The Immediate Myocardial Metabolic Enhancement During Initial Assessment and Treatment in Emergency care Trial of very early intravenous glucose-insulin-potassium (GIK) for acute coronary syndromes (ACS) in out-of-hospital emergency medical service (EMS) settings showed 80% reduction in infarct size at 30 days, suggesting potential longer-term benefits. Here we report 1-year outcomes. Prespecified 1-year end points of this randomized, placebo-controlled, double-blind, effectiveness trial included all-cause mortality and composites including cardiac arrest, mortality, or hospitalization for heart failure (HF). Of 871 participants randomized to GIK versus placebo, death occurred within 1 year in 11.6% versus 13.5%, respectively (unadjusted hazard ratio [HR] 0.83, 95% confidence interval [CI] 0.57 to 1.23, p = 0.36). The composite of cardiac arrest or 1-year mortality was 12.8% versus 17.0% (HR 0.71, 95% CI 0.50 to 1.02, p = 0.06). The composite of hospitalization for HF or mortality within 1 year was 17.2% versus 17.2% (HR 0.98, 95% CI 0.70 to 1.37, p = 0.92). The composite of mortality, cardiac arrest, or HF hospitalization within 1 year was 18.1% versus 20.4% (HR 0.85, 95% CI 0.62 to 1.16, p = 0.30). In patients presenting with suspected ST elevation myocardial infarction, HRs for 1-year mortality and the 3 composites were, respectively, 0.65 (95% CI 0.33 to 1.27, p = 0.21), 0.52 (95% CI 0.30 to 0.92, p = 0.03), 0.63 (95% CI 0.35 to 1.16, p = 0.14), and 0.51 (95% CI 0.30 to 0.87, p = 0.01). In patients with suspected acute coronary syndromes, serious end points generally were lower with GIK than placebo, but the differences were not statistically significant. However, in those with ST elevation myocardial infarction, the composites of cardiac arrest or 1-year mortality, and of cardiac arrest, mortality, or HF hospitalization within 1 year, were significantly reduced.
The Immediate Myocardial Metabolic Enhancement During Initial Assessment and Treatment in Emergency care (IMMEDIATE) Trial was a randomized, placebo-controlled, double-blind, clinical effectiveness trial of the administration of intravenous glucose-insulin-potassium (GIK) very early in the evolution of acute coronary syndromes (ACSs), by emergency medical service (EMS) paramedics, run in 13 cities across the United States. Before the IMMEDIATE Trial, experimental and clinical studies had shown GIK to protect myocardium from ischemic injury and lessen infarct size and thereby to preserve left ventricular (LV) function, but clinical trials of the use of GIK for patients seen in hospitals for established acute myocardial infarction (AMI) had not generally shown these benefits. In the IMMEDIATE Trial, rather than awaiting hospital diagnosis of AMI or ST elevation myocardial infarction (STEMI) to initiate GIK, as done in previous trials, GIK was delivered very early in the course of ACS, in the out-of-hospital setting. Paramedics, using electrocardiograph-based support by the acute cardiac ischemia time-insensitive predictive instrument (ACI-TIPI) and the thrombolytic predictive instrument (TPI), initiated GIK as soon as possible in the community, and the infusion was continued in the receiving hospital for 12 hours total. At 30 days, although the reduction in the primary end point of progression to AMI was not statistically significant, there were significant reductions in the composite end point of cardiac arrest or mortality and an 80% reduction in infarct size, suggesting possible longer-term benefits. One-year follow-up of trial participants for prespecified study end points, including mortality, heart failure (HF), and the composite of cardiac arrest, mortality, or HF, are reported here to address the degree to which such benefits might be sustained.
Methods
This study was a double-blind, randomized, controlled clinical effectiveness trial of intravenous GIK; its design and 30-day results, including the primary study end point, have been previously published. Here we report 1-year results on the prespecified clinical end points: all-cause mortality, the composite end point of mortality or hospitalization for HF within 1 year, and the composite end point of cardiac arrest, mortality, or HF hospitalization within 1 year.
