Author + information
- Received January 18, 2003
- Revision received March 20, 2003
- Accepted March 27, 2003
- Published online October 15, 2003.
- John G Webb, MD, FACC*,* (, )
- April M Lowe, MS†,
- Timothy A Sanborn, MD, FACC‡,
- Harvey D White, DSc§,
- Lynn A Sleeper, ScD†,
- Ronald G Carere, MD, FACC*,
- Christopher E Buller, MD, FACC∥,
- S.Chiu Wong, MD, FACC¶,
- Jean Boland, MD#,
- Vlad Dzavik, MD**,
- Mark Porway, MD, FACC††,
- Gordon Pate, MB*,
- Geoffrey Bergman, MD, FACC¶,
- Judith S Hochman, MD, FACC‡‡,
- for the SHOCK Investigators
- ↵*Reprint requests and correspondence:
Dr. John G. Webb, Director, Cardiac Catheterization and Interventional Cardiology, St. Paul's Hospital, 1081 Burrard Street, Vancouver, BC, Canada V6Z 1Y6.
Objectives We examined the clinical, angiographic, and procedural characteristics determining survival after percutaneous coronary intervention (PCI) for cardiogenic shock.
Background The SHOCK (SHould we emergently revascularize Occluded coronaries for Cardiogenic shocK?) trial prospectively enrolled patients with shock complicating acute myocardial infarction (MI). Patients were randomized to a strategy of early revascularization or initial medical stabilization.
Methods Patients randomized to early revascularization underwent PCI or bypass surgery on the basis of predefined clinical criteria. Patients randomized to early revascularization who underwent PCI and had angiographic films available for analysis are the subject of this report (n = 82).
Results The median time from MI to PCI was 11 h. The majority of patients had occluded culprit arteries (Thrombolysis In Myocardial Infarction [TIMI] grade 0 or 1 flow in 62%) and multivessel disease (81%). One-year mortality in PCI patients was 50%. Mortality was 39% if PCI was successful but 85% if unsuccessful (p < 0.001). Mortality was 38% if TIMI flow grade 3 was achieved, 55% with TIMI grade 2 flow, and 100% with TIMI grade 0 or 1 flow (p < 0.001). Mortality was 67% if severe mitral regurgitation was documented. Independent correlates of mortality were as follows: increasing age (p < 0.001), lower systolic blood pressure (p = 0.009), increasing time from randomization to PCI (p = 0.019), lower post-PCI TIMI flow (0/1 vs. 2/3) (p < 0.001), and multivessel PCI (p = 0.040).
Conclusions Restoration of coronary blood flow is a major predictor of survival in cardiogenic shock. Benefit appears to extend beyond the generally accepted 12-h post-infarction window. Surgery should be considered in shock patients with severe mitral insufficiency or multivessel disease not amenable to relatively complete percutaneous revascularization.
Retrospective studies suggest that early percutaneous coronary intervention (PCI) may improve outcome in patients with cardiogenic shock (1–5). The randomized SHOCK (SHould we emergently revascularize Occluded coronaries for Cardiogenic shocK?) trial showed that a strategy of early revascularization, either with surgery or PCI as deemed appropriate by the treating cardiologist, increased one-year survival from 34% to 47% (p = 0.025) compared with initial aggressive medical therapy in patients with shock due to left ventricular dysfunction complicating acute myocardial infarction (MI) (6). In this report, we examine the clinical, angiographic, and procedural characteristics and relationship to late mortality for patients undergoing PCI in the early revascularization arm of the SHOCK trial.
The SHOCK trial enrolled patients in 36 centers between 1993 and 1998 (6–8). Eligible patients had predominant left ventricular dysfunction resulting in cardiogenic shock within 36 h of MI. Patients were randomly assigned within 12 h of shock diagnosis to initial medical stabilization or emergency revascularization.
