Prevalence, Predictors, and In-Hospital Outcomes of Non-Infarct Artery Intervention During Primary Percutaneous Coronary Intervention for ST-Segment Elevation Myocardial Infarction (from the National Cardiovascular Data Registry)

Prevalence, Predictors, and In-Hospital Outcomes of Non-Infarct Artery Intervention During Primary Percutaneous Coronary Intervention for ST-Segment Elevation Myocardial Infarction (from the National Cardiovascular Data Registry) Matthew A. Cavender, MDa,*, Sarah Milford-Beland, MSb, Matthew T. Roe, MD, MHSb, Eric D. Peterson, MD, MPHb, William S. Weintraub, MDc, and Sunil V. Rao, MDb Guidelines support percutaneous coronary intervention (PCI) of the noninfarct-related artery during primary PCI for ST-segment elevation myocardial infarction (STEMI) in patients with hemodynamic compromise; however, in patients without hemodynamic compromise, PCI of the noninfarct-related artery is given a class III recommendation. We analyzed the National Cardiovascular Data Registry (n ⴝ 708,481 admissions, 638 sites) to determine the prevalence, predictors, and in-hospital outcomes of primary multivessel PCI from 2004 to 2007. Patients with STEMI and multivessel coronary artery disease who were undergoing primary PCI were identified (n ⴝ 31,681). After excluding the patients treated with staged PCI (n ⴝ 2,745), 10.8% (n ⴝ 3,134) of the remaining population (n ⴝ 28,936) were treated with multivessel PCI. Patients undergoing multivessel PCI were at higher risk and were more likely to be in cardiogenic shock. The overall in-hospital mortality rates were greater in patients undergoing multivessel PCI (7.9% vs 5.1%, p <0.01). Among patients with STEMI and cardiogenic shock (n ⴝ 3,087), those receiving multivessel PCI had greater in-hospital mortality (36.5% vs 27.8%; adjusted odds ratio 1.54, 95% confidence interval 1.22 to 1.95). In conclusion, these data suggest that performing multivessel PCI during primary PCI for STEMI does not improve short-term survival even for patients with cardiogenic shock. These findings suggest the need for definitive studies to evaluate the utility of noninfarct-related artery PCI among patients with STEMI. © 2009 Elsevier Inc. All rights reserved. (Am J Cardiol 2009;104:507–513) Previous studies examining the use of multivessel percutaneous coronary intervention (PCI) in the setting of STsegment elevation myocardial infarction (STEMI) have come to conflicting conclusions regarding its safety; howa Department of Medicine, Duke University Medical Center, Durham, North Carolina; bDuke Clinical Research Institute, Durham, North Carolina; and cChristiana Care Health System, Newark, Delaware. Manuscript received February 20, 2009; revised manuscript received and accepted April 2, 2009. This analysis was supported by a grant from the American College of Cardiology, Washington, DC, and the Society for Cardiac Angiography and Interventions, Washington, DC. This analysis was supported by the National Cardiovascular Data Registry. Dr. Roe was supported by research grants from Schering Plough, BMS/ Sanofi-Aventis, KAI Pharmaceuticals, and DeCODE Genetics; is on the consulting/advisory boards for Schering Plough, KAI Pharmaceuticals; and is on the Speakers Bureau or has received honoraria from Schering Plough and BMS/Sanofi-Aventis. Dr. Peterson has received research grants from Bristol-Myers Squibb, Bristol-Myers Squibb/Sanofi Pharmaceuticals Partnership, Bristol-Myers Squibb/Merck, and Schering (additional disclosure information for Dr. Peterson is available from: http://dcri.org/research/coi.jsp). Dr. Rao is a Consultant for Sanofi-Aventis and member of the Speakers’ Bureaus for Sanofi-Aventis, Bristol Myers Squibb, and the Medicines Company and has received research funding from Momenta Pharmaceuticals, Portola Pharmaceuticals, and Cordis. *Corresponding author: Tel: (919) 684-8111; fax: (919) 681-6448. E-mail address: [email protected] (M. Cavender).

