- Email: [email protected]
CABG—trial design
Acute Ischemic Heart Disease
SHould We Emergently Revascularize Occluded Coronaries for Cardiogenic ShocK: An international randomized trial of emergency PTCA/CABG—trial design Judith S. Hochman, MD,a Lynn A. Sleeper, ScD,b Emilie Godfrey, MS, RD,a Sonja M. McKinlay, PhD,b Timothy Sanborn, MD,c Jacques Col, MD,d and Thierry LeJemtel, MD,e for the SHOCK Trial Study Group New York, NY
Background Cardiogenic shock (CS) is the leading cause of death in patients hospitalized with acute myocardial infarction (MI). Nonrandomized studies suggest reduced mortality rate with revascularization.
Trial design The SHOCK trial is a multicenter, randomized, and unblinded study with a Registry for trial-eligible and ineligible nonrandomized patients. The trial is testing the hypothesis that a direct invasive strategy of emergency revascularization for patients with cardiogenic shock complicating acute MI will reduce 30-day all-cause mortality rate by 20 absolute percentage points compared with initial medical stabilization. Eligibility criteria include development of CS within 36 hours of an acute transmural MI as evidenced by ST elevation or new left bundle branch block MI; clinical criteria for CS with hemodynamic confirmation; absence of a mechanical, iatrogenic, or other cause of shock; and enrollment within 12 hours of CS diagnosis. Patients randomly assigned to emergency revascularization immediately undergo coronary angiography, with percutaneous transluminal coronary angioplasty or coronary artery bypass grafting depending on the coronary anatomy. Patients assigned to initial medical stabilization may undergo revascularization ≥54 hours after randomization.
End points The primary end point is all-cause 30-day mortality after randomization. Secondary end points include death at trial termination, changes in left ventricular dimensions and function measured by echocardiography at randomization and 2 weeks later, and changes in quality of life and physical functioning from 2 weeks after discharge to 6 months after MI. (Am Heart J 1999;137:313-21.)
Since the advent of the coronary care unit and the widespread availability of defibrillators, cardiogenic shock (CS) has been the leading cause of death for patients hospitalized with acute myocardial infarction (AMI).1,2 Over a 14-year period the incidence and mortality rate of CS has remained constant at 7% to 8% and at 80%, respectively.2 Although the occurrence of CS From athe St. Luke’s-Roosevelt Hospital, Columbia University, New York; bNew England Research Institutes, Watertown, Mass; cthe New York Hospital, Cornell Medical Center; dthe Cliniques Universitaires St. Luc, Brussels, Belgium; and the eAlbert Einstein College of Medicine, Bronx, NY. Study supported by R01 grants from the National Heart, Lung, and Blood Institute, #HL50020-018Z and HL49970, 1994-1999. Initial phase supported by a grant-in-aid from the American Heart Association New York Affiliate. Received July 22, 1998; accepted October 22, 1998. Reprint requests: Judith S. Hochman, MD, St Luke’s-Roosevelt Hospital, 1111 Amsterdam Ave, New York, NY 10025. E-mail: [email protected] Copyright © 1999 by Mosby, Inc. 0002-8703/99/$8.00 + 0 4/1/95352
complicating acute MI has declined with reperfusion therapy, it still remains the leading cause of death in patients with AMI, with a 60% to 70% mortality rate.1,3-7 Nonrandomized retrospective studies report a pooled 45% in-hospital mortality rate in patients who undergo percutaneous transluminal coronary angioplasty (PTCA)6,8-15 during hospitalization for CS complicating AMI and a 35% in-hospital mortality rate in patients who undergo coronary artery bypass graft (CABG) surgery during hospitalization.6,16-22 However, such low mortality rates in patients undergoing PTCA or CABG is partly from selection bias by highly skilled clinicians. The SHOCK prestudy Registry6 demonstrated that selection for cardiac catheterization, regardless of subsequent therapy, was associated with a better survival rate. Conversely, the only series of patients subjected to a direct invasive strategy for all (consecutive) patients with early shock4 reported 72%
American Heart Journal February 1999
314 Hochman et al
Figure 1
Sequence of events in the SHOCK trial, with allowable time windows. ERV and revascularization in IMS arm, if attempted, will be separated by minimum of 48 hours.
Table I. SHOCK trial inclusion criteria Index MI Prolonged (>30 minutes) chest pain or equivalent symptoms Electrocardiographic criteria at the time of MI diagnosis: ≥2 leads with ST elevation (≥2 mm precordial leads, ≥1 mm limb leads) or ≥1 mm precordial ST depression known to reflect posterior ST elevation MI Or new left bundle branch block (not known to be old) Or new pathologic Q waves in ≥2 related leads or posterior Q waves manifested by R/S ratio V1 or V2 >1. CS Clinical criteria Systolic blood pressure <90 mm Hg for 30 minutes before inotrope/vasopressor administration Or vasopressors or IABP are required to maintain systolic blood pressure ≥90 mm Hg Evidence of decreased organ perfusion (ie, urine output ≤30 mL/hr or cool and diaphoretic extremities) Heart rate ≥60 beats/min (including paced rhythms) Hemodynamic criteria Pulmonary capillary wedge pressure ≥15 mm Hg Cardiac index ≤2.2 L/min/m2
in-hospital mortality rate despite high rates of successful early reperfusion. In the United States, an estimated 85% of patients with AMI are admitted to hospitals that do not have the facilities to perform PTCA or CABG surgery.23 The GUSTO I Study5 and the SHOCK prestudy Registry,6 where the decision regarding an invasive strategy was left up to the individual clinician, reported use of early revascularization in only 11% and 24% of patients, respectively. If a direct invasive strategy with emergency PTCA or CABG is shown to significantly reduce mortality rate for CS complicating AMI, it will greatly affect practice patterns. The risks and resources associated with the emergent use of these procedures and transport of critically ill, intubated
patients receiving vasopressors and possibly intraaortic balloon pump (IABP) support would be justified. A randomized trial of a direct invasive strategy versus initial medical stabilization for CS caused by left ventricular (LV) failure complicating AMI is ongoing to test this hypothesis. The results of randomized trials of any therapy for patients with CS complicating AMI have never been published. The current report describes the design of the SHOCK Trial (SHould We Emergently Revascularize Occluded Coronaries for Cardiogenic ShocK).
Overview The trial is a multicenter, randomized, and unblinded study with a Registry for eligible and ineligible nonran-
American Heart Journal Volume 137, Number 2
Hochman et al 315
Figure 2
Overview of patient management by group assignment.
domized patients. The trial, which enrolled its first patient in April 1993, is funded by the National Heart, Lung and Blood Institute. It closed enrollment on November 30, 1998.
ria are present (Table II), and the designated time windows regarding shock onset and randomization are adhered to (Fig 1). An overview of trial patient management is shown in Fig 2.
End points
Selection criteria
Primary end point
There are 3 components of trial eligibility: Index myocardial infarction. Patients who have electrocardiographic evidence consistent with a recent total coronary occlusion, for example, ST elevations, new Q waves, and new left bundle branch block, are eligible. Patients with non-ST-segment elevation MI are excluded. CS characteristics. Patients must meet clinical criteria for CS, including hypotension and end-organ hypoperfusion that is not caused by bradycardia or a sustained ventricular arrhythmia. Hemodynamic confirmation of LV dysfunction is required before randomization except in patients with acute anterior MI and pulmonary congestion documented on chest radiograph, in whom right heart catheterization is mandatory but may be performed after randomization (Table I). Timing of enrollment (Fig 2). CS must develop within 36 hours of MI, and randomization must occur within 12 hours of shock diagnosis. If a reinfarction causes CS, the reinfarction is considered to be the “index” MI and the time of shock onset relates to this index infarction. As of October 1998, the median time from index MI onset to shock in the SHOCK Trial was
The primary end point is all-cause mortality at 30 days after randomization.
Secondary end points The secondary end points include all-cause mortality at hospital discharge, 6 months, and at trial termination; change in echocardiography regional wall motion score, LV dilation index, and other echo variables measured at randomization and 2 weeks after randomization or before hospital discharge, whichever comes first; and changes in quality of life and physical functioning from 2 weeks after discharge to 6 months after MI for survivors discharged within 2 months of MI.