The SHOCK trial protocol recommended PCI of the infarct-related artery if patients had one- or two-vessel coronary artery disease. Percutaneous coronary intervention was also recommended if patients had three-vessel disease but the non-culprit vessel stenosis was <90% or supplied small branch vessels. It was recommended that PCI be performed only on the infarct-related artery.
Emergency coronary artery bypass grafting was recommended for patients with left main coronary artery stenosis ≥50%, patients with stenoses >90% in two non-culprit major vessels, patients in whom PCI was unsuccessful or who had disease unsuitable for PCI. It was also recommended that patients with three-vessel coronary artery disease who had successful PCI of a culprit stenosis be evaluated for subsequent coronary bypass surgery during the same hospitalization. Although these recommendations were made, the choice of revascularization procedure was individualized and decided solely by the site physicians.
A total of 302 patients were enrolled in the SHOCK trial. Of these, 152 patients were randomly assigned to emergency revascularization (Fig. 1). Twenty of these patients did not undergo revascularization because of rapid death, minor disease, or unsuitable anatomy. Thus, revascularization was attempted in 132 patients: initial coronary artery bypass in 48 patients and initial PCI in 84. Eight PCI patients subsequently underwent bypass surgery within 24 h of PCI, which was either unsuccessful or performed as a bridge to surgery. A total of 82 patients randomly assigned to emergency revascularization and selected for PCI (excluding two additional patients for whom the angiographic film was unavailable for analysis) are the subject of this report.
Angiographic core laboratory
Angiograms were reviewed at the core laboratory (New York Hospital/Cornell Medical Center) by two independent, blinded readers. Discrepancies were resolved by a third reader. Thrombolysis In Myocardial Infarction (TIMI) criteria (9), an estimate of the amount of myocardium at risk based on the coronary artery jeopardy score (10)and other standard criteria were used (11–13). Significant coronary disease was defined as ≥50% diameter stenosis. A significant stenosis of the left main coronary artery was counted as two-vessel disease if the left coronary was non-dominant and as three-vessel disease if dominant, as well as left main disease. Percutaneous coronary intervention success was defined as the combination of a residual stenosis ≤50%, a ≥20% reduction in stenosis, and TIMI 2 or 3 flow.
The distributions of categorical variables were compared using Fisher exact test, and the distributions of ordered categorical variables, such as TIMI flow and the year PCI was performed, were compared using the Mantel-Haenszel test for linear trend. Group differences in continuous variables were compared using Student’s ttest for comparisons of age and heart rate, and the Wilcoxon rank sum test for the remaining variables. Thirty-day and one-year survival curves were estimated by the Kaplan-Meier method. The log-rank (to assess group differences in overall survival) and Wilcoxon test (to evaluate whether there was group difference in early deaths) were used to compare Kaplan-Meier survival estimates. One-year survival was also analyzed using Cox proportional hazards regression. The survival time of one patient was censored at the time of heart transplantation. Variables such as baseline characteristics, clinical management, hemodynamic and angiographic findings, and outcomes found in Tables 1, 2, and 3⇓⇓with a Cox regression univariate p value ≤0.20 were grouped and entered into intermediate models. The covariates with significant p values from these models were then combined to determine a final model with covariate p values <0.05. All analyses were conducted using the Statistical Analysis System (SAS, Cary, North Carolina) and S-Plus (Seattle, Washington).
The PCI patient and clinical characteristics by vital status at one year are presented in Table 1. The median time from MI to shock was 5.1 h and from MI to PCI was 11 h. Survivors were seven years younger, on average, than non-survivors. Only 15% (3/20) of patients who underwent PCI after an episode of cardiopulmonary resuscitation survived for one year.
