Prognostic Value of Blood Glucose in Patients with Cardiogenic Shock

Circ J 2006; 70: 1064–1069

Prognostic Value of Blood Glucose in Patients With Cardiogenic Shock

Katsushige Tada, MD; Ken Nagao, MD; Katsuhisa Tanjoh, MD; Nariyuki Hayashi, MD

Background Although an elevated blood glucose has prognostic value in cardiovascular disease, few data are available regarding its prognostic value for patients across the spectrum of cardiogenic shock. Method and Results A total of 81 patients with cardiogenic shock whose blood glucose and adrenaline were measured on arrival at the emergency room (ER) were enrolled in this prospective study. The primary endpoint was death from any cause in hospital. The rate of death was 12.3% (10/81), and the glucose level was lower among patients who were discharged alive than among those who died (8.7±3.7mmol/L vs 13.8±6.7mmol/L, p<0.001). The unadjusted rate of death increased in a stepwise fashion among patients in increasing quartiles of glucose level (p<0.05). The blood glucose level of 9.2mmol/L had the highest combined sensitivity and specificity for the identification of death. In the multiple logistic-regression analysis for the primary outcome, the adjusted odd ratio for a glucose level of 9.2mmol/L or more was 5.8 (95%confidence interval, 1.0–32.8, p=0.047). There was a significant positive correlation between the glucose and adrenaline levels (R=0.726, p<0.0001). Conclusion The measurement of blood glucose level on ER arrival provides predictive information for use in risk stratification across the spectrum of cardiac emergencies complicated by cardiogenic shock. (Circ J 2006; 70: 1064–1069) Key Words: Blood glucose; Cardiogenic shock; Prognosis

ver since Cruikshank et al reported elevated blood including LV dysfunction, right ventricular (RV) dysfunc- glucose levels in nondiabetic patients with acute tion, and serious arrhythmia. Therefore, we sought to eval- E myocardial infarction (AMI) in 1931,1 there have uate the prognostic implication of blood glucose level been reports of elevated blood glucose levels in a variety across the entire spectrum of cardiac emergencies compli- of critically ill patients.2–6 Hyperglycemia associated with cated by cardiogenic shock. insulin resistance7,8 is common in critically ill patients, even those who have not previously had diabetes. It has been reported that pronounced hyperglycemia may lead to Methods complications in such patients9,10 and increased risk of Patients death.2,11 Moreover, therapy to maintain the blood glucose A prospective clinical study was conducted. Patients level improves the outcome.2,12 Several studies have also who were transported directly by ambulance to the emer- reported a correlation between elevated blood glucose level gency room (ER) were enrolled when the following criteria on admission for AMI and the eventual outcome.13–16 A were fulfilled: age 18 years or older and cardiogenic shock, meta-analysis by Capes et al reported that outcomes were defined as (1) evidence of hypoperfusion in the field and/or poor in cases of stress hyperglycemia, even in AMI patients on ER arrival (cold clammy skin, especially the feet and without complications.17 In addition, a recent study showed hands, associated with peripheral cyanosis of the nail beds, there was a graded relation between elevated fasting glu- disordered mentation) and systolic blood pressure (BP) cose level and 30-day mortality in patients with AMI.18 <100mmHg in the field and/or ER arrival, (2) evidence of a However, few data are available for patients with AMI primary cardiac abnormality (LV dysfunction, RV dysfunc- complicated by cardiogenic shock caused by left ventricu- tion or serious arrhythmia). The diagnosis of cardiogenic lar (LV) dysfunction, although the SPRINT study identified shock was determined by attending cardiologists using blood glucose level at the time of admission as an indepen- clinical examination, electrocardiography, echocardio- dent prognostic factor for outcome in 89 patients with AMI graphy, X-ray and blood examination during the primary complicated by cardiogenic shock caused by LV dysfunc- care according to the American Heart Association (AHA) tion19 Furthermore, no data are available for patients with Guidelines.20,21 The hemodynamic findings were added to cardiac emergencies complicated by cardiogenic shock, the diagnosis when a pulmonary artery catheter was placed to guide therapy. (Received October 25, 2005; revised manuscript received April 26, Patients were excluded if they met any of the following 2006; accepted May 10, 2006) criteria: cardiac arrest prior to arrival at the ER, shock from Departments of Emergency and Critical Care Medicine, Nihon Uni- noncardiac causes (aortic disease, pulmonary embolism, versity School of Medicine, Tokyo, Japan Mailing address: Katsushige Tada, MD, Departments of Emergency drug overdose, exsanguination, cerebrovascular disease, and Critical Care Medicine, Nihon University Surugadai Hospital, 1-8- trauma, etc), blood samples could not be taken before 13 Kanda-Surugadai, Chiyoda-ku, Tokyo 101-8309, Japan. E-mail: administration of any medication, history or treatment for [email protected] diabetes mellitus before shock, glycol-hemoglobin (HbA1c)

