subgroup of 856 patients without signs of heart failure
the mortality reduction was 3.0% (1.2% versus 4.2%, P=0.01).
In two early studies with 30 and 104 patients the infusion of GIK in MI patients induced
hyperglycemia and hyperkalemia.8;9 The ECLA observed no significant differences
between mean glucose levels in the GIK and control group during 48 hours.10 The mean
potassium level was higher at 24 and 48 hours. The relation between these changes and
outcome were not investigated. We shought to determine the relation between metabolic
derangements induced by GIK infusion and 1-year mortality.
Methods
Study population
Patients with symptoms and signs of ST segment elevation MI eligible for primary PCI
were randomized to GIK or no infusion. GIK consisted of a continuous infusion of 80 mmol
potassium chloride in 500 mL 20% glucose with a rate of 3 mL/kg/hour in a peripheral
venous line and 50 units short-acting insulin in 50 mL 0.9% sodium chloride. The dose
and adjustments of insulin infusion were based on a modified algorithm, to obtain bloodglucose levels between 7.0 and 11.0 mmol/L (figure 1). The infusion was continued for a
maximum of 12 hours. All patients went to the catheterization laboratory where PCI was
performed if the coronary anatomy was suitable. The research protocol was reviewed and
87
Metabolic interventions in acute myocardial infarction
approved by the medical ethics committee, and patients were included after informed
consent.
Figure 1. Algorithm of insulin infusion rate in patients randomized to the GIK group.
Scheme 1: Starting dose of insulin infusion rate
Glucose (mmol/L)
Additional action
>15
8 IU short-acting insulin i.v.
Infusion rate (IU/hour)
30
11.0-15.0
30
10.0-10.9
24
9.0-9.9
18
8.0-8.9
12
7.0-7.9
6
<7
3
Scheme 2: Hourly adjustment of insulin infusion rate based on hourly measured glucose
Glucose level (mmol/L)
Additional action
>15
8 IU short-acting insulin i.v.
Infusion rate (IU/hour)
Increase with 6
11.0-15.0
Increase with 3
7.0-10.9
Leave unchanged
4.0-6.9
Decrease with 6
<4
1. Stop infusion for 15 minutes
2. Measure blood glucose level every 15 minutes until >6.9
mmol/L
3. Start infusion with infusion rate 6 mL/hour beneath rate
before stop
4. If symptoms of hypoglycemia are present than give 20 mL of
glucose 30% i.v.
Metabolic determinants
In all patients frequent measurements of glucose and potassium (Modular System,
Roche/Hitachi, Basel, Switzerland) were obtained. We defined hyperglycemia as a glucose
level >11.0 mmol/L11, hypoglycemia as a glucose ≤3.0 mmol/L12, hyperkalemia as a
potassium level of ≥5.5 mmol/L13, and hypokalemia as a potassium level ≤3.5 mmol/L14.
We compared the incidence of glucose or potassium values in GIK and control group, and
in addition determined the relation between these derangements and 1-year mortality.
88
GIK and metabolic derangements
Statistical analysis
Data were expressed as means with standard deviation unless otherwise indicated.
Differences between group means at baseline were assessed with the two-tailed Student’s
t-test. Chi-square analysis was used to test differences between proportions. Survival was
calculated by the Kaplan-Meier product-limit method. The Log-rank test was used to
evaluate differences in survival curves between the two treatment groups. The Cox
proportional-hazards regression model was used to calculate relative risks adjusted for
differences in baseline characteristics. Differences were considered significant for a twotailed P-value <0.05. The Statistical Package for the Social Sciences (SPSS Inc., Chicago,
IL, USA; version 11.0.1) was used for statistical analysis.
Figure 2. Mean glucose level over 96 hours after admission
25
GIK group
Glucose level (mmol/L)
20
Control group
15
10
5
0
0
10
20
30
40
50
60
70
80
90
100
Time after admission (hours)
Mean±SD glucose level in patients in the GIK group (unbroken line) and control group (dashed line).
Results
A total of 6991 glucose values and 7198 potassium values were analyzed in 940 patients.
Baseline characteristics of patients randomized to GIK (N=476) or no infusion (N=464) are
represented in table 1. Mean glucose level was 9.3±4.5 mmol/L in the GIK group
compared to 8.4±2.9 mmol/L in the control group (P<0.001), see table 2. The mean
glucose values were significantly higher in the GIK patients between 2 and 6 hours after
admission (figure 2). Hyperglycemia occurred in 337 GIK patients and in 157 control
89
Metabolic interventions in acute myocardial infarction
patients (70.8% versus 33.8%, P<0.001). A total of only 10 patients developed
hypoglycemia, 3 patients in the GIK group versus 7 patients in the control group (0.6%
versus 1.5%, P=0.34). Mean potassium level was 4.2±0.5 mmol/L in the GIK group versus
3.9±0.39 mmol/L in the control group (P<0.001), see table 2. The potassium levels were
significantly higher in the GIK group between 2 and 48 hours (figure 3). A total of 26
patients (5.5%) treated with GIK versus 15 patients (3.2%) who received no infusion had a
potassium level of 5.5 mmol/L or higher (P=0.11). Hypokalemia was observed less often
with 112 patients (23.5%) in the GIK group versus 191 patients (41.2%) in the control
group (P<0.001).
Figure 3. Mean potassium level over 96 hours after admission
4.9
GIK group
Potassium level (mmol/L)
4.7
Control group
4.5
4.3
4.1
3.9
3.7
3.5
0
10
20
30
40
50
60
70
80
90
100
Time after admission (hours)
Mean ± SD potassium level in patients in the GIK group (unbroken line) and control group (dashed
line).
Mortality at 1 year
Thirty-one patients (6.5%) in the GIK group died compared to 38 patients (8.2%) in the
control group (relative risk (RR) 0.79, 95% confidence interval 0.49–1.27, P=0.32). In the
856 patients (91.1%) without signs of heart failure (Killip class 1) 11 patients (2.6%) died in
the GIK group compared to 28 patients (6.5%) in the control group RR 0.38 (0.19–0.78,
P<0.01), see figure 4. In the 84 patients (8.9%) with signs of heart failure (Killip class ≥2)
20 patients (40%) in the GIK group versus 10 patients (29.4%) in the control group died RR
1.46 (0.68–3.11, P=0.32). After adjustment for baseline variables independently associated
with 1-year mortality in the overall population, the RR of mortality with GIK was 0.63
90
GIK and metabolic derangements
(0.39–1.04, P=0.07). In Killip class 1 patients the beneficial effect of GIK remained
significant with an adjusted RR of 0.38 (0.19–0.78, P<0.01). In Killip class ≥2 patients an
adjusted RR of 1.31 (0.55–3.14, P=0.54) was observed.
Table 1. Baseline characteristics of the Glucose Insulin Potassium Study
Characteristics
GIK group
Control group
Age, years (mean±SD)
59.9±11.9
60.8±12.0
Men
351 (73.7)
368 (79.3)
Referred patients
201 (42.3)
180 (38.8)
Body mass index (mean±SD)
26.7±3.8
27.0±4.0
Previous MI
52 (10.9)
53 (11.4)
Previous PCI
22 (4.7)
24 (5.2)
Previous CABG
14 (2.9)
12 (2.6)
History of stroke
17 (3.6)
15 (3.2)
Diabetes mellitus
50 (10.5)
49 (10.6)
Hypertension
134 (28.2)
130 (28.0)
Dyslipidemia
94 (19.7)
95 (20.5)
Currently smoker
225 (47.3)
237 (51.1)
Positive family history
195 (41.0)
179 (38.6)
Time to admission, min (IQR)*
150 (100-215)
150 (105-234)
Door to balloon time, min (IQR)
45 (31-64)
48 (34-69)
Killip class 1
426 (89.5)
430 (92.7)
Killip class ≥2
50 (10.5)
34 (7.3)
Anterior MI
250 (52.9)
224 (49.1)
Multi-vessel disease
249 (52.3)
242 (52.2)
PCI
436 (91.6)
424 (91.4)
Stent **
255 (58.5)
236 (55.7)
GP IIb/IIIa receptor blocker **
96 (22.1)
109 (25.7)
In-hospital CABG
19 (4.0)
19 (4.1)
TIMI 3 flow after PCI
459 (96.4)
435 (93.8)
Successful reperfusion‡
394 (82.8)
384 (82.8)
Intra-aortic balloon pump
45 (9.5)
42 (9.1)
Mechanical ventilation
12 (2.5)
4 (0.9)
Data are number (%) unless otherwise indicated. MI = myocardial infarction. IQR = interquartile
range. PCI = percutaneous coronary intervention. CABG = coronary artery bypass grafting. * Time
to admission denotes time between onset of symptoms and until admission. † In-hospital CABG =
CABG defined as patients treated initially conservative, followed by elective CABG within 7 days. ‡
Successful reperfusion denotes TIMI 3 flow and blush grade 2 or 3. ** Percentage defined as the
percentage of the patients that underwent PCI.
91
Metabolic interventions in acute myocardial infarction
Metabolic derangements and 1-year mortality
In 494 patients with hyperglycemia during admission 48 patients (9.4%) died versus 21
patients (4.7%) out of 446 patients without hyperglycemia RR 2.18 (1.28–3.70, P=0.004).
The number of patients who died with hyperglycemia in the GIK group were 28 (8.3%) out
of 337 patients compared to 20 patients (12.7%) out of 157 controls RR 0.62 (0.34–1.14,
P=0.14). Only 10 patients had a glucose level equal to or below 3.0 mmol/L and 2 of these
patients (20%) died. Both were randomized to infusion with GIK and both PCI procedures
were not successful. One died on day 2 and the other on day 4 after admission. The
lowest glucose value was 2.8 mmol/L and 3.0 mmol/L in these patients.
A total of 41 patients had hyperkalemia and 18 patients (43.9%) did not survive after 1
year. In the GIK group 26 patients had hyperkalemia and 9 patients (34.6%) died
compared to 9 patients (60%) out of 15 patients in the control group RR 0.35 (0.10–1.31,
P=0.19). Hypokalemia was present in 303 patients and 30 patients (9.9%) died. In the GIK
group 10 patients (8.9%) out of 112 patients compared to 20 (10.5%) out of 191 control
patients did not survive to 1 year, RR 0.84 (0.38–1.86, P=0.84).
Table 2. Metabolic data of GIK group and control group patients
Glucose level (mean±SD)
GIK group
Control group
P-value
9.3±4.5
8.4±2.9
<0.001
Range
2.2–39.0
1.7–41.0
5%-95% percentiles
5.3–18.9
5.6–13.9
Hyperglycemia (%)
337 (70.8)
157 (33.8)
<0.001
Hypoglycemia (%)
3 (0.6)
7 (1.5)
0.34
Potassium level (mean±SD)
4.2±0.5
3.9±0.4
<0.001
Range
2.3–6.9
2.4–7.3
5%-95% percentiles
3.5–4.9
3.4–4.6
Hyperkalemia (%)
26 (5.5)
15 (3.2)
0.11
Hypokalemia (%)
112 (23.5)
191 (41.2)
<0.001
Discussion
In ST segment elevation MI patients treated with primary PCI infusion of high-dose GIK is
related to significantly higher glucose levels during the first hours after admission.
Hyperglycemia is observed in at least 7 out of 10 patients treated with GIK. This is more
than double the rate of patients who received no infusion. The relation between GIK
infusion and hyperglycemia was found by others. Prather and colleagues found that the
mean glucose values were higher in 18 patients treated with GIK compared to 7 untreated
controls during the infusion period of 48 hours.8 A relation that was also present in a study
92
GIK and metabolic derangements
by Sysoeva and colleagues comparing 56 patients treated with GIK with 48 controls.9 In
the ECLA pilot trial including 407 patients no significant difference was observed in mean
glucose values during the 24 hour infusion between GIK and control patients10. In the lowdose Polish GIK (Pol-GIK) trial the mean glucose level in 494 GIK patients was even lower
in 460 control patients after 24 hours (5.9 mmol/L versus 6.2 mmol/L). In this study the
amount of insulin infusion was decreased because of a glucose level <3.4 mmol/L in 35
(9.4%) out of 369 patients. In our study the amount of GIK infused however induced
hyperglycemia and it has been suggested that hyperglycemia is related to impaired
preconditioning, increased infarct size, and no-reflow phenomenon.15;16 Hyperglycemia is
related to stress induced elevation of free fatty acids that may compromise myocardial
function as they reduce contractility and increase ischemic and reperfusion injury.
Possibly the potential beneficial effect of GIK could be diminished by the induction of
hyperglycemia. Hypoglycemia was uncommon, and only 2 patients with hypoglycemia
died, probably related to the sequel of unsuccessful reperfusion. In experimental studies
hypoglycemia induced myocardial damage.17 In clinical studies it was found that
hypoglycemia was related to increased in-hospital mortality.18
The infusion of potassium as part of the GIK infusion in our study had an only minor effect
on the mean potassium level of the GIK group. That the differences in potassium levels
between both groups persisted for 48 hours can be understood from the fact that
potassium needs to be cleared by the kidneys. In the Prather study the mean potassium
level was approximately 1.5 mmol/L higher in GIK patients after 24 hours till 96 hours after
admission.8 In the ECLA the potassium levels were also significantly higher (4.25 mmol/L
versus 4.04 mmol/L at 24 hours (P=0.0001) and 4.24 mmol/L versus 4.08 mmol/L at 48
hours (P=0.0027)). The number of GIK patients with hyperkalemia was somewhat higher
and with hypokalemia a bit reduced. In the Pol-GIK trial 4 GIK patients (0.8%) had
hyperkalemia and 3 patients (0.7%) had hypokalemia.13 The relation between
hyperkalemia and 1-year mortality was especially present in the control group. Avoiding
hypokalemia to be beneficial in several cardiovascular disease states including acute MI.19
The relation between hypokalemia and arrhythmias is found in some but not all
studies.14;20;21 Sodi-Pollares and colleagues were the first to suggest that GIK was
beneficial in the treatment of ischemic myocardial arrhythmias.22 It was observed that GIK
infusion limited changes on the electrocardiogram.
In our study GIK had no significant effect on 1-year mortality in the overall population. In
patients without heart failure at admission the beneficial effect that was found after 30days is still present.7 The ECLA pilot trial was the first to show a beneficial effect of GIK
and reperfusion on in-hospital mortality.10 This effect was observed in the subgroup of
93
Metabolic interventions in acute myocardial infarction
patients that also underwent thrombolysis, in-hospital mortality was 5.2% in 173 patients
who received GIK versus 15.2% in 79 controls (P<0.01).
Figure 4. Kaplan-Meier curve of Killip class 1 and Killip class ≥2 patients
100
Survival (%)
90
80
70
60
0
0
50
100
150
200
250
300
350
Time after randomization (days)
GIK - Killip 1
GIK - Killip ≥2
Control - Killip 1
Control - Killip ≥2
Kaplan-Meier curve showing cumulative survival of Killip 1 class and Killip class ≥2 patients
randomized to GIK group or control group. Differences between the groups in Killip class 1 were
significant (unadjusted P=0.01). Differences between the groups in Killip class ≥2 were nonsignificant (unadjusted P=0.32). P-value was determined with the use of the Log-rank test.
The main effect of the GIK infusion was considered to be the beneficial effect of
administration of glucose to the ischemic myocardium. The infusion of insulin during
myocardial ischemia was shown to have anti-inflammatory and anti-apoptotic effects.23-25
It is suggested that infusion of insulin (and glucose) would increase contractile force.4 In
contrast, others found that GIK made no difference in hemodynamic parameters in 110
patients.5 The results of intensive insulin therapy in critically ill patients to maintain blood
glucose below 6.1 mmol/l gives direction for metabolic interventions directed towards
strict glucose control.26 Although this study focussed on the effects of strict glucose
control as an important factor in explaining the positive results, the results may also be
interpreted differently. In most GIK studies there was a trend towards significance despite
the fact that normal glucose levels were not obtained.1;7;10 The beneficial effect of an
94
GIK and metabolic derangements
infusion of insulin during myocardial ischemia has been suggested by a small number of
clinical studies, although they lacked a randomised design and were not conclusive.27;28
The most favourable results of insulin treatment in acute MI were observed in the
Diabetes Insulin-Glucose in Acute Myocardial Infarction (DIGAMI) study. The DIGAMI
included 620 patients with diabetes mellitus or glucose at admission >11.0 mmol/L were
included.29;30 The combination of insulin-glucose infusion followed by intensive insulin
treatment resulted in both better glycemic control in the insulin-glucose group 24 hours
after randomization, with a blood-glucose level of 9.6 versus 11.7 mmol/L in the control
group (P<0.0001). The reduction of in hospital mortality from 11.1% in the control group
compared to 9.1% in the insulin-glucose group was non-significant.29;30 Data on the effect
in patients who received thrombolysis (50% of all patients) were not reported. At 1-year
follow-up, mortality was lower in the insulin-glucose group, with 18.6% versus 26.1% in
the control group (P=0.0273). In a retrospective study including 3554 diabetic patients
treated with coronary artery bypass grafting (CABG) glucose levels were lower in patients
regulated with continuous insulin infusion compared to subcutaneous insulin injections
(9.8 mmol/L versus 11.8 mmol/L, P <0.0001).31 The observed mortality was 2.5% in 2612
patients treated continuously versus 5.3% in 942 patients treated intermittent (P <0.0001).
Likewise, a recent study of 141 diabetic CABG patients found that treatment with GIK
resulted in lower glucose levels (7.7 mmol/L versus 14.4 mmol/L, P<0.0001) and after 2
years a reduction of mortality (P=0.04).32
Conclusion
This study shows that in ST segment elevation MI patients treated with primary PCI
infusion of high-dose GIK is related to both hyperglycemia and hyperkalemia, and
prevention of hypokalemia. Possibly the beneficial effect of GIK on mortality was reduced
by inducing these derangements. Substantial experimental evidence underscores the
potential beneficial effects of metabolic treatment with glucose, insulin, and potassium for
the heart exposed to the injuries of ischemia and reperfusion. Future trials have to
determine whether or not to aim for strict glucose regulation during acute MI, and which
algorithm allows a feasible and effective infusion of high-dose GIK without inducing
metabolic disturbances.
Contributors
This study was supported by a generous grant from the Netherlands Heart Foundation
(99.028). No other conflicts of interest or financial disclosure.
95
Metabolic interventions in acute myocardial infarction
References
1.
Fath-Ordoubadi F, Beatt KJ. Glucose-insulin-potassium therapy for treatment of acute
myocardial infarction: an overview of randomized placebo-controlled trials. Circulation
1997;96:1152-56.
2.
Heng MK, Norris RM, Singh BN, Barratt-Boyes C. Effects of glucose and glucose-insulinpotassium on haemodynamics and enzyme release after acute myocardial infarction. Br Heart
J 1977;39:748-57.
3.
Rogers WJ, Stanley AW, Jr., Breinig JB et al. Reduction of hospital mortality rate of acute
myocardial infarction with glucose-insulin-potassium infusion. Am Heart J 1976;92:441-54.
4.
Satler LF, Green CE, Kent KM, Pallas RS, Pearle DL, Rackley CE. Metabolic support during
coronary reperfusion. Am Heart J 1987;114:54-58.
5.
Stanley AW, Jr., Prather JW. Glucose-insulin-potassium, patient mortality and the acute
myocardial infarction: results from a prospective randomised study. Circulation 1978;57:61.
6. Rackley CE, Russell RO, Jr., Rogers WJ, Mantle JA, McDaniel HG, Papapietro SE. Clinical
experience with glucose-insulin-potassium therapy in acute myocardial infarction. Am Heart J
1981;102:1038-49.
7.
van der Horst IC, Zijlstra F, van't Hof AW et al. Glucose-insulin-potassium infusion in patients
treated with primary angioplasty for acute myocardial infarction: the glucose-insulinpotassium study: a randomized trial. J Am Coll Cardiol 2003;42:784-91.
8.
Prather JW, Russell RO, Jr., Mantle JA, McDaniel HG, Rackley CE. Metabolic consequences of
glucose-insulin-potassium infusion in treatment of acute myocardial infarction. Am J Cardiol
1976;38:95-99.
9.
Sysoeva NA, Oganov RG, Mikhailova IL. [Metabolic effects of a glucose-insulin-potassium
mixture in acute myocardial infarct]. Kardiologiia 1982;22:77-82.
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.
11.
Bolk J, van der Ploeg T, Cornel JH, Arnold AE, Sepers J, Umans VA. Impaired glucose
12.
Malmberg KA, Efendic S, Ryden LE. Feasibility of insulin-glucose infusion in diabetic patients
metabolism predicts mortality after a myocardial infarction. Int J Cardiol 2001;79:207-14.
with acute myocardial infarction. A report from the multicenter trial: DIGAMI. Diabetes Care
1994;17:1007-14.
13.
Ceremuzynski L, Budaj A, Czepiel A et al. Low-dose glucose-insulin-potassium is ineffective in
acute myocardial infarction: results of a randomized multicenter Pol-GIK trial. Cardiovasc
Drugs Ther 1999;13:191-200.
14.
Madias JE, Shah B, Chintalapally G, Chalavarya G, Madias NE. Admission serum potassium in
patients with acute myocardial infarction: its correlates and value as a determinant of inhospital outcome. Chest 2000;118:904-13.
15.
Iwakura K, Ito H, Ikushima M et al. Association between hyperglycemia and the no-reflow
16.
Kersten JR, Schmeling TJ, Orth KG, Pagel PS, Warltier DC. Acute hyperglycemia abolishes
phenomenon in patients with acute myocardial infarction. J Am Coll Cardiol 2003;41:1-7.
ischemic preconditioning in vivo. Am J Physiol 1998;275:H721-H725.
96
GIK and metabolic derangements
17.
Libby P, Maroko PR, Braunwald E. The effect of hypoglycemia on myocardial ischemic injury
18.
Fischer KF, Lees JA, Newman JH. Hypoglycemia in hospitalized patients. Causes and
19.
Macdonald JE, Struthers AD. What is the optimal serum potassium level in cardiovascular
20.
Boyd JC, Bruns DE, DiMarco JP, Sugg NK, Wills MR. Relationship of potassium and
21.
Higham PD, Adams PC, Murray A, Campbell RW. Plasma potassium, serum magnesium and
during acute experimental coronary artery occlusion. Circulation 1975;51:621-26.
outcomes. N Engl J Med 1986;315:1245-50.
patients? J Am Coll Cardiol 2004;43:155-61.
magnesium concentrations in serum to cardiac arrhythmias. Clin Chem 1984;30:754-57.
ventricular fibrillation: a prospective study. Q J Med 1993;86:609-17.
22.
Sodi-Pallares D, Testelli MR, Fishleder BL et al. Effects of an intravenous infusion of a
potassium-glucose-insulin solution on the electrocardiographic signs of myocardial infarction.
A preliminary clinical report. Am J Cardiol 1962;9:166-81.
23.
Chaudhuri A, Janicke D, Wilson MF et al. Anti-inflammatory and profibrinolytic effect of
24.
Gao F, Gao E, Yue TL et al. Nitric oxide mediates the antiapoptotic effect of insulin in
insulin in acute ST-segment-elevation myocardial infarction. Circulation 2004;109:849-54.
myocardial ischemia-reperfusion: the roles of PI3-kinase, Akt, and endothelial nitric oxide
synthase phosphorylation. Circulation 2002;105:1497-502.
25.
Jonassen AK, Sack MN, Mjos OD, Yellon DM. Myocardial protection by insulin at reperfusion
requires early administration and is mediated via Akt and p70s6 kinase cell-survival signaling.
Circ Res 2001;89:1191-98.
26.
van den Berghe G, Wouters P, Weekers F et al. Intensive insulin therapy in the critically ill
27.
Clark RS, English M, McNeill GP, Newton RW. Effect of intravenous infusion of insulin in
patients. N Engl J Med 2001;345:1359-67.
diabetics with acute myocardial infarction. Br Med J (Clin Res Ed) 1985;291:303-5.
28.
Gwilt DJ, Nattrass M, Pentecost BL. Use of low-dose insulin infusions in diabetics after
29.
Malmberg K, Ryden L, Efendic S et al. Randomized trial of insulin-glucose infusion followed
myocardial infarction. Br Med J (Clin Res Ed) 1982;285:1402-4.
by subcutaneous insulin treatment in diabetic patients with acute myocardial infarction
(DIGAMI study): effects on mortality at 1 year. J Am Coll Cardiol 1995;26:57-65.
30.
Malmberg K, Ryden L, Hamsten A, Herlitz J, Waldenstrom A, Wedel H. Effects of insulin
treatment on cause-specific one-year mortality and morbidity in diabetic patients with acute
myocardial infarction. DIGAMI Study Group. Diabetes Insulin-Glucose in Acute Myocardial
Infarction. Eur Heart J 1996;17:1337-44.
31.
Furnary AP, Gao G, Grunkemeier GL et al. Continuous insulin infusion reduces mortality in
patients with diabetes undergoing coronary artery bypass grafting. J Thorac Cardiovasc Surg
2003;125:1007-21.
32.
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;109:1497-502.
97
Metabolic interventions in acute myocardial infarction
98
GIK and 3-year outcome
Chapter 2.6
Long-term survival after glucose-insulin-potassium infusion
and primary angioplasty in ST segment elevation myocardial
infarction: the randomized Glucose Insulin Potassium Study
(GIPS)
Iwan CC van der Horst, Jorik R Timmer, Jan Paul Ottervanger, Arnoud WJ van ‘t Hof, Henk
JG Bilo, Menko-Jan de Boer, Jan CA Hoorntje, Jan-Henk E Dambrink, AT Marcel
Gosselink, Harry Suryapranata, Rijk OB Gans, Felix Zijlstra
Submitted
99
Metabolic interventions in acute myocardial infarction
Abstract
Introduction
A combined approach of mechanical reperfusion of the infarct-related coronary artery and
cardiac protection with glucose-insulin-potassium (GIK) infusion has been proposed to
protect the ischemic myocardium.
Methods
We investigated the long-term clinical outcome of GIK infusion in ST segment elevation
myocardial infarction (MI) patients treated with primary percutaneous coronary
intervention (PCI). We performed an open-label randomized controlled trial, conducted
from April 1998 to September 2001 in a single center. Nine hundred and forty patients
with ST segment elevation MI and eligible for primary PCI were enrolled and followed for
3 years. Patients received either a continuous infusion of 80 mmol potassium chloride in
500 mL 20% glucose with a rate of 3 mL/kg/hour with a continuous infusion of shortacting insulin with the use of a pump for 8 to 12 hours or no infusion.
Results
All-cause mortality was 98 (10.4%) out of 940 patients after 3-year. Factors related to longterm mortality were age, previous cardiovascular disease, anterior MI, Killip class ≥2, (all
P< 0.001), multi-vessel disease (P=0.001), and female gender (P=0.032). In the GIK group
46 (9.7%) out of 476 patients died compared to 52 (11.2%) out of 464 control patients
(P=0.44). In patients with Killip class 1 (N=856) randomization to GIK resulted in a lower 3year mortality, adjusted relative risk 0.56 (0.34-0.94, P=0.028). Other factors related with
long-term mortality in Killip class 1 patients were age (P<0.001), previous cardiovascular
disease (P<0.001), anterior MI location (P=0.05), and multi-vessel disease (P=0.03).
Conclusion
Cardiac protection with GIK infusion and primary PCI in acute MI did not result in a
significant mortality reduction in all patients. During the 3-years follow-up, the beneficial
effect observed after 30-days in the predefined subgroup of Killip class 1 patients was
sustained.
100
GIK and 3-year outcome
Introduction
High risk patients with early and sustained reperfusion of the infarct-related artery in acute
ST segment elevation myocardial infarction (MI) the mortality rate may still be up to 10
percent.1-4 To further improve the outcome for ST segment elevation MI patients therapies
directed toward cardiac protection have been suggested to be of additive value.5 One of
these strategies is the infusion of glucose, insulin and potassium (GIK).6 The concept of
metabolic protection of the ischemic myocardium hinges on the fact that glucose can be
metabolized anaerobically and can thereby provide glycolytic adenosine triphosphate
(ATP) in the cytosolic compartment.7 In both acute myocardial infarction and cardiac
surgery the effect of GIK infusion has been investigated. An overview of trials in the era
before reperfusion found that GIK resulted in improved short-term survival.8 New
promising results were found in a pilot trial of GIK, in particular in patients treated with
either thrombolysis or primary percutaneous coronary intervention (PCI).9 To investigate
the potential role of the combination of GIK and early reperfusion, we conducted a
randomized trial. We found no significant beneficial effect of GIK and primary PCI in the
overall population.10 In a study of insulin infusion in diabetic patients with acute MI no
effect was found on 30-day mortality.11 However, after 3.43 years follow-up the absolute
mortality reduction was 11% (P=0.011).12 Long-term effects of GIK in MI patients have not
been studied and described previously. We therefore investigated long-term clinical
outcome in all patients enrolled in the Glucose-Insulin-Potassium Study (GIPS).
Methods
Patients
The diagnosis acute MI included chest pain suggestive for myocardial ischemia for at least
30 minutes, a time from onset of symptoms less than 12 hours before hospital admission,
and an ECG recording with ST T segment elevation of more than 1 mV in two or more
contiguous leads or new onset left bundle-branch block on the electrocardiogram.13;14
Patients had to be eligible for acute angiography with a view to perform primary PCI.
Patients presenting at our center and patients referred for treatment of high-risk MI, from
nine referring hospitals without angioplasty facilities, were included. Patients were
excluded when pre-treated with thrombolysis or when an illness associated with a marked
restricted life expectancy was present. Before randomization baseline characteristics were
recorded. The research protocol was reviewed and approved by the medical ethics
committee of our hospital, and patients were included after informed consent.
101
Metabolic interventions in acute myocardial infarction
Study Protocol
After admission patients were randomly assigned, with the use of a computerized
randomization program, to either GIK infusion or no infusion. In the GIK group, a
continuous infusion of 80 mmol potassium chloride in 500 mL 20% glucose with a rate of
3 mL/kg/hour in a peripheral venous line was given over a period up to 12 hours. The
infusion was started as soon as possible. A continuous infusion of short-acting insulin (50
units Actrapid HM [Novo Nordisk, Copenhagen, Denmark] in 50 mL 0.9% sodium chloride
was started with the use of a pump (Perfusor-FM, B. Braun, Melsungen, Germany).
Baseline insulin-infusion dose and hourly adjustments of the insulin dose were based on a
modified algorithm, to obtain blood-glucose levels between 7.0 and 11.0 mmol/L, based
on measurements of whole-blood glucose (Modular System, Roche/Hitachi, Basel,
Switzerland).
After the infusion was started, all patients went to the catheterization laboratory where
both coronary arteries were visualized. PCI was performed with standard techniques if the
coronary anatomy was suitable for PCI. Additional standard treatment consisted of
heparin, nitroglycerin intravenously, and aspirin 500 mg intravenously or 300 mg orally.
Low molecular weight heparin was given after sheath removal.
Standard therapies after PCI included aspirin 80 mg, clopidogrel 75 mg, beta-blockers,
lipid lowering agents and angiotensin converting enzyme inhibitors or angiotensin II
receptor blockers and were used according to the guidelines of the European Society of
Cardiology/American College of Cardiology at that time.13;15 All patients were seen
approximately 6 weeks after discharge at the outpatient clinic. The additional treatment
was evaluated and changed if indicated, i.e. start or increase in dosing of antihypertensive medication if blood pressure was above the goal or of statins if lipid profile
did not meet guideline criteria.
Statistical analysis
The primary endpoint of this analysis was 3-year (long-term) mortality. The incidence of
death was evaluated in predefined subgroups including age <60 year versus ≥60 year,
men versus women, anterior MI versus non-anterior MI, diabetics versus non-diabetics,
and Killip class 1 versus Killip class ≥2, successful reperfusion versus no successful
reperfusion. Successful reperfusion was defined by TIMI blood flow grade 3 in
combination with myocardial blush grades 2 and 3.16;17 Analyses were performed
according to the intention-to-treat principle. Relative risks (RR) were calculated by dividing
102
GIK and 3-year outcome
Table 1. Baseline clinical and demographic characteristics and initial therapies
Characteristic
Number of patients
GIK
Control
476
464
Age, years (mean±SD)
59.9±11.9
60.8±12.0
Men
351 (73.7)
368 (79.3)
Referred patients
201 (42.3)
180 (38.8)
Body mass index
26.7±3.8
27.0±4.0
Previous MI
52 (10.9)
53 (11.4)
Previous PCI
22 (4.7)
24 (5.2)
Previous CABG
14 (2.9)
12 (2.6)
History of stroke
17 (3.6)
15 (3.2)
Diabetes mellitus
50 (10.5)
49 (10.6)
Hypertension
134 (28.2)
130 (28.0)
Dyslipidemie
94 (19.7)
95 (20.5)
Currently smoker
225 (47.3)
237 (51.1)
Positive family history
195 (41.0)
179 (38.6)
Time to admission, min (IQR)*
150 (100-215)
150 (105-234)
Door to balloon time, min (IQR)
45 (31-64)
48 (34-69)
Killip class 1
426 (89.5)
430 (92.7)
Killip class 2
24 (5.0)
14 (3.0)
Killip class 3
14 (2.9)
14 (3.0)
Killip class 4
12 (2.5)
6 (1.3)
Anterior MI
250 (52.9)
224 (49.1)
Multi-vessel disease
249 (52.3)
242 (52.2)
PCI
436 (91.6)
424 (91.4)
Stent **
255 (58.5)
236 (55.7)
GP IIb/IIIa receptor blocker **
96 (22.1)
109 (25.7)
In-hospital CABG
19 (4.0)
19 (4.1)
459 (96.4)
435 (93.8)
TIMI 3 flow after PCI
Data are number (%) unless otherwise indicated. MI = myocardial infarction. IQR = interquartile
range. PCI = percutaneous coronary intervention. CABG = coronary artery bypass grafting. * Time
to admission denotes time between onset of symptoms and until admission. † In-hospital CABG =
CABG defined as patients treated initially conservative, followed by elective CABG within 7 days. **
Percentage defined as the percentage of the patients that underwent PCI.
the cumulative incidence rate in the GIK group by the cumulative incidence rate in the
control group. The relative risks for 3-year mortality were also determined for clinical and
demographic characteristics and randomization group. Differences between group means
103
Metabolic interventions in acute myocardial infarction
at baseline were assessed with the two-tailed Student’s t-test. Chi-square analysis or
Fisher’s exact test was used to test differences between proportions. Mortality was
calculated by the Kaplan-Meier product-limit method. The Log-rank test was used to
evaluate differences in survival curves between the two treatment groups. The Cox
proportional-hazards regression model was used to calculate relative risks adjusted for
differences in baseline characteristics. Statistical significance was considered a two-tailed
P-value <0.05. The Statistical Package for the Social Sciences (SPSS Inc., Chicago, IL,
USA) version 11.5 was used for all statistical analysis. Sample size calculation and other
details of the statistical methods have been published.10
Table 2. Additional treatment at hospital discharge according to treatment group
Medication
GIK group
Control group
Aspirin
421 (88.4)
398 (85.8)
Clopidogrel
194 (40.8)
192 (41.4)
Coumarins
36 (7.6)
45 (9.7)
Beta-blockers
392 (82.3)
381 (82.1)
ACE-inhibitors
245 (51.5)
244 (52.6)
AII receptor antagonists
7 (1.5)
3 (0.6)
Calcium channel blockers
15 (3.2)
19 (4.1)
Diuretics
46 (9.7)
67 (14.4)
Nitrates
70 (14.7)
67 (14.4)
Statins
285 (51.3)
271 (48.7)
Data are number (%). ACE = angiotensin-converting enzyme. AII = angiotensin II.