Analyses were prespecified to be examined for the cohort defined by intent-to-treat for consenting participants with presumed ACS and for a subset of the cohort presenting with suspected STEMI.
From December 2006 through July 2011, paramedics in 36 EMS systems in 13 cities across the United States evaluated for study enrollment all patients aged ≥30 for whom a 12-lead electrocardiogram (ECG) was obtained for chest pain or other possible ACS symptoms. Paramedics were aided by electrocardiograph-based software that printed the ACI-TIPI patients’ predicted probabilities of ACS and that determined whether the TPI detected a likely STEMI. Paramedics were instructed to enroll patients who had an ACI-TIPI predicted probability of having ACS of ≥75%, had STEMI identified by the TPI, and/or who would have been identified as having STEMI by the community’s STEMI alert system. Excluded were patients with HF evidenced by more than basilar rales, those on dialysis for renal failure, or those unable to give assent. Assignment to study group was random by paramedic initiation of the blinded identical-appearing GIK or placebo study drug infusion packets.
As detailed elsewhere, the trial used exception from informed consent requirements for emergency research processes as in the Code of Federal Regulations (21CFR 50.24). This process included community consultation and notification, institutional review board approval, and having the paramedic read an information card to the patients before randomization and then asking for assent; full written consent was obtained once stabilized at the hospital. Oversight was provided by a Data and Safety Monitoring Board appointed by the National Heart, Lung, and Blood Institute; there were no interim efficacy analyses. Funding was from the National Heart, Lung, and Blood Institute; insulin was donated by Eli Lilly and Company (Indianapolis, Indiana); ACI-TIPI or TPI defibrillator-electrocardiograph software was provided by the manufacturers of EMS systems’ equipment, Philips Healthcare (Andover, Massachusetts), Physio-Control Inc. (Redmond, Washington), and Zoll Medical Corporation (Chelmsford, Massachusetts); no restrictions were imposed on study procedures or publication.
The study drug GIK solution had 30% glucose (300 g/L), 50 U/L of regular insulin, and 80 mEq of potassium chloride per liter, which had been shown to improve myocardial perfusion. The placebo was a 5% glucose solution in identical packaging. Per the study protocol, the study solution was administered at 1.5 ml/kg/hour (approximately 100 ml/hour for a patient weighing 70 kg) for 12 hours intravenously by way of an infusion pump.
Collected were demographic (including race and ethnicity by self-report) and presenting data, including detailed information on EMS, emergency department (ED), and hospital care (including ECGs, cardiac biomarkers, cardiac catheterization, and other tests pertaining to ACS). For safety purposes, as reported elsewhere, glucose and potassium blood tests were performed on ED arrival, at 6 hours after the start of the study drug, and once the study infusion was stopped. Also reported elsewhere are data collected on biological mechanism cohort participants.
A Clinical Events Committee adjudicated final diagnoses and the causes of all hospitalization end points used for analyses. The Committee members assigned diagnoses of AMI (including by Killip class), unstable angina pectoris (including by Canadian Cardiovascular Society Angina Classification), non-ACS cardiac disease, and noncardiac disease, based on out-of-hospital, ED, and 24-hour ECGs, biomarkers, and clinical data. In their review, they were blinded to study group, glucose and potassium tests, and whether the study infusion was stopped early. The analytical cohort of patients presenting with suspected STEMI was identified by having 3 cardiologists (blinded to study group) independently read the initial (out-of-hospital) ECG and determine whether the patient was sufficiently likely to have experienced an STEMI to deserve immediate referral for cardiac catheterization and reperfusion.
All analyses were done on the intent-to-treat cohort with treatment assignment as randomized. In addition, primary analyses were conducted for the subcohort of those patients presenting with STEMI, as defined previously. Analyses also were conducted on a modified intent-to-treat cohort, comprising those in the intent-to-treat cohort whom ED physicians confirmed as having ACS and therefore continued on the study drug, as GIK presumably would be used in clinical practice.