Core laboratory angiographic findings are shown in Table 2. The majority of patients had multivessel disease (81%). The culprit lesion was most frequently in the left anterior descending artery (50%) and was usually severely stenotic or occluded (mean diameter stenosis 95 ± 8%). Survivors had higher ejection fractions than non-survivors (34 ± 11 vs. 27 ± 7, p = 0.025). Initial culprit artery patency (TIMI flow grade 2 or 3) was present in 38%, with normal coronary flow (TIMI 3) in 21% of patients. Flow was absent or severely reduced (TIMI 0 or 1) in 62%. There was a 45% one-year survival rate for initial occlusion (TIMI 0 or 1), and a 58% one-year survival rate for initial patency (TIMI 2 or 3 flow) (p = 0.363).
The most common complication (8.5%) occurring in the catheterization laboratory was an arrhythmia requiring urgent treatment. Death in the catheterization laboratory occurred in 3.7% (3 patients). Although 81% of patients had multivessel disease, most PCI patients underwent single-vessel procedures (87%). During the study period there was an increase in the frequency of multivessel procedures from 0% in 1993 to 1994, to 10% in 1995 to 1996 and 23% in 1997 to 1998 (p = 0.018). Stents were implanted in the culprit artery in 28 patients (34%). The culprit artery stenting rate increased during the study from 0% in 1993 to 1994, to 10% in 1995 to 1996 to 74% in 1997 to 1998 (p < 0.001). Percutaneous coronary intervention was more often successful in stented than unstented patients (93% vs. 67%, p = 0.013), although one-year survival was similar (54% vs. 48%, p = 0.82) (Fig. 2). Glycoprotein (GP) IIb/IIIa (abciximab) use increased during the study period from 0% during 1993 to 1994, to 7% between 1995 to 1996 and 81% from 1997 to 1998 (p < 0.001).
Overall survival in the 82 patients selected for PCI was 54% (n = 44) at 30 days and 50% (n = 41) at one year. Percutaneous coronary intervention success was achieved in 76%. Thirty-day survival was 65% after successful PCI but only 20% if unsuccessful (p < 0.001). At one year, these values were 61% and 15%, respectively (p < 0.001). There were 17 patients with final post-PCI stenosis ≤50% who had reduced angiographic coronary flow (TIMI 0, 1, 2); this represents 53% of all patients with TIMI 0, 1, or 2 flow post-PCI. Their 30-day and one-year survival was 47%. Coronary flow after PCI was a significant determinant of survival (Fig. 3). All survivors had patent (TIMI 2 or 3 flow grade) infarct-related arteries immediately after PCI; however, patency did not guarantee survival (Table 3).
The PCI success rates by age were 58% in the 12 patients ≥75 years and 79% in patients <75 years (p = 0.15). One-year survival after PCI was lower in the elderly (17% vs. 56%, p = 0.026).
Despite increasing rates of multivessel PCI, stent placement, and GP IIb/IIIa antagonist use over the five-year course of the trial, the one-year survival rate for patients who underwent PCI was similar over time (Fig. 4). One-year survival was 55% after single-vessel PCI but only 20% after a single-stage multivessel procedure (p = 0.048). One-year survival was 38% for eight patients who underwent emergency coronary bypass surgery within 24 h of PCI. Among patients undergoing PCI, 18% had severe mitral regurgitation but did not undergo urgent valve surgery, and their survival was 33%. In PCI patients without severe mitral regurgitation, survival was 64% (p = 0.202).
Correlates of death
Independent correlates of death at one year based on multivariate Cox regression analysis are as follows: increasing age (p < 0.001), lower systolic blood pressure recorded on support measures (p = 0.009), increasing time from randomization to PCI (p = 0.019), lower final post-PCI TIMI flow grade (0/1 vs. 2/3) (p < 0.001), and multivessel PCI (p = 0.040) (Table 4). There was no association between time from MI or time from shock to PCI and outcome. Though time from randomization to PCI is a strong correlate of death, there were differences between patients who underwent rapid PCI (<0.87 h to PCI) and delayed PCI (≥0.87 h) signifying that severity of illness is contributing to this association. Patients with delayed PCI tended to be older than those with rapid PCI (67 ± 10 vs. 63 ± 10 years, p = 0.06), have more diabetes (37% vs. 15%, p = 0.038), a trend towards more left main disease (18% vs. 8%, p = 0.19) as well as a trend toward higher jeopardy score (7.6 ± 3.4 vs. 6.3 ± 3.0, p = 0.08), thus implying that they are a sicker group of patients in our cohort. When four outliers with long times from randomization to PCI (5.4 to 6.5 h) were removed, age, supported systolic blood pressure, and final TIMI flow remained significant, but multivessel PCI (p = 0.092) and time from randomization to PCI (p = 0.096) were no longer significant.