Circulation Journal Vol.70, August 2006 Blood Glucose in Cardiogenic Shock 1065 on arrival at the ER of 5.9% or more, chronic renal failure with hemodialysis before shock, or refusal by patient or families to participate in the study.

Treatment of Cardiogenic Shock Treatment was given in accordance with the guidelines 2000 for the management of cardiogenic shock.20 In cases of shock caused by a volume problem, such as RV infarction, large volumes of intravenous fluid were given rapidly, supplemented with vasopressors if circulatory collapse could not be corrected. In cases of shock because of a pump problem with LV dysfunction, vasopressors such as norepi- nephrine and dopamine were administered with reference to the systolic BP. In cases of shock because of a rate problem, tachyarrhythmias were treated promptly with an anti- arrythmic drug or electric defibrillation. Bradyarrhythmias were treated with atropine and/or transvenous pacing if Fig1. Blood glucose levels of all patients (Left), patients surviving circulatory collapse could not be corrected. If circulatory to hospital discharge (Middle), and patients who died (Right). Box show interquartile range, and “I” bars represent 10th and 90th per- collapse could not be corrected through these measures, centile values (*p<0.001 for the comparison between patients who then an assisted circulation device, such as intra-aortic survived to hospital discharge and those who died in hospital). balloon pumping or cardiopulmonary bypass, was used. Coronary reperfusion therapy was also initiated for any underlying conditions. If there were no contraindications, a according to quartiles of blood glucose level. Cut-off bolus intravenous injection of mutant tissue-type plasmino- values to differentiate survival and death were calculated gen activator was commenced as soon as possible after ER for blood glucose levels using receiver-operating character- arrival in cases of AMI with ST elevation or new bundle- istics (ROC) curves. A multiple logistic regression analysis branch block, then emergency coronary angiography for was performed for independent predictors of the primary coronary reperfusion therapy was performed following outcome of death in hospital, including age, gender, systolic intravenous thrombolysis. If TIMI grade22 0, 1 or 2 flow BP (≤70mmHg vs 71–90mmHg vs ≥91mmHg),20,21 primary was observed in the infract-related artery, emergency problems of shock (rate problem vs pump problem vs percutaneous coronary intervention (PCI) was immediately volume problems),20,21 and blood glucose level (<9.2mmol/L performed.23,24 vs ≥9.2 mmol/L). Finally, the relationship between the If hemodynamic improvement was not seen in cases of blood glucose and adrenaline levels was analyzed using acute coronary syndrome (ACS) with ST depression, emer- simple regression analysis. Values are expressed as mean± gency coronary angiography was performed immediately, 2SD, with p<0.05 as the level of statistical significance. followed by PCI or coronary arterial bypass graft surgery if necessary. If hemodynamic improvement did occur in cases of ACS with ST depression, coronary angiography Results was performed 12–48h later, followed by PCI or coronary Patient Characteristics arterial bypass graft surgery if necessary. Of the 2,801 patients who were transported by ambulance to the ER during the 4-year period from January 2001 Data Collection and Study Endpoints to December 2004, 1,154 were cardiogenic emergencies, On arrival at the ER, consent for emergency care was ob- including 140 that developed into cardiogenic shock. Of the tained from the patient, and blood taken for measurement 140 patients with cardiogenic shock, 81 met the criteria and of blood glucose, HbA1c and adrenaline. The immobilized were enrolled. The blood glucose level on ER arrival ranged glucose oxidase membrane method (GA-1160, Arkray Co, from 2.4 to 26.1 mmol/L, with a mean (± SD) of 9.3± Kyoto, Japan) was used for determination of blood glucose 4.4mmol/L, a median of 8.1mmol/L, 25th percentile value levels, and high-performance liquid phase chromatography of 6.7mmol/L, and 75th percentile value of 10.1mmol/L for determination of HbA1c and adrenaline levels (HbA1c: (Fig1). There were 2 patients whose blood glucose level HA-8160, Arkray Co; adrenaline: L7485, Hitachi High- was less than 3.8mmol/L: 1 had a long time interval from Technologies Ltd, Tokyo, Japan). The emergency life- the onset of AMI to blood sample collection, and had not saving technicians were asked to record the condition of eaten and drunk after the onset of AMI, and the other had a each patient up until arrival at the ER, including vital signs. medical history of gastrectomy for gastric cancer. The primary endpoint was death from any cause in hospital. Table1 shows the baseline characteristics of the patients The secondary endpoint was the cause of death. according to the quartile of blood glucose level. No significant differences were observed among the 4 groups in age, Data Analysis gender, systolic BP at the scene and on ER arrival, cause of Patients were divided into 4 groups according to quartiles shock, HbA1c levels, and primary problems of shock, of blood glucose level on ER arrival. The mean values and although there was a significant difference among the proportions of the baseline variables were compared among groups in pH at the initial arterial blood gas analysis. the 4 groups using a one-way analysis of variance for continuous variables, and the chi-square test for categorical Outcomes variables. The chi-square test was also used to assess the The rate of death in hospital was 12.3% (10/81) and the relationship between blood glucose level and the study blood glucose level was significantly higher in patients endpoints in patients who were divided into 4 groups who died than in those who survived to hospital discharge