Study design Patients are randomly assigned to emergency early revascularization or initial medical stabilization. Patients with CS complicating AMI are eligible for the randomized trial when shock is caused by predominant LV failure and eligibility is confirmed—that is, all inclusion criteria are met (Table I), no exclusion crite-
American Heart Journal February 1999
316 Hochman et al
Table II. SHOCK trial exclusion criteria Other severe systemic illness, including contraindications to coronary bypass surgery and including anoxic brain damage Active bleeding or other contraindication to heparin (eg, allergy) Acute mechanical cause of CS Ventricular septal rupture Acute severe mitral regurgitation Isolated right ventricular CS Shock from other causes (eg, hypovolemia, sepsis, anaphylaxis) Severe aortic stenosis or other severe valvular disease Known dilated cardiomyopathy (ischemic and other) Inability to obtain vascular access for catheterization Cardiac tamponade Recent hemorrhage (ie, hematocrit fall ≥15 percentage points within 24 hours of shock onset) Shock from β-blockade or calcium channel blockade Patient is not a candidate for any revascularization procedure (ie, known unsuitable coronary anatomy) Patient or surrogate refuses or cannot give informed consent Refusal by patient’s personal physician
5.8 hours for the 292 patients who have already been randomized, that is, one half of randomized patients have rapid onset of CS complicating AMI. The potential benefits of PTCA/CABG for CS that develops late after MI onset include interruption of the progressive necrosis seen in CS24-26 and limitation of infarct expansion.27 Furthermore, low mortality rates have been reported in patients with late shock when PTCA was performed several days after MI onset.9,12 It is strongly recommended that patients be enrolled as soon as possible after shock diagnosis. However, up to 12 hours is allowed to accommodate the transfer of patients from referring hospitals. The presence of any single exclusion criterion (Table II) renders a patient ineligible for the trial.
gency early revascularization (ERV) or initial medical stabilization (IMS). Randomization is stratified by center and uses a random permuted block design with blocks of varying size.
Trial operator certification All physicians who perform CABG on trial patients were certified according to the criteria used in the BARI trial28 and physicians who perform PTCA were certified according to criteria adapted from TIMI 2 and BARI Trials28,29: the operator must have performed at least 250 PTCA procedures, including at least 20 cases of acute MI with 85% success in the last 100 cases and a complication rate of ≥5% MI, CABG, and ≤2% procedure mortality rate.
Consent
Interventions
All participating centers have Institutional Review Board or Ethics Committee Consent approval. Written informed consent is obtained from the patient or the surrogate (if permitted by the Institional Review Board) before randomization. As of November 1, 1996, it became permissible for centers to use the National Institutes of Health/Federal Drug Administration guidelines for exception from informed consent if the local Board approved the waiver and all required procedures were followed. Application of these guidelines to the SHOCK Trial is appropriate because of the emergent nature of treatment. Two centers obtained approval to waive informed consent.
Trial patients receive routine intensive cardiac care with ventilatory support as needed. The only intervention to be tested is a direct invasive strategy with emergency early mechanical revascularization, if technically feasible, compared with IMS, including possible delayed revascularization. The use of other treatment modalities was left to the discretion of local investigators who were strongly recommended to adhere to routine care, including aspirin and therapeutic anticoagulation with heparin, as recommended by American Heart Association/American College of Cardiology guidelines on acute MI management30 and delineated in the SHOCK Trial protocol.
Emergency revascularization Randomization Patients are randomly assigned through an interactive voice randomization system computer to emer-
Patients randomly assigned to emergency revascularization undergo coronary angiography as soon as possible and within 6 hours of randomization. Revascular-
American Heart Journal Volume 137, Number 2
ization is attempted primarily with PTCA (including approved devices such as stents) but also CABG surgery. IABP insertion as soon as possible after shock onset, and before coronary angiography, is strongly recommended. The choice of revascularization modality and catheter-based techniques is individualized according to the patient’s coronary anatomy and the judgment of the local interventionalist and cardiac surgeon. The goal of catheter-based techniques is an optimal result, with use of stents and a glycoprotein IIb/IIIa receptor antagonist to achieve that goal and reflect clinical practice. The recommended algorithm for revascularization is PTCA of the infarct-related artery for single- and double-vessel disease, with multivessel PTCA as clinically determined. The decision to perform PTCA if TIMI 3 flow was present in the infarct-related artery was left to the investigator’s judgment. Emergency CABG is recommended for severe triple-vessel coronary artery disease, left main artery disease, and failed PTCA. Severe triplevessel disease is defined as 2 total or subtotal occlusions or >90% stenosis in 2 nonculprit major vessels. For triple-vessel coronary artery disease of moderate severity, PTCA of the infarct-related artery followed by CABG or direct CABG is recommended. When the patients are randomly assigned promptly after the onset of the index infarction, direct (without thrombolytic) PTCA is recommended. For patients undergoing emergency CABG, the time from completion of coronary angiography to total vented bypass is recommended to be less than 4 hours.
Initial medical stabilization Routine recommended care includes IABP support; inotropes, vasopressors, and subsequent vasodilators to maintain perfusion guided by hemodynamics; and mechanical ventilation, if necessary. Patients who are thrombolytic eligible and are seen within 24 hours after MI onset are recommended to receive thrombolytic therapy. Thrombolytic eligibility is defined as an absence of absolute contraindications to a thrombolytic, including active internal bleeding, recent (<6 months) cerebrovascular accident, recent (<2 months) intracranial or spinal surgery, intracranial neoplasm, arterial venous malformation, or aneurysm. Revascularization with PTCA or Cardiac Surgery may be performed 54 hours or more after randomization to the IMS arm at the discretion of the cardiologist caring for the patient. This time window ensures a 48-hour minimum separation between revascularization in the ERV and IMS groups. Based on the known high incidence of triple-ves-
Hochman et al 317
sel and left main disease with impaired LV function in patients with CS complicating acute MI, performance of coronary angiography ≥54 hours after randomization is encouraged. The only indications for PTCA or cardiac surgery <54 hours after randomization are for patients who have (1) a mechanical complication (ie, ventricular septal rupture or papillary muscle/chordal rupture/severe mitral regurgitation) or (2) angina at rest (≥30 minutes) that develops at least 24 hours after MI, with ST depression or elevation ≥1 mm, or T-wave inversions in at least 2 related leads that is refractory to medical management, including IABP.
Measurement Information on all aspects of the patient and the patient’s hospital stay is collected, including demographics; medical history; characteristics and timing of the MI and diagnosis of CS; medications; interventions; adverse events; and hemodynamics closest to the time of CS, randomization, 6 hours after randomization and 24 hours after randomization. Electrocardiograms are obtained for trial patients at the time of the index MI and at the time of shock onset, with 2 follow-up recordings demonstrating MI evolution. Creatine phosphokinase measurements with myocardial band fractions are obtained at 8, 16, and 24 hours after the index MI and at 8, 16, and 24 hours after randomization for those with delayed shock. Two-dimensional echocardiograms with color flow Doppler are obtained for trial patients as soon as possible, within 24 hours of randomization, and repeated 2 weeks after randomization or before discharge, whichever comes first. Baseline studies performed after shock onset but before randomization are permissible. It is strongly encouraged that Doppler echocardiography be performed before randomization for patients with a first inferior MI and for those with absence of extensive electrocardiographic abnormalities to rule out mechanical causes of shock and aortic dissection. Transesophageal echocardiograms performed when the transthoracic studies are technically suboptimal are permissible. Quality of life and physical function assessments are completed by telephone interview with trial patients at 2 weeks after hospital discharge and 6 months after MI. A one-year post-MI interview is also conducted with trial patients randomly assigned before the last year of the enrollment period. All patients alive at their last contact will have vital status assessed at trial termination, potentially providing 5.75 years of followup for the earliest enrollees.
American Heart Journal February 1999
318 Hochman et al
Quality of life is assessed with single-response questions including Andrews’ “Ladder of Life”31 and a 7point Likert scale regarding satisfaction with life as a whole.32 Quality of life is assessed with the Multidimensional Index of Life Quality (MILQ),33 developed specifically for patients with cardiovascular disease. Functional status is assessed with a modified form of the Sickness Impact Profile.34 New York Heart Association congestive heart failure class 1 through 4 functional status is also assessed during these interviews with 4 questions so that a judgment may be made using consistent criteria for all patients.
Angiography/angioplasty core laboratory All coronary angiograms and angioplasty films in both the Emergency Revascularization and Initial Medical Stabilization groups are sent to the Data Coordinating Center, where they are coded so that the core laboratory is blinded to site and patient identification number and treatment group. Each film is read independently by 2 angiographers with standardized data forms. A third reader resolves any discrepancies between the 2 primary readers using a third set of data forms. Data are collected on the extent and severity of coronary artery disease, TIMI flow, collateral flow, thrombus, and LV ejection fraction (when a left ventriculogram is performed). Electrocardiograms are provided to help identify the infarct-related artery. PTCA success is graded by change in infarct-related artery perfusion status (TIMI flow), post-intervention stenosis, and change in stenosis severity (at least 20% absolute decrease in stenosis). Quantitative coronary angiography is performed on the infarct-related artery before and after PTCA by the Arriflex QANSAD (1993-1996) and MEDIS CMS-QCA (19961988) systems. Mitral regurgitation, if present, is graded.
Echocardiography core laboratory Echocardiograms with color flow and continuous wave Doppler flow are sent to the Data Coordinating Center for preparation so that the trained echocardiographic readers are blinded to site and patient identification, treatment group, and timing of the echo (baseline vs follow-up). Each tape is read by 1 echocardiographer. Standardized data forms are completed, which include parameters of LV size and function, right ventricular size and function, the function of the infarct and remote (uninfarcted) myocardium, extent of mitral regurgitation, the Doppler transmitral filling pattern and wall motion score index.