Ejection fraction and mitral regurgitation grade were not found to be correlates of death when included in multivariate modeling, but these were only available for one-third of the patients. Cardiopulmonary resuscitation was not included as it was only available for two-thirds of the patients.
Emergency revascularization in the SHOCK Trial was associated with an increase in one-year survival, consistent with 132 lives saved per 1,000 patients treated. This represents a benefit larger than for any other therapy tested for acute MI (6,8). Percutaneous coronary intervention constituted the most frequent mode of revascularization chosen when a strategy of emergency revascularization was assigned by protocol. In-hospital survival after PCI in the SHOCK trial was similar to that reported in the SHOCK registry (54% vs. 56%, respectively) (7,14).
Infarct artery patency
Early reperfusion of an infarct-related artery is associated with improved outcome in shock (4,14,15). Thrombolysis results in relatively low rates of reperfusion in patients in whom shock is already established (3,16), and the likelihood of sustained reperfusion is higher with PCI (17–20). Post-PCI patency of the infarct artery is strongly associated with survival, even after adjustment for important clinical and hemodynamic characteristics. No patient who had an occluded infarct artery (TIMI flow grade 0 or 1) after PCI survived.
PCI was performed a median of 11 h after the onset of MI symptoms. It appears that, in the setting of shock, the benefit of PCI extends beyond the usually accepted 12-h post-MI window. Importantly, therapy appears effective for those transferred from primary care hospitals. The apparent lack of association between earlier times from MI or shock to PCI and survival may be explained by the use of intra-aortic balloon pump and thrombolytic agents in almost 90% and 50% of patients, respectively, who underwent PCI and survivor bias, whereby those who did not survive to the procedure are excluded.
The success rate of PCI was relatively low (76%) but consistent with previous retrospective shock studies (3,4), and not unexpected given that most patients had diffuse disease, occluded arteries, and were hemodynamically unstable. This study predates the advent of widespread stenting in the setting of acute MI. Stents were used in 34% of patients, mainly to salvage a failed balloon PCI, and were largely first-generation devices implanted without the benefit of current adjunctive techniques. Despite this, the rate of PCI success was higher in stented than unstented patients. Similarly, GP IIb/IIIa inhibitor and thienopyridine use was infrequent but increased during the nearly six-year trial period. Potentially, the increased use of stenting and adjunctive therapies that improve coronary blood flow might further extend the benefits of PCI (17–21).
Although multivessel disease was present in the majority of patients, most had single-vessel angioplasty. Percutaneous coronary intervention of the culprit lesion alone has been advocated in the management of acute MI (3,4). The prevalence of multivessel disease and the possibility of ischemia at a distance from the infarct zone and of progressive deterioration in left ventricular function may argue for a strategy of more complete revascularization with multivessel PCI or surgery (4,14,17,22).
Although the mode of revascularization was recommended by the protocol on the basis of the extent and severity of disease, the choice was at the discretion of the attending physician. It is notable that equivalent survival with surgery was achieved despite the selection of patients with more severe disease to undergo this means of revascularization (6). Survival after single-stage multivessel PCI procedures was disappointing, whether as a result of patient selection or less satisfactory revascularization. Where the likelihood of successful PCI is low or relatively complete revascularization cannot be achieved, emergency surgery may be preferable to PCI.