Circulation Journal Vol.70, August 2006 1066 TADA K et al.

Table 1 Baseline Characteristics According to the Quartile of Blood Glucose Level

Quartile-I Quartile-II Quartile-III Quartile-IV p value (2.4–6.7 mmlo/L) (6.8–8.1 mmlo/L) (8.1–10.1 mmlo/L) (10.2–26.1 mmlo/L) for trend No. of patients 20 21 20 20 Age, years 67.0±12.3 61.0±10.9 62.2±12.2 65.6±13.8 0.30 Male sex, % 85% 76.2% 85% 75% 0.39 Systolic blood pressure, mmHg At scene 90.7±23.3 100.7±22.5 84.9±18.6 83.9±21.5 0.16 On ER arrival 93.1±29.2 95.4±10.4 94.0±22.6 89.2±24.3 0.09 Cause of shock, % ACS 80% 85.7% 90% 75% 0.13 HbA1c, % 5.12±0.36 5.26±0.29 5.06±0.31 5.39±0.32 0.76 Primary cause of shock, % Pump 50% 66.7% 35% 45% Volume 20% 19.0% 60% 20% 0.60 Rate 30% 14.3% 5% 35% Initial arterial pH 7.41±0.1 7.42±0.06 7.40±0.07 7.32±0.20 0.003

ER, emergency room; ACS, acute coronary syndrome; HbA1c, glycol-hemoglobin.

Fig2. Association between the blood glucose level and the primary endpoint of death in hospital in all patients and in selected subgroups with pump problems and those with acute coronary syndrome (ACS). The range of blood glucose levels was: 2.4–6.7mmol/L (quartile 1), 6.8–8.1mmol/L (quartile 2), 8.1–10.1mmol/L (quartile 3), 10.2–26.1mmol/L (quartile 4). p values are for the trend within each group.

