Glucose-insulin-potassium for acute myocardial infarction: a perspective

Dr Carl S. Apstein
Boston University School of Medicine, Boston, MA, USA

Glucose-insulin-potassium (GIK) as metabolic therapy for acute myocardial infarction (MI) has received renewed interest after several clinical reports reported reduced mortality rates, and experimental studies have provided basic mechanisms that explain its beneficial clinical effects. These developments constitute the latest chapter in the long history of this therapy.

The long and controversial history of GIK
The slow recognition of GIK for metabolic treatment of acute MI is distressingly reminiscent of the sad history of streptokinase as thrombolytic therapy. First introduced in the late 1950s, the efficacy of streptokinase was not finally appreciated for three decades, too late for the many preventable acute MI deaths that occurred in the interim.
The use of GIK in acute MI has been inhibited by a history of inconclusive clinical trials, basic science controversies, and pharmaceutical industry indifference to sponsoring research without likelihood of patents and profits. In 1962, in a small, non-randomized trial, Sodi-Pallares et al.[1] reported that GIK improved some of the ECG abnormalities associated with acute MI, reduced ventricular arrhythmias, and improved early survival. Over the next 35 years inconclusive results were reported from clinical trials which were often of very poor design. Several trials initiated therapy as late as 48 h after the onset of chest pain, a timing too late to influence MI size. Others used oral glucose and subcutaneous injections of insulin or relatively low concentrations of glucose, which did not achieve the glucose and insulin plasma levels required to maximally decrease plasma free fatty acid levels. It is not surprising that such studies showed no benefit.[2]
Consideration of the glycolytic pathway raised concerns that GIK could starve the cell of energy (glucose phosphorylation requires ATP), and/or worsen ischemic injury by intensifying myocardial acidosis as a result of increased lactate production.[3] These concerns have been alleviated by recent quantitative perfusion imaging studies showing that substantial collateral flow provides significant residual perfusion in the acute MI region of most patients.[4,5] This residual flow appears adequate to support a significant level of oxidative metabolism and prevent excessive lactate accumulation from inhibiting glycolysis; direct measurements of ATP content and intracellular pH during a comparable degree of low-flow ischemia in isolated hearts have shown that a high glucose-insulin substrate increases ATP levels and the free energy status, and does not worsen tissue acidosis.[6–8] This observation is consistent with the analysis that during ischemia, proton generation from the hydrolysis of ATP exceeds that contributed by lactate;[9] thus by maintaining a higher level of ATP the glucose-insulin inhibited the development of acidosis. Furthermore, experimental studies of isolated hearts exposed to acute MI perfusion conditions have consistently shown improved function when high levels of glucose-insulin were provided as metabolic support.[6–8,10]

A resurgence of clinical interest in GIK
Interest in GIK was stimulated by the 1997 publication of a meta-analysis type of overview of randomized placebo-controlled acute MI trials.[2] Trials were included for analysis only if GIK (or placebo) was started within 48 h of chest pain. In nine such trials, involving 1932 patients, in-hospital mortality was reduced by 28% by the GIK therapy (P = 0.004). In the four studies in which the GIK was administered intravenously at high doses, the in-hospital mortality was reduced by 48% relative to the placebo group. However, the strength of these results was diminished by the intrinsic weaknesses and limitations of the retrospective meta-analysis technique. In addition, current relevance was potentially reduced because all of the cited studies were done before the advent of thrombolytic and percutaneous transluminal coronary angioplasty (PTCA).

Early GIK and subsequent reperfusion for acute MI
The benefits of GIK have now been demonstrated in two large, prospective, randomized, acute MI trials where GIK was given prior to subsequent urgent reperfusion.
In the Swedish DIGAMI trial[11] half of the patients received thrombolytic therapy for acute MI and were randomized to receive either glucose-insulin followed by multi-dose insulin therapy or standard care. In the glucose-insulin group there was a trend towards a decrease in mortality at 3 months post-MI, and this became significant at 1 year post-MI (29% relative mortality reduction, P = 0.027).
Although the DIGAMI trial specifically enrolled diabetics, its results may be applicable to non-diabetics as well, because the most dramatic benefit of the glucose-insulin therapy was seen in the patients with only ‘borderline’ or mild diabetes, i.e. patients who did not require insulin prior to their hospitalization for acute MI. In this subgroup of patients the in-hospitality mortality was reduced by glucose-insulin by 58% (P < 0.05) and the 1-year mortality was reduced by 52% (P < 0.02).
The strongest evidence for the benefits of GIK in the treatment of acute MI in the era of emergent reperfusion therapy comes from the recent ECLA (Estudios Cardiologicos Latinoamerica) study.[12,13] This was the largest prospective, randomized trial of GIK for the treatment of acute MI ever carried out. Relative to reperfusion alone, there was a remarkable 66% reduction (2P = 0.008) in the relative in-hospital mortality risk when GIK was given prior to reperfusion (95% of those reperfused had thrombolysis, 5% had primary PTCA); the absolute mortality risk decreased from 15.2 to 5.2% when GIK was part of the treatment.