0002-9149/09/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2009.04.016

ever, these studies were often underpowered or did not fully adjust for potential treatment selection biases.1–3 Given this, and that PCI is routinely performed electively in high-risk patients, a re-evaluation of multivessel PCI during primary PCI is relevant in the modern era.4 – 6 We analyzed contemporary data from the large multicenter national cardiovascular data registry (NCDR) to determine the prevalence and predictors of multivessel PCI during primary PCI for STEMI. In addition, we examined both the unadjusted and the adjusted outcomes in patients undergoing multivessel PCI and compared them with the outcomes in patients undergoing PCI of only the infarct-related artery. Methods The NCDR, which is co-sponsored by the American College of Cardiology and the Society for Cardiac Angiography and Interventions, has been previously described.7,8 The NCDR catalogs the clinical data and outcomes of PCI procedures gathered from ⬎600 sites across the United States. The data are entered into NCDR-certified databases at participating institutions and exported in a standard format to a common database at Heart House (Washington, DC). Only institutions whose submissions have met the quality criteria for data reporting are included. The definitions of all variables are prospectively defined by a committee of the NCDR and are available on the NCDR Web www.AJConline.org

508

The American Journal of Cardiology (www.AJConline.org)

Table 1 Baseline characteristics Variable Age (years) Median Interquartile range Men White Body mass index (kg/m2) Median Interquartile range Previous myocardial infarction Previous PCI Previous CABG Cerebrovascular disease Congestive heart failure Peripheral vascular diseases Diabetes mellitus Hypertension* Hypercholesterolemia† Chronic lung disease Renal failure/dialysis Nondialysis renal failure Dialysis Current or past smoker Interval from symptom onset to admission ⱕ6 hours ⬎6 and ⱕ12 hours ⬎12 and ⱕ24 hours ⬎24 and ⱕ48 hours ⬎48 hours and ⱕ7 days No time/silent MI LVEF ⬍30%‡ Cardiogenic shock on presentation Transfer for primary PCI Insurance payor Government Commercial Health Maintenance Organization None Non-United States insurance

Table 2 Baseline procedural characteristics Single-Vessel PCI Multivessel PCI p Value (n ⫽ 25,802) (n ⫽ 3,134) ⬍0.01 62 53–73 72.1% 86.0%

60 52–72 71.5% 83.9%

27.7 24.8, 31.3 19.3%

28.1 25.0, 31.6 17.4%

⬍0.01

17.4% 9.9% 7.7% 9.8% 7.4%

15.1% 5.1% 8.2% 13.2% 6.2%

⬍0.01 ⬍0.01 0.37 ⬍0.01 0.02

23.4% 63.2% 58.6% 12.1%

24.7% 60.4% 56.5% 11.8%

2.9% 0.8% 64.8%

2.6% 0.9% 63.2%

78.8% 10.1% 5.3% 2.6% 2.6% 0.5% 7.0% 10.3%

74.4% 10.6% 6.7% 3.8% 4.1% 0.6% 10.0% 13.8%

26.4%

30.1%

44.6% 31.5% 13.0%

40.6% 33.5% 13.8%

10.8% 0.1%

11.9% 0.2%

0.32 0.19 ⬍0.01

0.06 ⬍0.01 0.05 0.56 ⬎0.99

0.17 ⬍0.01

⬍0.01 ⬍0.01 0.01 ⬍0.01

* Defined as history of hypertension diagnosed and treated with medication, diet, and/or exercise, blood pressure ⬎140 mm Hg systolic or ⬎90 mm Hg diastolic on ⱖ2 occasions, or currently taking antihypertensive pharmacologic therapy. † Hyperlipidemia defined as history of hyperlipidemia diagnosed and/or treated by physician, documentation of total cholesterol ⬎200 mg/dl, low-density lipoprotein ⱖ130 mg/dl, high-density lipoprotein ⬍30 mg/dl, admission cholesterol ⬎200 mg/dl, or triglycerides ⬎150 mg/dl. ‡ LVEF not reported for 24% of patients. CABG ⫽ coronary artery bypass grafting; LVEF ⫽ left ventricular ejection fraction; MI ⫽ myocardial infarction.

site (available from: http://www.acc.org/ncdr/cathlab.htm). Currently, approximately 10% of the data are audited. For the purposes of the present analysis, we analyzed all PCI laboratory visits from April 1, 2004 to March 28, 2007.