(13.8±6.7mmol/L vs 8.7±3.7mmol/L, p=0.0004) (Fig1). (Fig3). The area under the ROC curve when blood glucose The primary endpoint of death in all patients increased in level was used to differentiate death from survival was 0.75 a stepwise fashion across the increasing quartiles of blood (95%confidence interval (CI) 0.58–0.93; p=0.01). A blood glucose level (with quartile 1 at 5% vs quartile 2 at 5% vs glucose level of 9.2mmol/L had a highest combined sen- quartile 3 at 10% vs quartile 4 at 30%, p=0.047). This asso- sitivity and specificity for identifying death (sensitivity of ciation remained significant in the subgroups of patients 70%, specificity of 77.5%, accuracy of 76.5% for differen- whose shock was caused by a pump problem or ACS tiating death from survival). In addition, higher values were (Fig2), although no significant difference was observed in associated with more accurate positive predictive values for the primary endpoint among each subgroup of patients differentiating death from survival (for a blood glucose level whose shock was caused by a volume problem, rate prob- of 21.6mmol/L, the positive predictive value was 100%). In lem or non-ACS. a multivariate logistic regression analysis for independent The secondary endpoint of cause of death was as fol- predictors of death, including age, gender, systolic BP on lows. Each patient in quartile 1 and 2 had a cardiac rupture ER arrival (≤70mmHg vs 71–90mmHg vs ≥91mmHg), and complicating AMI on day 2 and day 3, respectively. Of the primary problem of shock (rate, pump, volume), blood glu- 2 deaths in quartile 3, 1 was from cerebral infarction on day cose level (<9.2mmol/L vs ≥9.2mmol/L), a blood glucose 1, and the other from LV dysfunction complicating AMI on level of 9.2mmol/L or more was the strongest prognostic day 1. Of the 6 deaths in quartile 4, 3 were from LV dys- indicator of death, with an adjusted odds ratio of 5.80 (95% function complicating AMI on day 1, 1 was refractory low CI 1.03–32.8, p=0.047). Age, gender, systolic BP on ER output syndrome on day 5, 1 was papillary muscle rupture arrival, and the primary problem of shock were not inde- complicating AMI on day 1, and 1 was cardiac rupture pendent predictors of death (Fig4). When the initial arterial complicating AMI on day 1. No significant difference was pH after ER arrival was entered into the multiple logistic observed among the 4 groups for causes of death. analysis, the blood glucose level remained unchanged.

Cut-off Value of Blood Glucose Level for Death Blood Glucose Level and Adrenaline Concentration The capacity of blood glucose level to differentiate death The relationship between the blood glucose level and the from survival was assessed with a ROC curve analysis plasma adrenaline concentrations is shown in Fig5. There

Circulation Journal Vol.70, August 2006 Blood Glucose in Cardiogenic Shock 1067

1.00

0.75

Sensitivity Blood glucose = 8.2 mmol/L Blood glucose = 9.2 mmol/L 0.50 Blood glucose = 11.4 mmol/L Blood glucose = 12.1 mmol/L Area under curve = 0.75 0.25 (95%CI 0.58–0.93)

Blood glucose = 21.6 mmol/L 0.00 0.00 0.25 0.50 0.75 1.00 1-Specificity Blood glucose Positive Negative Sensitivity Specificity Accuracy level (mmol/L) predictive value predictive value 8.2 80 57.7 21.1 95.3 60.5 9.2 70 77.5 30.4 94.8 76.5 Fig3. Receiver-operating characteristics curve for cutoff 11.4 60 87.3 40.0 93.9 82.9 levels of blood glucose to differentiate survival and death. 12.1 50 90.1 41.7 92.8 85.2 CI, conﬁdence interval.

Survival Death Odds 0110 ratio 95%CI

· Age 1.01 (0.94–1.08) · Male sex 0.39 (0.06–2.56)

· Systolic blood pressure at ER ≤70 mmHg (reference) 71–90 mmHg 0.97 (0.13–7.19) ≥91 mmHg 0.31 (0.04–2.53) · Primary problem of shock Rate (reference) Pump 2.43 (0.18–33.56) Volume 6.85 (0.48–97.14) Fig4. Adjusted odds ratios for death (primary endpoint) according to selected factors, from the multiple logistic · Blood glucose ≥9.2 mmol/L 5.80 (1.03–32.8) regression analysis, Horizontal lines are 95%conﬁdence intervals (CI). ER, emergency room. was a significant positive correlation between them in patients with cardiogenic shock on ER arrival (R=0.726, p<0.0001).