As described in the report of the 30-day outcomes, the primary end point of the IMMEDIATE Trial was progression to myocardial infarction based on biomarkers and ECGs. It was calculated that a sample size of 800 evaluable study participants would provide 90% power to detect a relative 20.5% reduction from 55.7% to 44.3% between the placebo and GIK groups for this end point. To accommodate attrition, 880 study participants were planned for randomization.
Time-to-event outcomes were analyzed using Cox proportional hazards regressions and the assumption of proportional hazards checked using a Kolmogorov-type supremum test. Robust variance estimators were used to account for potential clustering across multiple enrollments by individual participant’s time-to-event analyses. For composite events, the time to the first occurrence of any component of the composite was used. All statistical testing used 2-sided 0.05 level of significance and were done using SAS software, version 9.2 (SAS Institute Inc., Cary, North Carolina).
Results
Of 911 patients randomized, there were 871 enrollments for 850 subjects, as shown in Figure 1 . (Of those randomized, 40 subjects initially agreed to have the study drug started in the ambulance but later declined written informed consent at the hospital and were excluded: 21 GIK group participants and 19 placebo group participants.) Enrollments were used as the unit of analysis. Demographic and clinical features are listed in Table 1 . The average age was 63 years, 71% were men, and 86% presented with a chief complaint of chest pain. Randomization into the trial occurred at a median of 90 minutes after ischemic symptom onset. STEMI was suspected in 41%; 47% of participants received percutaneous coronary intervention. Characteristics of GIK and placebo participants were similar.
| Characteristics | GIK (n = 411) | Placebo (n = 460) |
|---|---|---|
| Age (yrs) | 63.9 ± 13.9 | 63.3 ± 14.1 |
| Gender | ||
| Women | 113 (27.5) | 140 (30.4) |
| Men | 298 (72.5) | 320 (69.6) |
| Race ∗ | ||
| White | 332/403 (82.4) | 392/451 (86.9) |
| Black | 52/403 (12.9) | 42/451 (9.3) |
| Asian | 6/403 (1.5) | 3/451 (0.7) |
| American Indian or Alaskan Native | 8/403 (2.0) | 8/451 (1.8) |
| Native Hawaiian or other Pacific Islander | 2/403 (0.5) | 2/451 (0.4) |
| Other | 4/403 (1.0) | 4/451 (0.9) |
| Hispanic ethnic group | 44/402 (10.9) | 58/445 (13.0) |
| Chief complaint on presentation | ||
| Chest pain | 358/411 (87.1) | 391/460 (85.0) |
| Dyspnea | 15/411 (3.6) | 19/460 (4.1) |
| Initial out-of-hospital blood pressure (mm Hg) | ||
| Systolic | 143.3 ± 32.0 | 143.4 ± 34.9 |
| Diastolic | 84.4 ± 23.6 | 85.0 ± 25.1 |
| Initial out-of-hospital heart rate (beats/min) | 86.8 ± 24.7 | 86.6 ± 25.6 |
| Initial out-of-hospital respiratory rate (breaths/min) | 19.3 ± 4.2 | 19.5 ± 4.4 |
| Time from symptom onset to study drug (≤24 hours) | ||
| Median | 90.0 | 90.0 |
| Interquartile range | 50.0–159.3 | 52.0–159.3 |
| Time from symptom onset to study drug | ||
| 0–30 minutes | 24/401 (6.0) | 20/457 (4.4) |