Results
The baseline clinical and demographic characteristics of the 476 patients in the GIK group
and 464 patients in the control group are represented in table 1. GIK infusion was started
15 to 20 minutes after admission. Average door to balloon time was 45 minutes in the GIK
group and 48 minutes in the control group. After coronary angiography, 860 patients
(90.5%) underwent primary PCI, 38 patients (4.0%) were referred for CABG within 7 days
after initial stabilization, and 42 patients (4.5%) were treated conservatively. There were
no differences in additional medication after hospital discharge between the GIK and
control group (table 2).
After 3 years 98 patients (10.4%) had died. A total of 50 patients (5.3%) died within 30
days and 69 patients (7.3%) died within 1-year. Baseline characteristics that were related
104
GIK and 3-year outcome
to 3-year mortality were age (P<0.001), female gender (P=0.03), previous cardiovascular
disease (P<0.001), anterior MI (P<0.001), Killip class ≥2 (P<0.001), and multi-vessel
disease (P=0.001) (table 3). The relative risk (RR) per year was 1.04, female gender had a
RR of 1.62, patients with a history of cardiovascular disease had a RR of 2.27, patients with
an anterior MI had a RR of 2.56, patients with Killip class 2 a RR of 4.20, patients with Killip
class 3 a RR of 12.52, and patients with Killip class 4 a RR of 26.23 compared to patients
with Killip class 1.
Table 3. Adjusted relative risks for 3-year mortality of baseline demographic and clinical
characteristics in the overall population and Killip class 1 patients
Characteristics
Overall population
Killip class 1
RR 3-year mortality
P-value
RR 3-year mortality
P-value
Randomization
0.75 (0.50-1.12)
0.17
0.81 (0.54-1.22)
0.27
Age per year older
1.04 (1.02-1.06)
<0.001
1.05 (1.03-1.07)
<0.001
Female gender
1.62 (1.04-2.50)
0.032
1.49 (0.96-2.33)
0.07
Previous CVD*
2.73 (1.76-4.24)
<0.001
2.66 (1.72-4.12)
<0.001
Killip class ≥2†
6.64 (4.32-10.2)
<0.001
Anterior MI
2.31 (1.51-3.52)
<0.001
1.95 (1.27-2.97)
<0.001
Multi-vessel disease
2.32 (1.43-3.78)
0.001
2.18 (1.34-3.55)
0.002
Relative risk for 3-year mortality of baseline demographic and clinical characteristics. RR = relative
risk. MI = myocardial infarction. CVD = cardiovascular disease. * Previous cardiovascular disease is
a combination of previous myocardial infarction, previous percutaneous coronary intervention,
previous coronary artery bypass grafting, or previous cerebrovascular accident. † P-value of risk
compared to patients with Killip class 1.
3-year mortality was 46 of 476 patients (9.7%) randomized to GIK compared to 52 of 464
patients (11.2%) in the control group (P=0.44, using the Log-rank test) (figure 1). After
adjustment for characteristics related with 3-year mortality the RR was 0.75 (95%
confidence interval 0.50-1.12, P=0.17). In 856 patients (91.1%) without signs of heart
failure at admission (Killip class 1), 3-year mortality was 23 of 426 patients (5.4%) in the
GIK group versus 41 of 430 patients (9.5%) in the control group, (P=0.019 using the Logrank test) (figure 2). In Cox proportional-hazards regression model the RR was 0.56 (0.330.93, P=0.025). In patients without signs of heart failure at admission age (P<0.001),
previous cardiovascular disease (P<0.001), anterior MI (P=0.05), Killip class ≥2 (P<0.001),
and multi-vessel disease (P=0.03) were related to 3-year mortality (table 3). In 84 patients
(8.9%) with signs of heart failure (Killip class ≥2), 3-year mortality was 23 of 50 patients
(46%) in the GIK group and 11 of 34 patients (32%) in the control group (P=0.23, using
the Log-rank test). As in our previous analysis of 30-day outcome, we did additional
105
Metabolic interventions in acute myocardial infarction
analysis of predefined subgroups. Age, gender, ischemic time, infarct location, multivessel versus single vessel disease, and TIMI flow had no impact on the difference on
outcome between the GIK group and the control group.
Figure 1. Kaplan-Meier of the overall population
100
90
Survival (%)
80
GIK
70
Control
60
50
0
0
200
400
600
800
Time after randomization (days)
1000
P-value is 0.44 using the Log-rank test.
Discussion
In line with our 30-day mortality results we did not find a significant benefit on survival, 3year after GIK infusion in ST segment elevation MI patients treated with primary PCI, in the
overall population. The favourable effect in patients without signs of heart-failure at
admission was sustained. The absolute reduction in mortality in these patients was
approximately 3% during the entire follow-up period (Kaplan-Meier curve in Killip class 1
patients, figure 2).
Reasons for the absence of a survival benefit in all GIPS patients despite apparent
improvements in myocardial function and mortality in earlier trials remain unclear. In the
low-dose Polish-Glucose-Insulin-Potassium (Pol-GIK) trial with 954 patients 30-day
mortality in the GIK group was significantly higher than in the control group (8.9% versus
106
GIK and 3-year outcome
4.8%, P=0.01).18 It is possible that the mortality reduction observed in the Estudios
Cardiologicos Latinoamerica (ECLA) pilot trial was related to the mortality rate in the
control group (15.1% in the ECLA pilot versus 5.8% in the GIPS).9 In the Reevaluation of
Intensified Venous Metabolic Support for Acute Infarct Size Limitation (REVIVAL) trial with
312 MI patients treated with primary PCI, GIK did not result in a significant difference in
mortality after 6 months (5.8% in the GIK group versus 6.4% in the control group,
P=0.85).19
Figure 2. Kaplan-Meier in patients with Killip class 1 and Killip class ≥2
100
90
Survival (%)
80
70
60
50
0
0
200
400
600
800
Time after randomization (days)
GIK - Killip 1
GIK - Killip ≥2
Control - Killip 1
Control - Killip ≥2
1000
P-value in Killip class 1 patients is 0.019, using the Log-rank test. P-value in Killip class ≥2
patients is 0.23, using the Log-rank test.
With the low mortality rate in the control group of GIPS, as in the REVIVAL, the study may
have been underpowered to find an improvement in the mortality in the overall
population. Another explanation of our findings could be the high number of patients with
signs of heart failure in the GIK group. In these patients the volume load may be
detrimental.9 A high number of patients died during the first days after admission (figure
2). The number of patients with heart failure that died after 3 years was equal to the
number of patients in other trials, such as the SHould we emergently revascularize
107
Metabolic interventions in acute myocardial infarction
Occluded Coronaries for cardiogenic shocK (SHOCK) study.20 In the Diabetes InsulinGlucose in Acute Myocardial Infarction (DIGAMI) study with 620 patients with an acute MI
and diabetes mellitus or glucose at admission >11.0 mmol/L, at 1-year follow-up, absolute
mortality reduction was 7.5% (18.6% versus 26.1%, P=0.03).11 This increased to 11%
(33% versus 44%, P=0.011) after a mean follow-up of 3.4 years.12 We did not see a similar
pattern in our study population in mortality reduction. Possibly the post-discharge
treatment with insulin in the DIGAMI study for at least 3 months caused this effect. The
DIGAMI 2 will clarify this since diabetic MI patients have been randomized to acute insulinglucose infusion followed by multi-dose subcutaneous insulin, acute insulin-glucose
infusion followed by conventional hypoglycemic therapy, or conventional hypoglycemic
treatment. In 1548 patients admitted to a thoraco-surgical intensive care unit and
randomized to either strict glucose regulation or conventional glucose-metabolic
treatment short-term mortality decreased significantly (4.6% versus 8%, P<0.04).21 In two
studies including diabetic patients treated with coronary artery bypass grafting strict
glucose regulation with either insulin alone or GIK was related to a significant mortality
reduction.22;23
It has been postulated that through reduction in the extent of ischemic myocardial
damage and suppression of free fatty acid levels, GIK therapy prevents reperfusion injury
that may occur after successful revascularization.24 Protection of the cell membrane of
ischemic cardiac cells may also improve reflow after reperfusion and protect against noreflow phenomenon by reducing cell swelling and microvascular compression. It has been
estimated that GIK therapy has the potential to protect ischemic myocardium before
reperfusion for 10 hours or even longer.25 Our study offered the opportunity to analyse the
effect of GIK in patients with different ischemic times, i.e. time from symptom-onset to
admission and door to balloon time. We could not find a relation between time delays and
the effect of GIK on long-term outcome.
Should this therapy have any role in the contemporary management of acute MI? In MI
patients treated with primary PCI in the absence of signs of heart failure the absolute
mortality reduction of approximately 3% would imply that a number of 34 patients are
needed to treat to save one. Since this study had no exclusion criteria for elderly patients,
women, etcetera, the results can easily be translated into daily practice. In patients with
signs of heart failure infusion further research is needed on the infusion rate and quantity
before the effect on survival can be reinvestigated. Large trials are currently undertaken
such as the ECLA GIK 2/CREATE trial, Organization to Assess Strategies for Ischemic
Syndromes (OASIS)-6 trial, and the GIPS 2 to resolve various issues related to metabolic
interventions in acute MI.
108
GIK and 3-year outcome
Conclusion
In patients with ST segment elevation MI treated with primary PCI, infusion of GIK has no
significant long-term beneficial effect. In the large subgroup of patients without signs of
heart failure at admission GIK reduced short-term and long-term mortality.
Contributors
This study was supported by a generous grant from the Netherlands Heart Foundation
(99.028). No other conflicts of interest or financial disclosure.
References
1.
Rogers WJ, Canto JG, Lambrew CT et al. Temporal trends in the treatment of over 1.5 million
patients with myocardial infarction in the US from 1990 through 1999: the National Registry of
Myocardial Infarction 1, 2 and 3. J Am Coll Cardiol 2000;36:2056-63.
2.
Keeley EC, Grines CL. Primary coronary intervention for acute myocardial infarction. JAMA
2004;291:736-39.
3.
Grines CL, Browne KF, Marco J et al. A comparison of immediate angioplasty with
thrombolytic therapy for acute myocardial infarction. The Primary Angioplasty in Myocardial
Infarction Study Group. N Engl J Med 1993;328:673-79.
4.
Zijlstra F, Hoorntje JC, de Boer MJ et al. Long-term benefit of primary angioplasty as
compared with thrombolytic therapy for acute myocardial infarction. N Engl J Med
1999;341:1413-19.
5.
Kloner RA, Rezkalla SH. Cardiac protection during acute myocardial infarction: where do we
stand in 2004? J Am Coll Cardiol 2004;44:276-86.
6.
LaDisa JF, Jr., Krolikowski JG, Pagel PS, Warltier DC, Kersten JR. Cardioprotection by
glucose-insulin-potassium: dependence on KATP channel opening and blood glucose
concentration before ischemia. Am J Physiol Heart Circ Physiol 2004;287:H601-H607.
7.
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.
8.
Fath-Ordoubadi F, Beatt KJ. Glucose-insulin-potassium therapy for treatment of acute
myocardial infarction: an overview of randomized placebo-controlled trials. Circulation
1997;96:1152-56.
9.
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.
10.
van der Horst IC, Zijlstra F, van't Hof AW et al. Glucose-insulin-potassium infusion in patients
treated with primary angioplasty for acute myocardial infarction: the glucose-insulinpotassium study: a randomized trial. J Am Coll Cardiol 2003;42:784-91.
109
Metabolic interventions in acute myocardial infarction
11.
Malmberg K, Ryden L, Efendic S et al. 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 1995;26:57-65.
12.
Malmberg K. Prospective randomised study of intensive insulin treatment on long term
survival after acute myocardial infarction in patients with diabetes mellitus. DIGAMI (Diabetes
Mellitus, Insulin Glucose Infusion in Acute Myocardial Infarction) Study Group. BMJ
1997;314:1512-15.
13.
Ryan TJ, Antman EM, Brooks NH et al. 1999 update: ACC/AHA Guidelines for the
Management of Patients With Acute Myocardial Infarction: Executive Summary and
Recommendations: A report of the American College of Cardiology/American Heart
Association Task Force on Practice Guidelines (Committee on Management of Acute
Myocardial Infarction). Circulation 1999;100:1016-30.
14.
Shlipak MG, Lyons WL, Go AS, Chou TM, Evans GT, Browner WS. Should the
electrocardiogram be used to guide therapy for patients with left bundle-branch block and
suspected myocardial infarction? JAMA 1999;281:714-19.
15.
Van de Werf F, Ardissino D, Betriu A et al. Management of acute myocardial infarction in
patients presenting with ST segment elevation. The Task Force on the Management of Acute
Myocardial Infarction of the European Society of Cardiology. Eur Heart J 2003;24:28-66.
16.
The Thrombolysis in Myocardial Infarction (TIMI) trial. Phase I findings. TIMI Study Group N
17.
van 't Hof AW, Liem A, Suryapranata H, Hoorntje JC, de Boer MJ, Zijlstra F. Angiographic
Engl J Med 1985;312:932-36.
assessment of myocardial reperfusion in patients treated with primary angioplasty for acute
myocardial infarction: myocardial blush grade. Zwolle Myocardial Infarction Study Group.
Circulation 1998;97:2302-6.
18.
Ceremuzynski L, Budaj A, Czepiel A et al. Low-dose glucose-insulin-potassium is ineffective in
acute myocardial infarction: results of a randomized multicenter Pol-GIK trial. Cardiovasc
Drugs Ther 1999;13:191-200.
19.
Pache J, Kastrati A, Mehilli J et al. A randomized evaluation of the effects of glucose-insulinpotassium infusion on myocardial salvage in patients with acute myocardial infarction treated
with reperfusion therapy. Am Heart J 2004;148:105.
20.
Hochman JS, Sleeper LA, White HD et al. One-year survival following early revascularization
for cardiogenic shock. JAMA 2001;285:190-192.
21.
van den Berghe G, Wouters P, Weekers F et al. Intensive insulin therapy in the critically ill
22.
Furnary AP, Gao G, Grunkemeier GL et al. Continuous insulin infusion reduces mortality in
patients. N Engl J Med 2001;345:1359-67.
patients with diabetes undergoing coronary artery bypass grafting. J Thorac Cardiovasc Surg
2003;125:1007-21.
23.
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; 109:1497-502.
24.
Jonassen AK, Aasum E, Riemersma RA, Mjos OD, Larsen TS. Glucose-insulin-potassium
reduces infarct size when administered during reperfusion. Cardiovasc Drugs Ther
2000;14:615-23.
110
GIK and 3-year outcome
25.
Eberli FR, Weinberg EO, Grice WN, Horowitz GL, Apstein CS. Protective effect of increased
glycolytic substrate against systolic and diastolic dysfunction and increased coronary
resistance from prolonged global underperfusion and reperfusion in isolated rabbit hearts
perfused with erythrocyte suspensions. Circ Res 1991;68:466-81.
111
Metabolic interventions in acute myocardial infarction
112
Overview of GIK studies in acute MI
Chapter 3
Glucose-insulin-potassium infusion as adjunctive therapy in
myocardial
infarction:
current
evidence
and
potential
mechanisms
Jorik R Timmer, Iwan CC van der Horst, Jan Paul Ottervanger, Giuseppe de Luca, Arnoud
WJ van ‘t Hof, Henk JG Bilo, Felix Zijlstra
Italian Heart Journal 2004;5:727-31
113
Metabolic interventions in acute myocardial infarction
Abstract
Background
In ST segment elevation myocardial infarction (MI) there is conflicting evidence that
mortality, morbidity and infarct size is reduced by therapies influencing myocardial
metabolism, such as infusion of glucose-insulin-potassium (GIK). Several clinical trials
with GIK have already provided insight in the magnitude of this effect. Our aim was to
investigate current evidence on the potential beneficial effect of GIK infusion in ST
segment elevation MI.
Methods
Randomized trials comparing GIK with placebo or untreated controls in patients with ST
segment elevation MI were identified by electronic and manual searches. A systematic
analysis of all data was performed, with regard to inclusion criteria, dose of GIK and
additional use of reperfusion therapy.
Results
Thirteen trials, involving 5009 patients, were included. Overall, hospital mortality was
10.7% after GIK compared to 12.8% in controls (P=0.02). GIK infusions were in particular
effective when a high-dose was used and if given as adjunctive to reperfusion therapy.
Low-dose GIK was less beneficial, and may even increase hospital mortality. Also in
patients with heart failure on admission, GIK may have worse effects. In all analyzed trials,
GIK infusion caused only mild adverse effects, although fluid overload may be a problem
in certain patients.
Conclusion
GIK may reduce mortality in patients with STEMI, particularly if a high dose is used and
when GIK is administered as an adjunct to reperfusion therapy. However, all studies had a
relative small sample size and additional large randomized trials are certainly needed
before a definite conclusion can be made. The limited evidence currently available does
not warrant GIK therapy to be applied in patients at the present time.
114
Overview of GIK studies in acute MI
Introduction
Glucose-insulin-potassium (GIK) infusion in the treatment of ST segment elevation
myocardial infarction (MI) has been a field of interest for many decades.1;2 However, the
results of early studies investigating the effect of GIK on clinical outcome after myocardial
infarction
were
inconclusive
and
attention
focussed
on
reperfusion
therapies.
Furthermore, lack of financial interest of the pharmaceutical industry made it difficult to
perform a large clinical trial. Although early and sustained reperfusion is indeed the most
important initial treatment of ST segment elevation MI, agents that influence energy
substrate metabolism in the reperfused myocardium may have additional beneficial
effects. In a meta-analysis of earlier randomised trials of GIK in ST segment elevation MI, it
was shown that GIK has potential beneficial effects, but studies in the reperfusion era
were lacking.3 Furthermore, the meta-analysis suggested that beneficial effects of GIK are
particularly seen if high-dose GIK infusion was used.3;4 This potential dose dependent
effect of GIK was also observed before. After the meta-analysis was published, several
randomised trials have been reported, also with GIK as adjunctive therapy to reperfusion
therapies.5-7 However, most trials had a relative small sample size. A meta-analysis or
review of all relevant trials can provide sufficient power to demonstrate differences in
outcome. This article will review the currently available data concerning the clinical
benefits and potential mechanism of action of GIK. We also performed a stratified analysis
of the effects of high and low dose GIK infusion, including all recently published trials.
Methods
We attempted to obtain results from all completed, published, randomised trials of GIK in
acute MI. The literature was scanned by formal searches of electronic databases
(MEDLINE) and informal searches for studies concerning the potential mechanism of
action of GIK. Since experimental and clinical studies have suggested a dose-response
curve of GIK, we performed stratified analyses of the effects of low and high dose of GIK.
A high dose GIK was defined as an intravenous infusion of GIK in a dose equal to or
higher than used by Rackley et al, 30% glucose (300 mg/L), 50 IU/L regular insulin and 80
mmol/L potassium chloride at 1.5 mL/kg per hour infusion rate.4 Our primary efficacy
outcome of interest was hospital mortality. We calculated the relative risk (RR) for hospital
death of patients treated with GIK as compared to those treated with placebo or no
infusion. The RR and its 95% confidence interval (CI) were calculated for each trial and the
grand totals.
115
1998
1987
1977
1978
1976
1999
1968
1965
1968
1967
1971
1995
2003
ECLA
Satler
Heng
Stanley
Rogers
Pol-GIK
Pentecost
Mittra
MRC
Pilcher
Hjermann
DIGAMI
GIPS
476
464
306
314
104
100
49
53
480
488
85
85
100
100
494
460
61
73
55
55
12
15
10
7
135
133
139
Patients
(N)
50 IU/L
20 IU/L
-
Insulin
80 mmol/L
40 mmol/L
-
Potassium
50 IU
20%
5%
200 g or
-
240 g or
160 g or
240 g or
10%
50 IU
80 IU
16 IU
placebo
2x10 Usc
2x10 Usc
2x10 Usc
30 IU
10%
20 IU/L
0,9% NaCl/L
300g
NaCl
300g
50 IU
halfnormal saline
50%
0.3 IU/kg
normal saline
160 mmol/L
-
55 mEq
-
52-78 mEq
52 mEq
52-78 mEq
20 mEq
6.0 g/L
80 mEq/L
80 mEq/L
0.15mmol/kg
300 mg/L 50 IU/L
80 mEq/L
5% dextrose in water
25%
10%
-
Glucose
3 mL/kg/h
0.7L
-
-
in 1500
-
1.5 l/24h
42 mL/h
42 mL/h
1.5 ml/kg/h
1.5 ml/kg/h
1.5 ml/kg/h
1.5 ml/kg/h
1.5 mL/kg.h
1.0 mL/kg.h
-
Rate
8-12 h
24 h
10 d
14 d
14 d
14 d
48 h
24 h
24 h
48 h
48 h
6-12 h
1h
48 h
24 h
24 h
-
Period
4.8
5.8
9.1
11.1
11
20
6
12
103
115
10
24
15
16
44
22
4
9
4
9
1
0
0
0
10
8
16
Mortality
(%)
0.82 (0.46-1.46)
0.89 (0.67-1.19)
0.47 (0.19-1.11)
0.48 (0.13-1.54)
0.69 (0.65-1.21)
0.34 (0.13-0.81)
0.93 (0.40-2.14)
1.95 (1.12-3.47)
0.50 (0.11-1.92)
0.40 (0.08-1.57)
-
-
0.69 (0.27-1.70)
0.63 (0.24-1.52)
RR 95% CI
114
Time to
Treatment
43 min
9.85±1.07
7.5±1.21
2.5 (1.67-3.58) 45 min
2.5 (1.75-3.9) 48 min
13±7
Chest pain to treatment
11.8
13.3
unknown
11.3
11.6
11.94±1.23
11.3±1.26
9.82
5.0 (3.0-10.1)
5.0 (2.5-10.0)
<12
Chest pain to treatment
8 (4-11.5)
7.5 (4-14)
Chest pain to treatment
6.8
7.0
11.4 (0.56)
10.1 (0.53)
10.6 (0.50)
Chest pain to treatment
4.7 ± 2.7
Time to
Random.
RR denotes relative risk. CI denotes confidence interval. h denotes hour. d denotes days. min denotes minutes. or denotes oral. Usc denotes units subcutaneous.
Year
Study
Table 1. Diffent glucose-insulin-potassium infusion schemes in 13 trials in acute myocardial infarction
Metabolic interventions in acute myocardial infarction
Overview of GIK studies in acute MI
Results
Our search yielded 13 studies, involving 5009 patients. The meta-analysis of FathOrdoubadi and Beatt including 9 studies did not include three recently published trials.3
They also excluded the Diabetes Mellitus, Insulin Glucose Infusion in Acute Myocardial
Infarction (DIGAMI) in which GIK (insulin-glucose) was studied in patients with diabetes
mellitus or a glucose at admission of >11.0 mmol/L.3;8 Baseline characteristics of all
randomised trials are summarised in table 1. The studies used GIK in different doses, and
the time from onset of symptoms of myocardial infarction to treatment varied, with a large
proportion of patients being treated after more than 6 hours from onset of symptoms. It is
shown that in patients with ST segment elevation MI treated by primary coronary
intervention, symptom-onset-to-balloon time is related to mortality.9 This study found that
a symptom-onset-to-balloon time >4 hour was identified as independent predictor of oneyear mortality. The studies included in the meta-analysis of Fath-Ordoubadi and Beatt had
several other limitations.3 Firstly, randomization was not always optimal, resulting in
significant differences in baseline clinical characteristics between the two treatment
groups.1;10 Secondly, mortality rates were generally very high in both the treatment and in
placebo groups, with hospital mortality up to 28%.11 Furthermore, only the study by Satler
and colleagues including only 17 patients also received reperfusion therapy.12 The first
randomised prospective trial in the era of reperfusion, the DIGAMI study including only
patients with diabetes mellitus, found that the combination of insulin-glucose infusion with
an intensive insulin treatment for at least three months after discharge resulted in a
reduction in hospital mortality of 58% in patients without prior insulin use and a low
cardiovascular risk profile.8 Potassium was added to the infusion when necessary. The
Estudios Cardiologicos Latinoamerica (ECLA) pilot trial, was published in 1998 and also
showed a beneficial effect of GIK.6 This effect was in particular observed in patients who
also underwent reperfusion therapy. Hospital-mortality was 5.1% in patients treated with
GIK versus 15.1% in controls (P<0.01). Most patients were treated with thrombolysis as
reperfusion therapy, whereas only 3% were treated with primary angioplasty. In 1999, the
Polish-Glucose-Insulin-Potassium (Pol-GIK) trial was published, including 954 patients.5
This trial demonstrated no differences in cardiac mortality or the occurrence of cardiac
events at 35 days between GIK (6.5%) and control patients (4.6%). Total mortality at 35
days was even higher in the GIK group (8.9%) than in controls (4.8%, P<0.01). However,
in this trial low-dose GIK was used. The most recent study, the Glucose Insulin Potassium
Study (GIPS), included 940 ST segment elevation MI patients all treated with primary
angioplasty as reperfusion therapy.7 High-dose GIK resulted in a non-significant reduction
of mortality in GIK treated patients in the total patient group (4.8% versus 5.8%, P=0.50).
117
Metabolic interventions in acute myocardial infarction
However, a significant mortality reduction was found in patients without signs of heart
failure (1.2% versus 4.2%, P=0.01).
With regard to the presumed protective effect of GIK infusion on the myocardium, several
studies were performed investigating the benefit of GIK infusion during cardiac surgery.
Comparison of results is difficult because the studies differed in initiation of GIK (pre,
during or post cardiac surgery), duration and dosing of GIK13-15 and end points (including
enzymatic myocardial damage, need for inotropic support and mortality). Lazar et al.
found a benefit of GIK in two small sized studies.16;17 However, many other trials, including
a well sized randomised trial (N=1127), failed to show a clear benefit of GIK infusions in
cardiac surgery.18;19 So, no definite conclusions about the use of GIK infusions in cardiac
surgery can be drawn.
Table 2. 30-day mortality in patients treated with high-dose GIK and low-dose GIK versus
control patients of published trials
Included patient
Trial (ref)
GIK patients
Heng38
39
Stanley
10
Control patients
30-day mortality
GIK patients
Control patients
N
N
N (%)
N (%)
12
15
1 (8.3)
0 (-)
55
55
4 (7.3)
4 (7.3)
61
73
4 (6.5)
9 (12.3)
10
7
0 (-)
0 (-)
ECLA *
135
139
10 (7.4)
16 (11.5)*
GIPS7
476
464
23 (4.8)
27 (5.8)
8
306
314
28 (9.2)
35 (11.1)
Total (high-dose)
1065
1074
70 (6.6)
96 (8.9)
85
85
10 (11.8)
24 (23.5)
49
53
6 (6.7)
12 (22.7)
Rogers
Satler12
6
DIGAMI
Mittra11
Pilcher40
2
Pentecost
100
100
15 (15.0)
16 (16.0)
MRC1
480
488
103 (21.5)
115 (23.6)
Hjermann 21
104
100
11 (9.6)
20 (20)
6
ECLA *
133
139
10 (7.5)
16 (11.5)*
Pol-GIK5
494
460
44 (8.9)
22 (4.8)
Total (low-dose)
1445
1425
199 (13.8)
225 (15.8)
Total (all)
2510
2499
269 (10.7)
321 (12.8)
Data are number or number (%). Ref denotes reference. N denotes number of patients. ECLA =
Estudios Cardiologicos Latinoamerica. GIPS = Glucose-Insulin-Potassium Study. DIGAMI =
Diabetes and Insulin-Glucose Infusion in Acute Myocardial Infarction. MRC = Medical Research
Council. Pol-GIK = Polish Glucose Insulin Potassium. *In both comparisons the same control group
is used; in the total of all patients this groups is accounted for once.
118
Overview of GIK studies in acute MI
Time to treatment
Time between admission and initiation of GIK varied among the studies. Mittra mentioned
10 hours, whereas in the MRC study 70% of patients were treated within 30 minutes after
inclusion.1;11 In the ECLA pilot trial, time from onset of symptoms till the initiation of GIK
treatment was 10 to 11 hours.6 Whether the clinical effects of GIK on outcome are
influenced by the time of initiation of treatment is not yet known. However, reperfusion
may be obligatory as prolonged severe ischemia is followed by necrosis, even in the
presence of GIK. Furthermore, GIK might not reach adequate concentrations in nonperfused myocardium. It would be logical to assume that GIK infusion would be most
effective when initiated before reperfusion therapy has started, when glucose can
temporarily salvage the severely ischemic myocardium and offer protection to possible
reperfusion injury still to come.20
Figure 1. Odds ratios and confidence intervals of high-dose and low-dose GIK trials
Stanley
Rogers
GIPS
ECLA high-dose
DIGAMI
Total high-dose
Mittra
Pentecost
MRC
Pilcher
Hjermann
ECLA low-dose
Pol-GIK
Total low-dose
Total
0.0
0.5
GIK better
1.0
1.5
2.0
2.5
Control better
GIPS = Glucose-Insulin-Potassium Study. ECLA = Estudios Cardiologicos Latinoamerica. DIGAMI =
Diabetes Insulin-Glucose in Acute Myocardial Infarction study MRC = Medical Research Council.
Pol-GIK = Polish Glucose Insulin Potassium. In both comparisons of the ECLA the same control
group is used; in the total of all patients this groups is accounted for once.
119
Metabolic interventions in acute myocardial infarction
Dose
The above-mentioned meta-analysis showed that use of high-dose GIK is most effective.3
This was also demonstrated in dose-response studies, showing that increasing doses
were associated with more suppressed arterial FFA levels as well as increased myocardial
glucose uptake. Also the ECLA pilot-trial confirmed that high-dose GIK was superior to
low-dose.6 In the Pol-GIK trial, using a very low dose of GIK, no effect of GIK was
observed.5 Moreover, using this low dose regimen resulted in an increase in total
mortality rate. We performed a stratified analysis according to GIK-dose, including all
randomized trials investigating the influence of GIK on in-hospital mortality (table 2).
Administration of high dose GIK resulted in a reduction of hospital mortality (6.6% versus
8.9%, RR 0.7; 95%CI: 0.5–1.0, P=0.04). Administration of low dose GIK versus placebo
did not result in a significant difference of in-hospital mortality (13.8% versus 15.8%, RR
0.9; 95%CI: 0.7–1.1, P=0.13). Overall, GIK reduced statistically significant hospital
mortality from 13% versus 11% (RR 0.8; 95%CI: 0.7-1.0, P=0.02) (figure 1).
Adverse effects
The reported adverse effects of GIK treatment were in general mild. Withdrawal because
of side-effects was rare. As GIK was infused via peripheral intravenous catheters, phlebitis
was relative common with percentages up to 15%2, but severe phlebitis however was
rare.11 Infusion of large amounts of fluid in a short period may cause fluid overload. This
potential adverse reaction was however, not reported in the meta-analysis. Possibly,
because of inclusion of only few patients with Killip class >1 in earlier trials. Other studies
demonstrated a small increase of signs of congestive heart failure after GIK.21 In the GIPS
trial, there was a trend towards a higher mortality in GIK treated patients with Killip class
>1 (36% versus 27%), possibly due to volume overload.7 A potential side-effect of GIK
infusion is the induction of hypoglycemia.22 In the DIGAMI study, 15% of the patients who
received intensified therapy had a hypoglycemic event (glucose <3.0 mmol/L), compared
to none in the control group.8;23 In the Pol-GIK trial hypoglycemia occurred in 8% of the
patients, where after the insulin dose was decreased.5 They also found hyperglycemia in
less than 1% in 585 patients with no differences between GIK and control group. In the
GIPS hyperglycemia occurred also in the GIK group although the period was short-lived.
With a blood glucose level >16.8 mmol/L as criteria for interruption in the Pol-GIK; only
1% was interrupted. This is not surprising since a low dose GIK infusion was used.
Hjermann and colleagues used a reduced dose of insulin, which prevented the occurrence
of hypoglycemia.21
120
Overview of GIK studies in acute MI
Mode of action
There are several potential mechanisms by which GIK therapy might improve outcome
after acute myocardial infarction. Although the primary energy substrate for the nonischemic myocardium is free fatty acids (FFA), the myocardium uses various forms of
energy substrates, including glucose.24 During an acute myocardial infarction, circulating
FFA levels are elevated and insulin sensitivity is reduced, limiting cellular uptake of
glucose and promoting the use of FFA.25 These FFA are thought to have a detrimental
effect on ischemic myocardium through varying pathways. They are an energy supply
which is associated with relatively high oxygen consumption in comparison to the
utilisation of glucose.26 Moreover, in contrast with glucose, FFA can not be metabolised
without insulin. Indeed, experimental evidence suggests a negative influence of FFA on
myocardial mechanical performance in the setting of hypoxia.27 Furthermore, excess FFA
metabolism increases susceptibility to ventricular arrhythmia’s and reperfusion injury due
to disturbances in calcium homeostasis and accumulation of free radicals.28;29
Administration of GIK lowers the circulating levels of FFA through the inhibitory effect of
insulin on lipolyses.30 This decrease in FFA levels, in combination with an increase in
glucose and insulin availability, promotes the myocardial use of glucose over FFA.28
Glucose is less oxygen consuming and has beneficial effect on preservation of mechanical
function and membrane stability.31-33
Moreover, GIK therapy might also reduce
34
arrhythmias after successful reperfusion.