Patient registry To ensure that all potentially eligible patients are considered for the trial and to minimize the possibility that an enrolling center systematically excludes subgroups of patients from the trial, a prospective patient registry is maintained at all enrolling centers for patients with suspected CS complicating acute MI. All such patients are registered regardless of trial eligibility. Demographics, medical history, baseline and infarct characteristics, selected procedures, medications, and vital status at hospital discharge are recorded for Registry patients. Compliance with registration of patients is monitored by clinic site audits whereby a list of all patients with a discharge diagnosis of acute MI and CS are compared with those entered into the SHOCK Registry. Complete capture was attained by retrospectively registering any missed patients. The Registry closed August 31, 1997, with 1190 patients and was replaced with a screening log.
Sample size The SHOCK Trial was designed to detect an absolute group difference in all-cause 30-day mortality of 20% with 90% power. To monitor the data closely, a continuous sequential design was selected. A continuous plan was felt to be appropriate because the primary study end point (30-day mortality) is obtained relatively quickly and there was a strong sentiment to act promptly on an observation of improved survival in this critically ill population. A restricted sequential plan first proposed by Armitage35 was implemented. Under this design, paired observations are used as a monitoring tool. Observations are paired according to date of randomization, with pair members belonging to different treatment groups. When the outcomes of the 2 patients constituting the pair are discordant, this provides information regarding a treatment preference. The discordant pair results form a binary random walk and are plotted on a figure with boundaries for early stopping. The required sample size is a function of the desired statistical power, the overall type I error rate, and the anticipated mortality rates. Two sets of ERV versus IMS mortality rates were considered when designing the trial: (1) 55% versus 75% and (2) 30% versus 50%, respectively. The sample size required to ensure 90% power while restricting the overall type I error rate of the trial is 312 for the first setting and 328 for the latter. Therefore in the absence of unbiased information regarding the
American Heart Journal Volume 137, Number 2
absolute mortality rates likely to be observed, the trial was designed to accrue 328 patients (164 per arm). Selection of this higher target enrollment figure also allowed for high power as long as the final sample size exceeded 90% of the target (ie, enrollment of at least 296 patients). The power to detect an absolute 20% group difference based on the 2 sets of mortality rates described above is 88% and 87%, respectively, if 296 patients are enrolled.
Data and Safety Monitoring Board The data are continuously monitored by the Data Coordinating Center statistician and a summary report of 30-day all-cause mortality is reviewed by the Chair of the Data and Safety Monitoring Board on a monthly basis. The DSMB meets semiannually, is completely unblinded, and reviews study end points, trial safety, protocol violations, and data quality and completeness. The primary study end point, 30-day all-cause mortality, as well the secondary end points of the study, which include 6month survival rate and LV ejection fraction, are reviewed. Selected subgroup comparisons are also examined with tests of interaction. These include subgroups based on sex, age, and early versus late onset of CS. With regard to trial safety, the DSMB Chair is immediately notified of occurrences of deaths in the ERV group that are temporally associated with a coronary angiography or angioplasty procedure or instances of emergency bypass surgery required as a result of a procedural complication in the cardiac catheterization laboratory. Safety monitoring reports also include rates of adverse events occurring in each treatment group. Protocol violations are tracked with regard to type and origin (clinic site). These include failure to meet trial eligibility criteria and treatment crossover—for example, Initial Medical Stabilization group patients who undergo emergent revascularization outside of circumstances in which such treatment is clinically indicated per protocol criteria, and patients randomly assigned to the ERV group in which no attempt or a late attempt at revascularization is made. Trial investigators, including the Study Chair, are blinded to all primary and secondary outcomes, including overall trial mortality. Trial investigators (but not the Study Chair) are additionally blinded to the rates of coronary angiography and attempted revascularization in both trial arms.
Hochman et al 319
Analysis plan overview Primary analysis This analysis will be according to the intention-totreat principle for all study end points. The primary study end point, 30-day all cause mortality after randomization, will be evaluated with logistic regression. Clinic site differences in mortality rate between the 2 arms will be examined, as will differences in baseline patient characteristics. It is expected that these differences will be minimal because of randomization; however, a covariate-adjusted odds ratio for mortality will be reported as the primary trial result if significant group differences are identified. Cox proportional hazards modeling will also be used to examine the time course of deaths and to estimate a treatment effect for long-term survival (a secondary study end point).
Secondary analyses Four types of secondary analyses will be conducted on the primary study end point to better evaluate treatment efficacy. All involve the exclusion of patients determined to be trial ineligible or who had rapid death after randomization. Group 1A: Patients who were presumed eligible at the time of randomization but were determined shortly after randomization to have a disease process other than CS caused by acute MI with predominant LV failure, such as severe mitral regurgitation or aortic dissection. Group 1B: Patients who violated eligibility criteria at randomization (eg, shock diagnosis or randomization outside the prescribed time window or not meeting clinical or hemodynamic entry criteria defining CS). Group 2: Patients in either trial arm who die too rapidly to receive treatment, defined as within 30 minutes of randomization. The four secondary analyses proposed exclude the following groups: (1) group 1A, (2) groups 1A + 1B, (3) groups 2 + 1A, and (4) groups 2 + 1A + 1B.
Subgroup analyses There may be certain subgroups of patients in which the ERV versus IMS group mortality difference may be more or less pronounced than the difference observed in the overall analysis. Subgroup analyses of the primary study end point will be conducted with a logistic regression model with interaction terms between the subgroup identifier and the treatment group variable (ERV vs IMS). The study protocol details the scope of prespecified subgroup analyses to be conducted. Such analyses will focus on groups defined by patient age, sex, type of admission (transfer vs direct admission),
American Heart Journal February 1999
320 Hochman et al
geographic location, medical history and therapy, timing of CS and randomization, coronary anatomy and MI location, use of supportive therapies, and measures of cardiac function. An important subgroup is comprised of patients randomly assigned within 6 hours of MI symptom onset.
Timeline Thirty international centers are participating in the randomized trial and registry. Trial enrollment closed on November 30, 1998.
13.
14.
15.
16.
We thank Adriana Gianella for preparation of the manuscript and Venu Menon, MD, for his assistance.
17.
References
18.
1. Becker RC, Gore JM, Lambrew C, Weaver WD, Rubison RM, French WJ, et al. A composite view of cardiac rupture in the United States National Registry of Myocardial Infarction. J Amer Coll Cardiol 1996;27:1321-6. 2. Goldberg RJ, Gore JM, Alpert J, et al. Cardiogenic shock acute myocardial infarction. Incident and mortality from a community-wide perspective, 1975 to 1988. N Engl J Med 1991;325,1117-22. 3. Brodie B, Stuckey T, Weintraub R, et al. Timing and mechanism of death after direct angioplasty for acute myocardial infarction. J Amer Coll Cardiol 1995;25:295A-296A. 4. Himbert, D, Juliard JM, Steg, PG, Karrillon GJ, Aumont MC, Gourgon R. Limits of reperfusion therapy for immediate cardiogenic shock complicating acute myocardial infarction. Am J Cardiol 1994;74:492-4. 5. Holmes DR Jr, Bates ER, Kleiman NS, et al. Contemporary reperfusion therapy for cardiogenic shock: the GUSTO-I trial experience. The GUSTO-I Investigators. Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries. J Am Coll Cardiol 1995;26:668-74. 6. Hochman JS, Boland J, Sleeper LA, et al. Current spectrum of cardiogenic shock and effect of early revascularization on mortality: results of an international registry. Circulation 1995;91:873-81. 7. Menon V, Hochman JS, Holmes D, Bates E, Califf RM, Topol E, et al for the GUSTO III Investigators. Lack of progress in cardiogenic shock: lessons from the GUSTO trials. J Am Coll Cardiol 1998;31:1057-127. 8. Disler L. Haitas B, Benjamin J, Steingo L, McKibbin J. Cardiogenic shock in evolving myocardial infarction: treatment by angioplasty and streptokinase. Heart Lung 1987;16:649-52. 9. Lee L, Bates ER, Pitt B, Walton JA, Laufer N, WW Oneill. Percutaneous transluminal coronary angioplasty improves survival in acute myocardial infarction complicated by cardiogenic shock. Circulation 1988;78:1345-51. 10. Verna E, Repetto S, Boscarini M, Ghezzi I, Binaghi G. Emergency coronary angioplasty in patients with severe left ventricular dysfunction or cardiogenic shock after acute myocardial infarction. Eur Heart J 1989;10:958-66. 11. Meyer P, Blanc P, Baudouy M, Morand P. Treatment of primary cardiogenic shock by coronary transluminal angioplasty during the acute phase of myocardial infarction. Arch Mal Coeur Vaiss 1990; 83:329-34. 12. Moosvi AR, Khaja F, Villanueva L, Gheorghiade M, Douthat L, Gold-
19.
20.
21.
22.
23. 24.
25. 26.
27.
28.
29.
