Author + information
- Jorik R. Timmer, MD, PhD,
- Tone Svilaas, MD,
- Jan Paul Ottervanger, MD, PhD⁎ (, )
- Jose P.S. Henriques, MD, PhD,
- Jan-Henk E. Dambrink, MD, PhD,
- Stan A.J. van den Broek, MD, PhD,
- Iwan C.C. van der Horst, MD, PhD and
- Felix Zijlstra, MD, PhD, FACC
- ↵⁎Isala Klinieken, Department of Cardiology, Groot Wezenland 20, 8011 JW Zwolle, the Netherlands
To the Editor:Metabolic support through glucose-insulin-potassium (GIK), in adjunct to reperfusion therapy, might improve outcome in patients with ST-segment elevation myocardial infarction (STEMI) (1). The Glucose-Insulin-Potassium Study (GIPS)-I showed clinical benefit of GIK in STEMI patients without signs of heart failure (2). To verify these findings we performed a second study investigating GIK in patients without signs of heart failure treated with reperfusion therapy.
It concerns a randomized, open label study (3); STEMI patients without signs of heart failure on admission were randomized to traditional care or additional GIK infusion. Glucose-potassium (20% glucose with 80 mmol potassium/l) was infused at 2 ml/kg body weight per hour for 12 h through a peripheral line. Short-acting insulin was started according to admission glucose and adjusted on the basis of hourly measured glucose.
Primary end point was 30-day mortality. Enzymatic infarct size was estimated by the highest value of creatine kinase (peak CK). Multivariate analyses (including all univariate predictors and GIK intervention) were performed to identify independent predictors of 30-day mortality (Cox regression) and high enzymatic infarct size (logistic regression). A sample size of 1,044 patients was planned, on the basis of the mortality reduction in non-heart-failure patients as reported in GIPS-I (1.2% vs. 4.2%) (2).
Patient inclusion started August 2003. After planned interim analysis, the study was terminated in December 2004, because a significant difference was unlikely to be reached on further inclusion.
A total of 889 patients were included (Table 1).Anterior infarct location was more common in the GIK group, and previous percutaneous coronary intervention (PCI) more common in the control group. All other general characteristics were comparable between the two treatment groups. Reperfusion therapy consisted of primary PCI (n = 778, 88%) or thrombolysis (n = 65). Mean starting insulin dose in the GIK group was 9.6 ± 7.5 U/h. During 30 days’ follow-up, 13 patients (2.9%) died in the GIK group and 8 patients (1.8%) died in the control group (odds ratio [OR] 1.6; 95% confidence interval [CI]: 0.7 to 4.0, p = 0.27) (Fig. 1).Multivariate analyses showed that failed reperfusion, anterior infarct location, and higher admission glucose but not GIK infusion (OR 1.5; 95% CI: 0.6 to 3.9) were independent predictors of 30-day mortality.
Mean peak CK was 2,008 ± 1,930 U/l in patients treated with GIK compared with 1,932 ± 1,847 U/l in control subjects (p = 0.57). The prevalence of a high enzymatic infarct size (CK levels in the upper tertile [>2,149 U/l]) was comparable between patients treated with GIK and control subjects (34% vs. 32%, p = 0.57). Multivariate analysis did not change these findings.
We could not confirm the previously observed benefit of GIK in STEMI patients without signs of heart failure. Because the sample size of the GIPS-II trial was based on the results of the GIPS-I trial, it was not powered to demonstrate smaller benefits of GIK; however, not even a trend toward a therapeutic benefit was observed in our trial. Combined with the results of other trials, it can be concluded that, in the used regimen, GIK in adjunct to reperfusion therapy in STEMI patients does not lower mortality.
Previous reports on the effect of GIK in STEMI have been contradictory. The meta-analysis of Fath-Ordoubadi and Beatt (4) showed a benefit of GIK infusion; however, studies included in this meta-analysis were of older date, and none of the patients received reperfusion therapy. Most studies published after this meta-analysis only showed benefit of GIK in specific subgroups. The Estudios Cardiologicos Latinoamerica (ECLA) study demonstrated mortality reduction in the sub-group treated with reperfusion therapy (5). The Polish-Glucose-Insulin-Potassium (Pol-GIK) trial was prematurely stopped, because mortality was significantly higher in GIK patients (6). The GIPS-I trial, including only patients treated with primary PCI, found a mortality reduction in patients without heart failure on admission (2). The Clinical Trial of Reviparin and Metabolic Modulation in Acute Myocardial Infarction Treatment Evaluation (CREATE)-ECLA study, with a sample size of 20,201 patients, showed only a trend toward a better outcome in the patients (n = 1,831) treated with primary PCI (7). The only study to date showing a clinical benefit of (glucose-) insulin was the Diabetes Mellitus Insulin-Glucose Infusion in Acute Myocardial Infarction (DIGAMI) study; however, this study included hyperglycemic patients and was primarily aimed at reduction of glucose levels (8).
There might be several explanations for the lack of effects of GIK in STEMI. Experimental evidence suggests a beneficial effect of GIK by shifting myocardial free fatty acid metabolism toward glucose metabolism. This is less oxygen-consuming and produces less toxic side products (1). Furthermore, reperfusion injury might be reduced by limiting apoptosis of post-ischemic myocardial cells; however, once adequate reperfusion has been established, GIK might be of limited value, because maximal myocardial salvage has already taken place. Furthermore, glycometabolic support might only be beneficial when it is aimed at lowering glucose levels in patients with severe derangement of glucose metabolism, as in the DIGAMI study (8). In addition, side effects of GIK infusion include hyperglycemia, hyperkalemia, and fluid overload. This might diminish the potential benefits.
Although there is no role for GIK in routine practice of patients with STEMI, there remain several issues to be clarified. Administering concentrated GIK through a central line might diminish adverse effects of fluid overload. Starting GIK before reperfusion therapy, during ambulance transportation for example, and GIK regimens specifically aimed at euglycemia should be further investigated (8,9).
For a list of the Steering Committee and the participating hospitals and cardiologists, please see the online version of this article.
Please note: The GIPS-II study was made possible by a generous grant of the Netherlands Heart Foundation and from Guidant.
- American College of Cardiology Foundation
- Apstein C.S.,
- Opie L.H.
- Van der Horst I.C.,
- Zijlstra F.,
- van’t Hof A.W.J.,
- et al.
- Fath-Ordoubadi F.,
- Beatt K.J.
- Diaz R.,
- Paolasso E.A.,
- Piegas L.S.,
- et al.
- Malmberg K.,
- Ryden L.,
- Efendic S.,
- et al.
- Malmberg K.,
- Ryden L.,
- Wedel H.,
- et al.