As insulin itself induces coronary vasodilation,
myocardial metabolism could further be improved through enhanced myocardial
perfusion.35-37 SPECT analysis showed that GIK infusion improved regional myocardial
perfusion and function in segments adjacent to recently infarcted areas using in human
subjects.36 Recent evidence suggests that insulin might also inhibit reperfusion injury by
inhibiting apoptosis via activation of innate cell-survival pathways in the heart.20
Conclusion
Of 13 published randomized trials, 12 studies reported mortality reduction after GIK.
Despite the possibility of existence of a publication bias, i.e. small negative trials that not
have been published, this seems promising. Most effects were observed when GIK was
given in a high dose and when it was given as an adjunctive to reperfusion therapy.
Adverse effects were rare, fluid overload may be a problem in certain patients, such as
patients with signs of heart failure at admission. To achieve definite conclusions about the
place of GIK in acute myocardial infarction, more randomized trials in which GIK is
combined with optimal reperfusion therapy are needed. Currently large trials in different
populations are undertaken and within a few years more evidence will be available.
Possibly these data will change daily practice.
121
Metabolic interventions in acute myocardial infarction
References
1.
Medical Research Council Working Party on the Treatment of Myocardial Infarction.
Potassium, glucose, and insulin treatment for acute myocardial infarction. Lancet
1968;2:1355-60.
2.
Pentecost BL, Mayne NM, Lamb P. Controlled trial of intravenous glucose, potassium, and
insulin in acute myocardial infarction. Lancet 1968;1:946-48.
3.
Fath-Ordoubadi F, Beatt KJ. Glucose-insulin-potassium therapy for treatment of acute
myocardial infarction: an overview of randomized placebo-controlled trials. Circulation
1997;96:1152-56.
4.
Rackley CE, Russell RO, Jr., Rogers WJ, Mantle JA, McDaniel HG. Glucose-insulin-potassium
infusion in acute myocardial infarction. Review of clinical experience. Postgrad Med
1979;65:93-99.
5.
Ceremuzynski L, Budaj A, Czepiel A et al. Low-dose glucose-insulin-potassium is ineffective in
acute myocardial infarction: results of a randomized multicenter Pol-GIK trial. Cardiovasc
Drugs Ther 1999;13:191-200.
6.
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.
7.
van der Horst IC, Zijlstra F, van't Hof AW et al. Glucose-insulin-potassium infusion in patients
treated with primary angioplasty for acute myocardial infarction: the glucose-insulinpotassium study: a randomized trial. J Am Coll Cardiol 2003;42:784-91.
8.
Malmberg K, Ryden L, Efendic S et al. 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 1995;26:57-65.
9.
De Luca G, Suryapranata H, Zijlstra F et al. Symptom-onset-to-balloon time and mortality in
patients with acute myocardial infarction treated by primary angioplasty. J Am Coll Cardiol
2003;42:991-97.
10.
Rogers WJ, Stanley AW, Jr., Breinig JB et al. Reduction of hospital mortality rate of acute
11.
Mittra B. Potassium, glucose, and insulin in treatment of myocardial infarction. Lancet
myocardial infarction with glucose-insulin-potassium infusion. Am Heart J 1976;92:441-54.
1965;2:607-9.
12.
Satler LF, Green CE, Kent KM, Pallas RS, Pearle DL, Rackley CE. Metabolic support during
13.
Lazar HL, Zhang X, Rivers S, Bernard S, Shemin RJ. Limiting ischemic myocardial damage
14.
Lazar HL, Chipkin SR, Fitzgerald CA, Bao Y, Cabral H, Apstein CS. Tight glycemic control in
coronary reperfusion. Am Heart J 1987;114:54-58.
using glucose-insulin-potassium solutions. Ann Thorac Surg 1995;60:411-16.
diabetic coronary artery bypass graft patients improves perioperative outcomes and
decreases recurrent ischemic events. Circulation 2004;109:1497-502.
15.
Rao V, Christakis GT, Weisel RD, Ivanov J, Borger MA, Cohen G. The insulin cardioplegia trial:
myocardial protection for urgent coronary artery bypass grafting. J Thorac Cardiovasc Surg
2002;123:928-35.
122
Overview of GIK studies in acute MI
16.
Lazar HL, Philippides G, Fitzgerald C, Lancaster D, Shemin RJ, Apstein C. Glucose-insulinpotassium solutions enhance recovery after urgent coronary artery bypass grafting. J Thorac
Cardiovasc Surg 1997;113:354-60.
17.
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.
18.
Lell WA, Nielsen VG, McGiffin DC, Schmidt FE, Jr., Kirklin JK, Stanley AW, Jr. Glucose-insulinpotassium infusion for myocardial protection during off-pump coronary artery surgery. Ann
Thorac Surg 2002;73:1246-51.
19.
Smith A, Grattan A, Harper M, Royston D, Riedel BJ. Coronary revascularization: a procedure
in transition from on-pump to off-pump? The role of glucose-insulin-potassium revisited in a
randomized, placebo-controlled study. J Cardiothorac Vasc Anesth 2002;16:413-20.
20.
Sack MN, Yellon DM. Insulin therapy as an adjunct to reperfusion after acute coronary
ischemia: a proposed direct myocardial cell survival effect independent of metabolic
modulation. J Am Coll Cardiol 2003;41:1404-7.
21.
Hjermann I. A controlled study of peroral glucose, insulin and potassium treatment in
myocardial infarction. Acta Med Scand 1971;190:213-18.
22.
Sands MJ, Jr., McDonough MT, Neubauer SJ, Lemole GM, Spann JF. Hypoglycemia
23.
Malmberg KA, Efendic S, Ryden LE. Feasibility of insulin-glucose infusion in diabetic patients
complicating the use of solution of glucose, insulin and potassium. Chest 1975;67:363-65.
with acute myocardial infarction. A report from the multicenter trial: DIGAMI. Diabetes Care
1994;17:1007-14.
24.
Neely JR, Rovetto MJ, Whitmer JT, Morgan HE. Effects of ischemia on function and
25.
Rutenberg HL, Pamintuan JC, Soloff LA. Serum-free-fatty-acids and their relation to
26.
Vik-Mo H, Mjos OD. Influence of free fatty acids on myocardial oxygen consumption and
metabolism of the isolated working rat heart. Am J Physiol 1973;225:651-58.
complications after acute myocardial infarction. Lancet 1969;2:559-64.
ischemic injury. Am J Cardiol 1981;48:361-65.
27.
Severeid L, Connor WE, Long JP. The depressant effect of fatty acids on the isolated rabbit
28.
Oliver MF, Opie LH. Effects of glucose and fatty acids on myocardial ischaemia and
heart. Proc Soc Exp Biol Med 1969;131:1239-43.
arrhythmias. Lancet 1994;343:155-58.
29.
Tansey MJ, Opie LH. Relation between plasma free fatty acids and arrhythmias within the first
30.
McDaniel HG, Papapietro SE, Rogers WJ et al. Glucose-insulin-potassium induced alterations
twelve hours of acute myocardial infarction. Lancet 1983;2:419-22.
in individual plasma free fatty acids in patients with acute myocardial infarction. Am Heart J
1981;102:10-15.
31.
Cave AC, Ingwall JS, Friedrich J et al. ATP synthesis during low-flow ischemia: influence of
increased glycolytic substrate. Circulation 2000;101:2090-2096.
32.
Heng MK, Norris RM, Peter T, Nisbet HD, Singh BN. The effect of glucose-insulin-potassium
on experimental myocardial infarction in the dog. Cardiovasc Res 1978;12:429-35.
123
Metabolic interventions in acute myocardial infarction
33.
Rackley CE, Russell RO, Jr., Rogers WJ, Mantle JA, McDaniel HG, Papapietro SE. Glucoseinsulin-potassium administration in acute myocardial infarction. Annu Rev Med 1982;33:37583.
34.
Eberli FR, Weinberg EO, Grice WN, Horowitz GL, Apstein CS. Protective effect of increased
glycolytic substrate against systolic and diastolic dysfunction and increased coronary
resistance from prolonged global underperfusion and reperfusion in isolated rabbit hearts
perfused with erythrocyte suspensions. Circ Res 1991;68:466-81.
35.
Laine H, Nuutila P, Luotolahti M et al. Insulin-induced increment of coronary flow reserve is
not abolished by dexamethasone in healthy young men. J Clin Endocrinol Metab
2000;85:1868-73.
36.
Marano L, Bestetti A, Lomuscio A et al. Effects of infusion of glucose-insulin-potassium on
myocardial function after a recent myocardial infarction. Acta Cardiol 2000;55:9-15.
37.
Sundell J, Knuuti J. Insulin and myocardial blood flow. Cardiovasc Res 2003;57:312-19.
38.
Heng MK, Norris RM, Singh BN, Barratt-Boyes C. Effects of glucose and glucose-insulinpotassium on haemodynamics and enzyme release after acute myocardial infarction. Br Heart
J 1977;39:748-57.
39.
Stanley AW, Jr., Prather JW. Glucose-insulin-potassium, patient mortality and the acute
40.
Pilcher J, Etishamudin M, Exon P, Moore J. Potassium, glucose and insulin in myocardial
myocardial infarction: results from a prospective randomised study. Circulation 1978;57:61.
infarction. Lancet 1967;1:1109.
124
Admission glucose and long-term outcome
Chapter 4.1
Prognostic value of admission glucose in non-diabetic patients
with myocardial infarction
Jorik R Timmer, Iwan CC van der Horst, Jan Paul Ottervanger, José PS Henriques, Jan CA
Hoorntje, Felix Zijlstra, on behalf of the Zwolle Myocardial Infarction Study Group
American Heart Journal 2004;148:399-404
125
Metabolic interventions in acute myocardial infarction
Abstract
Introduction
Patients with acute myocardial infarction (MI) who have diabetes have an increased risk of
mortality. In non-diabetic patients, admission glucose levels may also be a predictor of
survival. However, data regarding admission glucose and long-term outcome in nondiabetic patients treated with reperfusion therapy for acute MI are limited.
Methods
We investigated long-term clinical outcome in 356 consecutive non-diabetic patients with
ST segment elevation MI, treated with primary percutaneous coronary intervention (PCI)
or thrombolysis as reperfusion therapy. Mean follow up was 8±2 years. The patients were
divided based on admission glucose level; group I: <7.8 mmol/L, group II: 7.8−11.0
mmol/L and group III: ≥11.1 mmol/L.
Results
Mortality in group I (N=163) was 19.0%, in group II (N=151) 26.5% and in group III (N=42)
35.7% (P<0.05). Higher glucose levels were associated with larger enzymatic infarct sizes
(P<0.01) and more reduced residual left ventricular function (P<0.05). Multivariate
analysis showed that Killip class >1 at admission, odds ration (OR) 2.9 (95% confidence
interval 1.7–5.0, P<0.001), age ≥60 years, OR 2.4 (1.5–4.0, P=0.001), thrombolysis as
compared to PCI, OR 1.7 (1.1–2.7, P=0.02), admission glucose category, OR 1.4 (1.0–1.9,
P=0.04) and anterior location, OR 1.6 (1.0–2.6, 0.03) were independent predictors of longterm clinical outcome.
Conclusion
Elevated admission glucose levels in non-diabetic patients treated with reperfusion
therapy for ST segment elevation MI are independently associated with larger infarct size
and higher long-term mortality.
126
Admission glucose and long-term outcome
Introduction
Prognosis after ST segment elevation myocardial infarction (MI) has improved markedly,
in particular due to reperfusion therapy by either thrombolytic therapy or primary
percutaneous coronary intervention (PCI).1-3 Despite these advantages in treatment
regimen, patients with diabetes mellitus (DM) still have a poor prognosis after ST segment
elevation MI.4;5 Interestingly, this increased risk is not confined to DM patients only, but
non DM patients with impaired glucose tolerance (IGT) may also have an increased
incidence of cardiovascular complications.6 Moreover, increased admission glucose levels
may be related to a higher mortality in patients with acute myocardial infarction (MI),
regardless of diabetic status.7;8 These elevated glucose levels are thought to reflect pre
existent IGT or increased physical stress. However, data regarding admission glucose and
long-term clinical outcome in non-diabetic patients after ST segment elevation MI are
limited and mainly based on patient groups in which only part of the patients received
reperfusion therapy.9;10 Therefore, we investigated the influence of admission glucose
levels on long-term clinical outcome in non-diabetic patients, all treated with reperfusion
therapy, either with thrombolysis or primary PCI, for acute ST segment elevation MI.
Methods
It concerns a sub-analysis of the so-called ‘Zwolle trial’, a randomized study comparing
primary PCI with thrombolysis, as described before.3 Baseline characteristics, clinical data,
angiographic data and outcomes were recorded prospectively in a dedicated database.
Patients were enrolled if they had no contraindications for thrombolytic therapy; had
symptoms of acute myocardial infarction lasting longer than 30 minutes, accompanied by
an electrocardiogram with ST segment elevation of more than 1 mm (0.1 mV) in two or
more contiguous leads; and presented within 6 hours, or between 6 to 24 hours if there
was evidence for continuing ischemia. After informed consent had been obtained,
patients were randomly assigned to undergo PCI or to receive streptokinase (SK) as
thrombolytic agent. All patients received heparin and aspirin. Additional revascularisation
procedures were performed if indicated for symptoms or signs of myocardial ischemia,
medication was according to the guidelines.11 Global left ventricular ejection fraction was
measured by equilibrium radionuclide ventriculography between days 4 and 10 after
treatment.12 Enzymatic infarct size was estimated by measurement of serial lactate
dehydrogenase (LDH) activity. Cumulative enzyme release from 5 to 7 serial
measurements up to 72 hours after symptom onset (LDH Q72) was calculated, without
knowledge of the randomization outcome or clinical data. Previous cardiovasculair
disease (CVD) was defined as a history of acute MI, coronary artery bypass grafting or PCI.
127
Metabolic interventions in acute myocardial infarction
Nonfatal recurrent myocardial infarction was defined as the combination of chest pain,
changes in the ST T segment, and a second increase in the serum creatine kinase level to
more than two times the upper limit of normal. If the creatine kinase level had not
decreased to normal levels, a second increase of more than 200 IU per liter over the
previous value was regarded as indicating a recurrent infarction.12 Patients with DM were
defined as patients with documented DM using either oral hypoglycemic agents or insulin
treatment at admission.
For the purpose of this sub-analysis, patients with DM were excluded. The remaining
patients were divided into 3 groups regarding to admission glucose levels. Group I
consisted of patients with no evidence of impaired glucose metabolism (glucose <7.8
mmol/L), group II consisted of patients with possible impaired glucose metabolism
(glucose 7.8–11.0 mmol/L) and group III consisted of patients with profound impaired
glucose metabolism (glucose ≥11.1 mmol/L). These cut-off values are based on diagnostic
criteria for IGT and DM advised by the American Diabetes Association.13 Follow-up
information was obtained in September 2000.14 All outpatients' reports were reviewed,
and general practitioners were contacted by phone. For patients who had clinical events
during follow-up, hospital records were reviewed. All subsequent hospital admissions (for
angina, recurrent infarction, additional intervention or heart failure) and medication used
during follow-up were recorded.
Statistical analysis
Statistical analysis was performed using Statistical Package for the Social Sciences (SPSS
Inc., Chicago, IL, USA) version 10.0. Differences between group means were tested by
two-tailed Student's t-test. A chi-square statistic was calculated to test differences
between proportions, with calculation of odds ratios (OR) and exact 95% confidence
intervals. The Fisher exact test was used when the expected value of cells was smaller
than 5. Statistical significance was defined as a P-value <0.05. Cumulative survival curves
were constructed according to the Kaplan–Meier method and differences between the
curves were tested for significance by the Log-rank statistic. Cox proportional-hazards
regression model was used to estimate the independent association between glucose
levels and long-term mortality. Correlations between numeric variables were calculated
using the bivariate Pearson correlation coefficient.
Results
Baseline characteristics
The Zwolle trial patient cohort consisted of 395 patients. Of these patients, 32 (8%) had
known DM and from 7 patients no laboratory data were available. Therefore, the present
128
Admission glucose and long-term outcome
sub-analysis included 356 patients. Of the 356 patients, the mean age was 59 ±10 year and
there were 293 male patients (82%). Residual ejection fraction of the left ventricle (LVEF)
was measured in 337 patients (95%) and enzymatic infarct size in 333 patients (94%).
During the total follow up period 86 patients died (24%).
Table 1. Baseline clinical and demographic characteristics of three patient groups
Group I
Group II
Group III
(<7.8 mmol/L)
(7.8−11.0 mmol/L)
(≥11.1 mmol/L)
163 (46)
151 (42)
42 (12)
58±11
60±10∗
64±9∗
Men
139 (85)
122 (81)
32 (76)
Anterior MI
57 (35)
58 (38)
16 (41)
Previous CVD
38 (23)
27 (18)
8 (19)
Multi-vessel disease
92 (58)
74 (49)
26 (62)
Primary PCI
84 (52)
66 (44)
24 (57)
Killip class >1
15 (9)
19 (13)
7 (17)
190±205
171±175
229±239
Death
31 (19)
40 (27)
15 (36)∗
MACE
53 (33 )
58 (38)
17 (41)
Glucose level
Number of patients
Age, years (mean±SD)
Time to admission, min (mean±SD)
Outcome
Data are number (%) unless otherwise indicated. ∗ Denotes statistical difference (P<0.05) compared
to group I. MI = myocardial infarction. CVD = cardiovascular disease. MACE = major adverse
cardiac event. PCI = percutaneous coronary intervention.
Admission glucose levels
The mean admission glucose level was 8.6±3.1 mmol/L (range 5.1 to 34.1 mmol/L). Mean
glucose levels were higher in females compared to males (9.3±4.1 versus 8.4±2.9 mmol/L,
P=0.048), in patients with age ≥60 year (9.0±3.5 versus 8.2±2.6 mmol/L, P=0.016) and in
those who had Killip class >1 at presentation (9.7±4.8 mmol/L versus 8.5±2.8 mmol/L,
P=0.019). There was no association between infarct location, reperfusion strategy,
previous CVD, multi-vessel disease and glucose levels. In table 1, the baseline
characteristics of patients in the 3 glucose categories are compared. Again, elevated
glucose was more often observed in females, in patients with a higher age and in those
with Killip class >1 at presentation. There were no significant differences with regard to
discharge medication between the three glucose categories
129
Metabolic interventions in acute myocardial infarction
Glucose and infarct size
There was a clear association between glucose levels, enzymatic infarct size and LVEF.
Higher admission glucose was strongly correlated with larger enzymatic infarct size (R =
0.2, P<0.01) and lower LVEF (R = 0.1, P<0.05). Subsequently, mean enzymatic infarct size
was highest in the patient group with the highest admission glucose (figure 1) and the
mean residual LVEF was lowest in this patient group (figure 2).
Table 2. Predictors of Long-term Mortality, multivariate analysis
OR
95% confidence interval
P-value
Killip class >1
2.9
1.7–5.0
<0.001
Age ≥60 year
2.4
1.5–4.0
0.001
Thrombolysis∗
1.7
1.1–2.7
0.02
Glucose category†
1.4
1.0–1.9
0.04
Anterior MI
1.6
1.0–2.6
0.03
Previous CVD
1.3
0.8–2.2
0.29
Multi-vessel disease
1.3
0.8–2.1
0.27
Female gender
1.2
0.7–2.1
0.53
∗ As compared to primary percutaneous coronary intervention. † Glucose < 7.8 mmol/L, 7.8 – 11.0
mmol/L and glucose ≥11.1 mmol/L; glucose <7.8 mmol/L was the reference group. MI = myocardial
infarction. CVD = cardiovascular disease.
Long-term mortality
Mean glucose level in patients who died during follow-up was 9.5±4.6 mmol/L compared
with 8.3±2.5 mmol/L in survivors (P=0.003). There was a gradual increase in mortality
between the three glucose categories. Thirty-one patients died in group I (19.0%), 40
patients in group II (26.5%) and 15 patients in group III (35.7%). Mortality was significantly
higher in the patient group with highest glucose compared to the group with normal
glucose levels (P<0.05). MACE endpoints were reached in 53 patients (32.5%) in group I,
in 58 patients (38.4%) in group II and in 17 (40.5%) patients in group III (table 1). KaplanMeier curves for cumulative mortality of patients in the 3 categories are shown in figure 3.
Log-rank statistics showed significant differences in overall mortality between the patient
groups (Log-rank 7.1; P= 0.029).
Reperfusion strategy
In order to investigate whether reperfusion strategy influences the association of
admission glucose with long-term outcome, we stratified patients according to their
randomization (primary PCI or thrombolysis). Regardless of reperfusion strategy, higher
glucose levels were associated with more reduced LVEF and higher mortality. In all three
130
Admission glucose and long-term outcome
patient groups treatment with PCI resulted in better preservation of LVEF, a smaller
enzymatic infarct size and lower mortality compared to treatment with thrombolysis. In
patients treated with primary PCI, no significant difference with regard to TIMI flow
(Thrombolyis In Myocardial Infarction) prior and post intervention was present between
the glucose categories.
Figure 1. Mean enzymatic infarct size (LDH Q72) in the three patient groups
P < 0.05
1600
1525
Mean LDH Q 72 (IU/L)
1400
1216
1200
1000
882
800
0
< 7.8
7.8 - 11.0
≥ 11.1
Glucose levels (mmol/L)
Multivariate analyses
To study the independent predictive value of admission glucose on long-term survival,
multivariate regression analysis was performed including age, reperfusion strategy,
gender and all variables that were significant predictors in univariate analysis. Unadjusted
predictors of long-term mortality were Killip class >1, increased age, anterior MI, previous
CVD and multi-vessel disease. After multivariate analysis, the presence of Killip class >1
at admission odds ratio (OR) 2.9 (95% confidence interval 1.7–5.0, P<0.001), age ≥60
years, OR 2.4 (1.5–4.0, P=0.001), thrombolysis as reperfusion strategy, OR 1.7 (1.1–2.7,
131
Metabolic interventions in acute myocardial infarction
P=0.02), admission glucose category, OR 1.4 (1.0–1.9, P=0.04) and anterior location, OR
1.6 (1.0–2.6, P=0.03) were independent predictors of long-term mortality (table 2).
Figure 2. Mean residual left ventricular ejection fraction (LVEF) in the three patient groups
50
49.2 %
49
P < 0.05
LVEF (%)
48
46.8%
47
46
45.7%
45
0
7.8 - 11.0
< 7.8
≥ 11.1
Glucose levels (mmol/L)
Discussion
Our study demonstrates that in non DM patients with ST segment elevation MI, elevated
glucose levels on admission are associated with larger infarct sizes and increased longterm mortality compared to normal glucose levels on admission. Although the
pathophysiological mechanism is unknown, this adverse relation of elevated glucose
levels on admission with increased mortality is evident, despite the use or method of
reperfusion therapy, and adjusting for other predictors of long-term mortality.
Impaired glucose tolerance
Patients with elevated glucose levels on admission may represent individuals who do not
meet the diagnostic criteria of DM but have a mild form of dysglycemia. This impairment
in glucose metabolism may lead to hyperglycemia during stressful events, such as acute
132
Admission glucose and long-term outcome
MI. This sub or pre diabetic state, also known as impaired glucose tolerance (IGT), is
associated with a higher incidence of cardiovascular events.15;16 As these patients also
appear to have an increased mortality after acute MI, specific risk reducing interventions
should be considered. Exercise training, dietary modifications and medical intervention
reduce the risk of subsequent DM in these patients and may be of value.17;18
Figure 3. Kaplan-Meier curve showing overall mortality of the three patient groups. Logrank P=0.029
80
Overall mortality (%)
70
I: glucose< 7.8 mmol/L
N = 163
II: glucose 7.8 – 11.0 mmol/L N = 151
III: glucose≥ 11.1 mmol/L
N = 42
60
50
40
Group III
Group II
30
Group I
20
10
0
1
2
3
4
5
6
7
8
9
Time after admission (years)
However, intervention during hospitalization may also be of benefit. Interestingly, a
stringent insulin regimen in DM patients with acute MI abolished the increase in mortality
associated with elevated admission glucose levels. Whether insulin therapy or
intervention through other hypoglycemic agents is also beneficial for IGT patients with
acute MI is unknown.
Stress related hyperglycemia and GIK
A stress response is accompanied by high levels of catecholamines such as cortisol and
adrenaline. These hormones increase glycogenolysis, lipolysis and reduce insulin
133
Metabolic interventions in acute myocardial infarction
sensitivity resulting in elevated glucose levels.19 Therefore, patients with elevated glucose
levels could represent patients with an increased response to stress, for example due to
more severe hemodynamic compromise or more extensive myocardial damage. Indeed,
in our study, patients with elevated glucose levels did more often have Killip class >1 at
admission, had larger enzymatic infarct size and more reduced LVEF. All these variables
are known to be predictors of long-term mortality.20-22 However, in our sub-analysis, after
adjusting for confounding variables (including Killip class), there was still an association
between admission glucose and long-term mortality. Furthermore, other evidence also
suggests that dysglycemia during acute MI is more than only an epiphenomenon. Stress
induced elevation of free fatty acids (FFA) may also compromise myocardial function as
they reduce contractility and increase ischemic and reperfusion injury.23 Moreover, the
adverse relation between admission glucose and clinical outcome is also present in
patients suffering from acute coronary events without myocardial damage.24 There is
some evidence that improvement of metabolic control through infusion of fluids
containing glucose, insulin and potassium (GIK) may reduce mortality in acute MI.25-27 The
beneficial effect of GIK is thought to result from a shift in primary energy substrate from
FFA to glucose during ischemia. Our study confirms the importance of dysglycemia
during acute MI, but further investigations regarding glycometabolic control should be
initiated and awaited for.
Reperfusion strategy
Patients treated with PCI had better survival and less reduced LVEF than patients treated
with thrombolysis, these results are in line with other studies.28;29 The impact of glucose
levels on outcome however, was independent of reperfusion strategy. Patients with
higher glucose levels had higher mortality, larger enzymatic infarct sizes and more
reduced LVEF whether treated with PCI or with thrombolysis. So, apart from primary PCI,
additional treatment of patients with elevated glucose is needed.
Study limitations
Our study included only a limited number of patients from a single center. We have no
data on glycolysated hemoglobin (HbA1C) in our patients, which could have given more
insight in the prevalence of IGT in our population. No routine tests were performed to
detect undiagnosed DM after admission. However, a significant interaction of diabetic
status with the association of hyperglycemia with adverse outcome is unlikely.24 Although
non-fasting admission glucose levels may be influenced by prior meals or diurnal
variations, the impact of a concomitant acute MI on glucose levels is probably much more
substantial. Furthermore, as admission glucose is readily available, it has the advantage
that immediate intervention can be instituted. Metabolic control through GIK infusions as
134
Admission glucose and long-term outcome
adjunctive treatment was not used in our patients and, especially in patients with more
severe glycometabolic derangement, this may have had influence on outcome.30 During
the study period, intracoronary stenting and treatment with glycoprotein IIb/IIIa receptor
blockers or clopidogrel were not available.
Conclusion
Elevated admission glucose levels in non-diabetic patients with admission MI are
independently associated with larger infarct sizes and a higher long-term mortality when
compared to patients with normal glucose levels. These findings warrant further
investigation regarding glycometabolic control during acute MI and secondary prevention
programs in patients with high admission glucose.
References
1.
Effectiveness of intravenous thrombolytic treatment in acute myocardial infarction. Gruppo
Italiano per lo Studio della Streptochinasi nell'Infarto Miocardico (GISSI) Lancet 1986;1:397402.
2.
Granger CB, Califf RM, Topol EJ. Thrombolytic therapy for acute myocardial infarction. A
review. Drugs 1992;44:293-325.
3.
Zijlstra F, Hoorntje JC, de Boer MJ et al. Long-term benefit of primary angioplasty as
compared with thrombolytic therapy for acute myocardial infarction. N Engl J Med
1999;341:1413-19.
4.
Mak KH, Moliterno DJ, Granger CB et al. Influence of diabetes mellitus on clinical outcome in
the thrombolytic era of acute myocardial infarction. GUSTO-I Investigators. Global Utilization
of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries. J Am Coll
Cardiol 1997;30:171-79.
5.
Waldecker B, Waas W, Haberbosch W et al. Type 2 diabetes and acute myocardial infarction.
Angiographic findings and results of an invasive therapeutic approach in type 2 diabetic
versus nondiabetic patients. Diabetes Care 1999;22:1832-38.
6.
Gerstein HC, Yusuf S. Dysglycaemia and risk of cardiovascular disease. Lancet 1996;347:94950.
7.
Fava S, Aquilina O, Azzopardi J, Agius MH, Fenech FF. The prognostic value of blood glucose
in diabetic patients with acute myocardial infarction. Diabet Med 1996;13:80-83.
8.
Oswald GA, Smith CC, Betteridge DJ, Yudkin JS. Determinants and importance of stress
hyperglycaemia in non-diabetic patients with myocardial infarction. Br Med J (Clin Res Ed)
1986;293:917-22.
9.
Norhammar AM, Ryden L, Malmberg K. Admission plasma glucose. Independent risk factor
for long-term prognosis after myocardial infarction even in nondiabetic patients. Diabetes
Care 1999;22:1827-31.
135
Metabolic interventions in acute myocardial infarction
10.
Wahab NN, Cowden EA, Pearce NJ, Gardner MJ, Merry H, Cox JL. Is blood glucose an
independent predictor of mortality in acute myocardial infarction in the thrombolytic era? J
Am Coll Cardiol 2002;40:1748-54.
11.
Van de Werf F, Ardissino D, Betriu A et al. Management of acute myocardial infarction in
patients presenting with ST-segment elevation. The Task Force on the Management of Acute
Myocardial Infarction of the European Society of Cardiology. Eur Heart J 2003;24:28-66.
12.
Zijlstra F, de Boer MJ, Hoorntje JC, Reiffers S, Reiber JH, Suryapranata H. A comparison of
immediate coronary angioplasty with intravenous streptokinase in acute myocardial
infarction. N Engl J Med 1993;328:680-684.
13.
Report of the expert committee on the diagnosis and classification of diabetes mellitus
14.
Henriques JP, Ottervanger JP, Zijlstra F. Prognostic Importance of Left Ventricular Function
Diabetes Care 2003;26 Suppl 1:S5-20.
after Angioplasty or Thrombolysis for Acute Myocardial Infarction: Long-Term Results of the
Zwolle Trial. Cardiovasc reviews and reports 2001;22:343-48.
15.
Coutinho M, Gerstein HC, Wang Y, Yusuf S. The relationship between glucose and incident
cardiovascular events. A metaregression analysis of published data from 20 studies of 95,783
individuals followed for 12.4 years. Diabetes Care 1999;22:233-40.
16.
Eschwege E, Richard JL, Thibult N et al. [Diabetes, hyperglycemia, hyperinsulinemia and risk
of cardiovascular mortality. Findings of the Paris Prospective Survey, 10 years later]. Rev
Epidemiol Sante Publique 1985;33:352-57.
17.
Buchanan TA, Xiang AH, Peters RK et al. Preservation of pancreatic beta-cell function and
prevention of type 2 diabetes by pharmacological treatment of insulin resistance in high-risk
hispanic women. Diabetes 2002;51:2796-803.
18.
Tuomilehto J, Lindstrom J, Eriksson JG et al. Prevention of type 2 diabetes mellitus by
changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med
2001;344:1343-50.
19.
Mizock BA. Alterations in fuel metabolism in critical illness: hyperglycaemia. Best Pract Res
Clin Endocrinol Metab 2001;15:533-51.
20.
Burns RJ, Gibbons RJ, Yi Q et al. The relationships of left ventricular ejection fraction, endsystolic volume index and infarct size to six-month mortality after hospital discharge
following myocardial infarction treated by thrombolysis. J Am Coll Cardiol 2002;39:30-36.
21.
DeGeare VS, Boura JA, Grines LL, O'Neill WW, Grines CL. Predictive value of the Killip
classification in patients undergoing primary percutaneous coronary intervention for acute
myocardial infarction. Am J Cardiol 2001;87:1035-38.
22.
Gosselink AT, Liem AL, Reiffers S, Zijlstra F. Prognostic value of predischarge radionuclide
ventriculography at rest and exercise after acute myocardial infarction treated with
thrombolytic therapy or primary coronary angioplasty. The Zwolle Myocardial Infarction
Study Group. Clin Cardiol 1998;21:254-60.
23.
Mjos OD. Effect of free fatty acids on myocardial function and oxygen consumption in intact
24.
Foo K, Cooper J, Deaner A et al. A single serum glucose measurement predicts adverse
dogs. J Clin Invest 1971;50:1386-89.
outcomes across the whole range of acute coronary syndromes. Heart 2003;89:512-16.
136
Admission glucose and long-term outcome
25.
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.
26. Fath-Ordoubadi F, Beatt KJ. Glucose-insulin-potassium therapy for treatment of acute
myocardial infarction: an overview of randomized placebo-controlled trials. Circulation
1997;96:1152-56.
27.
van der Horst IC, Zijlstra F, van't Hof AW et al. Glucose-insulin-potassium infusion in patients
treated with primary angioplasty for acute myocardial infarction: the glucose-insulinpotassium study: a randomized trial. J Am Coll Cardiol 2003;42:784-91.
28.
Grines C, Patel A, Zijlstra F, Weaver WD, Granger C, Simes RJ. Primary coronary angioplasty
compared with intravenous thrombolytic therapy for acute myocardial infarction: six-month
follow up and analysis of individual patient data from randomized trials. Am Heart J
2003;145:47-57.
29.
Nunn CM, O'Neill WW, Rothbaum D et al. Long-term outcome after primary angioplasty:
report from the primary angioplasty in myocardial infarction (PAMI-I) trial. J Am Coll Cardiol
1999;33:640-646.
30.
Malmberg K, Norhammar A, Wedel H, Ryden L. Glycometabolic state at admission: important
risk marker of mortality in conventionally treated patients with diabetes mellitus and acute
myocardial infarction: long-term results from the Diabetes and Insulin-Glucose Infusion in
Acute Myocardial Infarction (DIGAMI) study. Circulation 1999;99:2626-32.
137
Metabolic interventions in acute myocardial infarction
138
Admission glucose and myocardial perfusion
Chapter 4.2
Hyperglycemia reduces myocardial reperfusion after primary
percutaneous
coronary
intervention
for
ST
elevation
myocardial infarction
Jorik R Timmer, Iwan CC van der Horst, Giuseppe de Luca, Jan Paul Ottervanger, Jan CA
Hoorntje, Menko-Jan de Boer, Harry Suryapranata, Jan-Henk E Dambrink, AT Marcel
Gosselink, Felix Zijlstra, Arnoud WJ van ’t Hof
Submitted
139
Metabolic interventions in acute myocardial infarction
Abstract
Background
Patients with hyperglycemia on admission have an adverse prognosis after ST segment
elevation myocardial infarction (MI). However, the causal link between disturbances in
glucose metabolism and outcome is unclear. We sought to investigate whether
hyperglycemia is related to impaired myocardial perfusion in ST segment elevation MI
patients.