Variable Emergent or salvage PCI status Intra-aortic balloon pump use Right heart catheterization Culprit coronary lesion Proximal LAD artery Proximal right, midLAD, proximal circumflex artery Other Contrast volume (ml) Median Quartile 1, quartile 3 Fluoroscopy time (min) Median Quartile 1, quartile 3

Single-Vessel PCI Multivessel PCI p Value (n ⫽ 25,802) (n ⫽ 3,134) 91.3%

87.3%

⬍0.01

11.1%

16.0%

⬍0.01

3.9%

6.3%

⬍0.01 ⬍0.01

16.9% 35.5%

19.8% 32.2%

47.2%

47.6%

200 150, 262

255 200, 336

11.5 7.6, 17.9

16.2 11.3, 24.0

⬍0.01 ⬍0.01

LAD ⫽ left anterior descending.

Patients were only included if they presented with STEMI and were found to have coronary artery disease in ⬎1 major artery. Patients receiving PCI of the left main coronary artery, staged PCI (multiple PCI procedures before hospital discharge), or thrombolytics were excluded from the analysis. The patients were divided into groups according to whether they underwent PCI of only a single vessel or underwent multivessel PCI during the index catheterization. Our analysis included patients who had presented with STEMI and who were treated with primary PCI. Patients with a systolic blood pressure ⬍80 mm Hg and/or a cardiac index ⬍1.8 despite maximal treatment or requiring intravenous inotropes and/or an intra-aortic balloon pump to maintain the systolic blood pressure at ⬎80 mm Hg and/or the cardiac index ⬎1.8 L/min/m2 were classified as being in cardiogenic shock. The NCDR does not specifically define the infarct-related artery; therefore, the infarct-related artery was assumed to be the artery that was treated in patients undergoing single-vessel PCI. In patients with multivessel PCI, the infarct-related artery was considered the artery with the greatest preoperative stenosis (in percentages). In the event of a tie, the artery with the worst pre-PCI TIMI flow (i.e., none, slow, partial, complete) was considered the infarctrelated artery. The greatest lesion risk (i.e., high/C to nonhigh/non-C) was also used to break ties, and if necessary, the first lesion listed on the data collection form was selected as the infarct-related artery among those with equivalent stenosis, TIMI Flow, and lesion risk. The primary end point of our study was in-hospital mortality. The secondary end points included the occurrence of post-PCI stroke, renal failure, bleeding complications, and the use of blood products after left heart catheterization. The NCDR has defined stroke as a central neurologic deficit lasting ⬎72 hours. Renal failure was defined as a new requirement for dialysis or an increase in creatinine to ⬎2 mg/dl and 2 times the baseline creatinine. Bleeding complications included bleeding from any site that required transfusion, prolonged the

Coronary Artery Disease/Multivessel PCI in STEMI

509

Table 3 Factors associated with multivessel percutaneous coronary intervention Variable Previous CABG Interval from symptom onset to admission 6–12 hours (vs ⬍6 hours) 12–24 hours (vs ⱕ6 hours) 24–48 hours (vs ⱕ6 hours) 2–7 days (vs ⱕ6 hours) Silent MI (vs ⱕ6 hours) Cardiogenic shock Congestive heart failure Age (per 10 years) Hospital has teaching program Hospital PCI volume per yr (per 500 PCIs) Government insurance (vs other) Proximal LAD intervention White (vs nonwhite)

Unadjusted OR

95% CI

Adjusted OR

95% CI

p Value

0.48

0.41–0.56

0.51

0.43–0.59

⬍0.01 ⬍0.01

1.13 1.30 1.50 1.61 1.07 1.38 1.40 0.94 0.84 1.05 0.85 1.21 0.86

1.00–1.28 1.11–1.52 1.23–1.82 1.31–1.97 0.62–1.82 1.24–1.54 1.26–1.57 0.91–0.97 0.74–0.97 1.01–1.10 0.79–0.92 1.10–1.32 0.78–0.96

1.13 1.30 1.48 1.55 1.05 1.31 1.31 0.95 0.83 1.06 0.91 1.10 0.90

0.99–1.28 1.11–1.52 1.22–1.80 1.27–1.90 0.62–1.78 1.18–1.46 1.17–1.47 0.92–0.99 0.72–0.95 1.01–1.11 0.84–0.99 1.01–1.21 0.81–0.99

⬍0.01 ⬍0.01 0.01 0.01 0.01 0.03 0.03 0.04

Cardiac index ⫽ 0.58. Note, candidate variables evaluated but not statistically significant shown in Appendix A. Abbreviations as in Tables 1 and 2.