The SHOCK study excluded enrollment of patients in whom the investigator determined that mitral valve replacement was clinically indicated. Among enrolled patients who had severe mitral regurgitation and underwent PCI alone, the one- year survival rate was a disappointing 33%. Mitral insufficiency is easily missed in shock patients and should be specifically sought with echocardiography or ventriculography before PCI. Although a reduction in mitral insufficiency may occur with PCI, this is unpredictable and infrequent. Severe mitral insufficiency may warrant early surgical correction (4,14,23).
PCI in the elderly
The SHOCK trial suggested a lack of benefit for early revascularization in patients ≥75 years of age; however, numbers were small, with only 12 PCI patients ≥75 years of age. Although PCI tended to be less often successful in elderly patients, successful PCI is associated with increased survival. Dzavik et al. (24)reported higher survival rates for the 17% of patients ≥75 years of age in the SHOCK registry who were clinically selected to undergo early revascularization compared with those with late or no revascularization. This survival rate (52%) is comparable to the larger cohort of patients <75 years of age who underwent early revascularization in the registry. It seems reasonable to conclude that early revascularization may be appropriate for selected elderly patients and that management must be individualized.
Although patient selection bias is of concern (2,4,14), the SHOCK trial adhered to strict criteria for the diagnosis of shock and randomization-mandated early revascularization. Patients appeared representative of the larger, unselected population in the SHOCK registry (2,14), although by design those who were not suitable for an aggressive strategy (e.g., life-shortening illnesses) were not enrolled (7,8). Only 40% of patients assigned to emergency revascularization had a left ventricular angiogram performed, and therefore associations between survival and ejection fraction and mitral regurgitation are possibly biased. It is likely that the association of certain characteristics and survival was not significant because of the limited number of patients. This study largely predates the advent of widespread stenting and the use of GP IIb/IIIa antagonists (17,20,25).
In the SHOCK trial, successful early restoration of coronary blood flow is a major predictor of survival and an important therapeutic goal. Although very early intervention is preferable, in the setting of shock the benefit of reperfusion appears to extend beyond the generally accepted 12-h post-MI window. Surgery should be considered in shock patients with severe mitral insufficiency or multivessel disease not amenable to relatively complete percutaneous revascularization.
☆ Supported by R01 grants HL50020, HL49970, 1994 to 1999, from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland.
- myocardial infarction
- percutaneous coronary intervention
- SHould we emergently revascularize Occluded coronaries for Cardiogenic shocK
- Thrombolysis In Myocardial Infarction
- Received January 18, 2003.
- Revision received March 20, 2003.
- Accepted March 27, 2003.
- American College of Cardiology Foundation
- Antoniucci D.,
- Valenti R.,
- Santoro G.,
- et al.
- Hochman J.S.,
- Boland J.,
- Sleeper L.A.,
- et al.
- Webb J.G.,
- Hochman J.S.
- Califf R.M.,
- Phillips H.R.,
- Hindman M.C.,
- et al.
- Ryan T.J.,
- Faxon D.P.,
- Gunnary R.M.,
- et al.
- Ellis S.G.,
- Vandormael M.G.,
- Cowley M.J.,
- et al.
- Sheehan F.H.
- Berger P.,
- Holmes D.,
- Stebbins A.,
- Bates E.,
- Califf R.,
- Topol E.
- Bates E.,
- Topol E.
- Hasdai S.,
- Harrington R.A.,
- Hochman J.S.,
- et al.
- Stone G.W.,
- Grines C.L.,
- Cox D.A.,
- et al.
- Cotter G.,
- Kaluski E.,
- Blatt A.,
- et al.
- Webb J.G.,
- Sleeper L.A.,
- Buller C.,
- et al.
- Thompson C.R.,
- Buller C.,
- Sleeper L.A.,
- et al.
- ↵Dzavik V, Sleeper L, Cocke TP, et al. Early revascularization in associated with improved survival in elderly patients with acute myocardial infarction complicated with cardiogenic shock: a report from the SHOCK Trial Registry. Eur Heart J 2003;24:828–37