Methods
A total of 464 patients with ST segment elevation MI treated with primary percutaneous
coronary intervention (PCI) were included. To determine myocardial reperfusion we
determined ST segment elevation resolution on the electrocardiogram and myocardial
blush grade during PCI.
Results
A total of 93 patients (20%) had hyperglycemia (glucose ≥11.0 mmol/L). They had more
often a reduced myocardial blush grade (26% versus 15%, P=0.02) and incomplete ST
segment elevation resolution (59% versus 39%, P=0.01) compared to patients without
hyperglycemia on admission. Patients with hyperglycemia also had more reduced left
ventricular function and higher mortality. These observations were demonstrated in
patients with and without diabetes mellitus. Multivariate analyses did not change these
findings.
Conclusion
Hyperglycemia on admission is associated with reduced myocardial reperfusion after
primary PCI. This is associated with an adverse clinical outcome.
140
Admission glucose and myocardial perfusion
Introduction
It has been demonstrated that hyperglycemia on admission is associated with an adverse
prognosis after acute myocardial infarction (MI), both in patients with and without
diabetes mellitus.1;2 However, the causal link between disturbances in glucose metabolism
and outcome is not yet clear. Evidence suggests that hyperglycemia could be associated
with impaired microvascular myocardial reperfusion.3;4 We studied the association
between hyperglycemia and myocardial reperfusion in patients treated with primary
percutaneous coronary intervention (PCI) for acute ST segment elevation MI (MI). Both
electrocardiographic (ST segment elevation resolution) and angiographic (myocardial
blush grade) parameters of myocardial reperfusion were used. We also investigated the
impact of reduced myocardial reperfusion on clinical outcome.
Methods
Study population
All consecutive patients with symptoms consistent with acute MI of >30 min duration,
presenting within 24 hours after the onset of symptoms and with ST segment elevation of
more than 1 mm (0.1 mV) in two or more contiguous leads on the electrocardiogram were
evaluated for inclusion in this study. Patients were excluded when pre-treated with
thrombolysis or when an illness associated with a marked restricted life expectancy was
present. All patients went to the catheterization laboratory as soon as possible, where
both coronary arteries were visualized. PCI was performed with standard techniques if the
coronary anatomy was suitable for angioplasty. Additional treatment consisted of
intravenous heparin, nitroglycerin and aspirin. After sheath removal, low-molecularweight heparin was given for 1 to 3 days. Baseline characteristics, clinical data,
angiographic data and outcomes were recorded prospectively in a dedicated database.
Measurements and definitions
Core laboratory annalists (Diagram BV, Zwolle, The Netherlands) who were unaware of
the clinical history and outcome of the patients assessed ST segment elevation resolution,
TIMI (thrombolysis in myocardial infarction) flow and myocardial blush grade. TIMI flow
and myocardial blush grade were visually assessed on the angiogram. Myocardial blush
grade has been defined previously5: 0, no myocardial blush; 1, minimal myocardial blush
or contrast density; 2, moderate myocardial blush or contrast density but less than that
obtained during angiography of a contra or ipsilateral non-infarct related coronary artery;
and 3, normal myocardial blush or contrast density, comparable with that obtained during
141
Metabolic interventions in acute myocardial infarction
angiography of a contralateral or ipsilateral non-infarct-related coronary artery. When
myocardial blush persisted (“staining”), this phenomenon suggested leakage of contrast
medium into the extravascular space and was graded 0. ST segment elevation resolution
was analyzed as previously described in detail.6 In short, ST segment elevation resolution
was defined as complete when there was ≥70% ST segment elevation resolution, partial
when there was 30-70% resolution and absent if resolution was less than 30% as
measured 180 minutes after primary PCI compared to ST elevation on admission. Left
ventricular ejection fraction (LVEF) was measured before discharge by radionuclide
ventriculography or by echocardiography. Radionuclide ventriculography was performed
by using the multiple gated equilibrium method following the labeling of red blood cells of
the patient with technetium-99m-pertechnate. A General Electric 300 gamma camera with
a low-energy all-purpose parallel-hole collimator was used. Global ejection fraction was
calculated by a General Electric Star View computer using the fully automatic PAGE
program. Hyperglycemia was defined as whole blood glucose levels (Modular system –
Roche/Hitachi, Basel, Switserland) on admission of ≥11.1 mmol/L (≥200 mg/dL), as stated
by the American Diabetes Association.7
Statistical analysis
Differences between group means at baseline were assessed with the two-tailed Student’s
t-test. Chi-square analysis or Fisher’s exact test was used to test differences between
proportions. To study independent predictors of reduced myocardial reperfusion,
multivariate logistic regression analysis was performed. Statistical significance was
considered a two-tailed P-value <0.05. The Statistical Package for the Social Sciences
(SPSS Inc., Chicago, IL, USA) version 10.1 was used for all statistical analysis.
Results
Study population
A total of 464 patients were included in this analysis. Myocardial blush grade was
available in 401 patients (86%) and ST segment elevation resolution data were available in
302 patients (65%). Ejection fraction was measured in 404 patients (87%). Mean age was
61±12 years, 368 patients (79%) were men. At the end of the follow up period (mean
1.6±1.1 years) 38 patients (8%) had died. Baseline characteristics of the patient groups
according to the absence or presence of hyperglycemia are shown in table 1.
142
Admission glucose and myocardial perfusion
Table 1. Baseline characteristics of the patient groups according to admission glucose
Glucose <11.1mmol/L
Number of patients
Glucose ≥11.1mmol/L
P-value
371
93
60.3±12.0
62.8±11.7
0.07
Men
307 (83)
61 (66)
0.001
Previous MI
39 (11)
14 (15)
0.22
Previous PCI
20 (5)
4 (4)
0.67
Previous CABG
8 (2)
4 (4)
0.24
Diabetes mellitus
11 (3)
38 (41)
<0.001
Hypertension
99 (27)
31 (33)
0.20
Currently smoking
195 (53)
42 (45)
0.20
Age, years (mean±SD)
Positive family history
145 (39)
34 (37)
0.66
Ischemic time ≤3 hour
198 (60)
52 (61)
0.94
Heart rate (beats /minute)
73 ± 18
81 ± 26
0.01
Systolic BP ≤100 mmHg
37 (10)
14 (15)
0.16
Diastolic BP ≤60 mmHg
71 (19)
23 (25)
0.23
Killip class ≥2
20 (5)
14 (15)
0.001
Anterior infarction
176 (47)
48 (52)
0.47
Ischemic time ≤3 hour
198 (60)
52 (61)
0.94
Heart rate (beats /minute)
73±18
81±26
0.01
Admission glucose (mmol/L)*
8.1±1.4
14.4±3.9
<0.001
Glucose at 16 hours (mmol/L)*
7.9±1.9
12.0±3.7
<0.001
Cumulative ST elevation (mm)
9.4±8.7
9.0±6.9
0.77
30 (8)
12 (13)
0.15
IABP
Data are number (%) unless otherwise indicated. * Data are mean ± SD. MI = myocardial infarction.
PCI = percutaneous coronary intervention. CABG = coronary artery bypass grafting. BP = blood
pressure. IABP= intra aortic balloon pumping.
Hyperglycemia and myocardial reperfusion
Both complete ST segment elevation resolution (61% versus 42%, P=0.009) and optimal
myocardial blush (85% versus 74%, P=0.02) were higher in patients with normal glucose
levels compared to those with hyperglycemia. Patients with hyperglycemia had more
often a reduced LVEF (LVEF ≤30%) (28% versus 16%, P=0.01) and a higher mortality
compared to patients without hyperglycemia (14% versus 7%, P=0.02), table 2.
Diabetes mellitus and myocardial perfusion
In diabetic patients, the absence of hyperglycemia was associated with a higher presence
of complete ST segment elevation resolution (56% versus 41%, P=0.46) and optimal
143
Metabolic interventions in acute myocardial infarction
myocardial blush grade (91% versus. 74%, P=0.23). Also in non-diabetic patients with
normal glucose levels, both complete ST segment elevation resolution (61% versus 42%,
P=0.04) and optimal myocardial blush grade (85% versus 74%, P=0.07) were higher
compared to those without hyperglycemia.
Table 2. Myocardial reperfusion and outcome of the patient groups according to
admission glucose
Glucose <11.1mmol/L
Glucose ≥11.1 mmol/L
P-value
Myocardial blush grade (2-3)
272 (85)
59 (74)
0.02
Complete resolution
152 (61)
22 (42)
0.009
51 (16)
23 (28)
0.01
25 (7)
13 (14)
0.023
Myocardial reperfusion
Myocardial infarct size
LVEF ≤30%
Mortality
Data are number (%). LVEF = left ventricular ejection fraction.
Multivariate analyses - ST segment elevation resolution
To identify whether hyperglycemia was independently associated with ST segment
elevation resolution we performed multivariate analyses and included age and gender and
all clinical variables significantly different between patients with and without ST segment
elevation resolution (anterior MI P=0.04, hyperglycemia P=0.009 and cumulative ST
elevation P=0.002). After multivariate analyses, the variables that were independently
associated with incomplete ST segment elevation resolution were the presence of
admission hyperglycemia , relative risk (RR) 2.1 (95% confidence interval 1.1–4.0, P=0.02),
anterior MI, RR 2.4 (1.4–4.0, P=0.001) and a cumulative ST elevation on admission RR 1.1
per mm elevation (1.0–1.1, P=0.001). Gender (P=0.54) and age (P=0.94) were not
significantly associated with ST segment elevation resolution after multivariate analysis.
The addition of time to treatment to our analysis did not change these findings.
Multivariate analyses - myocardial blush grade
Univariate angiographic characteristics of patients with and without hyperglycemia are
shown in table 3. The only variable that was associated with hyperglycemia was a
myocardial blush grade 0-1. After multivariate analysis including all angiographic variables
(MVD, collaterals, infarct related vessel, TIMI flow and myocardial blush grade)
hyperglycemia at admission remained associated with an impaired blush grade of 0-1, RR
2.2 (1.2–4.2, P=0.02). A stratified analysis including only patients with TIMI 3 flow after PCI
also showed more often a reduced myocardial blush grade in patients with hyperglycemia
144
Admission glucose and myocardial perfusion
(20% versus 9%, P=0.01). Furthermore, there was a gradual increase (7%, 9%, 10%, 12%
and 19%) in reduced myocardial blush grade with increasing glucose levels (quintiles) in
these patients (figure 1).
Figure 1. Association between glucose at admission and myocardial blush grade
Reduced myocardial blush grade (%)
20
15
10
5
0
Quintile 1
Quintile 2
Quartile 3
Quartile 4
Quartile 5
Quintiles of glucose at admission
Myocardial reperfusion and outcome
Reduced LVEF was more often present in patients with incomplete ST segment elevation
resolution (25% versus 12%, P=0.01) and in those with reduced myocardial blush (33%
versus 14%, P<0.001). Mortality was also higher in patients with incomplete ST segment
elevation resolution (6% versus 3%, P=0.15) and in those with reduced myocardial blush
(16% versus 4%, P<0.001). These associations between reduced myocardial reperfusion
and worse outcome were observed in patients with and without hyperglycemia.
Discussion
This is the first study to report on the influence of hyperglycemia on myocardial
reperfusion as assessed by myocardial blush grade and ST segment elevation resolution.
Elevated glucose levels on admission were independently associated with both
incomplete ST segment elevation resolution and reduced myocardial blush. These
findings suggest microvascular dysfunction in hyperglycemic patients with ST segment
elevation MI and might explain their adverse prognosis. Whether meticulous regulation of
145
Metabolic interventions in acute myocardial infarction
glucose levels, prior or following restoration of epicardial flow, improves myocardial
reperfusion is unclear.
Table 3. Angiographic variables of the patient groups with regard to admission glucose
P-value
Glucose <11.1 mmol/L
Glucose ≥11.1 mmol/L
Multi-vessel disease
188 (51)
54 (58)
0.20
Collaterals
39 (12)
7 (8)
0.36
IRV (LAD)
164 (46)
42 (46)
0.91
TIMI 0 flow before PCI
222 (63)
58 (64)
0.86
TIMI 2-3 flow after PCI
232 (96)
82 (94)
0.44
Myocardial blush grade 0-1
49 (15)
21 (26)
0.02
Data are number (%). IRV = infarct related vessel. LAD = left descending artery. TIMI =
thrombolysis in myocardial infarction.
Increased age and diabetes were both more prevalent in hyperglycemic patients and their
presence has been associated with reduced myocardial flow after restoration of epicardial
flow.4;8 However, in our analysis, age was not associated with impaired ST segment
elevation resolution and also non-diabetic patients with hyperglycemia appeared to have
reduced myocardial reperfusion. There are several mechanisms that could clarify these
findings. Hyperglycemia is associated with both plugging of leukocytes in the
microvasculature of the myocardium and with increased procoaguable properties of
platelets.9-11 Furthermore, the ability of platelets to induce vasodilatation is reduced by
high levels of glucose.12 However, there may also be a link between stress, hyperglycemia
and reduced myocardial reperfusion. Killip class ≥2 was significantly higher in patients
with hyperglycemia (5% versus 15%, P=0.001). Stress, induced by hemodynamic
instability, increases both levels of free fatty acids and glucose.13 These free fatty acids
reduce endothelium derived vasodilatation and inhibit the myocardial use of glucose.14;15
Furthermore, improvement of myocardial flow by insulin is diminished, as it is dependent
on the concomitant use of glucose. Administration of solutions containing glucose –
insulin and potassium (GIK) may decrease the level of free fatty acids and promote the use
of glucose as myocardial energy substrate. There is some evidence that improvement of
metabolic control through infusion of fluids containing GIK may reduce mortality in acute
MI.16-18 Whether this improved prognosis results from improved myocardial reperfusion is
unclear. However, evidence suggests GIK infusion improves myocardial perfusion in
segments adjacent to the recently infarcted area.19 Besides GIK infusion, other specific
interventions, such as the use of glycoprotein IIb/IIIa receptor blockers, might be
beneficial. Also in our study, patients with incomplete ST segment elevation resolution
and reduced myocardial blush grade had significantly more reduced LVEF and higher
146
Admission glucose and myocardial perfusion
mortality. Therefore, the search for improvement of myocardial flow is pivotal and further
investigations are warranted.
Study limitations
As glycolysated hemoglobin levels, reflecting long-term glucose metabolism, were not
available, it remains unclear whether isolated acute hyperglycemia or more long-standing
disturbed glucose metabolism was associated with reduced myocardial reperfusion. No
routine tests were performed to diagnose previously unrecognized diabetes in all patients,
so some hyperglycemic patients may have had undiagnosed diabetes at discharge.
Conclusion
Hyperglycemia on admission is independently associated with impaired myocardial
reperfusion. Impaired reperfusion was associated reduced LVEF and a higher mortality.
Further investigation with regard to improvement of myocardial reperfusion in
hyperglycemia is warranted.
References
1.
Fava S, Aquilina O, Azzopardi J, Agius MH, Fenech FF. The prognostic value of blood glucose
in diabetic patients with acute myocardial infarction. Diabet Med 1996;13:80-83.
2.
Oswald GA, Smith CC, Betteridge DJ, Yudkin JS. Determinants and importance of stress
hyperglycaemia in non-diabetic patients with myocardial infarction. Br Med J (Clin Res Ed)
1986;293:917-22.
3.
Di Carli MF, Janisse J, Grunberger G, Ager J. Role of chronic hyperglycemia in the
pathogenesis of coronary microvascular dysfunction in diabetes. J Am Coll Cardiol
2003;41:1387-93.
4.
Iwakura K, Ito H, Ikushima M et al. Association between hyperglycemia and the no-reflow
phenomenon in patients with acute myocardial infarction. J Am Coll Cardiol 2003;41:1-7.
5. van 't Hof AW, Liem A, Suryapranata H, Hoorntje JC, de Boer MJ, Zijlstra F. Angiographic
assessment of myocardial reperfusion in patients treated with primary angioplasty for acute
myocardial infarction: myocardial blush grade. Zwolle Myocardial Infarction Study Group.
Circulation 1998;97:2302-6.
6.
van 't Hof AW, Liem A, de Boer MJ, Zijlstra F. Clinical value of 12-lead electrocardiogram after
successful reperfusion therapy for acute myocardial infarction. Zwolle Myocardial infarction
Study Group. Lancet 1997;350:615-19.
7.
Report of the expert committee on the diagnosis and classification of diabetes mellitus
Diabetes Care 2003;26 Suppl 1:S5-20.
8.
Angeja BG, de Lemos J, Murphy SA et al. Impact of diabetes mellitus on epicardial and
microvascular flow after fibrinolytic therapy. Am Heart J 2002;144:649-56.
9.
Winocour PD. Platelet abnormalities in diabetes mellitus. Diabetes 1992;41 Suppl 2:26-31.
147
Metabolic interventions in acute myocardial infarction
10.
Gresele P, Guglielmini G, De Angelis M et al. Acute, short-term hyperglycemia enhances
shear stress-induced platelet activation in patients with type II diabetes mellitus. J Am Coll
Cardiol 2003;41:1013-20.
11.
Hokama JY, Ritter LS, Davis-Gorman G, Cimetta AD, Copeland JG, McDonagh PF. Diabetes
enhances leukocyte accumulation in the coronary microcirculation early in reperfusion
following ischemia. J Diabetes Complications 2000;14:96-107.
12.
Oskarsson HJ, Hofmeyer TG. Platelets from patients with diabetes mellitus have impaired
13.
Rutenberg HL, Pamintuan JC, Soloff LA. Serum-free-fatty-acids and their relation to
ability to mediate vasodilation. J Am Coll Cardiol 1996;27:1464-70.
complications after acute myocardial infarction. Lancet 1969;2:559-64.
14.
Oliver MF, Opie LH. Effects of glucose and fatty acids on myocardial ischaemia and
15.
Lind L, Fugmann A, Branth S et al. The impairment in endothelial function induced by non-
arrhythmias. Lancet 1994;343:155-58.
esterified fatty acids can be reversed by insulin. Clin Sci (Lond) 2000;99:169-74.
16.
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.
17. Fath-Ordoubadi F, Beatt KJ. Glucose-insulin-potassium therapy for treatment of acute
myocardial infarction: an overview of randomized placebo-controlled trials. Circulation
1997;96:1152-56.
18.
van der Horst IC, Zijlstra F, van't Hof AW et al. Glucose-insulin-potassium infusion in patients
treated with primary angioplasty for acute myocardial infarction: the glucose-insulinpotassium study: a randomized trial. J Am Coll Cardiol 2003;42:784-91.
19.
Marano L, Bestetti A, Lomuscio A et al. Effects of infusion of glucose-insulin-potassium on
myocardial function after a recent myocardial infarction. Acta Cardiol 2000;55:9-15.
148
Hyperglycemia and myocardial reperfusion
Chapter 4.3
Hyperglycemia
and
no
reflow
phenomenon
in
acute
myocardial infarction
Iwan CC van der Horst, Arnoud WJ van ’t Hof, Henk JG Bilo, Felix Zijlstra
Journal of the American College of Cardiology 2003;41:2100
149
Metabolic interventions in acute myocardial infarction
To the editor:
Dr Iwakura and colleagues conclude that hyperglycemia might be associated with
impaired microvascular function after acute myocardial infarction (MI), this conclusion
was reached by a retrospective analysis of 146 patients.1 The authors were the first to find
an independent relation between elevated blood glucose levels at admission and no
reflow phenomenon. They also stated that this relation was independent of HbA1C level
or diabetes mellitus (DM), since there were no difference in HbA1c level or the frequency
of DM in the no reflow and reflow group.
We have some considerations about these results. First, the authors defined
hyperglycemia by the optimal cutoff to differentiate the patients showing no reflow with a
receiver-operating characteristic curve analysis. By using this cutoff in the same cohort
they found a relation between hyperglycemic patients and no reflow phenomenon. The
authors have generated an interesting hypothesis and new studies have to proof the
validity of the cutoff. Second, we are interested in more details on the methods used to
determine blood glucose and HbA1c levels. Information was lacking is the blood glucose
level was measured either with point of care or whole blood glucose measurement and if
the method is validated in critically ill patients. Clinically relevant differences in blood
glucose level have been observed with point of care and whole blood measurement in
patients with shock2, acidosis3, medication4, and different values due to differences in
hematocrit5. Finally, the absolute differences of 0.3% in HbA1c and 13% in frequency of
DM were indeed not statistically significant related to no reflow. It is our suggestion that
these clinically relevant differences were not significant due to the small number of
patients. Moreover, the authors found that 45.3% of the patients with hyperglycemia were
diabetic versus 9.9% in patients without hyperglycemia (P<0.0001).
The hypothesis that acute hyperglycemia (i.e. hyperglycemia at admission) is associated
with the no reflow phenomenon is intriguing. We however hypothesize that these relation
is not independent of chronic hyperglycemia (i.e. elevated HbA1c and/or DM). Therefore
new studies have to determine the effect of hyperglycemia on no reflow and preferentially
the effect of metabolic regulation.
References
1.
Iwakura K, Ito H, Ikushima M et al. Association between hyperglycemia and the no-reflow
phenomenon in patients with acute myocardial infarction. J Am Coll Cardiol 2003;41:1-7.
2.
Atkin SH, Dasmahapatra A, Jaker MA, Chorost MI, Reddy S. Fingerstick glucose
determination in shock. Ann Intern Med 1991;114:1020-1024.
3.
Tang Z, Du X, Louie RF, Kost GJ. Effects of pH on glucose measurements with handheld
glucose meters and a portable glucose analyzer for point-of-care testing. Arch Pathol Lab
Med 2000;124:577-82.
150
Hyperglycemia and myocardial reperfusion
4.
Tang Z, Du X, Louie RF, Kost GJ. Effects of drugs on glucose measurements with handheld
glucose meters and a portable glucose analyzer. Am J Clin Pathol 2000;113:75-86.
5.
Louie RF, Tang Z, Sutton, DV, Lee JH, Kost GJ. Point-of-care glucose testing: effects of critical
care variables, influence of reference instruments, and a modular glucose meter design. Arch
Pathol Lab Med 2000;124:257-66.
151
Metabolic interventions in acute myocardial infarction
152
Time-averaged glucose and outcome
Chapter 5.1
Glucose at admission or time-averaged glucose during
hospital stay: predict major adverse cardiac events in patients
with acute myocardial infarction?
Iwan CC van der Horst, Maarten WN Nijsten, Felix Zijlstra
Submitted
153
Metabolic interventions in acute myocardial infarction
Abstract
Background
Elevated blood glucose values are a prognostic factor in myocardial infarction (MI)
patients. The unfavorable relation between hyperglycemia and outcome is known for
admission glucose and fasting glucose after admission. These predictors are single
measurements and thus not indicative of overall hyperglycemia. Increased time-averaged
glucose may better predict adverse events in MI patients.
Methods
In a prospective study of MI patients treated with primary percutaneous coronary
intervention (PCI) frequent blood glucose measurements were obtained, to investigate the
relation between glucose and the occurrence of major adverse cardiac events (MACE) at
30 days follow-up. MACE was defined as death, recurrent infarction, repeat primary
coronary intervention, and left ventricular ejection fraction ≤30%.
Results
MACE occurred in 89 (21.3%) out 417 patients. In 17 patients (4.1%) it was a fatal event
and in 72 patients (17.3%) a non-fatal event. MACE occurred more often in patients who
were older, had previous cardiovascular disease, anterior infarction location, and multivessel disease. A mean of 7.4 glucose determinations were available per patient. Mean ±
SD admission glucose was 10.1±3.7 mmol/L in patients with a MACE versus 9.1±2.7
mmol/L in event free patients (P=0.005). Mean time-averaged glucose was 9.0±2.8
mmol/L in patients with MACE compared to 8.0±2.0 mmol/L in event free patients
(P<0.001). The area under the receiver operator characteristic curve was 0.64 for timeaveraged glucose and 0.59 for admission glucose. With Cox regression analysis including
factors related with MACE in univariate analysis time-averaged glucose emerged as a
significant independent predictor (P<0.001).
Conclusion
Elevated time-averaged glucose in MI has a stronger relation with 30-day MACE than
elevated glucose at admission.
154
Time-averaged glucose and outcome
Introduction
Acute myocardial infarction (MI) patients with hyperglycemia at admission have a worse
prognosis.1;2 This relation is found in both patients with diabetes mellitus (DM) and in
patients without DM.3-6 It has been argued that in patients without DM hyperglycemia may
be caused by undetected diabetes.7-9 Some patients who present with hyperglycemia are
indeed diabetic but many patients with an admission glucose above 11.0 mmol/L are not
diabetic.10 The early measurement of glycosylated hemoglobin (HbA1c) has been useful to
find unknown diabetics among MI patients.7;11-13 HbA1c reflects the level of glucose
regulation of the previous weeks. In diabetics it serves as a measure of glucose regulation
and is a strong predictor of the risk for coronary heart disease.14 In MI patients HbA1c also
showed a relation with worse outcome.5 In MI patients HbA1c and hyperglycemia at
admission independently predicted short term mortality.11 Elevated fasting glucose after
admission for MI also predicts unfavorable outcome.15;16 Admission glucose and fasting
glucose as predictors of outcome have the drawback that they are based on a single
measurement and thus are not indicative of persistent hyperglycemia.1-6;15-17 A first step to
define persistent hyperglycemia is to compute the mean of all glucose values, a simple
mean glucose ignores the unequal time distribution between measurements.17 The
calculation of the time-averaged glucose addresses this problem. In this post-hoc analysis
of data collected in a prospective study of MI patients we thought to investigate whether
elevated time-averaged glucose is a better predictor of major adverse cardiac events than
elevated glucose at admission.
Methods
Patients
Patients with symptoms consistent with acute MI of >30 min duration, presenting within
24 hour after the onset of symptoms and with ST segment elevation of more than 1 mm
(0.1 mV) in two or more contiguous leads on the electrocardiogram and treated with
primary percutaneous coronary intervention (PCI) were included in this study. At baseline
age, gender, previous cardiovascular disease defined as a history of coronary artery
bypass grafting (CABG), previous PCI, stroke and MI, existence of diabetes mellitus (DM),
smoking status, Killip class, electrocardiographic site of infarction, time of onset of
symptoms, and time of hospital admission were recorded. Patients were defined as
diabetic when treated with a diet, oral hypoglycemic drugs and/or insulin. The research
protocol was reviewed and approved by the medical ethics committee, and patients were
included after informed consent.
155
Metabolic interventions in acute myocardial infarction
Glucose measures
In all patients glucose levels were determined based on measurements of whole-blood
glucose (Modular System, Roche/Hitachi, Basel, Switzerland). The first glucose after
admission was defined as admission glucose. To determine the time-averaged glucose
level for an individual patient, all glucose measurements performed during admission
were analyzed with a dedicated computer algorithm. The first step was to interpolate all
glucose values obtained during the first 48 hours after admission into a curve. Then the
area under this glucose curve was calculated. This area under the curve was then divided
by 48 hours.
Enzymatic infarct size
Enzymatic infarct size was estimated by serial measurements of creatine kinase (CK)
fraction. CK was determined enzymatically on a Hitachi 717 automatic analyzer according
to the International Federation of Clinical Chemistry (IFCC) recommendation at 30 degrees
Celsius. Frequent CK determinations were performed according to a schedule that called
for 4 to 8 measurements in the first 96 hours to calculate the area under the CK curves.
Left ventricular function
Left ventricular ejection fraction (LVEF) was measured before discharge by radionuclide
ventriculography or by echocardiography. Radionuclide ventriculography was performed
with the multiple gated equilibrium method following the labeling of red blood cells of the
patient with technetium-99m-pertechnate. A General Electric 300 gamma camera with a
low-energy all-purpose parallel-hole collimator was used. Global ejection fraction was
calculated by a General Electric Star View computer using the fully automatic PAGE
program. Two-dimensional echocardiography was performed by observers who were
unaware of the clinical data. A LVEF ≤30% was defined as a poor left ventricular
function.18
Cardiac events
In all patients data were obtained with regard to mortality, recurrent myocardial infarction
or repeat PCI during the first 30 days after admission. The primary endpoint of the study
was the presence of a Major Adverse Cardiac Event (MACE) during the first 30 days after
admission for acute MI. MACE was defined as the composite incidence of death, recurrent
infarction, repeat PCI or a LVEF ≤30%. The combination of death and non-fatal short-term
events as recurrent infarction, repeat intervention, and heart failure has been shown to be
a valid predictor of 1-year mortality.19;20 Recurrent myocardial infarction was defined as
the occurrence of symptoms consistent with acute MI of >30 min duration, signs of
156
Time-averaged glucose and outcome
infarction on the electrocardiogram, and a second increase in the serum CK level to more
than two times the upper limit of normal. If the CK level had not decreased to normal
levels, a second increase of more than 200 IU per liter over the previous value was
regarded as indicating a recurrent infarction.21 Repeat PCI was defined as angioplasty
performed within 30 days due to repeat signs and symptoms of myocardial ischemia.
Table 1. Baseline characteristics patients with a major adverse cardiac events (MACE) and
without an adverse event within 30-days
Characteristics
MACE group
Event free group
89
338
Number of patients
Age, years (mean±SD)
62.5±12.6
60.1±11.8
Men
70 (78.7)
263 (80.2)
Previous cardiovascular events
17 (19.1)
53 (15.7)
Diabetes mellitus
11 (12.4)
33 (10.1)
11 (12.4)
28 (8.3)
6 (6.7)
17 (5.0)
Type 2 diabetes mellitus†
Tablets
2 (2.2)
13 (3.8)
Hypertension
Insulin
32 (36.0)
87 (26.5)
Dyslipidemia
16 (18.0)
75 (22.9)
Currently smoker
41 (46.1)
178 (54.3)
Positive family history
27 (30.3)
133 (40.5)
Data are numbers (%) unless otherwise indicated.
Statistical analysis
Differences between groups were assessed with the Student’s t-test or the Mann-Whitney
U test. The Chi-square test and the Chi-square test for trend were used to test differences
between proportions. Receiver operator characteristics (ROC) curves were computed to
assess the ability of glucose-derived parameters to predict MACE or mortality. We also
performed a Cox proportional-hazards regression model with factors at admission related
to MACE at least with a level of significance of 0.2 or less. The Statistical Package for the
Social Sciences (SPSS Inc., Chicago, IL, USA) version 11.5 was used for all statistical
analysis.
Results
Between April 1th 1998 and October 1th 2001, 417 patients were included. After 30 days a
MACE had occurred in 89 patients (21.3%), in 17 patients (4.1%) it was a fatal event and in
72 patients (17.3%) a non-fatal event. A repeat PCI was performed in 20 patients (4.8%).
157
Metabolic interventions in acute myocardial infarction
Left ventricular function data was available of 377 patients (90.4%). In 68 patients (18%)
the LVEF was equal or smaller than 30%. Baseline characteristics of patients with a MACE
and patients without an event are represented in table 1. Patients with a MACE were older,
more often female, and more likely to have a history of cardiovascular disease.
Figure 1. Association between admission glucose and time-averaged glucose and 30-day
major adverse cardiac events
Receiver operator characteristic curves that depict the ability of admission glucose (thin line) and
time-averaged glucose (thick line) to predict the occurrence of a major adverse coronary event
within 30 days after admission.
A mean of 7.4 glucose determinations were available per patient. Mean ± SD admission
glucose was 10.1±3.7 mmol/L in patients with a MACE versus 9.1±2.7 mmol/L in event
free patients (P=0.005). Mean time-averaged glucose was 9.0±2.8 mmol/L in patients with
MACE compared to 8.0 ± 2.0 mmol/L in event free patients (P<0.001). 44 patients had
diabetes mellitus and 34 patients (77.3%) had a glucose level at admission and 38 patients
(86.4%) had a time-averaged glucose in the highest quartile.
158
Time-averaged glucose and outcome
Table 2. Infarct characteristics, hemodynamic status and reperfusion treatment
Characteristics
MACE group
Number of patients
Event free group
89
338
Anterior MI
69 (77.5)
131 (39.9)**
Killip class 1
79 (88.8)
311 (92.0)
Killip class ≥2
10 (11.2)
27 (8.0)
Multi-vessel disease
58 (65.2)
156 (47.6)**
TIMI grade 0 flow before PCI
60 (67.4)
212 (64.6)
Stent
48 (53.9)
184 (56.1)
GP IIb/IIIa receptor blocker
28 (26.2)
79 (24.1)
TIMI grade 3 flow after PCI
75 (86.2)
305 (93.3)
Data are number (%) unless otherwise indicated. * P-value ≤0.2. ** P-value <0.01. MI = myocardial
infarction. IQR = interquartile range. †Time to admission denotes time between onset of symptoms
and until admission.
The ROC curves for admission glucose and glucose over the first 48 hours are shown in
figure 1. The area under the curve for admission glucose was 0.59 (95% confidence
interval 0.52-0.65) and for time-averaged glucose was 0.64 (0.57-0.70). After correction for
factors related with MACE in univariate analysis (tables 1, 2 and 3) anterior site of
infarction, Killip class ≥2 and time-averaged glucose were the only independent predictors
of 30-day MACE.
Table 3. Relation between glucose measures and major adverse cardiac events
MACE group
Event free group
P-value
Number of patients
89
338
Admission glucose
10.1±3.7
9.1±2.7
0.002
Time-averaged glucose
9.0±2.8
8.0±2.0
<0.001
Glucose data are mmol/L (mean ± SD).
The positive relations between admission glucose and 30-day MACE is illustrated in figure
2. MACE in the lowest quartile of glucose during admission was 11.5% compared to
33.3% in the highest quartile (P for trend <0.05). The stronger positive relation between
time-averaged glucose and 30-day MACE is also illustrated in figure 2. The stepwise
increase is more pronounced with time-averaged glucose than with admission glucose.
MACE in the lowest quartile of time-averaged glucose was 13.5% compared to 26.7% in
the highest quartile (P for trend <0.001).
159
Metabolic interventions in acute myocardial infarction
Figure 2. 30-day major cardiac adverse events (MACE) according to quartiles of
admission glucose and time-averaged glucose in MI patients
35
30
MACE (%)
25
20
15
10
5
0
Admission glucose
Quartile 1
Quartile 2
Time-averaged glucose
Quartile 3
Quartile 4
P-value for trend in admission glucose is <0.05 and for time-averaged glucose is <0.001.