| 31–60 minutes | 101/401 (25.2) | 121/457 (26.5) |
| 61–90 minutes | 60/401 (15.0) | 74/457 (16.2) |
| 91–180 minutes | 66/401 (16.5) | 82/457 (17.9) |
| 181–360 minutes | 46/401 (11.5) | 55/457 (12.0) |
| 361 minutes–24 h | 37/401 (9.2) | 36/457 (7.9) |
| Within 24 hours (unspecified) | 31/401 (7.7) | 34/457 (7.4) |
| >24 hours | 36/401 (9.0) | 35/457 (7.7) |
| STEMI on presenting out-of-hospital ECG (%) | 163/411 (39.7) | 194/460 (42.2) |
| ACI-TIPI score † | 74.6 ± 22.6 | 76.9 ± 20.6 |
| TPI triggered | 84/411 (20.4) | 116/460 (25.2) |
| Medical history | ||
| Diabetes mellitus | 121/411 (29.4) | 121/460 (26.3) |
| HF | 68/411 (16.5) | 77/460 (16.7) |
| Myocardial infarction | 152/411 (37.0) | 159/460 (34.6) |
| Hospital acute reperfusion treatment | ||
| Thrombolytic therapy | 3/411 (0.7) | 8/460 (1.7) |
| Percutaneous coronary intervention | 198/411 (48.2) | 208/460 (45.2) |
| Coronary artery bypass graft | 12/411 (2.9) | 13/460 (2.8) |
The main results are listed in Table 2 . In all participants, the 1-year mortality rates estimated from unadjusted Kaplan-Meier (K-M) curves were 11.6% with GIK versus 13.5% with placebo (unadjusted hazard ratio [HR] 0.83, 95% confidence interval [CI] 0.57 to 1.23, p = 0.36); the K-M rates of the composite of cardiac arrest or 1-year mortality were 12.8% for GIK and 17.0% for placebo (HR 0.71, 95% CI 0.50 to 1.02, p = 0.06); the K-M rates for the composite of hospitalization for HF or mortality within 1 year were 17.2% for both GIK and placebo (HR 0.98, 95% CI 0.70 to 1.37, p = 0.92); and the K-M rates for the composite of cardiac arrest, 1-year mortality, or HF hospitalization within 1 year were 18.1% with GIK versus 20.4% with placebo (HR 0.85, 95% CI 0.62 to 1.16, p = 0.30). The causes of death are listed in Table 3 .
| Outcome | Treatment | Unadjusted Rate, n/Total (%) | Unadjusted K-M Rate (%) | Unadjusted HR (GIK vs Placebo) | 95% CI of Unadjusted HR | Unadjusted Cox Regression p Value |
|---|---|---|---|---|---|---|
| 1-yr mortality | GIK | 45/411 (10.9) | 11.6 | 0.83 | 0.57–1.23 | 0.355 |
| Placebo | 60/460 (13.0) | 13.5 | ||||
| Cardiac arrest or mortality | GIK | 50/411 (12.2) | 12.8 | 0.71 | 0.50–1.02 | 0.062 |
| Placebo | 76/460 (16.5) | 17.0 | ||||
| In-hospital cardiac arrest | GIK | 15/411 (3.6) | 0.56 | 0.30–1.02 | 0.060 | |
| Placebo | 29/460 (6.3) | |||||
| 1-yr HF or mortality | GIK | 67/411 (16.3) | 17.2 | 0.98 | 0.70–1.37 | 0.915 |
| Placebo | 76/460 (16.5) | 17.2 | ||||
| 1-yr HF | GIK | 35/411 (8.5) | 9.6 | 1.22 | 0.74–2.02 | 0.439 |
| Placebo | 32/460 (7.0) | 7.8 | ||||
| Cardiac arrest, HF, or mortality | GIK | 71/411 (17.3) | 18.1 | 0.85 | 0.62–1.16 | 0.299 |
| Placebo | 91/460 (19.8) | 20.4 |
| Subgroup | No. of Deaths by Treatment | |
|---|---|---|
| GIK | Placebo | |
| All: total no. of deaths (1-yr analysis) | 45 | 60 |
| Death cardiac: arrhythmic | 3 | 2 |
| Death cardiac: HF | 6 | 10 |
| Death cardiac: ischemia | 13 | 22 |
| Death cardiac: not otherwise specified | 12 | 9 |
| Death non-cardiac | 11 | 17 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