30.
stein S. Early revascularization improves survival in cardiogenic shock complicating acute myocardial infarction. J Am Coll Cardiol 1992;19:907-14. Yamamoto H, Hayashi Y, Oka Y, et al. Efficacy of percutaneous transluminal coronary angioplasty in patients with acute myocardial infarction complicated by cardiogenic shock. Jpn Cir J 1992;56:815-21. Seydoux C, Goy JJ, Beuret P, et al. Effectiveness of percutaneous transluminal coronary angioplasty in cardiogenic shock during acute myocardial infarction. Am J Cardiol 1992;69:968-9. Hibbard MD, Holmes DR Jr, Baily KR, et al. Percutaneous transluminal coronary angioplasty in patients with cardiogenic shock. J Am Coll Cardiol 1992;19:639-46. Laks H, Rosenkranz E, Buckberg GD. Surgical treatment of cardiogenic shock after myocardial infarction. Circulation 1986;74:III-11-6. Guyton RA, Arcidi JM Jr, Langford DA, Morris DC, Liberman HA, Hatcher CR Jr. Emergency coronary bypass for cardiogenic shock. Circulation 1987;76:V22-7. Bolooki H. Emergency cardiac procedures in patients in cardiogenic shock due to complications of coronary artery disease. Circulation 1989;79:II37-48. Phillips SJ, Zeff RH, Skinner JR, Toon RS, Grignon A, Kongtahworn C. Reperfusion protocol and results in 738 patients with evolving myocardial infarction. Ann Thorac Surg 1986;41:119-25. Kirklin JK, Blackstone EH, Zorn GL Jr, et al. Intermediate-term results of coronary artery bypass grafting for acute myocardial infarction. Circulation 1985;72:II-175-8. Subramanian VA, Roberts AF, Zema MJ, et al. Cardiogenic shock following acute myocardial infarction: late functional results after emergency cardiac surgery. NY State J Med 1980;80:947-52. Keon WJ. New therapeutic approaches to acute myocardial infarction (MI) and its size: the surgical approach. Avd Exp Med Biol 1977;82:394-7. Facilities and services in the United States. In hospital statistics. Chicago: American Hospital association; 1992. p. 208-9. Alonso DR, Scheidt S, Post M, Killip T. Pathophysiology of cardiogenic shock: quantification of myocardial necrosis, clinical pathologic, and electrocardiographic correlations. Circulation 1993;48:588-96. Page DR, Caulfield BJ, Kastor JA, et al. Myocardial changes associated with cardiogenic shock. N Engl J Med 1971;285:133-7. Gutovitz AL, Sobel BE, Roberts R. Progressive nature of myocardial injury in selected patients with cardiogenic shock. J Am Coll Cardiol 1978;41:469-75. Hochman JS, Choo H. Limitations of myocardial infarct expansion by reperfusion independent of myocardial salvage. Circulation 1987;75:299-306. BARI Trial Investigators. Protocal for the Bypass Angioplasty Revascularization Investigation (BARI). Circulation 1991;84 (suppl):V-1-V-27. Passamani E, Morrison H, Herman M, Grose R, Chaitman B, Rogers W, et al for the TIMI Investigators. Thrombolysis in Myocardial Infarction (TIMI) phase II pilot study: tissue plasminigen activator followed by percutaneous transluminal coronary angioplasty. J Am Coll Cardiol 1987;10:51B-64B. Gunnar RM, Bourdillon PD, Dixon DW, Fuster V, Karp RB, Kennedy JW, et al. ACC/AHA guidelines for the early management of patients with acute myocardial infarction. A report of the American College of Cardiology/American Heart Association Task Force on Assessment of Diagnostic and Therapeutic Cardiovascular Procedures (subcommit-
American Heart Journal Volume 137, Number 2
31. 32. 33.
34
35.
tee to develop guidelines for the early management of patients with acute myocardial infarction. Circulation 1990;82:664-707. Andrews FM, Withey SB. Social indicators of well-being: Americans’ perceptions of life quality. New York: Plenum; 1976. Campbell A, Converse PE, Rogers WL. The quality of American life: Perceptions, evaluations, and satisfactions. New York: Russell Sage; 1976. Avis NE, Smith KW, Hambleton RK, Feldman HA, Selwyn A, Jacobs A. Development of the Multidimensional Index of Life Quality: a quality of life measure for cardiovascular disease. Medical Care 1996;34:1102-20. deBruin AF, Buys M, deWitte LP, Diederiks JPM. The Sickness Impact Profile: SIP68, a short generic version. First evaluation of the reliability and reproducibility. J Clin Epid 1994;47:863-71. Armitage P. Restricted sequential procedures. Biometrika 1957; 44:9-26.
Appendix The following are committee members, principal and co-principal investigators, and study coordinators in the SHOCK Trial. Executive Committee: Judith S. Hochman, Study Chair; Thierry LeJemtel, Study CoChair; Philip Aylward, Jean Boland, Jacques Col, O. Wayne Isom, Michael Picard, Timothy Sanborn, Lynn Sleeper, Sonjá M. McKinley, Harvey White, and Patrice Desvigne-Nickens (ex officio). Steering Committee: Executive Committe members and principal investigator from each clinical center. Publications Committee: Harvey White, Chair; James Abel, Judith Hochman, Thierry LeJemtel, Lynn Sleeper and John Webb. Data and Safety Monitoring Board: Eugene Braunwald, Chair; Floyd Loop, Charles McCarthy, Neal Scott, David Williams, and Janet Wittes. Clinical Centers (alphabetical by center): Mark Porway and Barbara Burkott, Baystate Medical Center (Springfield, Mass); Alice Jacobs and Mary Mazur, Boston University Medical Center (Boston); Susan Graham, Syed Raza, and Joie Celano, Buffalo General Hospital (Buffalo, NY); Jean Boland, Suzanne Pourbaix, and Mireille Massoz, C.H.R. Citadelle (Liege, Belgium); Jacques Col and Reine Lauwers, Cliniques Universitaires St. Luc (Brussels, Belgium); Philip Aylward and Caroline Thomas, Flinders Medical Centre (Adelaide, SA, Australia); Jo Dens, Frans Van de Werf, and Connie Luys, Gasthuisberg Universitaire Hospital (Leuven, Belgium); Harvey White, John French, and Barbara Williams, Green Lane Hospital (New Zealand); Pamela R. Paulsen and Elizabeth Miller, Hennepin County Medical Center (Minneapolis, Minn); Robert Forman, E. Scott Monrad, Marie Galvao, and Margaret Jones, J.D. Weiler Hospital of the AECOM (Bronx, NY); Jeffrey Brinker and Vicki Coombs, Johns Hopkins Hospital (Baltimore, Md); Mark Menegus, Mark
Hochman et al 321
Greenberg, and Barbara Levine, Montefiore/AECOM (Bronx, NY); John A. Ambrose, Samin Sharma, Thomas Cocke, Denise Ratner, and Eppie Brown, Mt Sinai Medical Center (New York); David H. Miller and Susan Hosat, New York Hospital/Cornell Medical Center (New York); S. Chiu Wong and Maisie Brown, New York Hospital Medical Center of Queens (Flushing, NY); Expedito Ribeiro, Antonio Carlos Carvalho, Antonia Petrizzo, and Claudia Rodrigues Alves, Paulista School of Medicine (Sao Paulo, Brazil); James Slater, Angela Palazzo, and Deborah Tormey, St Luke’s-Roosevelt Hospital Center (New York); Kenneth Baran, Patrick Koller, Patti Filkins, and Cheryl Iacarella, St Paul Heart Clinic (St Paul, Minn); John Webb, Christopher Thompson, and Elizabeth Buller, St Paul’s Hospital (Vancouver, BC, Canada); John Dervan, William Lawson, and Patricia Montes, State University of New York at Stony Brook; Matthias Pfisterer, Nikolaus Jander and Marcus Weinbacher, University Hospital Basel (Basel, Switzerland); Vladimir Dzavik and Cheryl Kee, University of Alberta Hospital (Edmonton, Alberta, Canada); J. David Talley, Mindy Dearen, Millie Rawert, and Rebecca Pacheo, University of Arkansas for Medical Sciences (Little Rock); Bonnie Weiner and Marie Borbone, University of Massachusetts Medical Center; Worcester, Mass; Abel Moreyra, Sebastian Palmeri, and Mary Helen Hosler, University of Medicine and Dentistry of New Jersey (New Brunswick, NJ); Eric Bates and Patsy Bruenger, University of Michigan Medical Center (Ann Arbor, Mich); L. David Hillis, Joaquin Cigarroa, and Jeanette Kissee, University of Texas SW Medical Center (Dallas); David Faxon and Rajinder Singh, USC Medical Center (Los Angeles); Christopher Buller, Ken Gin, and Rebecca Fox, Vancouver General Hospital (Vancouver, BC, Canada); Richard Steingart and Suzanne Parker, Winthrop University Hospital (Mineola, NY). Clinical Coordinating Center: Judith S. Hochman, Thierry H. LeJemtel, and Emilie Godfrey, St Luke’s/Roosevelt Hospital Center, Columbia University (New York). Data Coordinating Center: Sonja M. McKinlay and Lynn A. Sleeper, New England Research Institutes (Watertown, Mass). Echocardiography Core Laboratory: Michael H. Picard and Ravin Davidoff, Massachusetts General Hospital (Boston). Angiography Core Laboratory: Timothy Sanborn and Geoffrey Bergman, New York Hospital/Cornell Medical Center (New York). Program Administration: Patrice Desvigne-Nickens, National Heart, Lung, and Blood Institute, National Institutes of Health (Bethesda, Md).