Discussion We evaluated the prognostic implication of stress hyperglycemia across the entire spectrum of cardiogenic shock. In all types of cardiogenic shock, (volume, pump or rate problems), the blood glucose level on the ER arrival was an independent prognostic factor for the primary outcome of death in hospital. The rate of death increased in a stepwise fashion among the patients in increasing quartiles of the blood glucose level on ER arrival. The optimal cut-off point for death was a blood glucose level of 9.2mmol/L and none of the patients with a blood glucose level 21.6mmol/L or Fig5. Relationship between blood glucose level and adrenaline con- more survived to hospital discharge. Furthermore, a blood centration in patients with cardiogenic shock on arrival at the emer- glucose levels of 9.2mmol/L or more was the strongest gency room. prognostic indicator of death. Blood glucose levels also showed a significant positive correlation with levels of adrenaline, one of the stress hormones. These results show The AHA guideline reported that cardiogenic shock con- that the blood glucose level on ER arrival provides power- sisted of volume problems caused by RV dysfunction, ful information for use in risk stratiﬁcation across the entire pump problems caused by LV dysfunction, and rate prob- spectrum of cardiogenic shock. lems associated with tachyarrhythmias or bradyarrhythmias

Circulation Journal Vol.70, August 2006 1068 TADA K et al. without ACS, and that the primary care with hemodynamic Correction of hyperglycemia through early administration management differed from the primary problem.20,21 In the of insulin may also improve the outcome in patients with present study, hemodynamic management was performed cardiogenic shock by inhibiting the activation of the inflam- in accordance with the initial treatment manual according matory response, progression of cell apoptosis, and platelet to the AHA guideline.20,21 The severity of shock on ER activation. We did not investigate the relationship between arrival is generally assessed by the degree of systolic BP blood glucose control and cell damage, cytokines and and through arterial blood gas analysis. In this study, there immune function, or platelet activation, although blood was a significant difference among the 4 groups in the glucose levels were controlled to less than 11.1mmol/L initial arterial pH after ER arrival, although no significant during intensive care. difference was observed among the 4 groups in systolic BP. The pH is a useful marker of the severity of shock, but Study Limitations its measurement has several problems. When a patient is in First, this study was conducted at a single institution shock, it takes some time to collect an arterial blood with a small number of patients with cardiogenic shock. sample, there is a considerable risk of collecting a venous A prospective multicenter study of cardiogenic shock is blood sample in error, and furthermore some time will needed. Second, we were unable to thoroughly examine the elapse before results will be available. Furthermore, some relationship between blood glucose level and the time medical facilities are unable to examine arterial blood gas interval from the onset of cardiac disease to blood sample analysis at any time. On the other hand, blood glucose collection, or the relationship between the blood glucose determinations using whole venous blood are easily per- level and the time interval from the onset of shock to blood formed, and the results can be obtained instantly at the sample collection, because those time intervals may con- bedside in most medical facilities. Therefore, we evaluated tribute to hyperglycemia. Third, blood glucose level was the blood glucose level as a marker of the risk of death in measured in whole blood. If blood glucose is measured patients with cardiac emergencies complicated by cardio- using plasma, the cut-off point and upper limit of the blood genic shock. glucose level for death might change slightly. Finally, There are 2 mechanisms by which hyperglycemia might although the patients in this study were treated according to be an independent factor in determining the outcome of the standard guidelines,20,21 the precise managements of cardiogenic shock. One is the concept that blood glucose blood glucose level and body temperature may have influ- level increases according to the degree of biological