Discussion
In this study we observed that the time-averaged glucose in the first 48 hours after
admission for acute MI has a stronger relation with unfavourable short-term outcome than
glucose at admission. Previous studies primarily focused on the prognostic value of
admission hyperglycemia in both patients with and without diabetes mellitus.1-6 It has
been described that patients with diabetes mellitus have an impaired outcome after
myocardial infarction. Potentially, patients with persistent hyperglycemia are more likely
to have diabetes mellitus either known or unknown.9;10;16 In our study patients with
diabetes mellitus indeed had an elevated glucose at admission and a elevated timeaveraged glucose. The number of patients without diabetes mellitus in the highest quartile
of admission glucose and time-averaged glucose was approximately 2 out of 3 patients.
Some studies showed that an elevated fasting glucose after admission also predicts
unfavorable outcome.15;16 Only one study used a measure of persistent hyperglycemia in
the analysis of the relation between deregulation of the glucose metabolism during MI.17
In a study of 662 MI patients hyperglycemia was defined as a glucose level on admission
or a four day mean blood glucose level higher than 6.67 mmol/L. 457 patients (69.0%) had
160
Time-averaged glucose and outcome
hyperglycemia and only 195 (29.7%) had previously known diabetes mellitus. These
patients developed more complications, and had higher 28 day mortality.17
Mechanism of action
Several mechanisms can be postulated to explain the relation between acute and
persistent hyperglycemia and unfavourable outcome after reperfusion therapy for MI.
First, the relation between persistent hyperglycemia and outcome may be related to the
presence of ongoing stress in more severely ill patients. During myocardial ischemia an
increase in glucose levels is observed.22 It has been reported that glucose was related to
adrenaline and cortisol in patients with ST segment elevation MI.23 In the current study,
Killip class ≥2 on admission was more frequent and enzymatic infarct size was larger,
suggesting that hyperglycemia reflects extensive myocardial damage. The second
possible explanation is that patients with hyperglycemia are likely to have diabetes
mellitus, even if it has not been diagnosed.7;7-13 Diabetes mellitus is associated with
extensive coronary artery disease and adverse outcomes in MI patients. In the current
study, elevated admission glucose and elevated time-averaged glucose were related to a
higher prevalence of diabetes mellitus. Third, it is possible that hyperglycemia and
concomitant metabolic abnormalities may exacerbate myocardial damage in MI.
Table 4. Relation of quartiles of admission glucose and time-averaged glucose
respectively with major adverse cardiac events and infarct size
Quartile 1
Quartile 2
Quartile 3
Quartile 4
P-value*
MACE (%)
14 (13.5)
23 (22.1)
24 (23.1)
28 (26.7)
P<0.05
LVEF (% ± SEM)
44.6±1.1
41.9±1.2
41.8±1.2
40.7±1.2
CK AUC (IU ± SEM)
919±69
1430±155
1560±136
1579±146
MACE (%)
12 (11.5)
17 (16.3)
25 (24.0)
35 (33.3)
LVEF (% ± SEM)
45.3±1.0
43.3±1.0
40.8±1.3
39.4±1.4
CK AUC (IU ± SEM)
1001±80
1283±99
1505±143
1690±181
Admission glucose
Time-averaged glucose
P<0.001
MACE = major adverse cardiac events. LVEF = left ventricular ejection fraction. SEM = standard
error of the mean. CK = creatine kinase. AUC = area under the curve. *P-value denotes P-value for
trend.
Persistent hyperglycemia may reflect insulin resistance. Insulin resistance is more often
present in non-diabetic patients with an acute MI compared to matched controls.24 In 181
patients admitted with acute MI and no history of diabetes mellitus compared to 180
matched controls without previously known diabetes mellitus or cardiovascular disease
161
Metabolic interventions in acute myocardial infarction
glucose, HbA1c, proinsulin, proinsulin/insulin ratio, triglycerides, insulin resistance and
fibrinogen were all consistently higher in patients than controls (P<0.01). It has been
postulated
that
insulin
resistance,
and thereby
hyperglycemia,
decreases
the
responsiveness of leukocytes by inflammatory mediators. The question remains as to
whether this hyperglycemic state adds injury to the ischemic myocardium, or is merely a
marker for more severe disease or underlying metabolic disorder. Animal studies have
shown that hyperglycemia may impair cardiac contractile function, interfere with nitric
oxide pathways and promote apoptosis in the myocardium.25;26 Relative insulin deficiency
decreases glucose transporter translocation to the cell surface and increases free fatty
acids through increased lipolysis in adipose tissue. In this situation, glucose utilization is
reduced and free fatty acids are mainly used in myocardium. This results in increased
oxygen demand and, in ischemic myocardium, accumulation of unoxidized products of
free fatty acids. Free fatty acids inhibit glucose oxidation to a greater extent than glucose
uptake. In the presence of hyperglycemia, glycolytic intermediates therefore accumulate
in the myocyte. Furthermore, hyperglycemia can lead to activation of protein kinase C and
modification of macromolecules by forming advanced glycation end products, which can
augment the production of proinflammatory cytokines. Recently, in a retrospective study
it was shown that non-diabetic patients with signs of insulin resistance had an impaired
recovery of left ventricular function of coronary angioplasty compared to patients without
insulin resistance.27
Role for insulin (glucose-potassium) infusion
In the clinical setting, the ultimate proof that the hyperglycemic state may have harmful
effects on the ischemic myocardium is to demonstrate that cardiac outcomes can be
improved by strict glucose regulation. Previous studies have investigated the benefits of
insulin administration on outcomes after acute MI. The Diabetes and Insulin-Glucose
Infusion in Acute Myocardial Infarction (DIGAMI) study was a randomized trial of insulinglucose infusion therapy in the reperfusion era and demonstrated that insulin-glucose
therapy improved long-term survival after acute MI.28;29 In control patients, mortality was
higher in patients with higher admission blood glucose. However, in patients with insulinglucose therapy, reduction of mortality was more obvious in patients with higher blood
glucose and the relation between mortality and admission blood glucose disappeared.
These findings suggest that relative insulin deficiency and hyperglycemia may be at least
in part causally associated with increased mortality after acute MI. Evidence from animal
studies suggested that insulin therapy may reduce reperfusion injury and enhance the
cardioprotective effects of reperfusion therapy.30;31 On the other hand, hyperglycemia has
been associated with the no reflow phenomenon following successful reperfusion after ST
segment elevation MI.32 In the Estudio Cardiologicos Latinoamerica (ECLA) pilot trial, the
162
Time-averaged glucose and outcome
combination of reperfusion therapy and glucose–insulin–potassium infusion for 24 hours
following ST segment elevation MI resulted in a lower in-hospital mortality rate than the
group that received reperfusion therapy alone. In the DIGAMI study, a direct comparison
of the effect of thrombolysis in the treatment and control groups was not described. GIK
in reperfused patients seems promising, albeit that this observation comes from a subgroup of a pilot trial in which the mortality rate in the control group was high. The GIPS
and REVIVAL treated patients primarily with primary PCI.33;34 After 30-days the mortality
rates in GIK and control patients were almost similar. New trials as the ECLA GIK
2/CREATE trial, the Organization to Assess Strategies for Ischemic Syndromes (OASIS)-6
trial and GIPS-2 currently randomize thousands of patients to GIK or no GIK in adjunction
to either thrombolysis or primary PCI. These trials will determine the clinical benefit of GIK
in combination with different reperfusion strategies.
Conclusion
In patients with acute MI treated with primary PCI hyperglycemia defined by the timeaveraged glucose during hospital admission predicts unfavorable short-term outcome
better than admission hyperglycemia.
Contributors
This study was supported by a generous grant from the Netherlands Heart Foundation
(99.028). No other conflicts of interest or financial disclosure.
References
1.
Malmberg K, Norhammar A, Wedel H, Ryden L. Glycometabolic state at admission: important
risk marker of mortality in conventionally treated patients with diabetes mellitus and acute
myocardial infarction: long-term results from the Diabetes and Insulin-Glucose Infusion in
Acute Myocardial Infarction (DIGAMI) study. Circulation 1999;99:2626-32.
2.
Timmer JR, van der Horst IC, Ottervanger JP et al. Prognostic value of admission glucose in
non-diabetic patients with myocardial infarction. Am Heart J 2004;148:399-404.
3.
Bolk J, van der Ploeg T, Cornel JH, Arnold AE, Sepers J, Umans VA. Impaired glucose
metabolism predicts mortality after a myocardial infarction. Int J Cardiol 2001;79:207-14.
4.
Capes SE, Hunt D, Malmberg K, Gerstein HC. Stress hyperglycaemia and increased risk of
death after myocardial infarction in patients with and without diabetes: a systematic
overview. Lancet 2000;355:773-78.
163
Metabolic interventions in acute myocardial infarction
5.
Norhammar A, Tenerz A, Nilsson G et al. Glucose metabolism in patients with acute
myocardial infarction and no previous diagnosis of diabetes mellitus: a prospective study.
Lancet 2002;359:2140-2144.
6.
Stranders I, Diamant M, van Gelder RE et al. Admission blood glucose level as risk indicator
of death after myocardial infarction in patients with and without diabetes mellitus. Arch Intern
Med 2004;164:982-88.
7.
Oswald GA, Corcoran S, Yudkin JS. Prevalence and risks of hyperglycaemia and undiagnosed
diabetes in patients with acute myocardial infarction. Lancet 1984;1:1264-67.
8.
Husband DJ, Alberti KG, Julian DG. "Stress" hyperglycaemia during acute myocardial
infarction: an indicator of pre-existing diabetes? Lancet 1983;2:179-81.
9.
Madsen JK, Haunsoe S, Helquist S et al. Prevalence of hyperglycaemia and undiagnosed
diabetes mellitus in patients with acute myocardial infarction. Acta Med Scand 1986;220:32932.
10.
Tenerz A, Lonnberg I, Berne C, Nilsson G, Leppert J. Myocardial infarction and prevalence of
diabetes mellitus. Is increased casual blood glucose at admission a reliable criterion for the
diagnosis of diabetes? Eur Heart J 2001;22:1102-10.
11.
Hadjadj S, Coisne D, Mauco G et al. Prognostic value of admission plasma glucose and HbA
12.
Greci LS, Kailasam M, Malkani S et al. Utility of HbA(1c) levels for diabetes case finding in
13.
Oswald GA, Yudkin JS. Hyperglycaemia following acute myocardial infarction: the
14.
Selvin E, Marinopoulos S, Berkenblit G et al. Meta-analysis: glycosylated hemoglobin and
15.
O'Sullivan JJ, Conroy RM, Robinson K, Hickey N, Mulcahy R. In-hospital prognosis of patients
16.
Ravid M, Berkowicz M, Sohar E. Hyperglycemia during acute myocardial infarction. A six-year
in acute myocardial infarction. Diabet Med 2004;21:305-10.
hospitalized patients with hyperglycemia. Diabetes Care 2003;26:1064-68.
contribution of undiagnosed diabetes. Diabet Med 1987;4:68-70.
cardiovascular disease in diabetes mellitus. Ann Intern Med 2004;141:421-31.
with fasting hyperglycemia after first myocardial infarction. Diabetes Care 1991;14:758-60.
follow-up study. JAMA 1975;233:807-9.
17.
Sala J, Masia R, Gonzalez de Molina FJ et al. Short-term mortality of myocardial infarction
patients with diabetes or hyperglycaemia during admission. J Epidemiol Community Health
2002;56:707-12.
18.
Greenberg H, Case RB, Moss AJ, Brown MW, Carroll ER, Andrews ML. Analysis of mortality
events in the Multicenter Automatic Defibrillator Implantation Trial (MADIT-II). J Am Coll
Cardiol 2004;43:1459-65.
19.
Braunwald E, Cannon CP, McCabe CH. Use of composite endpoints in thrombolysis trials of
20.
Cannon CP, Sharis PJ, Schweiger MJ et al. Prospective validation of a composite end point in
acute myocardial infarction. Am J Cardiol 1993;72:3G-12G.
thrombolytic trials of acute myocardial infarction (TIMI 4 and 5). Thrombosis In Myocardial
Infarction. Am J Cardiol 1997;80:696-99.
21.
Zijlstra F, de Boer MJ, Hoorntje JC, Reiffers S, Reiber JH, Suryapranata H. A comparison of
immediate coronary angioplasty with intravenous streptokinase in acute myocardial
infarction. N Engl J Med 1993;328:680-684.
164
Time-averaged glucose and outcome
22.
Vetter NJ, Strange RC, Adams W, Oliver MF. Initial metabolic and hormonal response to acute
myocardial infarction. Lancet 1974;1:284-88.
23.
Oswald GA, Smith CC, Betteridge DJ, Yudkin JS. Determinants and importance of stress
hyperglycaemia in non-diabetic patients with myocardial infarction. Br Med J (Clin Res Ed)
1986;293:917-22.
24.
Bartnik M, Malmberg K, Hamsten A et al. Abnormal glucose tolerance--a common risk factor
in patients with acute myocardial infarction in comparison with population-based controls. J
Intern Med 2004;256:288-97.
25.
Ren J, Gintant GA, Miller RE, Davidoff AJ. High extracellular glucose impairs cardiac E-C
coupling in a glycosylation-dependent manner. Am J Physiol 1997;273:H2876-H2883.
26.
Du XL, Edelstein D, Dimmeler S, Ju Q, Sui C, Brownlee M. Hyperglycemia inhibits endothelial
nitric oxide synthase activity by posttranslational modification at the Akt site. J Clin Invest
2001;108:1341-48.
27.
Fujino T, Ishii Y, Takeuchi T et al. Recovery of BMIPP uptake and regional wall motion in
insulin resistant patients following angioplasty for acute myocardial infarction. Circ J
2003;67:757-62.
28.
Malmberg K, Ryden L, Efendic S et al. 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 1995;26:57-65.
29.
Malmberg K. Prospective randomised study of intensive insulin treatment on long term
survival after acute myocardial infarction in patients with diabetes mellitus. DIGAMI (Diabetes
Mellitus, Insulin Glucose Infusion in Acute Myocardial Infarction) Study Group. BMJ
1997;314:1512-15.
30.
Zhu P, Lu L, Xu Y, Greyson C, Schwartz GG. Glucose-insulin-potassium preserves systolic and
diastolic function in ischemia and reperfusion in pigs. Am J Physiol Heart Circ Physiol
2000;278:H595-H603.
31.
Doenst T, Richwine RT, Bray MS, Goodwin GW, Frazier OH, Taegtmeyer H. Insulin improves
functional and metabolic recovery of reperfused working rat heart. Ann Thorac Surg
1999;67:1682-88.
32.
Iwakura K, Ito H, Ikushima M et al. Association between hyperglycemia and the no-reflow
33.
Pache J, Kastrati A, Mehilli J et al. A randomized evaluation of the effects of glucose-insulin-
phenomenon in patients with acute myocardial infarction. J Am Coll Cardiol 2003;41:1-7.
potassium infusion on myocardial salvage in patients with acute myocardial infarction treated
with reperfusion therapy. Am Heart J 2004;148:105.
34.
van der Horst IC, Zijlstra F, van't Hof AW et al. Glucose-insulin-potassium infusion in patients
treated with primary angioplasty for acute myocardial infarction: the glucose-insulinpotassium study: a randomized trial. J Am Coll Cardiol 2003;42:784-91.
165
Metabolic interventions in acute myocardial infarction
166
Hyperglycemic index in critically ill patients
Chapter 5.2
The hyperglycemic index is a tool to assess glucose control: a
retrospective study
Mathijs Vogelzang, Iwan CC van der Horst, Maarten WN Nijsten
Critical Care 2004;8:R122-R127
167
Metabolic interventions in acute myocardial infarction
Abstract
Background
Critically ill patients may benefit from strict glucose regulation. An objective measure of
hyperglycemia to assess glucose regulation in acutely ill patients should reflect the
magnitude and duration of hyperglycemia, be independent of the number of
measurements and not be falsely lowered by hypoglycemic values. The time-average of
glucose values above the normal range meets these requirements. It was our aim to
investigate whether the hypeglycemic index (HGI) is related to outcome
Methods
A retrospective, single center study was performed at a 12-bed surgical ICU. From 1990
through 2001 all patients over 15 years, staying at least 4 days were included. Admission
type, sex, age, APACHE-II and outcome were recorded. The HGI was defined as the area
under the curve above the upper limit of normal (glucose level 6.0 mmol/L) divided by the
total length of stay. HGI, admission glucose, mean morning glucose, mean glucose and
maximal glucose were calculated for each patient. The relation of these measures with 30day mortality was determined.
Results
In 1779 patients with a median ICU-stay of 10 days the 30-day mortality was 17%. A total
of 65 528 glucose values were analyzed. Median HGI was 0.9 (0.3-2.1) in survivors
compared to 1.7 (0.7-3.3) mmol/L in non-survivors (p<0.001). Area under the receiver
operator characteristic curve was 0.64 for HGI, compared with 0.61 and 0.62 for mean
morning glucose and mean glucose. HGI was the only significant glucose measure in
binary logistic regression.
Conclusion
HGI showed a better relation with outcome than other glucose indices. HGI is an useful
measure of glucose regulation in critically ill patients.
168
Hyperglycemic index in critically ill patients
Introduction
Acute hyperglycemia is a prognostic factor for mortality in critically ill patients in the
presence and in the absence of diabetes mellitus.1-3 The benefit of strict glucose regulation
in the Intensive Care Unit (ICU) has been demonstrated by the Leuven study.4;5 A
remarkable reduction in morbidity and mortality was achieved in patients who were
treated according to a protocol that aimed for normoglycemia. Thus the logical aims of
glucose regulation are to eliminate hyperglycemia as rapidly as possible and to maintain
normoglycemia from then on, while avoiding hypoglycemia.6;7 Whereas this presents a
clear goal for algorithms, there is no obvious way to assess the performance of different
algorithms.8
In ICU patients we do not possess a measure like glycosylated hemoglobin A1c (HbA1c)
which has proven to be an important predictor of long-term complications and useful to
evaluate the quality of glucose regulation.9-11 Therefore glucose itself must be measured to
assess hyperglycemia during the stay at the ICU. In studies of acutely ill patients, regular
indices of glucose regulation that have been used are admission glucose, maximum
glucose, mean morning glucose and mean glucose.4;12-15 All these indices have specific
drawbacks. Admission glucose, maximum glucose and mean morning glucose are all
based on either a single measurement or a subset of measurements, and thus are not
indicative of overall hyperglycemia. A single mean glucose that uses all measurements
can be strongly biased by an unequal time distribution between measurements, as
commonly occurs in practice.16-18 Calculating a time-averaged glucose compensates for an
unequal time distribution of glucose measurements. However, hypoglycemic episodes
may still lower such an index, thus falsely suggesting normoglycemia, when in reality
hyperglycemia is present.
We hypothesized that an index that takes into account the unequal time distribution of
glucose sampling and that is not falsely lowered by low glucose values would be a better
index of glucose regulation. We defined the hyperglycemic index (HGI; figure 1) as the
area under the glucose curve above the normal range, divided by the length of stay. We
evaluated the association of HGI and conventional glucose indices of regulation with
mortality in a large group of ICU patients with a prolonged ICU stay.
Methods
Study population
In a retrospective analysis we included all patients more than 15 years of age admitted to
the surgical ICU of our tertiary teaching hospital from 1990 to the end of 2001. Since
glucose regulation appears to be especially relevant in patients with a prolonged stay at
169
Metabolic interventions in acute myocardial infarction
the ICU, we only studied patients who stayed four or more days at the ICU.4;5 Age, sex,
admission type and the acute physiology and chronic health evaluation II score (APACHEII) were obtained from case-records and electronic databases of all admitted patients to
our hospital. Blood glucoses values were obtained from the central laboratory database.
Figure 1. Calculation of the hyperglycemic index (HGI).
All measured glucose values (black dots) and their corresponding sampling times are taken into
account. The time-average is calculated for the area (shaded) under the glucose curve for
hyperglycemic values only. The normal glucose-range is indicated by the hatched area, with 6.0
mmol/L (dotted line) as the cut-off. HGI is the shaded area divided by the total length of stay. In this
case HGI is 0.73 mmol/L, as indicated by the dashed line. Note that normal or hypoglycemic
measurements do not affect HGI, and thus do not falsely lower this index.
Therapeutic protocol
Patients were fed enterally as soon as possible. Total parenteral nutrition was only given
when enteral nutrition failed. Concentrated glucose infusion was not routinely used.
Insulin was only administrated to patients with diabetes mellitus or patients with glucose
levels exceeding 10.0 mmol/L, and was never dosed above 10 IU/hour. Whole blood
samples were taken from arterial or central lines and sent to the central laboratory for
glucose measurement.
Glucose indices
Admission glucose was defined as the first measurement after ICU-admission. Morning
glucose was calculated as the arithmetic mean of all measurements done between 6 and 8
170
Hyperglycemic index in critically ill patients
AM.4;5 Mean glucose was calculated as the arithmetic mean of all measurements.
Maximum glucose was the highest glucose determined for the entire ICU-stay.
To determine the hyperglycemic index (HGI) of a patient, all glucose measurements
performed during the ICU stay were analyzed. As indicated in figure 1, the first step was to
interpolate all glucose values. Then the area between this glucose curve and the upper
normal range was calculated. HGI was defined as this area under the curve divided by the
total length of stay, thus making HGI independent of length of stay.
As the Leuven proved improved outcome by lowering glucose levels under 6.0 mmol/L,
we chose this value as our upper range of normal in all tests unless otherwise noted.
Since the Leuven study others have hypothesized that 6.0 mmol/L might not be the best
target to aim for.19 Therefore we also performed an analysis of the performance of HGI at
other cut-off levels than 6.0.
Table 1. Characteristics for surviving and non-surviving patients and results of univariate
analysis of glucose indices
Survivors
Non-survivors
Number of patients
1484 (83)
295 (17)
Men
982 (66)
182 (61)
0.18
53±19
63±16
<0.001
10 (6-20)
10 (6-17)
Age, years (mean±SD)
Length of ICU stay, hour*
Reason for ICU admission
P-value
0.12
<0.001
Trauma
372 (25)
30 (10)
Abdominal surgery
443 (30)
101 (34)
Liver transplant
219 (15)
38 (13)
Vascular surgery
164 (11)
51 (17)
Miscellaneous
286 (19)
75 (25)
APACHE-II score
18 (14-23)
25 (20-28)
<0.001
Number of glucose measurements*
20 (10.5-41)
27 (15-45)
<0.001
Mean glucose, mmol/L*
6.9 (6.0-8.4)
7.7 (6.4-9.5)
<0.001
Morning glucose, mmol/L*
6.6 (5.9-7.9)
7.5 (6.2-8.8)
<0.001
Admission glucose, mmol/L*
7.2 (5.8-9.5)
7.9 (6.0-10.9)
0.07
Maximum glucose, mmol/L*
10.2 (8.0-14.2)
12.3 (9.5-16.4)
<0.001
0.9 (0.3-2.1)
1.8 (0.7-3.4)
<0.001
HGI, mmol/L*
Data are number (%) unless otherwise indicated. *Values are medians (interquartile range) unless
otherwise noted. ICU denotes intensive care unit. HGI denotes hyperglycemic index.
As the other measures of glucose regulation, HGI is expressed in mmol/L. Thus a patient
in whom all glucose values happen to be 8.5 mmol/L will have an HGI of 2.5 mmol/L. A
patient who is normoglycemic with all glucoses ≤6.0 mmol/L will have an HGI of 0.0
171
Metabolic interventions in acute myocardial infarction
mmol/L. Even though this study's primary focus is hyperglycemia, hypoglycemia cannot
be omitted. We determined the number of hypoglycemic episodes per patient as the
number of intervals of 4 hours with at least one measurement <2.7 mmol/L.6;7
Statistical analysis
Data were expressed as medians and interquartile ranges (IQR) unless otherwise
indicated. Differences between groups were assessed with the Mann-Whitney U test. Chisquare analysis was used to test differences between proportions. The primary endpoint
was 30-day mortality. In univariate analysis we assessed the performance of HGI and
other glucose-derived measures in relation with 30-day mortality. Patients were divided in
survivors (patients alive at 30 days) and non-survivors. Receiver operator characteristics
(ROC) curves were computed. We performed a multivariate binary logistic regression
analysis with age, sex, type of admission, APACHE-II and all glucose-derived measures as
independent parameters and 30-day mortality as the dependent parameter. Differences
were considered significant for a two-tailed P-value <0.05. The Statistical Package for the
Social Sciences (SPSS Inc., Chicago, IL, USA; version 11.0.1) was used for statistical
analysis.
Results
In the 12-year period 6 885 patients were admitted to the ICU. A total of 1 779 patients
(26%) stayed for a period of at least four days and were included in the study. Mean age
was 55 years (SD±19) and 65% were males. Table 1 lists the demographic data and
glucose related measures for survivors and non-survivors. APACHE-II scores were
available for the years 1992-1999, for all other parameters there were no missing data.
Abdominal surgery and trauma were the most frequent reasons for ICU admission. A total
of 65528 glucose measurements were performed in the 1 779 included patients, with a
median number of glucose measurements of 21 (IQR 11-42). In less than 1% of the
patients not a single glucose measurement was performed. The median mean glucose
value of all patients was 7.0 (IQR 6.1-8.6), median morning glucose was 6.7 (IQR 5.9-8.1),
median admission glucose was 7.3 (IQR 5.8-9.7), median maximum glucose was 8.7 (IQR
6.9-11.6) and median HGI was 1.0 (IQR 0.4-2.4). Severe hypoglycemia (glucose <2.7
mmol/L) occurred in 177 (6.6%) patients. The median duration of such hypoglycemic
episodes was 1.5 (0.6-3.4) hours.
Survivors and non-survivors both stayed at the ICU for a median of 10 days, IQR was 6 to
20 days for survivors and 6 to 17 days for non-survivors. A total of 295 patients (17%) died
within 30 days after ICU admission.
172
Hyperglycemic index in critically ill patients
Figure 2. Receiver operator characteristic (ROC) curves for different glucose measures
In the univariate analysis, median mean glucose level was 7.7 mmol/L in non-survivors
compared to 6.8 mmol/L in survivors (P<0.001). Median HGI was 1.8 in non-survivors,
twice as high as in survivors (P<0.001). The ROC curves for all glucose derived
parameters are shown in figure 2. HGI had the highest area under the curve (0.64).
Figure 3 depicts the relation of HGI-quartiles with mortality. Mortality in the lowest HGIquartile was 8.6% compared to 25.1% in the highest HGI-quartile (P<0.001). Figure 4
shows the area under the ROC curve for HGI when cut-off values other than 6.0 mmol/L
are used. In multivariate analysis with APACHE-II, sex and age, HGI remained as the only
significant glucose index in the binary logistic model (P<0.001). P-values of mean
glucose, morning glucose, admission glucose and maximum glucose were 0.08, 0.17,
0.43 and 0.49, respectively. With regard to mortality the results of regression analysis did
not differ between the cohort of patients whose APACHE-II scores were available and the
cohort of which the APACHE-II scores were not available.
Discussion
Of all measures of hyperglycemia evaluated, HGI correlated best with 30-day mortality in
this population of critically ill patients. This supports our hypothesis that HGI is a useful
173
Metabolic interventions in acute myocardial infarction
index to quantify glucose regulation. Therefore, assuming that normoglycemia is the aim,
the goal of a glucose-insulin algorithm is clear. The algorithm should obtain an HGI as
close to zero as possible. Both in the univariate analysis and in the multivariate binary
logistic regression analysis - where severity of illness, age and sex were included - HGI
emerged as the best indicator of hyperglycemia. We assume this reflects the fact that HGI
takes better account of the factor time, and avoids the phenomenon that alternating high
and low values average out to a normal value.
Figure 3. Relation between HGI - divided in quartiles - and mortality
30-day mortality (%)
30
20
10
0
0-0.4 mmol/L
0.4-1.0 mmol/L
1.0-2.4 mmol/L
2.4-8.9 mmol/L
HGI quartiles
In the highest quartile mortality is nearly three times higher than mortality in the lowest quartile
(P<0.001). HGI denotes hyperglycemic index.
The principle of calculating the area under the glucose curve is not new. Brown and
Dodek determined the area under the curve above a glucose threshold of 11.5 mmol/L.8
The area under the curve was used to assess the speed of initial normalization of glucose
with an insulin algorithm. Other recent studies have also used glucose thresholds above
10.0 mmol/L to separate good control from poor control.20;21 However, such thresholds
are now considered as high with regard to the Leuven study, since that study
demonstrated improved outcome if normoglycemia (4.4 to 6.1 mmol/L) was pursued.4;5
Regarding the fact that the vast majority of glucoses in our patients were <10 mmol/L,
crucial information would have been lost with a cut-off of 10.0 mmol/L as reflected by a
decreased area under the ROC-curve at a cut-off value for HGI of 10.0 mmol/L (figure 4).
The cut-off of 6.0 mmol/L was based on the upper limit of the group of patients with strict
174
Hyperglycemic index in critically ill patients
regulation in the Leuven study. Recently Finney and colleagues found that glucose
regulation below 8.3 mmol/L was not related to a better outcome.19 This agrees with our
observations, as figure 4 shows that the optimal HGI-cut-off lies between 6.0 and 8.0
mmol/L.
Study limitations
Some limitations of this study should be mentioned. Our study concerned a retrospective,
single ICU study that covers a period when strict glucose regulation was not a major
issue. Mortality at 30 days was used as an outcome measure to identify the best glucose
index. HGI was the best measure of glucose regulation, but with a ROC-area of 0.64 HGI
alone can not serve as a useful predictor of mortality. The relative contributions of
endogenous glucose production, exogenous glucose supply and insulin to HGI and other
some other measures could not be identified since our study did not include glucose
infusion or (par)enteral feeding, nor did it include intensive treatment with insulin.
Figure 4. HGI for various glucose cut-offs
ROC area
0,7
0,6
0,5
4
6
8
10
12
14
16
Cut-off for HGI mmol/l
The cut-off in all other analyses was chosen at 6 mmol/L as it was the upper limit of the intensive
treatment group in the Leuven study. To see how the performance of HGI is at other cut-off values
the area under the ROC curve was determined for HGI cut-offs from 4.0 to 15.0 mmol/L. In the
patients studied, a cut-off between 6.0 and 8.0 mmol/L was associated with the highest area.
It should be stressed that HGI was designed to quantify hyperglycemia, not hypoglycemia.
Prevention of hypoglycemia is a critical requirement for any algorithm for glucose
regulation.5-7 However, unlike hyperglycemia, hypoglycemia is a phenomenon that tends
to be relatively short-lived as our results show, and might be quantified by more
175
Metabolic interventions in acute myocardial infarction
straightforward measures, such as the lowest glucose. An elevated admission glucose
level is associated with a worse outcome, as has been found by many investigators in
various patient categories, as was also found in our study.1-3;12;13;22-30 In our study however
the area under the ROC curves was smaller for all conventional measures of glucose
regulation than it was for HGI.
Like other indices of glucose regulation, HGI is related to outcome. However in contrast to
admission glucose, HGI is also amenable to therapy. HGI involves additional computation
(figure 1) compared with more straightforward indices, but HGI does not require more
information. Calculating HGI should be feasible in ICU's that possess a patient database
management system (PDMS) that can provide automated input for the HGI calculation.
The fact that HGI expresses glucose regulation as a single value, has methodological
advantages. The performance of glucose-insulin algorithms could be compared with HGI.
Therefore it is important to regularly measure glucose. A major advantage of HGI is that
periods of very frequent sampling (e.g. during hyperglycemia or hypoglycemia) are
compensated for as HGI is based on a time average. HGI must be reassessed in the era of
tighter glucose regulation. Moreover, the value of HGI needs confirmation in other ICUs.
Since HGI has not been used by other investigators, it would be of interest how HGI
compares with other glucose indices in observational or intervention studies. Existing
glucose patient databases could be reanalyzed to determine HGI. The use of glucose
measures to predict outcome independently of other parameters such as age and severity
scores is interesting but lacks power as is proven by the area under the ROC. More in
general HGI may be more useful to relate hyperglycemia with organ failure scores such as
the Sequential Organ Failure Assessment (SOFA) score
31
, or parameters of systemic
inflammation. Continuous measurements of blood glucose will allow to calculate and
compare HGI and the value of other glucose measures to a degree not possible with
intermittent measurements.32-35 Currently available glucose sensors are promising but
have not yet proven to be sufficiently reliable in critically ill patients and do not allow
continuous measurements over prolonged periods.32-35
Conclusion
In conclusion, HGI quantifies the impact of hyperglycemia in critically ill patients better
than other glucose indices. HGI may thus be a useful measure of glucose regulation.
References
1.
Capes SE, Hunt D, Malmberg K, Gerstein HC. Stress hyperglycaemia and increased risk of
death after myocardial infarction in patients with and without diabetes: a systematic
overview. Lancet 2000;355:773-78.
176
Hyperglycemic index in critically ill patients
2.
Capes SE, Hunt D, Malmberg K, Pathak P, Gerstein HC. Stress hyperglycemia and prognosis
of stroke in nondiabetic and diabetic patients: a systematic overview. Stroke 2001;32:242632.
3.
Umpierrez GE, Isaacs SD, Bazargan N, You X, Thaler LM, Kitabchi AE. Hyperglycemia: an
independent marker of in-hospital mortality in patients with undiagnosed diabetes. J Clin
Endocrinol Metab 2002;87:978-82.
4.
van den Berghe G, Wouters P, Weekers F et al. Intensive insulin therapy in the critically ill
patients. N Engl J Med 2001;345:1359-67.
5. 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.
6.
Fischer KF, Lees JA, Newman JH. Hypoglycemia in hospitalized patients. Causes and
outcomes. N Engl J Med 1986;315:1245-50.
7. Stagnaro-Green A, Barton MK, Linekin PL, Corkery E, deBeer K, Roman SH. Mortality in
hospitalized patients with hypoglycemia and severe hyperglycemia. Mt Sinai J Med
1995;62:422-26.
8.
Brown G, Dodek P. Intravenous insulin nomogram improves blood glucose control in the
critically ill. Crit Care Med 2001;29:1714-19.
9.
Tenerz A, Lonnberg I, Berne C, Nilsson G, Leppert J. Myocardial infarction and prevalence of
diabetes mellitus. Is increased casual blood glucose at admission a reliable criterion for the
diagnosis of diabetes? Eur Heart J 2001;22:1102-10.