SHould We Emergently Revascularize Occluded Coronaries for Cardiogenic ShocK: An international randomized trial of emergency PTCA/CABG—trial design Judith S. Hochman, MD,a Lynn A. Sleeper, ScD,b Emilie Godfrey, MS, RD,a Sonja M. McKinlay, PhD,b Timothy Sanborn, MD,c Jacques Col, MD,d and Thierry LeJemtel, MD,e for the SHOCK Trial Study Group New York, NY
Background Cardiogenic shock (CS) is the leading cause of death in patients hospitalized with acute myocardial infarction (MI). Nonrandomized studies suggest reduced mortality rate with revascularization.
Trial design The SHOCK trial is a multicenter, randomized, and unblinded study with a Registry for trial-eligible and ineligible nonrandomized patients. The trial is testing the hypothesis that a direct invasive strategy of emergency revascularization for patients with cardiogenic shock complicating acute MI will reduce 30-day all-cause mortality rate by 20 absolute percentage points compared with initial medical stabilization. Eligibility criteria include development of CS within 36 hours of an acute transmural MI as evidenced by ST elevation or new left bundle branch block MI; clinical criteria for CS with hemodynamic confirmation; absence of a mechanical, iatrogenic, or other cause of shock; and enrollment within 12 hours of CS diagnosis. Patients randomly assigned to emergency revascularization immediately undergo coronary angiography, with percutaneous transluminal coronary angioplasty or coronary artery bypass grafting depending on the coronary anatomy. Patients assigned to initial medical stabilization may undergo revascularization ≥54 hours after randomization.
End points The primary end point is all-cause 30-day mortality after randomization. Secondary end points include death at trial termination, changes in left ventricular dimensions and function measured by echocardiography at randomization and 2 weeks later, and changes in quality of life and physical functioning from 2 weeks after discharge to 6 months after MI. (Am Heart J 1999;137:313-21.)
Since the advent of the coronary care unit and the widespread availability of defibrillators, cardiogenic shock (CS) has been the leading cause of death for patients hospitalized with acute myocardial infarction (AMI).1,2 Over a 14-year period the incidence and mortality rate of CS has remained constant at 7% to 8% and at 80%, respectively.2 Although the occurrence of CS From athe St. Luke’s-Roosevelt Hospital, Columbia University, New York; bNew England Research Institutes, Watertown, Mass; cthe New York Hospital, Cornell Medical Center; dthe Cliniques Universitaires St. Luc, Brussels, Belgium; and the eAlbert Einstein College of Medicine, Bronx, NY. Study supported by R01 grants from the National Heart, Lung, and Blood Institute, #HL50020-018Z and HL49970, 1994-1999. Initial phase supported by a grant-in-aid from the American Heart Association New York Affiliate. Received July 22, 1998; accepted October 22, 1998. Reprint requests: Judith S. Hochman, MD, St Luke’s-Roosevelt Hospital, 1111 Amsterdam Ave, New York, NY 10025. E-mail: [email protected] Copyright © 1999 by Mosby, Inc. 0002-8703/99/$8.00 + 0 4/1/95352
complicating acute MI has declined with reperfusion therapy, it still remains the leading cause of death in patients with AMI, with a 60% to 70% mortality rate.1,3-7 Nonrandomized retrospective studies report a pooled 45% in-hospital mortality rate in patients who undergo percutaneous transluminal coronary angioplasty (PTCA)6,8-15 during hospitalization for CS complicating AMI and a 35% in-hospital mortality rate in patients who undergo coronary artery bypass graft (CABG) surgery during hospitalization.6,16-22 However, such low mortality rates in patients undergoing PTCA or CABG is partly from selection bias by highly skilled clinicians. The SHOCK prestudy Registry6 demonstrated that selection for cardiac catheterization, regardless of subsequent therapy, was associated with a better survival rate. Conversely, the only series of patients subjected to a direct invasive strategy for all (consecutive) patients with early shock4 reported 72%
American Heart Journal February 1999
314 Hochman et al
Figure 1
Sequence of events in the SHOCK trial, with allowable time windows. ERV and revascularization in IMS arm, if attempted, will be separated by minimum of 48 hours.
Table I. SHOCK trial inclusion criteria Index MI Prolonged (>30 minutes) chest pain or equivalent symptoms Electrocardiographic criteria at the time of MI diagnosis: ≥2 leads with ST elevation (≥2 mm precordial leads, ≥1 mm limb leads) or ≥1 mm precordial ST depression known to reflect posterior ST elevation MI Or new left bundle branch block (not known to be old) Or new pathologic Q waves in ≥2 related leads or posterior Q waves manifested by R/S ratio V1 or V2 >1. CS Clinical criteria Systolic blood pressure <90 mm Hg for 30 minutes before inotrope/vasopressor administration Or vasopressors or IABP are required to maintain systolic blood pressure ≥90 mm Hg Evidence of decreased organ perfusion (ie, urine output ≤30 mL/hr or cool and diaphoretic extremities) Heart rate ≥60 beats/min (including paced rhythms) Hemodynamic criteria Pulmonary capillary wedge pressure ≥15 mm Hg Cardiac index ≤2.2 L/min/m2
in-hospital mortality rate despite high rates of successful early reperfusion. In the United States, an estimated 85% of patients with AMI are admitted to hospitals that do not have the facilities to perform PTCA or CABG surgery.23 The GUSTO I Study5 and the SHOCK prestudy Registry,6 where the decision regarding an invasive strategy was left up to the individual clinician, reported use of early revascularization in only 11% and 24% of patients, respectively. If a direct invasive strategy with emergency PTCA or CABG is shown to significantly reduce mortality rate for CS complicating AMI, it will greatly affect practice patterns. The risks and resources associated with the emergent use of these procedures and transport of critically ill, intubated
patients receiving vasopressors and possibly intraaortic balloon pump (IABP) support would be justified. A randomized trial of a direct invasive strategy versus initial medical stabilization for CS caused by left ventricular (LV) failure complicating AMI is ongoing to test this hypothesis. The results of randomized trials of any therapy for patients with CS complicating AMI have never been published. The current report describes the design of the SHOCK Trial (SHould We Emergently Revascularize Occluded Coronaries for Cardiogenic ShocK).
Overview The trial is a multicenter, randomized, and unblinded study with a Registry for eligible and ineligible nonran-
American Heart Journal Volume 137, Number 2
Hochman et al 315
Figure 2
Overview of patient management by group assignment.
domized patients. The trial, which enrolled its first patient in April 1993, is funded by the National Heart, Lung and Blood Institute. It closed enrollment on November 30, 1998.
ria are present (Table II), and the designated time windows regarding shock onset and randomization are adhered to (Fig 1). An overview of trial patient management is shown in Fig 2.
End points
Selection criteria
Primary end point
There are 3 components of trial eligibility: Index myocardial infarction. Patients who have electrocardiographic evidence consistent with a recent total coronary occlusion, for example, ST elevations, new Q waves, and new left bundle branch block, are eligible. Patients with non-ST-segment elevation MI are excluded. CS characteristics. Patients must meet clinical criteria for CS, including hypotension and end-organ hypoperfusion that is not caused by bradycardia or a sustained ventricular arrhythmia. Hemodynamic confirmation of LV dysfunction is required before randomization except in patients with acute anterior MI and pulmonary congestion documented on chest radiograph, in whom right heart catheterization is mandatory but may be performed after randomization (Table I). Timing of enrollment (Fig 2). CS must develop within 36 hours of MI, and randomization must occur within 12 hours of shock diagnosis. If a reinfarction causes CS, the reinfarction is considered to be the “index” MI and the time of shock onset relates to this index infarction. As of October 1998, the median time from index MI onset to shock in the SHOCK Trial was
The primary end point is all-cause mortality at 30 days after randomization.
Secondary end points The secondary end points include all-cause mortality at hospital discharge, 6 months, and at trial termination; change in echocardiography regional wall motion score, LV dilation index, and other echo variables measured at randomization and 2 weeks after randomization or before hospital discharge, whichever comes first; and changes in quality of life and physical functioning from 2 weeks after discharge to 6 months after MI for survivors discharged within 2 months of MI.