stress enced outcomes. and is associated with the risk of complications and death. In conclusion, determination of blood glucose level When an organism experiences stress, the stress hormones, using venous blood on arrival at the ER can be performed including catecholamines, cortisol, and various inflamma- quickly and easily, and provides predictive information for tory cytokines, are secreted25 and promote glycogenolysis use in risk stratification across the spectrum of cardiogenic and gluconeogenesis, thereby elevating the blood glucose shock. level.26 Growth hormone and free fatty acids also induce insulin resistance, resulting in stress hyperglycemia.9,27 In References addition, stress hyperglycemia has been reported to pro- 28 1. Cruikshank N. Coronary thrombosis and myocardial infarction with mote platelet activation, impairing the microcirculation glycosuria. BMJ 1931; 1: 618–619. 29 and exacerbating cell damage. In the present study, the 2. Van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx result that blood glucose levels correlated positively with F, Schetz M, et al. Intensive insulin therapy in critically ill patients. adrenaline levels, and were associated with the risk of N Engl J Med 2001; 345: 1359–1367. death reflect this mechanism. 3. Laird AM, Miller PR, Kilgo PD, Meredith JW, Chang MC. Rela- tionship of early hyperglycemia to mortality in trauma patients. J The other possible mechanism is that hyperglycemia Trauma 2004; 56: 1058–1062. itself causes cell damage. Although long-term elevation of 4. Holm C, Horbrand F, Mayr M, Von Donnersmarck GH, Muhlbauer blood glucose in diabetics has been established to cause dia- W. Acute hyperglycaemia following thermal injury: Friend or foe? betic cardiomyopathy,30–32 patients with diabetes mellitus Resuscitation 2004; 60: 71–77. 5. Van der Horst IC, Zijlstra F, van’t Hof AW, Doggen CJ, de Boer MJ, were excluded from the current study. Recent studies have Suryapranata H, et al; Zwolle Infarct Study Group. Glucose-insulin- shown that acute hyperglycemia induced apoptosis in per- potassium infusion in patients treated with primary angioplasty for fused rat hearts,33 and activation of inflammatory cytokines acute myocardial infarction. J Am Coll Cardiol 2003; 42: 784–791. and the immune system, which might cause further damage 6. Furnary AP, Gao G, Grunkemeier GL, Wu Y, Zerr KJ, Bookin SO, 34 et al. Continuous insulin infusion reduces mortality in patients with to the myocardium. Miyazaki et al reported that the coro- diabetes undergoing coronary artery bypass grafting. J Thorac Car- nary blood flow was increased when the blood glucose diovasc Surg 2003; 125: 1007–1021. level was managed after the onset of AMI.35 We were una- 7. Wolfe RR, Allsop JR, Burke JF. Glucose metabolism in man: Re- ble to prove that hyperglicemia itself was associated with sponses to intravenous glucose infusion. Metabolism 1979; 28: 210– 220. myocardial damage. 8. Shangraw RE, Jahoor F, Miyoshi H, Neff WA, Stuart CA, Herndon Many reports have confirmed that following severe DN, et al. Differentiation between septic and postburn insulin resis- stress, such as cardiac surgery, trauma or burns, correction tance. Metabolism 1989; 38: 983–989. of hyperglycemia through the administration of insulin 9. McCowen KC, Malhotra A, Bistrian BR. Stress-induced hyper- alone or insulin and potassium, improves outcomes.2–5,36 glycemia. Crit Care Clin 2001; 17: 107–124. 10. Scott JF, Robinson GM, French JM, O’Connell JE, Alberti KG, Gray The treatment of hyperglycemia in patients with cardiac CS. Glucose potassium insulin infusions in the treatment of acute disease also improves outcomes.37 Insulin has a beneficial stroke patients with mild to moderate hyperglycemia: The Glucose effect by suppressing inflammatory cytokines released as Insulin in Stroke Trial (GIST). Stroke 1999; 30: 793–799. part of the activation of the inflammatory response asso- 11. Van den Berghe G, Wouters P, Weekers F, Mohan S, Baxter RC, 38,39 Veldhuis JD, et al. Reactivation of pituitary hormone release and ciated with stress hyperglycemia. Insulin has been metabolic improvement by infusion of growth hormone-releasing reported to also inhibit activated oxygen,40 and to correct peptide and thyrotropin-releasing hormone in patients with protracted the abnormal mitochondrial function of hepatic cells.41 critical illness. J Clin Endocrinol Metab 1999; 84: 1311–1323.