10.
Tenerz A, Norhammar A, Silveira A et al. Diabetes, insulin resistance, and the metabolic
syndrome in patients with acute myocardial infarction without previously known diabetes.
Diabetes Care 2003;26:2770-2776.
11.
Woo J, Lam CW, Kay R, Wong AH, Teoh R, Nicholls MG. The influence of hyperglycemia and
diabetes mellitus on immediate and 3-month morbidity and mortality after acute stroke. Arch
Neurol 1990;47:1174-77.
12.
Bolk J, van der Ploeg T, Cornel JH, Arnold AE, Sepers J, Umans VA. Impaired glucose
metabolism predicts mortality after a myocardial infarction. Int J Cardiol 2001;79:207-14.
13.
Gore DC, Chinkes D, Heggers J, Herndon DN, Wolf SE, Desai M. Association of
14.
Krinsley JS. Association between hyperglycemia and increased hospital mortality in a
hyperglycemia with increased mortality after severe burn injury. J Trauma 2001;51:540-544.
heterogeneous population of critically ill patients. Mayo Clin Proc 2003;78:1471-78.
15.
Yendamuri S, Fulda GJ, Tinkoff GH. Admission hyperglycemia as a prognostic indicator in
16.
Bonnier M, Lonnroth P, Gudbjornsdottir S, Attvall S, Jansson PA. Validation of a glucose-
trauma. J Trauma 2003;55:33-38.
insulin-potassium infusion algorithm in hospitalized diabetic patients. J Intern Med
2003;253:189-93.
17. van der Horst IC, Gans RO, Nijsten MW, Ligtenberg JJ. Beneficial effect of glucose-insulinpotassium infusion in noncritically ill patients has to be proven. J Intern Med 2003;254:513.
18.
Bonnier M, Lonnroth P, Gudbjornsdottir S, Attvall S, Jansson PA. Beneficial effect of glucoseinsulin- potassium infusion in noncritically ill patients has to be proven - reply. J Intern Med
2003;254:514.
177
Metabolic interventions in acute myocardial infarction
19.
Finney SJ, Zekveld C, Elia A, Evans TW. Glucose control and mortality in critically ill patients.
20.
Furnary AP, Gao G, Grunkemeier GL et al. Continuous insulin infusion reduces mortality in
JAMA 2003;290:2041-47.
patients with diabetes undergoing coronary artery bypass grafting. J Thorac Cardiovasc Surg
2003;125:1007-21.
21.
Pomposelli JJ, Baxter JK, III, Babineau TJ et al. Early postoperative glucose control predicts
nosocomial infection rate in diabetic patients. JPEN J Parenter Enteral Nutr 1998;22:77-81.
22.
Bruno A, Levine SR, Frankel MR et al. Admission glucose level and clinical outcomes in the
23.
Foo K, Cooper J, Deaner A et al. A single serum glucose measurement predicts adverse
24.
Goldberg PA, Siegel MD, Sherwin RS et al. Implementation of a Safe and Effective Insulin
25.
Lam AM, Winn HR, Cullen BF, Sundling N. Hyperglycemia and neurological outcome in
26.
Paret G, Barzilai A, Lahat E et al. Gunshot wounds in brains of children: prognostic variables
NINDS rt-PA Stroke Trial. Neurology 2002;59:669-74.
outcomes across the whole range of acute coronary syndromes. Heart 2003;89:512-16.
Infusion Protocol in a Medical Intensive Care Unit. Diabetes Care 2004;27:461-67.
patients with head injury. J Neurosurg 1991;75:545-51.
in mortality, course, and outcome. J Neurotrauma 1998;15:967-72.
27.
Penney DG. Hyperglycemia exacerbates brain damage in acute severe carbon monoxide
poisoning. Med Hypotheses 1988;27:241-44.
28.
Rovlias A, Kotsou S. The influence of hyperglycemia on neurological outcome in patients with
severe head injury. Neurosurgery 2000;46:335-42.
29.
Young LH, Dahl DM, Rauner D, Barrett EJ. Physiological hyperinsulinemia inhibits myocardial
30.
Dorhout Mees SM, Van Dijk GW, Algra A, Kempink DR, Rinkel GJ. Glucose levels and
protein degradation in vivo in the canine heart. Circ Res 1992;71:393-400.
outcome after subarachnoid hemorrhage. Neurology 2003;61:1132-33.
31.
Vincent JL, Moreno R, Takala J et al. The SOFA (Sepsis-related Organ Failure Assessment)
score to describe organ dysfunction/failure. On behalf of the Working Group on SepsisRelated Problems of the European Society of Intensive Care Medicine. Intensive Care Med
1996;22:707-10.
32.
Chee F, Fernando T, van Heerden PV. Closed-loop control of blood glucose levels in critically
33.
Chee F, Fernando T, van Heerden PV. Closed-loop glucose control in critically ill patients
ill patients. Anaesth Intensive Care 2002;30:295-307.
using continuous glucose monitoring system (CGMS) in real time. IEEE Trans Inf Technol
Biomed 2003;7:43-53.
34.
Jungheim K, Wientjes KJ, Heinemann L, Lodwig V, Koschinsky T, Schoonen AJ.
Subcutaneous continuous glucose monitoring: feasibility of a new microdialysis-based
glucose sensor system. Diabetes Care 2001;24:1696-97.
35.
Kapitza C, Lodwig V, Obermaier K et al. Continuous glucose monitoring: reliable
measurements for up to 4 days with the SCGM1 system. Diabetes Technol Ther 2003;5:60914.
178
Hyperglycemia and outcome in critically ill patients
Chapter 5.3
Relationship of baseline glucose and mortality during medical
critical illness?
Jack JM Ligtenberg, Sofie Meijering, Iwan CC van der Horst, Jan G Zijlstra, Mathijs
Vogelzang, Maarten WN Nijsten
Chest in press
179
Metabolic interventions in acute myocardial infarction
To the editor:
We read with interest the study by Cely and colleagues (September 2004) on the
relationship of baseline glucose homeostasis to hyperglycemia in critically ill patients.1
The authors show that hyperglycemia generally is present in critical illness and that
patients with low ‘baseline’ HbA1c levels are less inclined to hyperglycemia than patients
with higher HbA1c. Furthermore, they mention that ‘patients with normal and abnormal
baseline glucose control had similar survival rates’. We don’t think that this statement has
statistical validity, concerning the limited number of evaluated patients (75 patients; ICU
mortality rate 29%).
To determine the relation between blood glucose regulation and mortality, we conducted
an analysis among all patients admitted to our medical ICU over a 2 years period. 1209
consecutive patients were included. ICU mortality was 19.4%. Total number of glucose
measurements was 10954. Mortality was significantly lower in the group with a mean
glucose of 4.0-6.0 mmol/L versus the group with a mean glucose of 6.0-10.0 mmol/L: 12%
versus 17.6%, P=0.03. Patients that died in the ICU (N= 235) also had significantly higher
baseline glucose values than patients that left the ICU alive (N=974): 9.0±5.3 versus
7.6±4.3 mmol/L. We also calculated mortality rate after dividing patients in groups, based
on average glucose level during ICU admission. Mortality rate in the group with an
average glucose of 4.4-6.1 mmol/L (N= 300) was 12%, whilst mortality with an average
glucose of >11.1 mmol/L (N= 96) was 37.5%. Maximal glucose values during admission
appeared to be lower in survivors than in non-survivors: 9.6±5.2 versus 11.7±6.0 mmol/L.
Acute hyperglycemia is frequently found in stress situations.2-8 Since it is so common, it
could be explained as a physiologic adaptation. On the other hand, hyperglycemia is
associated with complications and there is increasing evidence that strict regulation of
glucose could have beneficial effects on morbidity and even mortality.9-11
Based on our own findings and those of others, we assume that hyperglycemia is
correlated with mortality in critically ill patients. The next important step is to design and
start up feasible glucose regulation protocols and study the effect of strict glucose
regulation also in medical intensive care patients.
References
1.
Cely CM, Arora P, Quartin AA, Kett DH, Schein RM. Relationship of baseline glucose
homeostasis to hyperglycemia during medical critical illness. Chest 2004;126:879-87.
2.
Bonnier M, Lonnroth P, Gudbjornsdottir S, Attvall S, Jansson PA. Validation of a glucoseinsulin-potassium infusion algorithm in hospitalized diabetic patients. J Intern Med
2003;253:189-93.
3.
Goldberg PA, Siegel MD, Russell RR et al. Experience with the continuous glucose monitoring
system in a medical intensive care unit. Diabetes Technol Ther 2004;6:339-47.
180
Hyperglycemia and outcome in critically ill patients
4.
McCowen KC, Malhotra A, Bistrian BR. Stress-induced hyperglycemia. Crit Care Clin
2001;17:107-24.
5.
Mizock BA. Alterations in fuel metabolism in critical illness: hyperglycaemia. Best Pract Res
Clin Endocrinol Metab 2001;15:533-51.
6.
Umpierrez GE, Isaacs SD, Bazargan N, You X, Thaler LM, Kitabchi AE. Hyperglycemia: an
independent marker of in-hospital mortality in patients with undiagnosed diabetes. J Clin
Endocrinol Metab 2002;87:978-82.
7.
van der Horst IC, Gans RO, Nijsten MW, Ligtenberg JJ. Beneficial effect of glucose-insulinpotassium infusion in noncritically ill patients has to be proven. J Intern Med 2003;254:513.
8.
van der Horst IC, Ligtenberg JJ, Bilo HJ, Zijlstra F, Gans RO. Glucose-insulin-potassium
infusion in sepsis and septic shock: no hard evidence yet. Crit Care 2003;7:13-15.
9.
Krinsley JS. Association between hyperglycemia and increased hospital mortality in a
heterogeneous population of critically ill patients. Mayo Clin Proc 2003;78:1471-78.
10.
van den Berghe G, Wouters P, Weekers F et al. Intensive insulin therapy in the critically ill
11.
van den Berghe G, Wouters PJ, Bouillon R et al. Outcome benefit of intensive insulin therapy
patients. N Engl J Med 2001;345:1359-67.
in the critically ill: Insulin dose versus glycemic control. Crit Care Med 2003;31:359-66.
181
Metabolic interventions in acute myocardial infarction
182
Rationale and design of GIPS-2
Chapter 6
Glucose-insulin-potassium and reperfusion in acute myocardial
infarction: rationale and design of the Glucose-InsulinPotassium Study-2 (GIPS-2)
Iwan CC van der Horst, Jorik R Timmer, Jan Paul Ottervanger, Henk JG Bilo, Rijk OB Gans,
Menko-Jan de Boer, Felix Zijlstra, on behalf of the GIPS investigators
American Heart Journal in press
183
Metabolic interventions in acute myocardial infarction
Abstract
Background
The combination of reperfusion therapy and high-dose glucose-insulin-potassium infusion
(GIK) seems beneficial in acute myocardial infarction (MI). Current evidence however is
not considered conclusive.
Study design
The Glucose-Insulin-Potassium Study-2 (GIPS-2) will investigate whether GIK, in
adjunction to reperfusion therapy, is beneficial in MI patients without signs of heart failure
at admission. A total of at least 1044 patients with an acute MI treated with either
thrombolysis or primary percutaneous coronary intervention (PCI) will be randomized to
an infusion of high-dose GIK or no infusion. The primary endpoint of the study is 30-day
mortality. Secondary endpoints are mortality at 1-year, recurrence of MI, repeat
intervention and infarct size.
Implications
If high-dose GIK significantly reduces mortality at 30-days in all patients, the adjunction of
this treatment to reperfusion therapy may become part of standard regimen for patients
with acute MI without heart failure.
184
Rationale and design of GIPS-2
Introduction
Mortality, morbidity and infarct size following acute myocardial infarction (MI) are
considerably reduced by reperfusion therapy, i.e. thrombolysis or primary percutaneous
coronary intervention (PCI), and adequate medical treatment.1-3 Further reduction may be
obtained by therapies influencing cardiac metabolism, such as the infusion of glucoseinsulin-potassium (GIK).
Background
GIK therapy may improve outcome after MI through several potential mechanisms.4 GIK
reduces the amount of circulating free fatty acids (FFA) and promotes the use of glucose
as primary myocardial energy substrate.5-8 Glucose is less oxygen consuming in
comparison to free fatty acids (FFA) and has a beneficial effect on myocardial function and
membrane stability.9;10 Furthermore, insulin improves myocardial perfusion and may
reduce reperfusion injury.11 Insulin activates intracellular signaling pathways that promote
cell survival and inhibit apoptosis related events. As apoptosis may play a significant role
in reperfusion injury, GIK infusion may limit additional myocardial damage after adequate
perfusion has been restored 11-13
Previous studies
Treatment with GIK was first mentioned in the early sixties by Sodi-Pollares.14 Thereafter,
GIK infusion has been advocated in small trials as the therapy of choice in the early hours
following acute myocardial infarction.15-21 An systematic overview involving nine of these
early studies with 1932 patients concluded that GIK might play an important role in
reducing in-hospital mortality after acute myocardial infarction.22 This benefit was
particularly present in studies that used a high dose GIK scheme, with a reduction of inhospital mortality from 11.8% to 7.0% (p=0.05). This is in accordance with experimental
studies showing that high-dose GIK (≥5 units insulin/hour) maximally suppresses
circulating FFA levels and induces maximum myocardial glucose uptake.5 The first clinical
study of GIK patients with MI in the era of reperfusion was the Diabetes Mellitus, Insulin
Glucose Infusion in Acute Myocardial Infarction (DIGAMI) study.23 This study included 620
patients with diabetes mellitus or hyperglycemia (glucose > 11.0 mmol/l) on admission
and randomly assigned them to either conventional therapy or adjunctive therapy through
a glucose-insulin infusion followed by a multidose insulin regimen for at least 3 months.
Patients treated with GIK had a significantly lower 1-year mortality (18.6% versus 26.1%,
p=0.03). A few years later the Estudios Cardiologicos Latinoamerica (ECLA) pilot trial,
which included 407 patients was published.24 It showed that after 30 days mortality in the
group of patients randomized to GIK (both high and low dose) was lower than in the
185
Metabolic interventions in acute myocardial infarction
control group (6.7% versus 11.5%, NS). The effect was most pronounced in patients in
which GIK had been combined with reperfusion treatment, mostly thrombolysis (5.1%
versus 15.1%, P=0.01). Again, high dose GIK proofed to be most beneficial compared to
low dose GIK. In the Polish-Glucose-Insulin-Potassium (Pol-GIK) trial with 954 patients, the
mortality in patients treated with low dose GIK was higher than in the control group (8.9%
versus 4.8%, P=0.01).25 Mortality due to a cardiovascular incident was not significantly
different, and as such the cause of this difference remained unclear. In the GlucoseInsulin-Potassium Study 1 (GIPS-1) 940 patients were randomized to high-dose GIK or no
infusion in combination with primary PCI.26
Figure 1. The odds ratio and confidence interval of all randomized trials with GIK stratified
according to the use of high versus low dose GIK
Stanley
Rogers
GIPS
ECLA high-dose
DIGAMI
Total high-dose
Mittra
Pentecost
MRC
Pilcher
Hjermann
ECLA low-dose
Pol-GIK
Total low-dose
Total
0.0
0.5
GIK better
1.0
1.5
2.0
2.5
Control better
Overall there is a significant benefit of GIK in comparison to the control group. High dose GIK seems
more beneficial in comparison to low dose. HD denotes high-dose GIK. LD denotes low-dose GIK.
GIPS = Glucose-Insulin-Potassium Study. ECLA = Estudios Cardiologicos Latinoamerica. DIGAMI =
Diabetes Insulin-Glucose in Acute Myocardial Infarction study MRC = Medical Research Council.
Pol-GIK = Polish Glucose Insulin Potassium. In both comparisons of the ECLA the same control
group is used; in the total of all patients this groups is accounted for once.
186
Rationale and design of GIPS-2
In the overall population the beneficial effect of GIK did not reach significance (4.8%
versus 5.8%, P=0.50). In 856 of the 940 patients (91.6%) admitted without signs of heart
failure (Killip class 1) a significant difference was observed (1.2% versus 4.2%, P=0.01).
Results of all randomized trials are summarized in figure 1.
Figure 2. Trial design of GIPS-2
Taken all these results together, a recent editorial concluded that current evidence for the
use of GIK in acute MI is not conclusive and whether it is, is a crucial issue to resolve.27
The multicenter GIPS-2 is designed to address this issue, i.e. whether the infusion of highdose GIK has a beneficial effect on 30-day and 1-year mortality and morbidity in patients
eligible for reperfusion therapy, without signs of heart failure.
187
Metabolic interventions in acute myocardial infarction
Methods
Overview
The planned inclusion calls for a total of at least 1044 patients with an acute MI treated
with thrombolysis or primary PCI which will be randomly assigned to high-dose GIK
infusion or no infusion (ratio 1:1). It is a multicenter trial performed in the Netherlands,
with participation of hospitals of varying sizes, both with and without coronary
intervention facilities. Randomization and GIK administration will take place in the hospital
that will initiate the reperfusion therapy. A flow chart of the study design is shown in
figure 2.
Patients have to meet the stated inclusion criteria. Patients with signs of heart failure and
reduced life expectancy due to other diseases will be excluded. Patients eligible for
randomization will either receive conventional care with reperfusion with GIK or no
infusion. Primary endpoint is 30-day mortality. Patients will at least be followed until 1year after randomization. MI denotes myocardial infarction. GIK denotes glucose-insulinpotassium infusion.
Endpoints
The primary endpoint is 30-day mortality. Secondary endpoints are 1-year mortality,
recurrent myocardial infarction and repeat intervention. Furthermore, left ventricular
function, enzymatic infarct size and ST segment elevation resolution on the
electrocardiogram will be measured. Systemic metabolism will be derived from specific
laboratory measurements. At baseline kreatinin, K, glucose, HbA1c, and free fatty acids
(FFA) will be determined. Glucose and FFA measurements will be repeated at 1-2 hour
intervals in all patients during hours of infusion. In earlier studies administration of GIK
lowered the circulating levels of FFA through the inhibitory effect of insulin on lipolyses.8
This decrease in FFA levels, in combination with an increase in glucose and insulin
availability, promotes the myocardial use of glucose over FFA.28
To obtain insight in the effect of GIK infusion on left ventricular function, we will determine
left ventricular ejection fraction (LVEF). LVEF will be measured before discharge by
radionuclide ventriculography or by echocardiography. Radionuclide ventriculography will
be performed by using the multiple gated equilibrium method following the labeling of
red blood cells of the patient with 99mTc-pertechnate [General Electric 300 gamma
camera with a low-energy all-purpose parallel-hole collimator]. Global ejection fraction will
be calculated by a General Electric Star View computer using the fully automatic PAGE
program.
188
Rationale and design of GIPS-2
Electrocardiography (ECG) will be performed at admission, and at 3 hours after PCI. All
ECGs will be analyzed as pairs by an independent core laboratory (Diagram BV, Zwolle,
The Netherlands) and graded for ST segment elevation resolution by 2 investigators who
are unaware of the data and outcome. ST segment resolution will be defined as complete
(>70% resolution), partial (30-70% resolution) or no resolution (<30% resolution).29
Core laboratory annalists who are unaware of the clinical history and outcome of the
patients will assess TIMI (Thrombolysis In Myocardial Infarction) flow and myocardial
blush grade (MBG). TIMI flow and MBG will be visually assessed on the angiogram. MBG
has been defined previously: 0, no myocardial blush; 1, minimal myocardial blush or
contrast density; 2, moderate myocardial blush or contrast density but less than that
obtained during angiography of a contra or ipsilateral non-infarct related coronary artery;
and 3, normal myocardial blush or contrast density, comparable with that obtained during
angiography of a contralateral or ipsilateral non-infarct-related coronary artery.30;31 When
myocardial blush persists, this phenomenon suggests leakage of contrast medium into
the extravascular space and is graded 0. With these measures more detailed information
about the microcirculation will be available. We can determine whether GIK is beneficial in
patients with disturbed or normal blush grade or in patients with TIMI 3 flow after primary
PCI.
Number of patients
Based on the results of the GIPS-1 it can be anticipated that the primary endpoint in the
two groups of randomization will be the following. A sample size of 1044 patients, with
522 patients in each treatment group is planned. The number of patients included in this
study is based on a power-analysis. A study of this size has a statistical power of 80% at
an alpha level of 0.05 to show a mortality reduction as previously reported in patients
without signs of heart failure on admission. These mortality rates were 1.2% in the GIK
group and 4.2% in the control group.26 We did hypothesize that a risk reduction of 30% in
patients with Killip class 1 would be more realistic.26 We therefore will perform an interim
analysis using Snapinn’s method after approximately 731 patients.32 This method
describes a conditional probability procedure which attempts to maintain the overall
significance level by balancing the probabilities of false early rejection and false early
acceptance. At the time of the interim analysis it will be considered to prolong the
inclusion and thereby including the sufficient number of patients.
Patient selection, patient consent and randomization
All patients admitted to one of the participating hospitals with acute MI are considered for
the study and prospectively evaluated. The purpose and details of the study are explained
189
Metabolic interventions in acute myocardial infarction
to all patients and their consent will be obtained prior to participation in the study. Patients
are enrolled in an emergency situation, and therefore consent is given verbally in the
presence of an independent witness. Written informed consent has to be given by the
patient as soon as the clinical situation of the patient allows it, i.e. after the patient has
recovered from the acute phase of the MI.
After verbal consent is obtained patients will be randomized by telephone using a
computerized voice response system. Randomization will be performed by means of a
computer program in blocks of 4-8 patients (randomly changing block size) to achieve a
balanced allocation. All patients will receive a unique randomization number.
Study population
The diagnosis acute MI will include chest pain suggestive for myocardial ischemia for at
least 30 minutes, the time from onset of this symptoms should be less than 6 hours before
hospital admission, and an ECG recording with ST T segment elevation of more than 1
mV in 2 or more leads. Patients have to be eligible for either thrombolysis or primary PCI.
Patients with signs of heart failure at admission (Killip class ≥2) will be excluded. Heart
failure at admission is defined as the existence of either one of the following symptoms,
i.e. a heart rate over 90 beats/min, a systolic blood pressure beneath 100 mmHg with
anterior myocardial infarction, a third heart sound, or rales more than one hand-wide.
Further exclusion criteria are the unwillingness to participate and the existence of a lifethreatening disease with a life-expectancy of less than 6 months. There will be no upper
age limit.
Medication
The mode of reperfusion therapy given is chosen by the individual cardiologist. If the
cardiologist decides for thrombolysis, this may consist of any of the appropriate
regimens, as stated in the recently updated guidelines of the European Society of
Cardiology/American College of Cardiology.33;34 If the cardiologist decides for primary PCI
the patient is transported to the catheterisation laboratory for emergency coronary
angiography and if indicated primary angioplasty is performed. Additional standard
treatment consist of 5000 IU heparin, nitroglycerin intravenously, and aspirin 500 mg
intravenously or 300 mg orally. During angioplasty 10,000 IU heparin is administered as
bolus. Low molecular weight heparin is given for 24 hours thereafter.
Standard therapies including aspirin 80 mg, clopidogrel 75 mg, beta-blockers, lipid
lowering agents and angiotensin converting enzyme inhibitors or angiotensin II receptor
blockers are likely to be protective in most patients, and should be used according to the
guidelines of the European Society of Cardiology/American College of Cardiology.33;34
190
Rationale and design of GIPS-2
Figure 3. Algorithm of GIK administration
Scheme 1: Starting dose of insulin infusion rate
Glucose (mmol/L)
Additional action
> 15
6 IU short-acting insulin i.v.
Infusion rate (IU/hour)
23
11.0-15.0
23
10.0-10.9
18
9.0-9.9
14
8.0-8.9
9
7.0-7.9
5
< 7
3
Scheme 2: Hourly adjustment of insulin infusion rate based on hourly measured glucose
Glucose level (mmol/L) Additional action
> 15
6 IU short-acting insulin i.v.
Infusion rate (IU/hour)
Increase with 5
11.0-15.0
Increase with 3
10.0-10.9
Increase with 2
6.0-9.9
Leave unchanged
4.0-5.9
Decrease with 8
<4
1. Stop infusion for 15 minutes
2. Measure blood glucose level every 15 minutes until
>6,9 mmol/L
3. Start infusion with infusion rate 7 ml/hour beneath rate
before stop
4. If symptoms of hypoglycemia are present than give 20
ml of glucose 30% i.v.
An infusion of 500 ml 20% glucose with 40 mmol potassium chloride is started with a fixed infusion
rate of 2 ml/kg body weight/hour. In addition, a variable infusion of short acting insulin is started.
Initial insulin dose is according to the admission glucose (Scheme 1). For example: a glucose level
at admission of 9.5 mmol/L dictates an initial insulin dose of 14 units/hour. Based on subsequent
hourly measured glucose levels the insulin infusion rate will be changed according to Scheme 2. As
can be seen in scheme 2; insulin infusion is increased when glucose levels are ≥ 10 mmol/L and
decreased when glucose levels are < 6.0 mmol/L. If blood glucose levels drop more than 30% and
glucose is still > 11.0 mmol/L the rate should not be increased. If blood glucose levels drops more
than 30% and the eventual glucose is between 6.0 and 11.0 mmol/L, than the rate should be
decreased with 5 units. Infusion rate must not exceed 38 ml/hour. If the infusion rate should be
increased above 38 ml/hour than add 6 IU of short-acting insulin. ∗ GIK denotes glucose-insulinpotassium infusion.
191
Metabolic interventions in acute myocardial infarction
Patients randomly assigned to the GIK group receive an infusion of 40 mmol potassium
chloride in 500 ml 20% glucose, which is administered with a fixed rate of 2.0 ml/kilogram
body weight/hour over a 12 hour period in a peripheral venous line. The infusion is started
as soon as possible after admission at the hospital (i.e. after blood-glucose level is
determined), preferably before reperfusion therapy is started. Thereby a continuous
infusion of short-acting insulin (50 units Actrapid HM [Novo Nordisk, Copenhagen,
Denmark]) in 49.5 ml of 0.9 percent sodium chloride with the use of a perfusor-pump will
also be started. This insulin infusion rate is variable. The first measured glucose level,
prior to the glucose-potassium infusion, determines the dosage at which the insulin pump
is started (figure 3 – scheme 1). Hereafter, glucose levels will be measured every hour.
Subsequent adjustment of the insulin dose will be based on these hourly measurements
of glucose, aiming at blood-glucose levels of 6.0 to 10.0 mmol/l (figure 3 – scheme 2).
After the glucose-potassium infusion is stopped insulin may be continued based on
glucose measurements according to the conventional care. In the control group blood
glucose regulation will be according to the conventional care. In general, the initial insulin
infusion rate will be at least 5 units/hour, as most patients will have an admission glucose
≥7.0 mmol/L. The insulin dose is comparable with the DIGAMI study (4.8 units/hour) and
the high dose regimen of the ECLA trial (5.6 units/hour).35;36 Whereas it is higher than the
low dose regimen of the ECLA (1.5 units/hour) or the Pol-GIK trial (1.3 units/hour).24;25 Most
studies employed insulin as a fixed dose with glucose delivery titrated to maintain target
glucose concentrations. In our studies we use a fixed glucose infusion combined with
potassium chloride and titrate insulin to obtain strict control of blood-glucose levels. It is
known that hyperglycemia is a prognostic factor in acute MI in both patients with and
without diabetes mellitus.37;38 Treating MI patients with insulin in the DIGAMI has shown to
have beneficial effects on both glucose control and survival.23 The reduction of mortality
by intensive insulin therapy in critically ill patients to maintain blood glucose below 6.1
mmol/l gives also direction for metabolic interventions, such as GIK infusion, that are
directed towards prevention of hyperglycemia.39 In two studies including diabetic patients
treated with coronary artery bypass grafting strict glucose control with either insulin alone
or GIK was related to a significant mortality reduction.40;41
Baseline demographic and clinical characteristics
Baseline characteristics that will be collected include age, gender, time of symptom onset,
time of admission, history of previous myocardial infarction, hypertension, diabetes
mellitus (and type and medication, i.e. diet, tablets, insulin), positive family history for
cardiovascular diseases. Physical diagnosis at admission includes heart rate, systolic and
diastolic blood pressure, weight, length, heart sounds and lung sounds.
192
Rationale and design of GIPS-2
Inhospital and follow-up assessment
All randomized patients should undergo trial follow-up at discharge, 30 days, and 1 year.
Follow-up information will be obtained by a telephone interview with the patient.
Additional information on events that occurred will be retrieved from hospital records.
The hospital in which the patient initially was included and randomized is responsible for
complementation of follow-up.
Statistical analysis
The primary endpoint of the study is 30-day mortality. Secondary endpoints are one-year
mortality, recurrent infarction, repeat intervention and infarct size. The incidence of death
will be evaluated in predefined subgroups with regard to age, gender, time to treatment,
location of the infarction, diabetic status and glycometabolic state on admission. Analyses
will be performed according to the intention-to-treat principle. Differences between group
means will be assessed with the two-tailed Student’s t-test. Chi-square analysis or Fisher’s
exact test was used to test differences between proportions. Survival will be calculated by
the Kaplan-Meier product-limit method. The Mantel-Cox (or Log-rank) test will be used to
evaluate differences in survival between the two treatment groups. The Cox proportionalhazards regression model will be used to calculate relative risks and to adjust for
differences in baseline characteristics. Statistical significance is considered as a two-tailed
P-value <0.05. The Statistical Package for the Social Sciences (SPSS Inc., Chicago, IL,
USA) version 11.0.1 will be used for all statistical analysis.
Study organisation
Data on mortality will be collected according to the Case Record Form (CRF). The
participating centers will send copies of the CRF. All data will be recorded in a dedicated
database at the data-management centre. For entering the data into the database a double
data entry system will be used. Data will be checked through an error check program,
queries generated out of that program will be solved by the research personnel of the
participating centers. Monitors will periodically check a sample of the recorded patient
data against source documents at the study site.
Future GIK trials
Other promising trials investigating GIK in acute MI are underway. The multinational
DIGAMI–2 will provide data on potential effects of glycometabolic control in
hyperglycemic patients. It was planned to randomize approximately 3000 patients with
suspected MI to acute insulin-glucose infusion followed by multidose subcutaneous
insulin, acute insulin-glucose infusion followed by conventional hypoglycemic therapy, or
conventional hypoglycemic treatment. The main comparison will be the mortality rate
193
Metabolic interventions in acute myocardial infarction
between patients randomized to insulin-glucose infuson. The combined multinational
ECLA GIK 2/CREATE trial, investigating high-dose GIK, has already included a tremendous
number of patients (>27000) and therefore has optimal statistical power. Primary
endpoint will be to compare 30-day mortality rate in acute MI patients presenting within
12 hours from symptoms onset and eligible for any reperfusion strategy (thrombolysis or
primary PCI) allocated to a high dose GIK infusion versus the control treatment. The
multinational Organization to Assess Strategies for Ischemic Syndromes (OASIS)-6 trial
will include at least 10000 patients presenting with an acute MI treated with a range of
thrombolytic agents, primary PCI and those not eligible for reperfusion therapy (e.g. late
presentation or contra-indication to reperfusion therapy). The study has a side arm to
investigate the effect of GIK infusion on death and nonfatal cardiac arrest up to day 30.
Although the GIPS-2 trial will only include a fraction of patients compared to the latter
trials, extensive laboratory and angiographic data and measurement of residual left
ventricular function will provide additional insight into the therapeutic mechanisms of GIK.
The time between onset of symptoms and initiation of the GIK infusion will be short and
secondary prevention will be initiated in all patients. Finally, the GIPS-2 will give the
opportunity to complete follow-up over a long time.
Current status
Enrolment started at August 4th 2003.
Conclusion
GIPS-2 is a multicenter, randomized, controlled trial that will enroll patients during the
acute phase of MI to determine whether GIK should be added to the standard regimen in
patients without heart failure at admission.
Acknowledgements
Research of ICC van der Horst was made possible by a generous grant of the Netherlands
Heart Foundation (99.028). The GIPS-2 is recently accepted for an independent grant of
the Netherlands Heart Foundation.
194
Rationale and design of GIPS-2
Appendix
Steering committee: Dr F. Zijlstra (chair), I.C.C. van der Horst, Dr R.O.B. Gans, Groningen
University Hospital, Dr J.P. Ottervanger, J.R. Timmer, Dr M.J. de Boer, Dr H.J.G. Bilo, Isala
Klinieken, Locatie Weezenlanden
Participating centers: Amsterdam, Academic Medical Center, Dr J.P.S. Henriques Apeldoorn, Ziekhuiscentrum Apeldoorn, Dr E.G. Faber - Assen, Wilhelmina Ziekenhuis, Dr
M. Pentinga - Delfzijl, Delftzicht Ziekenhuis, Dr J.N. Spanjaard - Deventer, Deventer
Ziekenhuis, Dr H. Groeneveld - Dokkum, Talma Sionsberg, Dr H. The - Emmen,
Scheperziekenhuis, Dr T. Vet – Groningen, University Hospital, Dr S.A.J. van den Broek Hardenberg, Streekziekenhuis, locatie Röpcke-Zweers, Dr A.J. Schaap - Harderwijk,
Ziekenhuis St. Jansdal, Dr R. Dijkgraaf - Meppel, Diaconessenhuis, Dr R. Brons - Sneek,
Antonius Ziekenhuis, Dr. A. Oomen - Winschoten, Sint Lucas Ziekenhuis, Dr T.R.
Bouwmeester - Zwolle, Isala Klinieken, locatie Sophia, Dr A.H.E.M. Maas - Zwolle, Isala
Klinieken, locatie Weezenlanden, Dr J.H.E. Dambrink -.
References
1.
Grines CL, Browne KF, Marco J et al. A comparison of immediate angioplasty with
thrombolytic therapy for acute myocardial infarction. The Primary Angioplasty in Myocardial
Infarction Study Group. N Engl J Med 1993;328:673-79.
2.
Zijlstra F, de Boer MJ, Hoorntje JC, Reiffers S, Reiber JH, Suryapranata H. A comparison of
immediate coronary angioplasty with intravenous streptokinase in acute myocardial
infarction. N Engl J Med 1993;328:680-684.
3.
Zijlstra F, Hoorntje JC, de Boer MJ et al. Long-term benefit of primary angioplasty as
compared with thrombolytic therapy for acute myocardial infarction. N Engl J Med
1999;341:1413-19.
4.