Study design Patients are randomly assigned to emergency early revascularization or initial medical stabilization. Patients with CS complicating AMI are eligible for the randomized trial when shock is caused by predominant LV failure and eligibility is confirmed—that is, all inclusion criteria are met (Table I), no exclusion crite-
American Heart Journal February 1999
316 Hochman et al
Table II. SHOCK trial exclusion criteria Other severe systemic illness, including contraindications to coronary bypass surgery and including anoxic brain damage Active bleeding or other contraindication to heparin (eg, allergy) Acute mechanical cause of CS Ventricular septal rupture Acute severe mitral regurgitation Isolated right ventricular CS Shock from other causes (eg, hypovolemia, sepsis, anaphylaxis) Severe aortic stenosis or other severe valvular disease Known dilated cardiomyopathy (ischemic and other) Inability to obtain vascular access for catheterization Cardiac tamponade Recent hemorrhage (ie, hematocrit fall ≥15 percentage points within 24 hours of shock onset) Shock from β-blockade or calcium channel blockade Patient is not a candidate for any revascularization procedure (ie, known unsuitable coronary anatomy) Patient or surrogate refuses or cannot give informed consent Refusal by patient’s personal physician
5.8 hours for the 292 patients who have already been randomized, that is, one half of randomized patients have rapid onset of CS complicating AMI. The potential benefits of PTCA/CABG for CS that develops late after MI onset include interruption of the progressive necrosis seen in CS24-26 and limitation of infarct expansion.27 Furthermore, low mortality rates have been reported in patients with late shock when PTCA was performed several days after MI onset.9,12 It is strongly recommended that patients be enrolled as soon as possible after shock diagnosis. However, up to 12 hours is allowed to accommodate the transfer of patients from referring hospitals. The presence of any single exclusion criterion (Table II) renders a patient ineligible for the trial.
gency early revascularization (ERV) or initial medical stabilization (IMS). Randomization is stratified by center and uses a random permuted block design with blocks of varying size.
Trial operator certification All physicians who perform CABG on trial patients were certified according to the criteria used in the BARI trial28 and physicians who perform PTCA were certified according to criteria adapted from TIMI 2 and BARI Trials28,29: the operator must have performed at least 250 PTCA procedures, including at least 20 cases of acute MI with 85% success in the last 100 cases and a complication rate of ≥5% MI, CABG, and ≤2% procedure mortality rate.
Consent
Interventions
All participating centers have Institutional Review Board or Ethics Committee Consent approval. Written informed consent is obtained from the patient or the surrogate (if permitted by the Institional Review Board) before randomization. As of November 1, 1996, it became permissible for centers to use the National Institutes of Health/Federal Drug Administration guidelines for exception from informed consent if the local Board approved the waiver and all required procedures were followed. Application of these guidelines to the SHOCK Trial is appropriate because of the emergent nature of treatment. Two centers obtained approval to waive informed consent.
Trial patients receive routine intensive cardiac care with ventilatory support as needed. The only intervention to be tested is a direct invasive strategy with emergency early mechanical revascularization, if technically feasible, compared with IMS, including possible delayed revascularization. The use of other treatment modalities was left to the discretion of local investigators who were strongly recommended to adhere to routine care, including aspirin and therapeutic anticoagulation with heparin, as recommended by American Heart Association/American College of Cardiology guidelines on acute MI management30 and delineated in the SHOCK Trial protocol.
Emergency revascularization Randomization Patients are randomly assigned through an interactive voice randomization system computer to emer-
Patients randomly assigned to emergency revascularization undergo coronary angiography as soon as possible and within 6 hours of randomization. Revascular-
American Heart Journal Volume 137, Number 2
ization is attempted primarily with PTCA (including approved devices such as stents) but also CABG surgery. IABP insertion as soon as possible after shock onset, and before coronary angiography, is strongly recommended. The choice of revascularization modality and catheter-based techniques is individualized according to the patient’s coronary anatomy and the judgment of the local interventionalist and cardiac surgeon. The goal of catheter-based techniques is an optimal result, with use of stents and a glycoprotein IIb/IIIa receptor antagonist to achieve that goal and reflect clinical practice. The recommended algorithm for revascularization is PTCA of the infarct-related artery for single- and double-vessel disease, with multivessel PTCA as clinically determined. The decision to perform PTCA if TIMI 3 flow was present in the infarct-related artery was left to the investigator’s judgment. Emergency CABG is recommended for severe triple-vessel coronary artery disease, left main artery disease, and failed PTCA. Severe triplevessel disease is defined as 2 total or subtotal occlusions or >90% stenosis in 2 nonculprit major vessels. For triple-vessel coronary artery disease of moderate severity, PTCA of the infarct-related artery followed by CABG or direct CABG is recommended. When the patients are randomly assigned promptly after the onset of the index infarction, direct (without thrombolytic) PTCA is recommended. For patients undergoing emergency CABG, the time from completion of coronary angiography to total vented bypass is recommended to be less than 4 hours.
Initial medical stabilization Routine recommended care includes IABP support; inotropes, vasopressors, and subsequent vasodilators to maintain perfusion guided by hemodynamics; and mechanical ventilation, if necessary. Patients who are thrombolytic eligible and are seen within 24 hours after MI onset are recommended to receive thrombolytic therapy. Thrombolytic eligibility is defined as an absence of absolute contraindications to a thrombolytic, including active internal bleeding, recent (<6 months) cerebrovascular accident, recent (<2 months) intracranial or spinal surgery, intracranial neoplasm, arterial venous malformation, or aneurysm. Revascularization with PTCA or Cardiac Surgery may be performed 54 hours or more after randomization to the IMS arm at the discretion of the cardiologist caring for the patient. This time window ensures a 48-hour minimum separation between revascularization in the ERV and IMS groups. Based on the known high incidence of triple-ves-
Hochman et al 317
sel and left main disease with impaired LV function in patients with CS complicating acute MI, performance of coronary angiography ≥54 hours after randomization is encouraged. The only indications for PTCA or cardiac surgery <54 hours after randomization are for patients who have (1) a mechanical complication (ie, ventricular septal rupture or papillary muscle/chordal rupture/severe mitral regurgitation) or (2) angina at rest (≥30 minutes) that develops at least 24 hours after MI, with ST depression or elevation ≥1 mm, or T-wave inversions in at least 2 related leads that is refractory to medical management, including IABP.
Measurement Information on all aspects of the patient and the patient’s hospital stay is collected, including demographics; medical history; characteristics and timing of the MI and diagnosis of CS; medications; interventions; adverse events; and hemodynamics closest to the time of CS, randomization, 6 hours after randomization and 24 hours after randomization. Electrocardiograms are obtained for trial patients at the time of the index MI and at the time of shock onset, with 2 follow-up recordings demonstrating MI evolution. Creatine phosphokinase measurements with myocardial band fractions are obtained at 8, 16, and 24 hours after the index MI and at 8, 16, and 24 hours after randomization for those with delayed shock. Two-dimensional echocardiograms with color flow Doppler are obtained for trial patients as soon as possible, within 24 hours of randomization, and repeated 2 weeks after randomization or before discharge, whichever comes first. Baseline studies performed after shock onset but before randomization are permissible. It is strongly encouraged that Doppler echocardiography be performed before randomization for patients with a first inferior MI and for those with absence of extensive electrocardiographic abnormalities to rule out mechanical causes of shock and aortic dissection. Transesophageal echocardiograms performed when the transthoracic studies are technically suboptimal are permissible. Quality of life and physical function assessments are completed by telephone interview with trial patients at 2 weeks after hospital discharge and 6 months after MI. A one-year post-MI interview is also conducted with trial patients randomly assigned before the last year of the enrollment period. All patients alive at their last contact will have vital status assessed at trial termination, potentially providing 5.75 years of followup for the earliest enrollees.
American Heart Journal February 1999
318 Hochman et al
Quality of life is assessed with single-response questions including Andrews’ “Ladder of Life”31 and a 7point Likert scale regarding satisfaction with life as a whole.32 Quality of life is assessed with the Multidimensional Index of Life Quality (MILQ),33 developed specifically for patients with cardiovascular disease. Functional status is assessed with a modified form of the Sickness Impact Profile.34 New York Heart Association congestive heart failure class 1 through 4 functional status is also assessed during these interviews with 4 questions so that a judgment may be made using consistent criteria for all patients.
Angiography/angioplasty core laboratory All coronary angiograms and angioplasty films in both the Emergency Revascularization and Initial Medical Stabilization groups are sent to the Data Coordinating Center, where they are coded so that the core laboratory is blinded to site and patient identification number and treatment group. Each film is read independently by 2 angiographers with standardized data forms. A third reader resolves any discrepancies between the 2 primary readers using a third set of data forms. Data are collected on the extent and severity of coronary artery disease, TIMI flow, collateral flow, thrombus, and LV ejection fraction (when a left ventriculogram is performed). Electrocardiograms are provided to help identify the infarct-related artery. PTCA success is graded by change in infarct-related artery perfusion status (TIMI flow), post-intervention stenosis, and change in stenosis severity (at least 20% absolute decrease in stenosis). Quantitative coronary angiography is performed on the infarct-related artery before and after PTCA by the Arriflex QANSAD (1993-1996) and MEDIS CMS-QCA (19961988) systems. Mitral regurgitation, if present, is graded.
Echocardiography core laboratory Echocardiograms with color flow and continuous wave Doppler flow are sent to the Data Coordinating Center for preparation so that the trained echocardiographic readers are blinded to site and patient identification, treatment group, and timing of the echo (baseline vs follow-up). Each tape is read by 1 echocardiographer. Standardized data forms are completed, which include parameters of LV size and function, right ventricular size and function, the function of the infarct and remote (uninfarcted) myocardium, extent of mitral regurgitation, the Doppler transmitral filling pattern and wall motion score index.