Circulation Journal Vol.70, August 2006 Blood Glucose in Cardiogenic Shock 1069

12. Malmberg K, Norhammar A, Wedel H, Ryden L. Glycometabolic 25. Marfella R, Siniscalchi M, Esposito K, Sellitto A, De Fanis U, state at admission: Important risk marker of mortality in convention- Romano C, et al. Effects of stress hyperglycemia on acute myocar- ally treated patients with diabetes mellitus and acute myocardial dial infarction: Role of inflammatory immune process in functional infarction: Long-term results from the Diabetes and Insulin-Glucose. cardiac outcome. Diabetes Care 2003; 26: 3129–3135. Circulation 1999; 99: 2626–2632. 26. Wolfe RR, Herndon DN, Jahoor F, Miyoshi H, Wolfe M. Effect of 13. Hadjadj S, Coisne D, Mauco G, Ragot S, Duengler F, Sosner P, et al. severe burn injury on substrate cycling by glucose and fatty acids. N Prognostic value of admission plasma glucose and HbA1c in acute Engl J Med 1987; 317: 403–408. myocardial infarction. Diabet Med 2004; 21: 305–310. 27. Mizock BA. Alterations in carbohydrate metabolism during stress: A 14. Suleiman M, Hammerman H, Boulos M, Kapeliovich MR, Suleiman review of the literature. Am J Med 1995; 98: 75–84. A, Agmon Y, et al. Fasting glucose is an important independent risk 28. Gresele P, Guglielmini G, De Angelis M, Ciferri S, Ciofetta M, factor for 30-day mortality in patients with acute myocardial infarc- Falcinelli E, et al. Acute, short-term hyperglycemia enhances shear tion: A prospective study. Circulation 2005; 111: 754–760. stress-induced platelet activation in patients with type II diabetes 15. Zeller M, Cottin Y, Brindisi MC, Dentan G, Laurent Y, Janin- mellitus. J Am Coll Cardiol 2003; 41: 1013–1020. Manificat L, et al; The RICO survey working group. Impaired fasting 29. Iwakura K, Ito H, Ikushima M, Kawano S, Okamura A, Asano K, et glucose and cardiogenic shock in patients with acute myocardial al. Association between hyperglycemia and the no-reflow phenome- infarction. Eur Heart J 2004; 25: 308–312. non in patients with acute myocardial infarction. J Am Coll Cardiol 16. Kosuge M, Kimura K, Kojima S, Sakamoto T, Matsui K, Ishihara M, 2003; 41: 1–7. et al. Effects of glucose abnormalities on in-hospital outcome after 30. Fiordaliso F, Bianchi R, Staszewsky L, Cuccovillo I, Doni M, coronary intervention for acute myocardial infarction. Circ J 2005; Laragione T, et al. Antioxidant treatment attenuates hyperglycemia- 69: 375–379. induced cardiomyocyte death in rats. J Mol Cell Cardiol 2004; 37: 17. Capes SE, Hunt D, Malmberg K, Gerstein HC. Stress hyperglycaemia 959–968. and increased risk of death after myocardial infarction in patients 31. Cai L, Li W, Wang G, Guo L, Jiang Y, Kang YJ. Hyperglycemia-in- with and without diabetes: A systematic overview. Lancet 2000; duced apoptosis in mouse myocardium: Mitochondrial cytochrome 355: 773–778. C-mediated caspase-3 activation pathway. Diabetes 2002; 51: 1938– 18. Suleiman M, Hammerman H, Boulos M, Kapeliovich MR, Suleiman 1948. A, Agmon Y, et al. Fasting glucose is an important independent risk 32. Xu G, Takashi E, Kudo M, Ishiwata T, Naito Z. Contradictory effects factor for 30-day mortality in patients with acute myocardial infarc- of short- and long-term hyperglycemias on ischemic injury of myo- tion: A prospective study. Circulation 2005, 111: 754–760. cardium via intracellular signaling pathway. Exp Mol Pathol 2004; 19. Leor J, Goldbourt U, Reicher-Reiss H, Kaplinsky E, Behar S. Car- 76: 57–65. diogenic shock complicating acute myocardial infarction in patients 33. Ceriello A, Quagliaro L, D’Amico M, Di Filippo