Apstein CS. Increased glycolytic substrate protection improves ischemic cardiac dysfunction
and reduces injury. Am Heart J 2000;139:S107-S114.
5.
McDaniel HG, Papapietro SE, Rogers WJ et al. Glucose-insulin-potassium induced alterations
in individual plasma free fatty acids in patients with acute myocardial infarction. Am Heart J
1981;102:10-15.
6.
Opie LH. The glucose hypothesis: its relation to acute myocardial ischemia. J Mol Cell Cardiol
1970;1:107-15.
7.
Prather JW, Russell RO, Jr., Mantle JA, McDaniel HG, Rackley CE. Metabolic consequences of
glucose-insulin-potassium infusion in treatment of acute myocardial infarction. Am J Cardiol
1976;38:95-99.
8. Rackley CE, Russell RO, Jr., Rogers WJ, Mantle JA, McDaniel HG, Papapietro SE. Clinical
experience with glucose-insulin-potassium therapy in acute myocardial infarction. Am Heart J
1981;102:1038-49.
9.
Cave AC, Ingwall JS, Friedrich J et al. ATP synthesis during low-flow ischemia: influence of
increased glycolytic substrate. Circulation 2000;101:2090-2096.
195
Metabolic interventions in acute myocardial infarction
10.
Korvald C, Elvenes OP, Myrmel T. Myocardial substrate metabolism influences left ventricular
11.
Legtenberg RJ, Houston RJ, Oeseburg B, Smits P. Physiological insulin concentrations
energetics in vivo. Am J Physiol Heart Circ Physiol 2000;278:H1345-H1351.
protect against ischemia-induced loss of cardiac function in rats. Comp Biochem Physiol A
Mol Integr Physiol 2002;132:161-67.
12.
McNulty PH, Pfau S, Deckelbaum LI. Effect of plasma insulin level on myocardial blood flow
and its mechanism of action. Am J Cardiol 2000;85:161-65.
13.
Marano L, Bestetti A, Lomuscio A et al. Effects of infusion of glucose-insulin-potassium on
14.
Sodi-Pallares D, Testelli MR, Fishleder BL et al. Effects of an intravenous infusion of a
myocardial function after a recent myocardial infarction. Acta Cardiol 2000;55:9-15.
potassium-glucose-insulin solution on the electrocardiographic signs of myocardial infarction.
A preliminary clinical report. Am J Cardiol 1962;9:166-81.
15.
Hjermann I. A controlled study of peroral glucose, insulin and potassium treatment in
myocardial infarction. Acta Med Scand 1971;190:213-18.
16.
Medical Research Council Working Party on the Treatment of Myocardial Infarction.
Potassium, glucose, and insulin treatment for acute myocardial infarction. Lancet
1968;2:1355-60.
17.
Mittra B. Potassium, glucose, and insulin in treatment of myocardial infarction. Lancet
18.
Pentecost BL, Mayne NM, Lamb P. Controlled trial of intravenous glucose, potassium, and
1965;2:607-9.
insulin in acute myocardial infarction. Lancet 1968;1:946-48.
19.
Rogers WJ, Stanley AW, Jr., Breinig JB et al. Reduction of hospital mortality rate of acute
20.
Stanley AW, Jr., Prather JW. Glucose-insulin-potassium, patient mortality and the acute
myocardial infarction with glucose-insulin-potassium infusion. Am Heart J 1976;92:441-54.
myocardial infarction: results from a prospective randomised study. Circulation 1978;57:61.
21.
Pilcher J, Etishamudin M, Exon P, Moore J. Potassium, glucose and insulin in myocardial
infarction. Lancet 1967;1:1109.
22. Fath-Ordoubadi F, Beatt KJ. Glucose-insulin-potassium therapy for treatment of acute
myocardial infarction: an overview of randomized placebo-controlled trials. Circulation
1997;96:1152-56.
23.
Malmberg K, Ryden L, Efendic S et al. 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 1995;26:57-65.
24.
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.
25.
Ceremuzynski L, Budaj A, Czepiel A et al. Low-dose glucose-insulin-potassium is ineffective in
acute myocardial infarction: results of a randomized multicenter Pol-GIK trial. Cardiovasc
Drugs Ther 1999;13:191-200.
26.
van der Horst IC, Zijlstra F, van't Hof AW et al. Glucose-insulin-potassium infusion in patients
treated with primary angioplasty for acute myocardial infarction: the glucose-insulinpotassium study: a randomized trial. J Am Coll Cardiol 2003;42:784-91.
196
Rationale and design of GIPS-2
27.
Apstein CS. The benefits of glucose-insulin-potassium for acute myocardial infarction (and
28.
Eberli FR, Weinberg EO, Grice WN, Horowitz GL, Apstein CS. Protective effect of increased
some concerns). J Am Coll Cardiol 2003;42:792-95.
glycolytic substrate against systolic and diastolic dysfunction and increased coronary
resistance from prolonged global underperfusion and reperfusion in isolated rabbit hearts
perfused with erythrocyte suspensions. Circ Res 1991;68:466-81.
29.
't Hof AW, Liem A, de Boer MJ, Zijlstra F. Clinical value of 12-lead electrocardiogram after
successful reperfusion therapy for acute myocardial infarction. Zwolle Myocardial infarction
Study Group. Lancet 1997;350:615-19.
30.
Henriques JP, Zijlstra F, van 't Hof AW et al. Angiographic assessment of reperfusion in acute
31.
van 't Hof AW, Liem A, Suryapranata H, Hoorntje JC, de Boer MJ, Zijlstra F. Angiographic
myocardial infarction by myocardial blush grade. Circulation 2003;107:2115-19.
assessment of myocardial reperfusion in patients treated with primary angioplasty for acute
myocardial infarction: myocardial blush grade. Zwolle Myocardial Infarction Study Group.
Circulation 1998;97:2302-6.
32.
Snapinn SM. Monitoring clinical trials with a conditional probability stopping rule. Stat Med
1992;11:659-72.
33.
Ryan TJ, Antman EM, Brooks NH et al. 1999 update: ACC/AHA Guidelines for the
Management of Patients With Acute Myocardial Infarction: Executive Summary and
Recommendations: A report of the American College of Cardiology/American Heart
Association Task Force on Practice Guidelines (Committee on Management of Acute
Myocardial Infarction). Circulation 1999;100:1016-30.
34.
Van de Werf F, Ardissino D, Betriu A et al. Management of acute myocardial infarction in
patients presenting with ST-segment elevation. The Task Force on the Management of Acute
Myocardial Infarction of the European Society of Cardiology. Eur Heart J 2003;24:28-66.
35.
Sack MN, Yellon DM. Insulin therapy as an adjunct to reperfusion after acute coronary
ischemia: a proposed direct myocardial cell survival effect independent of metabolic
modulation. J Am Coll Cardiol 2003;41:1404-7.
36.
Yellon DM, Baxter GF. Protecting the ischaemic and reperfused myocardium in acute
37.
Bolk J, van der Ploeg T, Cornel JH, Arnold AE, Sepers J, Umans VA. Impaired glucose
myocardial infarction: distant dream or near reality? Heart 2000;83:381-87.
metabolism predicts mortality after a myocardial infarction. Int J Cardiol 2001;79:207-14.
38.
Capes SE, Hunt D, Malmberg K, Gerstein HC. Stress hyperglycaemia and increased risk of
death after myocardial infarction in patients with and without diabetes: a systematic
overview. Lancet 2000;355:773-78.
39.
van den Berghe G, Wouters P, Weekers F et al. Intensive insulin therapy in the critically ill
40.
Furnary AP, Gao G, Grunkemeier GL et al. Continuous insulin infusion reduces mortality in
patients. N Engl J Med 2001;345:1359-67.
patients with diabetes undergoing coronary artery bypass grafting. J Thorac Cardiovasc Surg
2003;125:1007-21.
41.
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; 109:1497-502.
197
Metabolic interventions in acute myocardial infarction
198
Summary and future perspectives
Chapter 7
Summary and future perspectives
199
Metabolic interventions in acute myocardial infarction
This thesis primarily describes the results of high-dose glucose-insulin-potassium (GIK)
infusion in ST segment elevation MI patients treated with primary percutaneous coronary
intervention (PCI).1 In the introduction (Chapter 1) it has already been stated that the
experimental and clinical evidence accumulated over the better part of a century
underscores the importance of glucose metabolism during ischemia/reperfusion of the
heart. In 1912, Goulston suggested that treatment with glucose could be beneficial in
several categories of heart disease.2 Sodi-Pallares and colleagues were the first to report
the use of GIK in patients with acute myocardial infarction and found that it limited
electrocardiographic changes.3 After the promising results of nine early randomized trials
with 1932 patients and publication of the first results in patients treated with both
reperfusion therapy and GIK the Glucose-Insulin-Potassium Study (GIPS-1) was set up.4;5
GIPS-1 included 940 patients randomized to primary PCI with or without GIK (glucose 20%
with 80 mmol potassium in 500 ml sodiumchloride 0.9% at a rate of 3 ml/kg/ per hour and
50 IU insulin in 50 ml water administered according to serum glucose concentrations).
Baseline insulin-infusion dose and adjustments of the insulin dose, to obtain bloodglucose levels between 7.0 and 11.0 mmol/L, were based on a modified algorithm. In
Chapter 2.1 we describe the results of GIPS-1 on 30-day mortality and major adverse
cardiac events (MACE).6 We observed that the mortality rate was lower in the GIK group
(4.8%) compared to the control group (5.8%), albeit not statistically significant. After
adjustment for differences at baseline, the relative risk of mortality with GIK became 0.61
(0.34–1.10, P=0.09). In the large predefined subgroup of 856 patients without signs of
heart failure (Killip class 1), a significant reduction of mortality was seen (1.2% in the GIK
compared to 4.2% in the control group, P=0.01). In 84 patients (8.9%) with signs of heart
failure (Killip class ≥2), the effect of GIK is uncertain since a (non-significant) higher
mortality rate was observed.
In Chapters 2.2 and 2.3, we analyzed the effects of GIK on myocardial function and
determined enzymatic infarct size and left ventricular function. There was no difference in
the time course or magnitude of creatine kinase MB (CK-MB) release between the two
groups: peak CK-MB level was 249±228 IU/L in the GIK group versus 240±200 IU/L in the
control group (NS). The mean left ventricular ejection fraction (LVEF) was higher in the
GIK group (43.7±1.0% versus 42.4±11.7%, P=0.12). Poor left ventricular function (LVEF
≤30%) was observed in 18% in the controls and in 12% in the GIK group (P=0.01). We
observed a relation between enzyme release and left ventricular function; the mean peak
CK-MB was 463±318 IU/L in patients with LVEF ≤30% versus 215±166 IU/L in patients with
LVEF >30% (P<0.001). We therefore hypothesized that the effect of GIK on left ventricular
function is mediated by factors unrelated to enzymatic infarct size.
Chapter 2.4 describes the effect of GIK on the resolution of ST segment elevation. The
electrocardiograms at admission and after 3 hours were analyzed in 612 patients.
200
Summary and future perspectives
Combined complete (>70%) and partial (30-70%) resolution was more commonly
observed in the GIK group (87%) when compared to the control group (78%), relative risk
1.72 (95% confidence interval 1.2-2.46, P<0.001). One-year mortality was lower in patients
with combined complete and partial resolution compared to patients without resolution
(3.8% versus 10.3%, P=0.011). ST segment elevation resolution after reperfusion therapy
reflects myocardial flow rather than epicardial flow and predicts clinical outcome better
than epicardial vessel patency alone in patients treated with reperfusion therapy7-9 One of
the effects of GIK in reperfused MI patients may be the restoration/preservation of
myocardial perfusion.
The goal of the infusion of GIK during acute MI is to modulate the (myocardial)
metabolism in such a manner that it is related to the preservation of myocardial function
and a favorable clinical outcome. In Chapter 2.5, we describe the effect of GIK infusion on
glucose and potassium levels. Mean serum glucose was 9.3±4.5 mmol/L in the GIK group
compared to 8.4±2.9 mmol/L in the control group (P<0.001). Hyperglycemia (glucose
>11.0 mmol/L) occurred in 337 patients (70.8%) treated with GIK and in 157 controls
(33.8%) (P<0.001). A total of 48 hyperglycemic patients (9.4%) died versus 21 patients
(4.7%) without hyperglycemia (P=0.004). In patients with hyperglycemia, 1-year mortality
was 8.3% in the GIK group versus 12.7% in controls (P=0.14). Hypoglycemia (glucose
≤3.0 mmol/L) occurred in 10 patients of whom 2 patients (20%) died, both treated with
GIK. Hyperkalemia (potassium >5.5 mmol/L) was present in 26 GIK patients (5.5%) and in
15 controls (3.2%) (P=0.11); of these patients 9 (34.6%) died in the GIK group and 9 (60%)
in the control group (P=0.19). Hypokalemia (potassium ≤3.5 mmol/L) occurred in 112
patients (23.5%) in the GIK group and in 191 patients in the control group (41.2%)
(P<0.001). A total of 30 patients (9.9%) with hypokalemia died, 10 GIK patients (8.9%) and
20 controls (10.5%) (P=0.84). The beneficial effect of GIK on mortality may be diminished
by these derangements.
The 3-year results are represented in Chapter 2.6. All-cause mortality occurred in 98
(10.4%) out of 940 patients after 3 years. Factors related with long-term mortality were
age, previous cardiovascular disease, anterior MI, Killip class ≥2, and multi-vessel disease
(all P<0.001). In the GIK group, 46 (9.7%) out of 476 patients died compared to 52 (11.2%)
out of 464 patients (P=0.44, using the Log-rank test). In patients with Killip class 1 (N=856)
randomization to GIK had an independent relation with 3-year mortality, with an adjusted
relative risk of 0.56 (0.34-0.94, P=0.028). Other factors related with long-term mortality in
Killip class 1 patients were age (P<0.001), previous cardiovascular disease (P<0.001) and
multi-vessel disease (P=0.046). Cardiac protection with GIK infusion and primary PCI in
acute MI did not result in a significant reduction in mortality in all patients. During the 3year follow-up, the beneficial effect observed after 30-days in the predefined subgroup of
Killip class 1 patients was sustained.
201
Metabolic interventions in acute myocardial infarction
In Chapter 3 we describe the results of an overview off all studies on glucose-insulinpotassium solution. Of 13 published randomized trials, 12 studies reported mortality
reduction after GIK. Most promising effects were observed when GIK was given in a high
dose and when it was given as an adjunctive to reperfusion therapy.
In the second part of this thesis we describe our observational studies regarding the
relation between hyperglycemia and outcome without controlled interventions on glucose
metabolism. In Chapter 4.1, we determined whether admission glucose predicted longterm outcome in MI patients without diabetes mellitus that were treated with
streptokinase or primary PCI. We observed that elevated admission glucose levels in nondiabetic patients treated with reperfusion therapy for ST segment elevation MI are
independently associated with larger infarct size and higher long-term mortality.
In Chapter 4.2, we describe the results of an analysis of the effect of admission
hyperglycemia on myocardial perfusion in MI patients. A total of 93 patients (20%) out of
464 patients had hyperglycemia (glucose ≥11.0 mmol/L). They had more often a reduced
myocardial blush grade (26% versus 15%, P=0.02) and incomplete ST segment elevation
resolution (59% versus 39%, P=0.01) compared to patients without hyperglycemia on
admission. Also, in patients with hyperglycemia there was a greater reduction in left
ventricular function and a higher mortality. This was observed in patients with and without
diabetes mellitus. Hyperglycemia on admission is associated with reduced myocardial
reperfusion after primary PCI, which was associated with adverse clinical outcome.
In Chapter 4.3, we respond to a study that stated that hyperglycemia might be associated
with impaired microvascular function after acute myocardial infarction (MI). The authors
reached this conclusion by a retrospective analysis of 146 patients.10 The hypothesis that
acute hyperglycemia (i.e., hyperglycemia at admission) is associated with the ‘no reflow’
phenomenon is intriguing. However, we hypothesize that ‘no reflow’ is related to acute
hyperglycemia and chronic hyperglycemia (i.e., elevated HbA1c and/or DM). Therefore,
new studies will have to determine the effect of hyperglycemia on ‘no reflow’ and
preferentially on the effect of metabolic regulation.
Although admission glucose is a predictor of unfavorable outcome in MI patients, the
predictive value is not strong. We hypothesized that persistent hyperglycemia (an
elevated time-averaged glucose during admission) in critically ill patients could be a better
predictor. In Chapter 5.1, we describe the relation between an elevated glucose at
admission and an elevated time-averaged glucose with short-term MACE in MI patients
treated with primary PCI. We included 417 patients of which 89 patients (21.3%) had had
at least one MACE. In 17 patients (4.1%) it had been a fatal event and in 72 patients
(17.3%) a non-fatal event. MACE occurred more often in patients who were older, had
previous cardiovascular disease, anterior infarction location, and multi-vessel disease. The
area under the receiver operator characteristic curve was 0.64 for time-averaged glucose
202
Summary and future perspectives
and 0.59 for admission glucose. Time-averaged glucose emerged as a significant
independent predictor after multivariate analysis (P<0.01). We concluded that elevated
time-averaged glucose in MI has a stronger relation with short-term MACE than elevated
admission glucose.
In Chapter 5.2, the relation between hyperglycemia during admission and outcome was
analyzed in a retrospective study of critically ill patients admitted to a surgical Intensive
Care Unit. An objective measure of hyperglycemia to assess glucose regulation was
defined. The so-called, hyperglycemic index (HGI) denotes the area under the curve as
‘above the upper limit of normal’ (glucose level 6.0 mmol/L) divided by ‘the total length of
stay’. In 1779 patients with a median ICU-stay of 10 days, 30-day mortality was 17%.
Median HGI was 0.9 (0.3-2.1) in survivors compared to 1.7 (0.7-3.3) mmol/L in nonsurvivors (p<0.001). Area under the receiver operator characteristic curve was 0.64 for
HGI, compared with 0.61 and 0.62 for mean morning glucose and mean glucose. HGI was
the only significant glucose measure in binary logistic regression. We concluded that HGI
quantifies the impact of hyperglycemia in critically ill patients better than other glucose
indices. HGI may thus be a useful measure of glucose regulation. We also determined
whether the relation between HGI and outcome was present in patients admitted to a
medical Intensive Care Unit (Chapter 5.3).
Finally, in Chapter 6, we describe the protocol of the GIPS-2 study. This chapter is already
the implementation of future perspectives based on the results described in this thesis.
GIPS-2 is an ongoing multi-center trial on the effect of GIK infusion in ST segment
elevation MI patients without signs of heart-failure and eligible for reperfusion therapy.
Future perspectives
The definite role for metabolic modulation of the ischemic myocardium with GIK infusion
has to be determined yet. Studies with GIK in acute MI were primarily based on the
premises that infusion of glucose is beneficial for the ischemic myocardium.11;12 Decades
ago the glucose hypothesis was proposed: enhanced uptake and metabolism of glucose
delays cellular damage.13;14 Glucose utilization during ischemia prevents the breakdown of
glycogen stores and leads to increased net intramyocardial glycogen synthesis, thereby
limiting enzymatic infarct size and contracture.15-17 Recently, Opie stated that new
concepts (table 1) have updated and outdated this hypothesis although the balance of
glucose versus fatty acid oxidation remains crucial in ischemia.18;19
We agree, that the potential positive effects of insulin during stress situations are
multifarious and open for investigation in different patient groups.37;38 Insulin is involved in
the uptake of glucose by several tissues including the myocardium.39;40 It is also involved
203
Metabolic interventions in acute myocardial infarction
in gene transcription, expression of various metabolic enzymes, and activation of various
pathways with mitogenic activity.41 Insulin inhibits postischemic apoptosis, energetic
failure and damage to cardiac tissue, possibly through reduced oxidative stress mediated
by KATP channel activation.33;42-44 Insulin stimulates protein synthesis and inhibits
intracellular protein breakdown in cardiac tissue.45-47 Insulin potentiates ischemic
preconditioning.48;49 Insulin has an anti-inflammatory effect.50;51 Insulin improves the
fibrinolytic profile during myocardial ischemia.51;52 Insulin/strict-glucose-control restores
the abnormalities in the serum lipid profile.53;54 Finally, insulin has vasodilating capacities
in the blood vessels of muscle tissue as well as in the myocardium.55;56
Table 1. Key new concepts on the role of insulin and glucose in myocardial ischemia
[partly adapted from Opie LH, Jonassen A, Yellon DM. Cardiovasc J S Afr 2004;15:S1]18
•
Reset of glycolysis-oxidation mismatch20
•
Fatty acid oxidation↓, glucose↑21
•
Low flow induces myocardial glucose uptake and is related to ATP↑22;23
•
Insulin is related to increased myocardial flow24
•
Insulin stimulates Akt/PKB25
•
Inhibition of malonyl CoA decarboxylase: fatty acid oxidation↓, glucose↑26
•
PPAR & nucleus↑27-29
•
Fatty acid-induced genes30
•
Prosurvival pathway stimulation31;32
•
Activation of KATP channel related to reduced oxidative stress33
•
Mitochondrial uncoupling proteins 2 and 3 regulatory role on myocardial
metabolism34
•
AMP-activated protein kinase mediates ischemic glucose uptake35;36
The results of a landmark study on intensive insulin therapy in critically ill patients in order
to maintain blood glucose between 4.4 and 6.1 mmol/L and the results of the Diabetes
Mellitus, Insulin Glucose Infusion in Acute Myocardial Infarction study emphasize the
beneficial effect of insulin.57-60 Although, these studies focussed on the effects of more
strict glucose regulation as an important factor in explaining the positive results, the
results may also be interpreted in a different manner. In studies regarding acute MI there
is a trend towards a significant beneficial effect of insulin or glucose-insulin-potassium
infusion despite the fact that normoglycemia was not obtained.5;6;61 Therefore, insulin
could
be
both
a
moderator
for
transmembrane
glucose
transport
to
obtain
normoglycemia and a moderator of the effects that are mentioned above. It is to be
expected that in future studies the role of insulin or GIK, with or without meticulous
glucose regulation, will be established in the treatment of various conditions.37 The GIPS-2
204
Summary and future perspectives
is designed to investigate the effect of GIK infusion in ST segment elevation MI patients
without signs of heart-failure and eligible for reperfusion therapy. For patients with signs
of heart failure on admission, a more tailored approach seems warranted, such as the
admission of glucose 30% or an infusion of no more than 500 ml, or a period of infusion
of 12 to 24 hours.62 A first step is to investigate the effect of GIK infusion on the
hemodynamics after admission with acute MI. Already we performed a pilot-study on this
subject in approximately 90 MI patients. The results of this study will be used to
determine a tailored approach.
Only three small prospective studies, including 12 to 30 patients, have been published on
the possible beneficial effect of GIK infusion in stable patients with ischemic cardiac
dysfunction.63-65 We designed a prospective study to address whether high-dose lowvolume infusion of GIK administered through a central line will lead to improved
myocardial function without causing hemodynamic disturbances in stable patients
suffering from severe ischemic myocardial dysfunction.
To investigate the potential benefit of strict glucose regulation by insulin in patients
admitted to Intensive or Coronary Care Units, a feasible algorithm and reliable glucose
monitor are mandatory.66 To obtain strict regulation in critically ill patients, it is necessary
to measure glucose levels frequently and then change the rate of insulin infusion
accordingly. These frequent measurements also protect against adverse hypoglycemic
episodes.67 The infusion of either insulin alone or GIK can be adjusted according to an
algorithm.68
Table 2. Characteristics of algorithms with proved effective glucose regulation (the most
effective form of administration is mentioned first)
•
Continuous versus intermittent infusion
•
Intravenous versus subcutaneous administration
•
Adjusting the rate of infusion to the last two blood-glucose values prevents
hypoglycemic episodes and promotes stricter regulation
•
Starting with a bolus expedites reaching target values
•
Hourly adjustments lead to stricter regulation
•
Starting at lower values reduces the period of infusion
•
Combining insulin treatment with a supply of energy (i.e. parenteral or enteral feeding,
glucose infusion)
At the moment, an algorithm in which the infusion of insulin has been determined on the
basis of the last two measured blood-glucose values seems to be the most effective to
obtain strict glucose regulation. Other features of an algorithm which seem more
205
Metabolic interventions in acute myocardial infarction
favorable are represented in table 2. After an algorithm is implemented, it has to be
evaluated and adjusted until satisfying results are achieved.
To prevent delay due to long turnaround times of glucose measurements at the
laboratory, bedside glucometry, and an accepted method for estimating blood glucose
concentrations among ambulatory and hospital ward patients would be useful. Because
the accuracy of this system among the critically ill has not been properly evaluated,
although it is used, we performed a prospective audit of bedside glucometry in our ICU
setting.59;60 We installed point-of-care bloodgas/glucose analyzers – ABL 715 [Radiometer
Medical, Copenhagen, Denmark] – in our Intensive and Coronary Care Units.59;60 In 27
male and 19 female Medical Intensive Care Unit patients, aged 32 years to 88 years, we
performed a prospective audit in which the reference laboratory instrument [YSI, Yellow
Springs Instruments, Ohio, USA] were compared with the bloodgas analyzer. Pearson’s
correlation coefficient was calculated, which was fairly good: 0.95. A tested handheld
device [Precision PCx, Abbott Laboratories, Illinois, USA] proofed also to be good with
100% of all paired measures within the set ranges. A continuous display of blood glucose
levels in critically ill patients for optimal titration of insulin therapy could be a next step.69-72
The performance of the Continuous Glucose Monitoring System [Medtronic MiniMed,
Northridge, USA] was recently found to be clinically acceptable.73 In a pilot-study of 20
patients, we also investigated the reliability and accuracy of the Continuous Glucose
Monitoring System which was fairly good, with a Pearson’s correlation coefficient of 0.88.
We found that glucometry with the bloodgas analyzer, and the continuous monitoring
system provided a reasonable estimate of conventional (reference) laboratory glucose
assessment. These systems could play an important role in future investigations in both
the effect and potential beneficial mechanisms of metabolic modulation in critically ill
patients.
At the moment, clinical studies support the effectiveness of metabolic interventions in
several groups of critically ill patients.5;6;57-61;74;75 We will have to wait for new clinical
studies that will show that experimentally obtained results, such as the effect of insulin on
immune function, inflammation, lipid profile, and apoptosis, can be effectively translated
in every day practice. In conclusion to this thesis: substantial experimental evidence
underscores the potential benefit of treatment with insulin, glucose and potassium for the
heart exposed to ischemia and reperfusion.
References
1. Rackley CE, Russell RO, Jr., Rogers WJ, Mantle JA, McDaniel HG, Papapietro SE. Clinical
experience with glucose-insulin-potassium therapy in acute myocardial infarction. Am Heart J
1981;102:1038-49.
206
Summary and future perspectives
2.
Goulston A. West Indian cane sugar in treatment of certain forms of heart diseases. British
Medical Journal 1912;2:693-95.
3.
Sodi-Pallares D, Testelli MR, Fishleder BL et al. Effects of an intravenous infusion of a
potassium-glucose-insulin solution on the electrocardiographic signs of myocardial infarction.
A preliminary clinical report. Am J Cardiol 1962;9:166-81.
4.
Fath-Ordoubadi F, Beatt KJ. Glucose-insulin-potassium therapy for treatment of acute
myocardial infarction: an overview of randomized placebo-controlled trials. Circulation
1997;96:1152-56.
5.
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.
6.
van der Horst IC, Zijlstra F, van't Hof AW et al. Glucose-insulin-potassium infusion in patients
treated with primary angioplasty for acute myocardial infarction: the glucose-insulinpotassium study: a randomized trial. J Am Coll Cardiol 2003;42:784-91.
7.
Barbash GI, Roth A, Hod H et al. Rapid resolution of ST elevation and prediction of clinical
outcome in patients undergoing thrombolysis with alteplase (recombinant tissue-type
plasminogen activator): results of the Israeli Study of Early Intervention in Myocardial
Infarction. Br Heart J 1990;64:241-47.
8.
Schroder R, Dissmann R, Bruggemann T et al. Extent of early ST segment elevation
resolution: a simple but strong predictor of outcome in patients with acute myocardial
infarction. J Am Coll Cardiol 1994;24:384-91.
9.
van 't Hof AW, Liem A, de Boer MJ, Zijlstra F. Clinical value of 12-lead electrocardiogram after
successful reperfusion therapy for acute myocardial infarction. Zwolle Myocardial infarction
Study Group. Lancet 1997;350:615-19.
10.
Iwakura K, Ito H, Ikushima M et al. Association between hyperglycemia and the no-reflow
11.
Rackley CE, Russell RO, Jr., Rogers WJ, Mantle JA, McDaniel HG. Glucose-insulin-potassium
phenomenon in patients with acute myocardial infarction. J Am Coll Cardiol 2003;41:1-7.
infusion in acute myocardial infarction. Review of clinical experience. Postgrad Med
1979;65:93-99.
12.
Henderson MJ, Morgan HE, Park CR. Regulation of glucose uptake in muscle. IV. The effect of
hypophysectomy on glucose transport, phosphorylation, and insulin sensitivity in the
isolated, perfused heart. J Biol Chem 1961;236:273-77.
13.
Opie LH. Hypothesis: Glycolytic rates control cell viability in ischemia. J Appl Cardiol
14.
Opie LH. The glucose hypothesis: its relation to acute myocardial ischemia. J Mol Cell Cardiol
15.
Owen P, Dennis S, Opie LH. Glucose flux rate regulates onset of ischemic contracture in
16.
Eberli FR, Weinberg EO, Grice WN, Horowitz GL, Apstein CS. Protective effect of increased
1988;3:407-14.
1970;1:107-15.
globally underperfused rat hearts. Circ Res 1990;66:344-54.
glycolytic substrate against systolic and diastolic dysfunction and increased coronary
resistance from prolonged global underperfusion and reperfusion in isolated rabbit hearts
perfused with erythrocyte suspensions. Circ Res 1991;68:466-81.
207
Metabolic interventions in acute myocardial infarction
17.
Depre C, Vanoverschelde JL, Taegtmeyer H. Glucose for the heart. Circulation 1999;99:578-
18.
Opie LH, Jonnesen A, Yellon DM. The glucose hypothesis: Updated or outdated? Role of
19.
Van Rooyen J, McCarthy J, Opie LH. Increased glycolysis during ischaemia mediates the
88.
insulin. Cardiovasc J S Afr 2004;15:S1.
protective effect of glucose and insulin in the isolated rat heart despite the presence of
cardiodepressant exogenous substrates. Cardiovasc J S Afr 2002;13:103-9.
20.
Lopaschuk GD, Wambolt RB, Barr RL. An imbalance between glycolysis and glucose
oxidation is a possible explanation for the detrimental effects of high levels of fatty acids
during aerobic reperfusion of ischemic hearts. J Pharmacol Exp Ther 1993;264:135-44.
21.
Stanley WC, Lopaschuk GD, Hall JL, McCormack JG. Regulation of myocardial carbohydrate
metabolism under normal and ischaemic conditions. Potential for pharmacological
interventions. Cardiovasc Res 1997;33:243-57.
22.
Cave AC, Ingwall JS, Friedrich J et al. ATP synthesis during low-flow ischemia: influence of
23.
Sondergaard HM, Bottcher M, Schmitz O, Nielsen TT, Botker HE. Insulin-stimulated
increased glycolytic substrate. Circulation 2000;101:2090-2096.
myocardial glucose uptake and the relation to perfusion and the nitric oxide system. J Vasc
Res 2004;41:38-45.
24.
Sundell J, Luotolahti M. Association between insulin resistance and reduced coronary
25.
Jonassen AK, Sack MN, Mjos OD, Yellon DM. Myocardial protection by insulin at reperfusion
vasoreactivity in healthy subjects. Can J Cardiol 2004;20:691-95.
requires early administration and is mediated via Akt and p70s6 kinase cell-survival signaling.
Circ Res 2001;89:1191-98.
26.
Dyck JR, Cheng JF, Stanley WC et al. Malonyl coenzyme a decarboxylase inhibition protects
the ischemic heart by inhibiting fatty acid oxidation and stimulating glucose oxidation. Circ
Res 2004;94:e78-e84.
27.
Young ME, Laws FA, Goodwin GW, Taegtmeyer H. Reactivation of peroxisome proliferatoractivated receptor alpha is associated with contractile dysfunction in hypertrophied rat heart.
J Biol Chem 2001;276:44390-44395.
28.
Young ME, Guthrie PH, Razeghi P et al. Impaired long-chain fatty acid oxidation and
29.
Young LH. Insulin resistance and the effects of thiazolidinediones on cardiac metabolism. Am
30.
Razeghi P, Young ME, Alcorn JL, Moravec CS, Frazier OH, Taegtmeyer H. Metabolic gene
contractile dysfunction in the obese Zucker rat heart. Diabetes 2002;51:2587-95.
J Med 2003;115 Suppl 8A:75S-80S.
expression in fetal and failing human heart. Circulation 2001;104:2923-31.
31.
Hausenloy DJ, Duchen MR, Yellon DM. Inhibiting mitochondrial permeability transition pore
opening at reperfusion protects against ischaemia-reperfusion injury. Cardiovasc Res
2003;60:617-25.
32.
Zhang HF, Fan Q, Qian XX et al. Role of insulin in the anti-apoptotic effect of glucose-insulinpotassium in rabbits with acute myocardial ischemia and reperfusion. Apoptosis 2004;9:77783.
208
Summary and future perspectives
33.
LaDisa JF, Jr., Krolikowski JG, Pagel PS, Warltier DC, Kersten JR. Cardioprotection by
glucose-insulin-potassium: dependence on KATP channel opening and blood glucose
concentration before ischemia. Am J Physiol Heart Circ Physiol 2004;287:H601-H607.
34.
Essop MF, Razeghi P, McLeod C, Young ME, Taegtmeyer H, Sack MN. Hypoxia-induced
decrease of UCP3 gene expression in rat heart parallels metabolic gene switching but fails to
affect mitochondrial respiratory coupling. Biochem Biophys Res Commun 2004;314:561-64.
35.
Russell RR, III, Li J, Coven DL et al. AMP-activated protein kinase mediates ischemic glucose
uptake and prevents postischemic cardiac dysfunction, apoptosis, and injury. J Clin Invest
2004;114:495-503.
36.
Altarejos JY, Taniguchi M, Clanachan AS, Lopaschuk GD. Myocardial ischemia differentially
37.
Das UN. Insulin: an endogenous cardioprotector. Curr Opin Crit Care 2003;9:375-83.
38.
Groeneveld AB, Beishuizen A, Visser FC. Insulin: a wonder drug in the critically ill? Crit Care
regulates LKB1 and an alternate 5'AMP-activated protein kinase kinase. J Biol Chem 2004.