Patient registry To ensure that all potentially eligible patients are considered for the trial and to minimize the possibility that an enrolling center systematically excludes subgroups of patients from the trial, a prospective patient registry is maintained at all enrolling centers for patients with suspected CS complicating acute MI. All such patients are registered regardless of trial eligibility. Demographics, medical history, baseline and infarct characteristics, selected procedures, medications, and vital status at hospital discharge are recorded for Registry patients. Compliance with registration of patients is monitored by clinic site audits whereby a list of all patients with a discharge diagnosis of acute MI and CS are compared with those entered into the SHOCK Registry. Complete capture was attained by retrospectively registering any missed patients. The Registry closed August 31, 1997, with 1190 patients and was replaced with a screening log.
Sample size The SHOCK Trial was designed to detect an absolute group difference in all-cause 30-day mortality of 20% with 90% power. To monitor the data closely, a continuous sequential design was selected. A continuous plan was felt to be appropriate because the primary study end point (30-day mortality) is obtained relatively quickly and there was a strong sentiment to act promptly on an observation of improved survival in this critically ill population. A restricted sequential plan first proposed by Armitage35 was implemented. Under this design, paired observations are used as a monitoring tool. Observations are paired according to date of randomization, with pair members belonging to different treatment groups. When the outcomes of the 2 patients constituting the pair are discordant, this provides information regarding a treatment preference. The discordant pair results form a binary random walk and are plotted on a figure with boundaries for early stopping. The required sample size is a function of the desired statistical power, the overall type I error rate, and the anticipated mortality rates. Two sets of ERV versus IMS mortality rates were considered when designing the trial: (1) 55% versus 75% and (2) 30% versus 50%, respectively. The sample size required to ensure 90% power while restricting the overall type I error rate of the trial is 312 for the first setting and 328 for the latter. Therefore in the absence of unbiased information regarding the
American Heart Journal Volume 137, Number 2
absolute mortality rates likely to be observed, the trial was designed to accrue 328 patients (164 per arm). Selection of this higher target enrollment figure also allowed for high power as long as the final sample size exceeded 90% of the target (ie, enrollment of at least 296 patients). The power to detect an absolute 20% group difference based on the 2 sets of mortality rates described above is 88% and 87%, respectively, if 296 patients are enrolled.
Data and Safety Monitoring Board The data are continuously monitored by the Data Coordinating Center statistician and a summary report of 30-day all-cause mortality is reviewed by the Chair of the Data and Safety Monitoring Board on a monthly basis. The DSMB meets semiannually, is completely unblinded, and reviews study end points, trial safety, protocol violations, and data quality and completeness. The primary study end point, 30-day all-cause mortality, as well the secondary end points of the study, which include 6month survival rate and LV ejection fraction, are reviewed. Selected subgroup comparisons are also examined with tests of interaction. These include subgroups based on sex, age, and early versus late onset of CS. With regard to trial safety, the DSMB Chair is immediately notified of occurrences of deaths in the ERV group that are temporally associated with a coronary angiography or angioplasty procedure or instances of emergency bypass surgery required as a result of a procedural complication in the cardiac catheterization laboratory. Safety monitoring reports also include rates of adverse events occurring in each treatment group. Protocol violations are tracked with regard to type and origin (clinic site). These include failure to meet trial eligibility criteria and treatment crossover—for example, Initial Medical Stabilization group patients who undergo emergent revascularization outside of circumstances in which such treatment is clinically indicated per protocol criteria, and patients randomly assigned to the ERV group in which no attempt or a late attempt at revascularization is made. Trial investigators, including the Study Chair, are blinded to all primary and secondary outcomes, including overall trial mortality. Trial investigators (but not the Study Chair) are additionally blinded to the rates of coronary angiography and attempted revascularization in both trial arms.
Hochman et al 319
Analysis plan overview Primary analysis This analysis will be according to the intention-totreat principle for all study end points. The primary study end point, 30-day all cause mortality after randomization, will be evaluated with logistic regression. Clinic site differences in mortality rate between the 2 arms will be examined, as will differences in baseline patient characteristics. It is expected that these differences will be minimal because of randomization; however, a covariate-adjusted odds ratio for mortality will be reported as the primary trial result if significant group differences are identified. Cox proportional hazards modeling will also be used to examine the time course of deaths and to estimate a treatment effect for long-term survival (a secondary study end point).
Secondary analyses Four types of secondary analyses will be conducted on the primary study end point to better evaluate treatment efficacy. All involve the exclusion of patients determined to be trial ineligible or who had rapid death after randomization. Group 1A: Patients who were presumed eligible at the time of randomization but were determined shortly after randomization to have a disease process other than CS caused by acute MI with predominant LV failure, such as severe mitral regurgitation or aortic dissection. Group 1B: Patients who violated eligibility criteria at randomization (eg, shock diagnosis or randomization outside the prescribed time window or not meeting clinical or hemodynamic entry criteria defining CS). Group 2: Patients in either trial arm who die too rapidly to receive treatment, defined as within 30 minutes of randomization. The four secondary analyses proposed exclude the following groups: (1) group 1A, (2) groups 1A + 1B, (3) groups 2 + 1A, and (4) groups 2 + 1A + 1B.
Subgroup analyses There may be certain subgroups of patients in which the ERV versus IMS group mortality difference may be more or less pronounced than the difference observed in the overall analysis. Subgroup analyses of the primary study end point will be conducted with a logistic regression model with interaction terms between the subgroup identifier and the treatment group variable (ERV vs IMS). The study protocol details the scope of prespecified subgroup analyses to be conducted. Such analyses will focus on groups defined by patient age, sex, type of admission (transfer vs direct admission),
American Heart Journal February 1999
320 Hochman et al
geographic location, medical history and therapy, timing of CS and randomization, coronary anatomy and MI location, use of supportive therapies, and measures of cardiac function. An important subgroup is comprised of patients randomly assigned within 6 hours of MI symptom onset.
Timeline Thirty international centers are participating in the randomized trial and registry. Trial enrollment closed on November 30, 1998.
13.
14.
15.
16.
We thank Adriana Gianella for preparation of the manuscript and Venu Menon, MD, for his assistance.
17.
References
18.
1. Becker RC, Gore JM, Lambrew C, Weaver WD, Rubison RM, French WJ, et al. A composite view of cardiac rupture in the United States National Registry of Myocardial Infarction. J Amer Coll Cardiol 1996;27:1321-6. 2. Goldberg RJ, Gore JM, Alpert J, et al. Cardiogenic shock acute myocardial infarction. Incident and mortality from a community-wide perspective, 1975 to 1988. N Engl J Med 1991;325,1117-22. 3. Brodie B, Stuckey T, Weintraub R, et al. Timing and mechanism of death after direct angioplasty for acute myocardial infarction. J Amer Coll Cardiol 1995;25:295A-296A. 4. Himbert, D, Juliard JM, Steg, PG, Karrillon GJ, Aumont MC, Gourgon R. Limits of reperfusion therapy for immediate cardiogenic shock complicating acute myocardial infarction. Am J Cardiol 1994;74:492-4. 5. Holmes DR Jr, Bates ER, Kleiman NS, et al. Contemporary reperfusion therapy for cardiogenic shock: the GUSTO-I trial experience. The GUSTO-I Investigators. Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries. J Am Coll Cardiol 1995;26:668-74. 6. Hochman JS, Boland J, Sleeper LA, et al. Current spectrum of cardiogenic shock and effect of early revascularization on mortality: results of an international registry. Circulation 1995;91:873-81. 7. Menon V, Hochman JS, Holmes D, Bates E, Califf RM, Topol E, et al for the GUSTO III Investigators. Lack of progress in cardiogenic shock: lessons from the GUSTO trials. J Am Coll Cardiol 1998;31:1057-127. 8. Disler L. Haitas B, Benjamin J, Steingo L, McKibbin J. Cardiogenic shock in evolving myocardial infarction: treatment by angioplasty and streptokinase. Heart Lung 1987;16:649-52. 9. Lee L, Bates ER, Pitt B, Walton JA, Laufer N, WW Oneill. Percutaneous transluminal coronary angioplasty improves survival in acute myocardial infarction complicated by cardiogenic shock. Circulation 1988;78:1345-51. 10. Verna E, Repetto S, Boscarini M, Ghezzi I, Binaghi G. Emergency coronary angioplasty in patients with severe left ventricular dysfunction or cardiogenic shock after acute myocardial infarction. Eur Heart J 1989;10:958-66. 11. Meyer P, Blanc P, Baudouy M, Morand P. Treatment of primary cardiogenic shock by coronary transluminal angioplasty during the acute phase of myocardial infarction. Arch Mal Coeur Vaiss 1990; 83:329-34. 12. Moosvi AR, Khaja F, Villanueva L, Gheorghiade M, Douthat L, Gold-
19.
20.
21.
22.
23. 24.
25. 26.
27.
28.
29.