Figure 1. Variability in use of multivessel PCI in primary PCI for STEMI. Prevalence of multivessel coronary interventions varied greatly among centers.

hospital stay, or caused a decrease in hemoglobin ⬎3.0 g/dl. In addition, bleeding complications included external bleeding from the percutaneous access site or hematoma (⬎10 cm for femoral access, ⬎2 cm for radial access, and ⬎5 cm for brachial access). The complete definitions of all variables and clinical conditions have been prospectively defined by a committee of the American College of Cardiology and are available from the American College of Cardiology Web site (http://www.accncdr.com/WebNCDR/ ELEMENTS.ASPX#1). The demographic and baseline characteristics, treatment patterns, angiographic status, and in-hospital outcomes were compared between those patients undergoing single-vessel versus multivessel PCI. Continuous variables are reported as the median (with interquartile ranges), and categorical variables are reported as frequencies. Continuous and ordinal categorical variables were compared using stratum-adjusted Wilcoxon rank-sum tests, and nominal categorical variables were compared using stratum-adjusted chi-square tests, for which stratification was by hospital. In determining the factors independently associated with undergoing multivessel PCI, we considered a comprehen-

sive list of patient and hospital characteristics as candidates (see Appendix A). Using backward elimination, a multivariate logistic regression model in which all variables were significantly associated with undergoing multivessel PCI was developed, and the adjusted odds ratios were calculated for each of these variables. To study the variability in the prevalence of multivessel PCI across the United States, we considered the distribution of multivessel PCI rates across centers. We also calculated the Pearson correlation coefficient between the hospital annual STEMI volume and multivessel PCI rate and tested whether this linear association was significant. For these hospital level analyses, 166 sites with ⬍20 PCI admissions were excluded, because sites with such a small denominator could have had unstable multivessel PCI rates. In examining the relation between multivessel PCI (vs single-vessel PCI) and mortality, unadjusted comparisons were initially performed. The mortality and outcome rates were compared among all patients, as well as among those with cardiogenic shock and those without cardiogenic shock. Multivariate logistic regression models were constructed to adjust for a broad range of potentially confound-

510

The American Journal of Cardiology (www.AJConline.org)

Table 4 Unadjusted in-hospital outcomes Variable

Single-Vessel Multivessel p Value PCI PCI

All patients with STEMI (n) In-hospital mortality Death in laboratory Cerebrovascular accident/stroke Bleeding complications Blood products transfused after laboratory visit Renal failure

25,802 5.12% 0.36% 0.56% 5.30% 9.00%

3,134 7.85% 1.24% 0.72% 6.71% 8.86%

⬍0.01 ⬍0.01 0.31 ⬍0.01 0.59

1.81%

2.31%

0.09

Patients with STEMI without cardiogenic shock (n) In-hospital mortality Death in laboratory Cerebrovascular accident/stroke Bleeding complications Blood products transfused after laboratory visit Renal failure

23,146

2,701

2.53% 0.10% 0.46% 4.53% 7.31%

3.26% 0.52% 0.45% 5.67% 6.60%

0.09 0.02 0.90 0.02 0.13

1.21%

1.23%

0.96

Patients with STEMI and cardiogenic shock (n) In-hospital mortality Death in laboratory Cerebrovascular accident/stroke Bleeding complications Blood products transfused after laboratory visit Renal failure

2,654

433

27.77% 2.64% 1.49% 12.48% 24.72%

36.49% 5.77% 2.56% 13.81% 24.30%

ⱕ0.01 0.25 0.18 0.44 0.72

7.41%

9.72%

0.03

ing patient characteristics in both the cardiogenic shock population and the population without cardiogenic shock (see Appendix B). Because patients within a hospital were more likely to be similar, all adjusted analyses were performed using generalized estimating equation models to account for correlations among clustered responses (i.e., within-hospital correlations). The p values ⬍0.05 were established as the level of statistical significance for all tests. All analyses were performed by the Duke Clinical Research Institute using Statistical Analysis Systems software, version 8.2 (SAS Institute, Cary, North Carolina).