C, Marfella R, without heart failure on admission: Incidence, risk factors, and out- Nappo F, et al. Acute hyperglycemia induces nitrotyrosine formation come: SPRINT Study Group. Am J Med 1993; 94: 265–273. and apoptosis in perfused heart from rat. Diabetes 2002; 51: 1076– 20. The American Heart Association in Collaboration with the Interna- 1082. tional Liaison Committee on Resuscitation. Guidelines 2000 for 34. Kosuge M, Kimura K, Ishikawa T, Shimizu T, Hibi K, Toda N, et al. Cardiopulmonary Resuscitation and Emergency Cardiovascular Persistent hyperglycemia is associated with left ventricular dysfunc- Care: An international consensus on science. Circulation 2000; 102: tion in patients with acute myocardial infarction. Circ J 2005; 69: I-1–I-390. 23–28. 21. Antman EM, Anbe DT, Armstrong PW, Bates ER, Green LA, Hand 35. Miyazaki C, Takeuchi M, Yoshitani H, Otani S, Sakamoto K, M, et al; American College of Cardiology; American Heart Associa- Yoshikawa J. Optimum hyperglycemic therapy can improve coro- tion Task Force on Practice Guidelines. ACC/AHA guidelines for nary flow velocity reserve in diabetic patients: Demonstration by the management of patients with ST-elevation myocardial infarction. transthoracic Doppler echocardiography. Circ J 2003; 67: 945–950. Executive summary: A report of the American College of Cardiolo- 36. Pittas AG, Siegel RD, Lau J. Insulin therapy for critically ill hospital- gy/American Heart Association Task Force on Practice Guidelines ized patients: A meta-analysis of randomized controlled trials. Arch (Writing Committee to Revise the 1999 Guidelines for the Manage- Intern Med 2004; 164: 2005–2011. ment of Patients With Acute Myocardial Infarction). Circulation 37. Mehta SR, Yusuf S, Diaz R, Zhu J, Pais P, Xavier D; CREATE- 2004; 110: 588–636. ECLA Trial Group Investigators. Effect of glucose-insulin-potassium 22. Chesebro JH, Knatterud G, Roberts R, Borer J, Cohen LS, Dalen J, infusion on mortality in patients with acute ST-segment elevation et al. Thrombolysis in Myocardial Infarction (TIMI) Trial. Phase I: A myocardial infarction: The CREATE-ECLA randomized controlled comparison between intravenous tissue plasminogen activator and in- trial. JAMA 2005; 293: 437–446. travenous streptokinase: Clinical findings through hospital discharge. 38. Johan Groeneveld AB, Beishuizen A, Visser FC. Insulin: A wonder Circulation 1987; 76: 142–154. drug in the critically ill? Crit Care 2002; 6: 102–105. 23. Nagao K, Hayashi N, Kanmatsuse K, Kikuchi S, Ohuba T, Takahashi 39. Das UN. Is insulin an endogenous cardioprotector? Crit Care 2002; H; FAST Trial. An early and complete reperfusion strategy for acute 6: 389–393. myocardial infarction using fibrinolysis and subsequent transluminal 40. Bertuglia S, Giusti A, Fedele S, Picano E. Glucose-insulin-potassium therapy: The FAST trial. Circ J 2002; 66: 576–582. treatment in combination with dipyridamole inhibits ischaemia-re- 24. Watanabe K, Nagao K, Watanabe I, Kanmatsuse K. Relationship be- perfusion-induced damage. Diabetologia 2001; 44: 2165–2170. tween the door-to-TIMI-3 flow time and the infarct size in patients 41. Vanhorebeek I, De Vos R, Mesotten D, Wouters PJ, De Wolf-Peeters suffering from acute myocardial infarction: Analysis based on the C, Van den Berghe G. Protection of hepatocyte mitochondrial ultra- fibrinolysis and subsequent transluminal (FAST-3) trial. Circ J 2004; structure and function by strict blood glucose control with insulin in 68: 280–285. critically ill patients. Lancet 2005; 365: 53–59.

Circulation Journal Vol.70, August 2006