Optimal GIK dosage
The ECLA study also compared high-dose GIK (the Rackley regimen[14]) with a lower dose. During the 1-year follow-up period the high-dose GIK group had a statistically significant survival advantage relative to the control group, but the low-dose GIK group did not, suggesting a greater degree of myocardial salvage by the high-dose GIK. There was no difference in the in-hospital mortality risk between the high- and low-dose GIK groups, but this result is not conclusive because the small group sizes (n = 133–135), with 8–10 deaths per group, provided little statistical power for ruling out a dose-related difference.
The superiority of the high-dose GIK in the ECLA study is consistent with the recent meta-analysis of GIK usage in acute MI. In the nine trials that used a variety of GIK regimens, the acute MI mortality risk was reduced by 28% by GIK relative to controls, but in the four trials which used high-dose GIK (i.e. the Rackley regimen), the relative acute MI mortality reduction was 48%.2 Furthermore, a recent Polish study of low-dose GIK for acute MI showed no beneficial effect.[15,16] This study’s regimen delivered only approximately 15% of the glucose of the Rackley regimen. Thus the Rackley GIK regimen appears to the current best choice.

Reservations about the ECLA study
Some unusual aspects of the ECLA study suggest that its conclusions be considered cautiously. There was a relatively long gap of 10–11 h between the onset of symptoms and the initiation of treatment. Whether such a long delay favors the finding of a beneficial effect of GIK is not known. Since the initiation of acute MI therapy is often faster, it is important to determine the relative benefits of GIK when treatment is started more quickly than in the ECLA study. For example, in experimental studies, there was no difference in glycolytic flux rates between the control and glucose-insulin groups during the early ischemic period; a difference emerged only with more prolonged ischemia. During relatively brief periods of ischemia, myocardial glycogen stores may be able to provide a level of substrate adequate to support maximal glycolytic flux, and glucose-insulin may become important only after glycogen is depleted.[6]
The control (non-GIK) patients who underwent reperfusion in the ECLA study had a relatively high mortality risk of 15.2%, approximately twice as high as many recent large trials of thrombolysis for acute MI.[17] The relatively long time to treatment may partially account for this higher mortality risk in the ECLA study. Nonetheless, the dramatically beneficial result with GIK relative to the control group may have been partly due the surprising and unusually high mortality risk of the control group.
Also surprising is the result that the non-reperfused, non-GIK patients had a mortality risk of only 6.7%. In other words, the non-GIK reperfused patients had more than twice the mortality risk (15.2%) of the nonreperfused patients, a result which is not consistent with numerous randomized trials of thrombolytic therapy of acute MI.[17] A likely explanation for this surprising result is the small subgroup size and the fact that selection of patients for reperfusion therapy was not randomized, but left to the physician’s discretion; thus it is likely that the reperfusion group comprised sicker patients. Nonetheless, these unusual aspects of the ECLA study suggest caution, and argue for replication, before the GIK results on mortality reduction can be accepted definitively.

GIK-reperfusion interaction
An important interaction between GIK and reperfusion therapy for acute MI was observed in the ECLA study.[12,13] The reduction of acute MI mortality by GIK was observed only in the group of 252 patients who received reperfusion therapy; the patients not reperfused received no benefit from GIK. However, this result is not conclusive because the ECLA study’s non-reperfused group contained only 155 patients of whom 13 (8.4%) died; this sample size and mortality rate provided little statistical power. However, consistent with this ECLA result are ischemia-reperfusion experiments in isolated hearts. In these studies glucose-insulin had a relatively small beneficial effect on function during ischemia, but a large effect to improve post-ischemic recovery.[6,10] This result suggests that GIK may slow the rate of ischemic necrosis, so that reperfusion can salvage a larger amount of tissue; however if reperfusion does not occur, the potential benefits of GIK may not be observed.
In contrast with ECLA, the meta-analysis of pre-thrombolytic era trials of GIK for acute MI included 1932 patients, and demonstrated a 28–48% reduction in acute MI mortality by GIK.[2] Thus, while the ECLA results might argue for a strategy of using GIK only in patients destined for reperfusion therapy, the meta-analysis results derived from a 10-fold larger sample size lead to an opposite, statistically stronger conclusion. A possible explanation for the discrepancy between the ECLA and meta-analysis regarding the differing GIK effects in the non-reperfused patients may lie in the phenomenon of spontaneous reperfusion. Spontaneous thrombolysis and reperfusion occur in a significant fraction of acute MI patients who do not receive pharmacologic thrombolytic therapy. If GIK were beneficial in such patients, such an effect might be observed in a large sample size, such as that considered in the meta-analysis overview,[2] but it might not be observed in the smaller sample size of the ECLA study. Clearly, more studies are required to resolve the important question of whether GIK is beneficial for acute MI patients in the absence of reperfusion therapy. 