30.
stein S. Early revascularization improves survival in cardiogenic shock complicating acute myocardial infarction. J Am Coll Cardiol 1992;19:907-14. Yamamoto H, Hayashi Y, Oka Y, et al. Efficacy of percutaneous transluminal coronary angioplasty in patients with acute myocardial infarction complicated by cardiogenic shock. Jpn Cir J 1992;56:815-21. Seydoux C, Goy JJ, Beuret P, et al. Effectiveness of percutaneous transluminal coronary angioplasty in cardiogenic shock during acute myocardial infarction. Am J Cardiol 1992;69:968-9. Hibbard MD, Holmes DR Jr, Baily KR, et al. Percutaneous transluminal coronary angioplasty in patients with cardiogenic shock. J Am Coll Cardiol 1992;19:639-46. Laks H, Rosenkranz E, Buckberg GD. Surgical treatment of cardiogenic shock after myocardial infarction. Circulation 1986;74:III-11-6. Guyton RA, Arcidi JM Jr, Langford DA, Morris DC, Liberman HA, Hatcher CR Jr. Emergency coronary bypass for cardiogenic shock. Circulation 1987;76:V22-7. Bolooki H. Emergency cardiac procedures in patients in cardiogenic shock due to complications of coronary artery disease. Circulation 1989;79:II37-48. Phillips SJ, Zeff RH, Skinner JR, Toon RS, Grignon A, Kongtahworn C. Reperfusion protocol and results in 738 patients with evolving myocardial infarction. Ann Thorac Surg 1986;41:119-25. Kirklin JK, Blackstone EH, Zorn GL Jr, et al. Intermediate-term results of coronary artery bypass grafting for acute myocardial infarction. Circulation 1985;72:II-175-8. Subramanian VA, Roberts AF, Zema MJ, et al. Cardiogenic shock following acute myocardial infarction: late functional results after emergency cardiac surgery. NY State J Med 1980;80:947-52. Keon WJ. New therapeutic approaches to acute myocardial infarction (MI) and its size: the surgical approach. Avd Exp Med Biol 1977;82:394-7. Facilities and services in the United States. In hospital statistics. Chicago: American Hospital association; 1992. p. 208-9. Alonso DR, Scheidt S, Post M, Killip T. Pathophysiology of cardiogenic shock: quantification of myocardial necrosis, clinical pathologic, and electrocardiographic correlations. Circulation 1993;48:588-96. Page DR, Caulfield BJ, Kastor JA, et al. Myocardial changes associated with cardiogenic shock. N Engl J Med 1971;285:133-7. Gutovitz AL, Sobel BE, Roberts R. Progressive nature of myocardial injury in selected patients with cardiogenic shock. J Am Coll Cardiol 1978;41:469-75. Hochman JS, Choo H. Limitations of myocardial infarct expansion by reperfusion independent of myocardial salvage. Circulation 1987;75:299-306. BARI Trial Investigators. Protocal for the Bypass Angioplasty Revascularization Investigation (BARI). Circulation 1991;84 (suppl):V-1-V-27. Passamani E, Morrison H, Herman M, Grose R, Chaitman B, Rogers W, et al for the TIMI Investigators. Thrombolysis in Myocardial Infarction (TIMI) phase II pilot study: tissue plasminigen activator followed by percutaneous transluminal coronary angioplasty. J Am Coll Cardiol 1987;10:51B-64B. Gunnar RM, Bourdillon PD, Dixon DW, Fuster V, Karp RB, Kennedy JW, et al. ACC/AHA guidelines for the early management of patients with acute myocardial infarction. A report of the American College of Cardiology/American Heart Association Task Force on Assessment of Diagnostic and Therapeutic Cardiovascular Procedures (subcommit-
American Heart Journal Volume 137, Number 2
31. 32. 33.
34
35.
tee to develop guidelines for the early management of patients with acute myocardial infarction. Circulation 1990;82:664-707. Andrews FM, Withey SB. Social indicators of well-being: Americans’ perceptions of life quality. New York: Plenum; 1976. Campbell A, Converse PE, Rogers WL. The quality of American life: Perceptions, evaluations, and satisfactions. New York: Russell Sage; 1976. Avis NE, Smith KW, Hambleton RK, Feldman HA, Selwyn A, Jacobs A. Development of the Multidimensional Index of Life Quality: a quality of life measure for cardiovascular disease. Medical Care 1996;34:1102-20. deBruin AF, Buys M, deWitte LP, Diederiks JPM. The Sickness Impact Profile: SIP68, a short generic version. First evaluation of the reliability and reproducibility. J Clin Epid 1994;47:863-71. Armitage P. Restricted sequential procedures. Biometrika 1957; 44:9-26.
Appendix The following are committee members, principal and co-principal investigators, and study coordinators in the SHOCK Trial. Executive Committee: Judith S. Hochman, Study Chair; Thierry LeJemtel, Study CoChair; Philip Aylward, Jean Boland, Jacques Col, O. Wayne Isom, Michael Picard, Timothy Sanborn, Lynn Sleeper, Sonjá M. McKinley, Harvey White, and Patrice Desvigne-Nickens (ex officio). Steering Committee: Executive Committe members and principal investigator from each clinical center. Publications Committee: Harvey White, Chair; James Abel, Judith Hochman, Thierry LeJemtel, Lynn Sleeper and John Webb. Data and Safety Monitoring Board: Eugene Braunwald, Chair; Floyd Loop, Charles McCarthy, Neal Scott, David Williams, and Janet Wittes. Clinical Centers (alphabetical by center): Mark Porway and Barbara Burkott, Baystate Medical Center (Springfield, Mass); Alice Jacobs and Mary Mazur, Boston University Medical Center (Boston); Susan Graham, Syed Raza, and Joie Celano, Buffalo General Hospital (Buffalo, NY); Jean Boland, Suzanne Pourbaix, and Mireille Massoz, C.H.R. Citadelle (Liege, Belgium); Jacques Col and Reine Lauwers, Cliniques Universitaires St. Luc (Brussels, Belgium); Philip Aylward and Caroline Thomas, Flinders Medical Centre (Adelaide, SA, Australia); Jo Dens, Frans Van de Werf, and Connie Luys, Gasthuisberg Universitaire Hospital (Leuven, Belgium); Harvey White, John French, and Barbara Williams, Green Lane Hospital (New Zealand); Pamela R. Paulsen and Elizabeth Miller, Hennepin County Medical Center (Minneapolis, Minn); Robert Forman, E. Scott Monrad, Marie Galvao, and Margaret Jones, J.D. Weiler Hospital of the AECOM (Bronx, NY); Jeffrey Brinker and Vicki Coombs, Johns Hopkins Hospital (Baltimore, Md); Mark Menegus, Mark
Hochman et al 321
Greenberg, and Barbara Levine, Montefiore/AECOM (Bronx, NY); John A. Ambrose, Samin Sharma, Thomas Cocke, Denise Ratner, and Eppie Brown, Mt Sinai Medical Center (New York); David H. Miller and Susan Hosat, New York Hospital/Cornell Medical Center (New York); S. Chiu Wong and Maisie Brown, New York Hospital Medical Center of Queens (Flushing, NY); Expedito Ribeiro, Antonio Carlos Carvalho, Antonia Petrizzo, and Claudia Rodrigues Alves, Paulista School of Medicine (Sao Paulo, Brazil); James Slater, Angela Palazzo, and Deborah Tormey, St Luke’s-Roosevelt Hospital Center (New York); Kenneth Baran, Patrick Koller, Patti Filkins, and Cheryl Iacarella, St Paul Heart Clinic (St Paul, Minn); John Webb, Christopher Thompson, and Elizabeth Buller, St Paul’s Hospital (Vancouver, BC, Canada); John Dervan, William Lawson, and Patricia Montes, State University of New York at Stony Brook; Matthias Pfisterer, Nikolaus Jander and Marcus Weinbacher, University Hospital Basel (Basel, Switzerland); Vladimir Dzavik and Cheryl Kee, University of Alberta Hospital (Edmonton, Alberta, Canada); J. David Talley, Mindy Dearen, Millie Rawert, and Rebecca Pacheo, University of Arkansas for Medical Sciences (Little Rock); Bonnie Weiner and Marie Borbone, University of Massachusetts Medical Center; Worcester, Mass; Abel Moreyra, Sebastian Palmeri, and Mary Helen Hosler, University of Medicine and Dentistry of New Jersey (New Brunswick, NJ); Eric Bates and Patsy Bruenger, University of Michigan Medical Center (Ann Arbor, Mich); L. David Hillis, Joaquin Cigarroa, and Jeanette Kissee, University of Texas SW Medical Center (Dallas); David Faxon and Rajinder Singh, USC Medical Center (Los Angeles); Christopher Buller, Ken Gin, and Rebecca Fox, Vancouver General Hospital (Vancouver, BC, Canada); Richard Steingart and Suzanne Parker, Winthrop University Hospital (Mineola, NY). Clinical Coordinating Center: Judith S. Hochman, Thierry H. LeJemtel, and Emilie Godfrey, St Luke’s/Roosevelt Hospital Center, Columbia University (New York). Data Coordinating Center: Sonja M. McKinlay and Lynn A. Sleeper, New England Research Institutes (Watertown, Mass). Echocardiography Core Laboratory: Michael H. Picard and Ravin Davidoff, Massachusetts General Hospital (Boston). Angiography Core Laboratory: Timothy Sanborn and Geoffrey Bergman, New York Hospital/Cornell Medical Center (New York). Program Administration: Patrice Desvigne-Nickens, National Heart, Lung, and Blood Institute, National Institutes of Health (Bethesda, Md).