2002;6:102-5.
39.
McNulty PH, Jacob R, Deckelbaum LI, Young LH. Effect of hyperinsulinemia on myocardial
40.
Ferrannini E, Santoro D, Bonadonna R, Natali A, Parodi O, Camici PG. Metabolic and
41.
Sun XJ, Wang LM, Zhang Y et al. Role of IRS-2 in insulin and cytokine signalling. Nature
42.
Aikawa R, Nawano M, Gu Y et al. Insulin prevents cardiomyocytes from oxidative stress-
43.
Jonassen AK, Brar BK, Mjos OD, Sack MN, Latchman DS, Yellon DM. Insulin administered at
amino acid uptake in patients with coronary artery disease. Metabolism 2000;49:1365-69.
hemodynamic effects of insulin on human hearts. Am J Physiol 1993;264:E308-E315.
1995;377:173-77.
induced apoptosis through activation of PI3 kinase/Akt. Circulation 2000;102:2873-79.
reoxygenation exerts a cardioprotective effect in myocytes by a possible anti-apoptotic
mechanism. J Mol Cell Cardiol 2000;32:757-64.
44.
Kim B, Feldman EL. Insulin-like growth factor I prevents mannitol-induced degradation of
45.
Madibally SV, Solomon V, Mitchell RN, Van De WL, Yarmush ML, Toner M. Influence of
46.
Solomon V, Madihally S, Mitchell RN, Yarmush M, Toner M. Antiproteolytic action of insulin
47.
Young LH, Dahl DM, Rauner D, Barrett EJ. Physiological hyperinsulinemia inhibits myocardial
48.
Kersten JR, Schmeling TJ, Orth KG, Pagel PS, Warltier DC. Acute hyperglycemia abolishes
focal adhesion kinase and Akt. J Biol Chem 2002;277:27393-400.
insulin therapy on burn wound healing in rats. J Surg Res 2003;109:92-100.
in burn-injured rats. J Surg Res 2002;105:234-42.
protein degradation in vivo in the canine heart. Circ Res 1992;71:393-400.
ischemic preconditioning in vivo. Am J Physiol 1998;275:H721-H725.
49.
Kersten JR, Montgomery MW, Ghassemi T et al. Diabetes and hyperglycemia impair
activation
of
mitochondrial
K(ATP)
channels.
Am
J
Physiol
Heart
Circ
Physiol
2001;280:H1744-H1750.
50.
Dandona P, Aljada A, Mohanty P et al. Insulin inhibits intranuclear nuclear factor kappaB and
stimulates IkappaB in mononuclear cells in obese subjects: evidence for an anti-inflammatory
effect? J Clin Endocrinol Metab 2001;86:3257-65.
51.
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.
209
Metabolic interventions in acute myocardial infarction
52.
Melidonis A, Stefanidis A, Tournis S et al. The role of strict metabolic control by insulin
infusion on fibrinolytic profile during an acute coronary event in diabetic patients. Clin Cardiol
2000;23:160-164.
53.
Mesotten D, Swinnen JV, Vanderhoydonc F, Wouters PJ, Van den BG. Contribution of
circulating lipids to the improved outcome of critical illness by glycemic control with intensive
insulin therapy. J Clin Endocrinol Metab 2004;89:219-26.
54.
Dai T, Abou-Rjaily GA, Al Share' QY et al. Interaction between altered insulin and lipid
55.
Legtenberg RJ, Houston RJ, Oeseburg B, Smits P. Physiological insulin concentrations
metabolism in CEACAM1-inactive transgenic mice. J Biol Chem 2004;279:45155-61.
protect against ischemia-induced loss of cardiac function in rats. Comp Biochem Physiol A
Mol Integr Physiol 2002;132:161-67.
56.
McNulty PH, Pfau S, Deckelbaum LI. Effect of plasma insulin level on myocardial blood flow
57.
Malmberg K, Ryden L, Efendic S et al. Randomized trial of insulin-glucose infusion followed
and its mechanism of action. Am J Cardiol 2000;85:161-65.
by subcutaneous insulin treatment in diabetic patients with acute myocardial infarction
(DIGAMI study): effects on mortality at 1 year. J Am Coll Cardiol 1995;26:57-65.
58.
Malmberg K, Ryden L, Hamsten A, Herlitz J, Waldenstrom A, Wedel H. Effects of insulin
treatment on cause-specific one-year mortality and morbidity in diabetic patients with acute
myocardial infarction. DIGAMI Study Group. Diabetes Insulin-Glucose in Acute Myocardial
Infarction. Eur Heart J 1996;17:1337-44.
59.
van den Berghe G, Wouters P, Weekers F et al. Intensive insulin therapy in the critically ill
60.
van den Berghe G, Wouters PJ, Bouillon R et al. Outcome benefit of intensive insulin therapy
61.
Pache J, Kastrati A, Mehilli J et al. A randomized evaluation of the effects of glucose-insulin-
patients. N Engl J Med 2001;345:1359-67.
in the critically ill: Insulin dose versus glycemic control. Crit Care Med 2003;31:359-66.
potassium infusion on myocardial salvage in patients with acute myocardial infarction treated
with reperfusion therapy. Am Heart J 2004;148:105.
62.
Heng MK, Norris RM, Singh BN, Barratt-Boyes C. Effects of glucose and glucose-insulinpotassium on haemodynamics and enzyme release after acute myocardial infarction. Br Heart
J 1977;39:748-57.
63.
Alan S, Ulgen MS, Dedeoglu I, Kaya H, Toprak N. Long-term glucose insulin potassium
infusion improves systolic and diastolic function in patients with chronic ischemic
cardiomyopathy. Swiss Med Wkly 2003;133:419-22.
64.
Cottin Y, Lhuillier I, Gilson L et al. Glucose insulin potassium infusion improves systolic
function in patients with chronic ischemic cardiomyopathy. Eur J Heart Fail 2002;4:181-84.
65.
Khoury VK, Haluska B, Prins J, Marwick TH. Effects of glucose-insulin-potassium infusion on
66.
Brown G, Dodek P. Intravenous insulin nomogram improves blood glucose control in the
chronic ischaemic left ventricular dysfunction. Heart 2003;89:61-65.
critically ill. Crit Care Med 2001;29:1714-19.
67.
Libby P, Maroko PR, Braunwald E. The effect of hypoglycemia on myocardial ischemic injury
68.
Goldberg PA, Siegel MD, Sherwin RS et al. Implementation of a Safe and Effective Insulin
during acute experimental coronary artery occlusion. Circulation 1975;51:621-26.
Infusion Protocol in a Medical Intensive Care Unit. Diabetes Care 2004;27:461-67.
210
Summary and future perspectives
69.
Chee F, Fernando T, van Heerden PV. Closed-loop control of blood glucose levels in critically
70.
Chee F, Fernando T, van Heerden PV. Closed-loop glucose control in critically ill patients
ill patients. Anaesth Intensive Care 2002;30:295-307.
using continuous glucose monitoring system (CGMS) in real time. IEEE Trans Inf Technol
Biomed 2003;7:43-53.
71.
Jungheim K, Wientjes KJ, Heinemann L, Lodwig V, Koschinsky T, Schoonen AJ.
Subcutaneous continuous glucose monitoring: feasibility of a new microdialysis-based
glucose sensor system. Diabetes Care 2001;24:1696-97.
72.
Kapitza C, Lodwig V, Obermaier K et al. Continuous glucose monitoring: reliable
measurements for up to 4 days with the SCGM1 system. Diabetes Technol Ther 2003;5:60914.
73.
Goldberg PA, Siegel MD, Russell RR et al. Experience with the continuous glucose monitoring
74.
Furnary AP, Gao G, Grunkemeier GL et al. Continuous insulin infusion reduces mortality in
system in a medical intensive care unit. Diabetes Technol Ther 2004;6:339-47.
patients with diabetes undergoing coronary artery bypass grafting. J Thorac Cardiovasc Surg
2003;125:1007-21.
75.
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;109:1497-502.
211
Metabolic interventions in acute myocardial infarction
212
Dutch summary
Chapter 8
Dutch summary
213
Metabolic interventions in acute myocardial infarction
Dit proefschrift beschrijft bovenal de resultaten van onderzoek naar de effecten van het
glucose-insuline-kalium (GIK) infuus bij patiënten met een acuut myocardinfarct (MI) die
behandeld zijn met een primaire percutane coronair interventie (PCI).1 In het tweede deel
van dit proefschrift wordt nader ingegaan op het voorkomen van hyperglykemie en de
morbiditeit en mortaliteit bij patiënten waarbij strikte glucose regulatie niet is nagestreefd.
Reeds in 1912 suggereerde Goulston dat glucose gunstig kon zijn bij verschillende
soorten hartziekten.2 Het beoogde effect van GIK infusie tijdens ischemie en reperfusie
van het myocard is dat de opname en verbranding van glucose wordt bevorderd en de
opname en verbranding van vrije vetzuren wordt beperkt (de glucose hypothese van
Opie).3 Hierdoor zou het metabolisme zo te beïnvloeden zijn dat de functie van het
myocard wordt behouden en ritmestoornissen worden voorkomen, hetgeen tot een
betere prognose leidt. Bij verbranding van vrije vetzuren komen immers schadelijke
produkten vrij en meer zuurstof is nodig om een zelfde hoeveelheid energie te verkrijgen
dan bij de verbranding van glucose.
Sodi-Pallares en zijn collega’s waren de eerste die rapporteerden over het gebruik van een
GIK infuus bij patiënten met een MI. De auteurs concludeerden dat GIK ritme- en
geleidingsstoornissen gedurende ischemie van het myocard beperkte.4 Na het verschijnen
van een overzicht van 9 onderzoeken met 1932 MI patiënten naar het effect van GIK op
overleving, werd in een begeleidend redactioneel schrijven geschreven dat de tijd
gekomen was voor een grote studie.5
Een eerste pilot-onderzoek (Estudios Cardiologicos Latinoamerica) met 407 patiënten dat
daarop volgde, concludeerde dat met name patiënten, die behandeld werden met zowel
een GIK infuus als thrombolyse, van het GIK infuus bleken te profiteren.6 Dit betrof echter
een beperkt aantal patiënten en de sterfte in de groep patiënten die met thrombolyse en
niet met GIK behandeld werden, was hoog (15.2%). Het vormde echter wel de basis voor
het opzetten van de Glucose-Insulin-Potasssium Study (GIPS-1).7
In de GIPS-1 zijn 940 patiënten met een acuut MI geïncludeerd en gerandomiseerd naar
een behandeling van het afgesloten vat door middel van een primaire PCI met GIK
(N=476) of zonder GIK (N=464). Het GIK infuus bestond uit glucose 20% met 80 mmol
kalium in 500 ml natriumchloride 0.9% dat met een snelheid van 3 ml/kg/per uur werd
toegediend. Daarnaast werd afhankelijk van de serum glucose concentratie insuline (50 IU
insuline in 50 ml water) gegeven. Het doel van de infusie van insuline was om de serum
glucose concentratie tussen de 7.0 mmol/L en de 11.0 mmol/L te krijgen en vervolgens te
houden. In vergelijking met de reeds uitgevoerde studies werd in de GIPS-1 reperfusie
nagestreefd middels PCI (en niet middels thrombolyse) en werd een hoge dosis GIK
infuus gebruikt. Een hoge dosis GIK infuus is een dosis die qua samenstelling gelijk of
hoger ligt met een dosis waarvan is aangetoond dat het de opname en verbranding van
vrije vetzuren onderdrukt (glucose hypothese).8
214
Dutch summary
In hoofdstuk 2.1 wordt het resultaat op de mortaliteit en cardiale complicaties na 30 dagen
beschreven. De mortaliteit in de GIK groep was lager (4.8%) vergeleken met de controle
groep (5.8%). Het relatieve risico op overlijden in de GIK groep was 0.61. Het gevonden
verschil was echter niet statistisch significant. Analyse van vooraf vastgestelde
subgroepen liet zien dat patiënten (N=856) zonder tekenen van hartfalen bij binnenkomst
(Killip klasse 1), een statistisch significante reductie van de mortaliteit (1.2% in de GIK
groep versus 4.2% in de controle groep, P=0.01) hadden. Bij patiënten die bij
binnenkomst wel tekenen van hartfalen vertonen (Killip klasse ≥2) (N=84), werd echter
een hogere mortaliteit gezien. Deze laatste groep is echter te klein om duidelijke
conclusies uit te trekken. Wel lijkt voor deze groep het in toekomstig onderzoek van
belang de samenstelling van GIK aan te passen.
In Hoofdstukken 2.2 en 2.3, wordt het effect van het GIK infuus op de functie van het
myocard, de enzymatische infarctgrootte en de linker ventrikel functie, beschreven. Geen
verschil werd gevonden in het beloop of de hoogte van het creatine kinase MB (CK-MB)
(als maat voor enzymatische infarctgrootte) tussen beide groepen. Het behoud van de
linker ventrikel ejectiefractie (LVEF) bleek wel beter te zijn in de GIK groep vergeleken met
de controle groep. In de GIK groep had 12% van de patiënten een slechte LVEF (LVEF
≤30%), in de controle groep had 18% een slechte LVEF (P=0.01). Aangezien er een
verband is aangetoond tussen de hoogte van het CK-MB en de LVEF wordt gesuggereerd
dat het effect van GIK mogelijk bewerkstelligd wordt door factoren die niet samenhangen
met de enzymatische infarct grootte.
Hoofdstuk 2.4 beschrijft het effect van het GIK infuus op de ST segment elevatie resolutie
(afname). ST segment elevatie resolutie blijkt een goede weerspiegeling van de perfusie
van het myocard te zijn en tevens een prognostische factor.9-11 Bij 612 patiënten
opgenomen in GIPS-1 werd het ECG bij binnenkomst en het ECG 3 uur na binnenkomst
geanalyseerd. Een complete (>70%) of partiële (30-70%) resolutie werd vaker bereikt in
de GIK groep (87%) vergeleken met de controle groep (78%) (P<0.001). Omdat ST
segment elevatie resolutie na reperfusie therapie een afspiegeling van bloeddoorstroming
in het myocard is wordt gesuggereerd dat het effect van GIK mogelijk voor een deel kan
samenhangen met een toename van de myocardiale perfusie.
In Hoofdstuk 2.5, wordt het effect van het GIK infuus op de serum concentraties van
glucose en kalium beschreven. De gemiddelde glucose waarde bij patiënten behandeld
met GIK lag hoger (9.3 mmol/L) vergeleken met de controle groep (8.4 mmol/L) (P<0.001).
Hyperglykemie gedefinieerd als een glucose waarde van >11.0 mmol/L werd bij 70.8%
van de patiënten in de GIK groep geobserveerd vergeleken met 33.8% van de patiënten in
de controle groep (P<0.001). De mortaliteit van de patiënten met een hyperglykemie lag
hoger vergeleken met de patiënten zonder hyperglykemie, respectievelijk 9.4% en 4.7%
(P=0.004). Hypoglykemie (glucose ≤3.0 mmol/L) werd 10 maal geobserveerd, 2 patiënten
overleden, beiden werden behandeld met het GIK infuus. Hyperkaliëmie (kalium >5.5
215
Metabolic interventions in acute myocardial infarction
mmol/L) kwam bij 5.5% van de GIK patiënten en 3.2% van de controle patiënten voor
(p=0.11). Hypokaliëmie (kalium ≤3.5 mmol/L) werd vaker gevonden bij controle patiënten
(23.5% versus 41.2%, P<0.001). Hyperglykemie blijkt iets vaker voor te komen bij
patiënten die worden behandeld met het GIK infuus, hypokaliëmie minder vaak. Het is
heel goed mogelijk dat de voordelige effecten van het GIK infuus, besproken in de eerdere
hoofdstukken verminderd worden door het optreden van bovengenoemde metabole
effecten.
De 3-jaars resultaten van de GIPS-1 worden gerapporteerd in Hoofdstuk 2.6. De mortaliteit
na 3 jaar in de totale groep bedroeg 10.4% (98 van de 940 patiënten). Factoren
samenhangend met mortaliteit waren leeftijd, cardiovasculaire aandoeningen in de
voorgeschiedenis, een voorwand MI, Killip klasse ≥2, en meervatslijden. Na 3 jaar waren
in de GIK groep, 46 van de 476 patiënten (9.7%) overleden, vergeleken met 52 van de 464
patiënten (11.2%) in de controle groep (NS). In de subgroep van patiënten die bij
binnenkomst geen tekenen van hartfalen hadden (Killip klasse 1, N=856) bleek het GIK
infuus een onafhankelijke voorspeller te zijn voor de 3-jaars overleving, met een relatief
risico op overlijden van 0.56 (P=0.028). Het beschermende effect van het GIK infuus
resulteerde niet in een significante reductie van mortaliteit na 3 jaar bij alle patiënten.
Echter het positieve effect in de subgroep van patiënten met een Killip klasse 1 bleek ook
na 3 jaar nog aanwezig te zijn.
In Hoofdstuk 3 wordt een overzicht gegeven van alle studies naar het effect van glucoseinsuline(-kalium) infusie bij patiënten met een acuut MI. Van de 13 gepubliceerde
onderzoeken laten 12 onderzoeken een gunstig effect zien op mortaliteit gedurende de
eerste 30 dagen na opname. De meest gunstige resulaten zijn verkregen in onderzoeken
waarin een hoge dosis GIK gebruikt werd en waarin GIK gecombineerd werd met een
vorm van reperfusie. Omdat de onderzoeken verschillen in opzet is een definitieve
conclusie omtrent het toevoegen van GIK aan de behandeling nog niet te trekken.
In het tweede deel van dit proefschrift worden de resultaten van observationele studies
beschreven, met betrekking op het vóórkomen van hyperglykemie en de morbiditeit en
mortaliteit bij patiënten waarbij niet gestreeft is naar strikte glucose regulatie.
Hoofdstuk 4.1, beschrijft de invloed van het glucose bij binnenkomst en de morbiditeit en
mortaliteit bij patiënten met een MI die niet bekend met diabetes mellitus (DM). De
onderzochte patiënten werden behandeld met streptokinase (thrombolyse) of primaire
PCI. Een verhoogde glucose waarde bij binnenkomst blijkt onafhankelijk geassocieerd te
zijn met een grotere infarctgrootte en een hogere mortaliteit.
Hoofdstuk 4.2, beschrijft het effect van de glucose bij binnenkomst op de myocardiale
perfusie bij patiënten met een MI behandeld met primaire PCI. Van 464 patiënten hadden
93 patiënten (20%) hyperglykemie (glucose ≥11.0 mmol/L). Vergeleken met patiënten die
geen hyperglykemie bij binnenkomst hadden, bleken zij vaker een verminderde
216
Dutch summary
myocardiale blush (maat voor perfusie van het myocard) te hebben (26% versus 15%,
P=0.02) en vaker een incomplete ST segment elevatie resolutie te hebben (59% versus
39%, P=0.01). Bovendien was de mortaliteit in de groep van patiënten met een
hyperglykemie bij binnenkomst hoger en was de reductie van de LVEF in deze groep
groter. Dit effect werd zowel geobserveerd bij patiënten die wel en niet bekend met DM
waren. Concluderend is hyperglykemie bij binnenkomst geassocieerd met een
verminderde myocardiale reperfusie en geassocieerd met een slechtere beloop.
Hoofdstuk 4.3, is een reactie op een publicatie waarin geconcludeerd wordt dat
hyperglykemie mogelijk samenhangt met een gestoorde microvasculaire functie na een
acuut MI. De auteurs concludeerden dit naar aanleiding van een retrospectieve analyse bij
146 patiënten.12 Hun hypothese dat acute hyperglykemie (hyperglykemie bij binnenkomst)
geassocieerd is met het ‘no reflow’ fenomeen is intrigerend. Echter, het is heel goed
mogelijk dat ‘no reflow’ niet alleen gerelateerd is aan acute hyperglykemie maar ook aan
chronische hyperglykemie (een verhoogd HbA1c en/of DM). Nieuwe studies zullen het
effect van hyperglykemie (acuut en chronisch) op het ‘no reflow’ fenomeen moeten
vaststellen.
Ondanks dat hyperglykemie bij binnenkomst een voorspeller is voor een minder gunstig
beloop bij patiënten met een MI is de voorspellende waarde niet erg sterk. Persisterende
hyperglykemie (een verhoogde glucose waarde gedurende de opname gedeeld door de
factor tijd) is mogelijk een betere voorspeller. In Hoofdstuk 5.1, wordt de relatie
beschreven tussen hyperglykemie bij binnenkomst, persisterende hyperglykemie en
cardiale complicaties (fataal/niet fataal) op korte termijn bij patiënten met een MI
behandeld met primaire PCI. Er werden 417 patiënten geïncludeerd, hiervan hadden 89
patiënten (21.3%) een cardiale complicatie, waarvan 17 maal (4.1%) fataal. Voorspellers
voor deze complicaties waren een oudere leeftijd, een cardiovasculaire aandoening in de
voorgeschiedenis, een voorwandinfarct en meervatslijden. In een multivariate analyse is
persisterende hyperglykemie maar niet hyperglykemie bij binnenkomst een onafhankelijke
voorspeller voor deze complicaties. Persisterende hyperglykemie blijkt dus een betere
voorspeller te zijn voor het optreden van cardiale complicatie vergeleken met
hyperglykemie bij binnenkomst.
In Hoofdstuk 5.2, wordt het begrip hyperglycemic index (HGI) geïntroduceerd. HGI is een
objectieve maat voor persisterende hyperglykemie. Het beschrijft de area under the curve
(AUC) boven een glucose van 6.0 mmol/L gedurende de gehele opname (mmol/L). In een
retrospectieve studie op de chirurgisch Intensive Care werd bij 1779 patiënten de HGI als
glucoseparameter bepaald, met als uitkomstmaat mortaliteit na 30 dagen. Deze mortaliteit
na 30 dagen was 17%. De mediane HGI was 0.9 mmol/L (0.3-2.1) bij patiënten die
overleefden versus 1.7 (0.7-3.3) mmol/L bij patiënten die overleden (P<0.001). De area
onder de receiver operator characteristic (ROC) curve was 0.64 voor HGI, vergeleken met
0.61 and 0.62 voor andere glucoseparameters (gemiddelde ochtend glucose waarde en
217
Metabolic interventions in acute myocardial infarction
de gemiddeld glucose tijdens de opname respectievelijk). HGI was de enige glucose
parameter die een onafhankelijke voorspeller was voor de mortaliteit. Concluderend is
HGI als maat voor persisterende hyperglykemie een betere voorspeller dan andere
glucoseparameters. Naast de retrospectieve studie op de chirurgische Intensive Care,
werd dit nogmaals bevestigd bij patiënten die werden opgenomen op de algemene
Intensive Care (Hoofdstuk 5.3).
Hoofdstuk 6 beschrijft het protocol van de GIPS-2. GIPS-2 is een multi-center onderzoek
naar het effect van het GIK infuus bij ST segment elevatie MI patiënten, die in aanmerking
komen voor reperfusie therapie echter zonder tekenen van hartfalen bij binnenkomst. In
deze GIPS-2 zijn de resultaten en de bevindingen van dit proefschrift geïmplementeerd.
Zo wordt gestreefd naar het voorkomen van hyperglykemie (streefwaarde serum glucose
concentratie tussen de 6.0 mmol/L en 10.0 mmol/L).
In Hoofdstuk 7 wordt ten slotte beschreven dat toekomstig onderzoek de rol van
scherpere glucoseregulatie tijdens myocardiale ischemie zou dienen uit te zoeken. Ten
tijde van de GIPS-1 werd in een gerandomiseerd onderzoek op een chirurgische Intensive
Care een (onverwacht) grote afname van complicaties en mortaliteit waargenomen door
patiënten zeer scherp te reguleren.13 Het doel van de infusie van insuline was om de
serum glucose concentratie tussen de 4.4 mmol/L en de 6.1 mmol/L te krijgen en
vervolgens te houden. Na het verschijnen van deze resultaten zijn velen er toe overgegaan
om bij ernstig zieke patiënten scherpere glucoseregulatie na te streven dan in de GIPS-1
en 2. In de dagelijkse praktijk blijkt het niet eenvoudig om tot scherpere regulatie te
komen. Bovendien zijn na het bovenbeschreven onderzoek meer resultaten beschikbaar
gekomen waardoor het niet geheel duidelijk is welke mate van regulatie optimaal is. Van
belang lijkt dat glucoseregulatie een vast onderdeel van de behandeling van ernstig zieke
patiënten vormt.
Met het beschikbare nieuwe bewijs uit zowel experimenteel en klinisch onderzoek lijkt niet
het toedienen van glucose maar het toedienen van insuline met name van belang.14-18
Recent schreef Opie dat de glucose hypothese misschien aangepast zou moeten worden
in een insuline hypothese.19 Daar het toedienen van insuline het risico heeft van het
optreden van een hypoglykemie en/of een hypokaliemie is het waarschijnlijk dat een
combinatie van glucose, insuline en kalium gegeven zal moeten blijven worden.
Onderzoeken
zullen
moeten
aantonen
of
voor
het
beïnvloeden
van
het
glucosemetabolisme tijdens myocardiale ischemie het noodzakelijk is om GIK toe te
dienen in een tot nu toe gebruikelijke samenstelling. Op basis van de huidige resultaten
zoals beschreven in dit proefschrift lijkt een rol voor GIK in de behandeling van patiënten
met een acuut MI van belang.
218
Dutch summary
Referenties
1. Rackley CE, Russell RO, Jr., Rogers WJ, Mantle JA, McDaniel HG, Papapietro SE. Clinical
experience with glucose-insulin-potassium therapy in acute myocardial infarction. Am Heart J
1981;102:1038-49.
2.
Goulston A. West Indian cane sugar in treatment of certain forms of heart diseases. British
Medical Journal 1912;2:693-95.
3.
Opie LH. The glucose hypothesis: its relation to acute myocardial ischemia. J Mol Cell Cardiol
1970;1:107-15.
4.
Sodi-Pallares D, Testelli MR, Fishleder BL et al. Effects of an intravenous infusion of a
potassium-glucose-insulin solution on the electrocardiographic signs of myocardial infarction.
A preliminary clinical report. Am J Cardiol 1962;9:166-81.
5.
Fath-Ordoubadi F, Beatt KJ. Glucose-insulin-potassium therapy for treatment of acute
myocardial infarction: an overview of randomized placebo-controlled trials. Circulation
1997;96:1152-56.
6.
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.
7.
van der Horst IC, Zijlstra F, van't Hof AW et al. Glucose-insulin-potassium infusion inpatients
treated with primary angioplasty for acute myocardial infarction: the glucose-insulinpotassium study: a randomized trial. J Am Coll Cardiol 2003;42:784-91.
8.
Rackley CE, Russell RO, Jr., Rogers WJ, Mantle JA, McDaniel HG. Glucose-insulin-potassium
infusion in acute myocardial infarction. Review of clinical experience. Postgrad Med
1979;65:93-99.
9.
Barbash GI, Roth A, Hod H et al. Rapid resolution of ST elevation and prediction of clinical
outcome in patients undergoing thrombolysis with alteplase (recombinant tissue-type
plasminogen activator): results of the Israeli Study of Early Intervention in Myocardial
Infarction. Br Heart J 1990;64:241-47.
10.
Schroder R, Dissmann R, Bruggemann T et al. Extent of early ST segment elevation
resolution: a simple but strong predictor of outcome in patients with acute myocardial
infarction. J Am Coll Cardiol 1994;24:384-91.
11.
van 't Hof AW, Liem A, de Boer MJ, Zijlstra F. Clinical value of 12-lead electrocardiogram after
successful reperfusion therapy for acute myocardial infarction. Zwolle Myocardial infarction
Study Group. Lancet 1997;350:615-19.
12.
Iwakura K, Ito H, Ikushima M et al. Association between hyperglycemia and the no-reflow
phenomenon in patients with acute myocardial infarction. J Am Coll Cardiol 2003;41:1-7.
13.
van den Berghe G., Wouters P, Weekers F et al. Intensive insulin therapy in the critically ill
14.
Jonassen AK, Brar BK, Mjos OD, Sack MN, Latchman DS, Yellon DM. Insulin administered at
patients. N Engl J Med 2001;345:1359-67.
reoxygenation exerts a cardioprotective effect in myocytes by a possible anti-apoptotic
mechanism. J Mol Cell Cardiol 2000;32:757-64.
15.
LaDisa JF, Jr., Krolikowski JG, Pagel PS, Warltier DC, Kersten JR. Cardioprotection by
glucose-insulin-potassium: dependence on KATP channel opening and blood glucose
concentration before ischemia. Am J Physiol Heart Circ Physiol 2004;287:H601-H607.
219
Metabolic interventions in acute myocardial infarction
16.
Young ME, Guthrie PH, Razeghi P et al. Impaired long-chain fatty acid oxidation and
17.
Zhang HF, Fan Q, Qian XX et al. Role of insulin in the anti-apoptotic effect of glucose-insulin-
contractile dysfunction in the obese Zucker rat heart. Diabetes 2002;51:2587-95.
potassium in rabbits with acute myocardial ischemia and reperfusion. Apoptosis 2004;9:77783.
18.
Sundell J, Luotolahti M. Association between insulin resistance and reduced coronary
vasoreactivity in healthy subjects. Can J Cardiol 2004;20:691-95.
19.
Opie LH, Jonnesen A, Yellon DM. The glucose hypothesis: Updated or outdated? Role of
insulin. Cardiovasc J S Afr 2004;15:S1.
220
List of abbreviations
List of Abbreviations
ACC
American College of Cardiology
ADA
American Diabetes Association
ADP
Adenosine diphosphate
AHA
American Heart Association
AMP
Adenosine monophosphate
APACHE-II
Acute Physiology And Chronic Health Evaluation II
ATP
Adenosine triphosphate
AUC
Area under the curve
CABG
Coronary artery bypass grafting
CADILLAC
Controlled Abciximab and Device Investigation to Lower Late
Angioplasty Complications
CCU
Coronary care unit
CI
Confidence interval
CK
creatine kinase
CoA
Coenzyme A
CRF
Case Record Form
CVD
Cardiovascular event
DIGAMI
Diabetes Insulin-Glucose in Acute Myocardial Infarction study
ECLA
Estudios Cardiologicos Latinoamerica pilot trial
ESC
European Society of Cardiology
FFA
Free fatty acid
GCS
Glasgow Coma Scale
GIK
Glucose-insulin-potassium infusion
GIPS
Glucose-Insulin-Potassium Study
GLUT
Glucose transporter
GP IIb/IIIa
Glycoprotein IIb/IIIa
GRACE
Global Registry of Acute Coronary Events
HbA1C
Glycosylated hemoglobin A1c
HGI
Hyperglycemic index
IABP
Intra-aortic balloon pump
ICU
Intensive care unit
IFCC
International Federation of Clinical Chemistry
IGF
Insulin-like growth factor
IGT
Impaired glucose tolerance
IRA
Infarct related artery
221
Metabolic interventions in acute myocardial infarction
IRV
Infarct related vessel
IQR
Interquartile range
LAD
Left descending artery
LDH
Lactate dehydrogenase
MACE
Major Adverse Cardiac Event
MBG
Myocardial blush grade
MI
Myocardial infarction
MRC
Medical Research Council
MVD
Multi-vessel disease
NF-κB
Nuclear factor-κB
OASIS
Organization to Assess Strategies for Ischemic Syndromes
OR
Odds ratio
PDMS
Patient database management system
PCI
Primary coronary intervention
PI3-K
Phosphatidylinositol 3-kinase
Pol-GIK
Polish-Glucose-Insulin Potassium trial
PTCA
REVIVAL
Primary transluminal coronary angioplasty
Reevaluation of Intensified Venous Metabolic Support for Acute
Infarct Size Limitation
ROC
Receiver operator characteristics
RR
Relative risk
SD
Standard deviation
SEM
Standard error of the mean
SHOCK
Should we emergenly revascularize Occluded Coronaries for
cardiogenic shocK
SK
Streptokinase
SOFA
Sequential Organ Failure Assessment
SPSS
Statistical Package for the Social Sciences
TG
Triglyceride
TIMI
Thrombolysis in myocardial infarction
222
Acknowledgements
Acknowledgements
Thank you
Anouk (love you), Kasper (love you too); B or J; Papa & Mama (Ω); family and friends;
Henk Bilo; Menko-Jan de Boer; Greet van den Berghe; Dirk-Jan van Veldhuisen; Bruce
Wolffenbuttel; Carl Apstein; Jan-Paul Ottervanger (my third co-promotor); Arnoud van ‘t
Hof; Jan Hoorntje; Harry Suryapranata; Jan-Henk Dambrink; Marcel Gosselink; Jorik
Timmer; Nicolette Ernst; José Henriques; Giuseppe de Luca; Stoffer Reiffers; Kor
Miedema; Mathijs Vogelzang; Jack Ligtenberg; Jan Zijlstra; Rob Spanjersberg; Sofie
Meijering; Carine Doggen; Gerrit Zijlstra; Peter van der Voort; Ferdinand Snellen; Bas
Houweling; Harma Israël; Thea Schenk (√); Vera Derks; Anneke Stegeman (many thanks);
everyone at Diagram BV, Zwolle, the Netherlands; Edwin Nibbering; Joep Dille; Olga
Klompstra & Alma Guikema (coffee time); Cardiologist and CCU personnel at the
University Medical Center Groningen (see you again very soon); Brigitte Klinkenbijl; David
Stocker; Christine Gesseney; Coldplay & Keane (for accompanying me at night); Maarten
Nijsten (utmost); Jolanda & Rémon ()
Very special thanks
Felix Zijlstra; Rijk Gans
Last but not least
All Cardiologists, (former) residents & CCU, laboratory and nuclear medicine personnel at
the Isala Klinieken, location Weezenlanden (without you this was never achieved!)
223
Metabolic interventions in acute myocardial infarction
Generous contributions and unrestricted grants from
Abbott Laboratories BV, Amgen BV, Astra-Zeneca BV, Aventis BV, Bristol-Myers Squibb
BV, Eli Lilly BV, Groningen University Institute for Drug Exploration (GUIDE), Medtronic
Europe Sarlé, Medtronic Minimed BV, Merck Sharp & Dohme BV, Novartis Pharma BV,
Novo-Nordisk Farma BV, Otsuka Pharmaceutical Factory Inc Japan, Pfizer Nederland BV,
Sanofi-Synthelabo BV, Servier Farma Nederland BV, and Solvay Pharma BV
224