Future directions in cardiac metabolic therapy
Limitations of GIK therapy are the requirement of intravenous administration and the restriction to a relatively short course of treatment. Several pharmacologic agents that can be taken orally and chronically also have the potential to favorably alter cardiac energy metabolism. These compounds include trimetazidine, ranolazine, etomoxir, dichloroacetate, carnitine and propionyl-L-carnitine. Future studies in cardiac energy metabolism should compare such agents to GIK and also determine whether a given drug’s effects are additive to those of GIK.

Conclusion
What should the clinician do now? I am frequently asked whether the results reported to date constitute adequate evidence to justify the routine use of GIK for acute MI. I believe that they do not; more trials are needed, and quickly. Hopefully, the ECLA II study will soon provide the necessary information.

REFERENCES
1. Sodi-Pallares D, Testelli M, Fishleder F. Effects of an intravenous infusion of a potassium-insulin-glucose solution on the electrocardiographic signs of myocardial infarction. Am J Cardiol 1962; 9: 166–181.
2. Fath-Ordoubadi F, Beatt KJ. Glucose-insulin-potassium (GIK) therapy for treatment of acute myocardial infarction: an overview of randomized placebo controlled trials. Circulation 1997; 96: 1152–1156.
3. Neely J, Grotyohann L. Role of glycolytic products in damage to myocardium: dissociation of adenosine triphosphate levels and recovery of function of reperfused myocardium. Circ Res 1984; 55: 816–824.
4. Milavetz JJ, Giebel DW, Christian TF et al. Time to therapy and salvage in myocardial infarction. J Am Coll Cardiol 1998: 31: 1246–1251.
5. Christian TF, O’Connor MK, Schwartz RS et al. Technetium-99m MIBI to assess coronary collateral flow during acute myocardial infarction in two closed chest animal models. J Nucl Med 1997; 38: 1840–1846.
6. Eberli FR, Weinberg EO, Grice WN et al. 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: 466–481.
7. Cave A, Eberli FR, Ngoy S et al. Increased glycolytic substrate protects against ischemic diastolic dysfunction: 31P-NMR studies in the isolated blood perfused rat heart. Circulation 1993; 88 (suppl I): I-43.
8. Cave A, Ingwall JS, Friedrich J et al. ATP synthesis during low-flow ischemia: influence of increased glycolytic substrate. Circulation. In press. 
9. Dennis SC, Gevers W, Opie LH. Protons in ischemia: where do they come from; where do they go to? J Mol Cell Cardiol 1991; 23: 1077–1086.
10. 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: 515–526.
11. Malmberg K, Ryden L, Hamsten A et al. 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: 1337–1344.
12. Díaz R, Paolasso EC, Piegas LS et al., on behalf of the ECLA (Estudios Cardiologicos Latinoamerica) Collaborative Group. Metabolic modulation of acute myocardial infarction. The ECLA Glucose-Insulin-Potassium Pilot Trial. Circulation 1998; 98: 2227–2234.
13. Apstein CS. Glucose-insulin-potassium for acute myocardial infarction. Remarkable results from a new, prospective randomized trial. Circulation 1998; 98: 2223–2226.
14. Rackley CE, Russell RO, Rogers WJ et al. Clinical experience with glucose-insulin-potassium therapy in acute myocardial infarction. Am Heart J 1981; 102: 1038–1049.
15. Ceremuzynski L, Budaj A, Czepiel A et al., for Pol-GIK Trial Investigators. Low-dose glucose-insulin-potassium is ineffective in acute myocardial infarction. Results of randomized multicenter Pol-GIK trial. Cardiovasc Drugs Ther 1999; 13: 191–200.
16. Apstein CS, Opie L. Glucose-insulin-potassium (GIK) for acute myocardial infarction: a negative study with a positive value. Cardiovasc Drugs Ther 1999; 13: 185–189.
17. Fibrinolytic Therapy Trialists’ Collaborative Group. 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. Lancet 1994; 343: 311–322.