PCI of the infarct-related artery alone (Table 1). In addition, patients who underwent multivessel PCI were at a greater risk than patients who underwent single-vessel PCI, as reflected by a greater incidence of cardiogenic shock, congestive heart failure, left ventricular ejection fraction ⬍30%, and having the proximal left anterior descending as the infarct-related artery (Table 2). Table 3 lists the factors associated with treatment of STEMI with multivessel PCI, including a later presentation from the time of symptom onset, cardiogenic shock, congestive heart failure, and younger age. Variability in the use of multivessel PCI in centers with ⬎20 PCI admissions was high, with single-center use ranging from 0% to 38% (Figure 1). No evidence was found of any unadjusted linear association between STEMI volume and the use of multivessel PCI (Pearson correlation coefficient ⫺0.02712, p ⫽ 0.56); however, after adjustment for various patient characteristics, volume was a predictor of the use of multivessel PCI (Table 3). Table 4 lists the unadjusted rates of the primary and secondary end points between the 2 groups. The unadjusted rate of in-hospital mortality was significantly greater for patients treated with multivessel PCI. Patients treated with multivessel PCI also had an increased rate of bleeding compared with patients treated with single-vessel PCI (6.7% vs 5.3%, p ⱕ0.01). In patients who did not present in cardiogenic shock, mortality was greater in the group treated with multivessel PCI. After adjusting for potential confounders, this trend persisted, although it was not statistically significant (odds ratio of mortality for multivessel PCI vs single-vessel PCI 1.23, 95% confidence interval 0.94 to 1.61; Figure 2). For the patients with cardiogenic shock, the unadjusted rate of in-hospital mortality was significantly greater for patients treated with multivessel PCI (36.49% vs 27.77%, p ⱕ0.01). Patients treated with multivessel PCI also had an increased incidence of renal failure compared with patients treated with single-vessel PCI (9.72% vs 7.4%, p ⫽ 0.03). After adjustment, a statistically significant association was found between multivessel PCI and increased in-hospital mortality in patients with cardiogenic shock (adjusted odds ratio 1.54, 95% confidence interval 1.22 to 1.95; Figure 2). Discussion

Results A total of 708,481 admissions from 638 sites were entered into the NCDR from 2004 to 2007. We identified 31,681 patients undergoing primary PCI for STEMI who had multivessel coronary artery disease. Of these patients, 8.7% (n ⫽ 2,745) underwent staged PCI and were excluded from the analysis, leaving 28,936 patient encounters included in the analysis. Of these, 10.8% (n ⫽ 3,134) were treated with multivessel PCI. Patients who underwent multivessel PCI had a lower incidence of previously documented myocardial infarction, previous PCI or coronary artery bypass grafting, and peripheral vascular disease, but they had a greater body mass index and an increased rate of diabetes that was statistically nonsignificant compared with patients who had received

Our analysis of a large contemporary PCI database has shown that multivessel coronary interventions during primary PCI for STEMI occurs in ⱕ10% of PCI cases in clinical practice; however, its use varied considerably among centers. Although no consistency was found in its use, we did find that multivessel PCI in the setting of primary PCI was associated with worse outcomes. Among patients with STEMI without hemodynamic compromise, multivessel PCI did not improve in-hospital outcomes, supporting the current American College of Cardiology/American Heart Association class III guideline recommendations. For patients with cardiogenic shock, increased deaths occurred among patients treated with multivessel PCI compared with those treated with single-vessel PCI of the infarct-related artery alone. These findings conflict with the

Coronary Artery Disease/Multivessel PCI in STEMI

511

Figure 2. Odds ratios for mortality with multivessel PCI. Patients in cardiogenic shock treated with multivessel PCI had increased odds of death compared with patients treated with PCI of infarct-related artery only, even after adjustment for confounding variables. Patients not in cardiogenic shock showed trend toward increased odds of death that was not statistically significant. Note, candidate variables controlled for are listed in Appendix B.

current guideline recommendations and support the need for future definitive investigation.9 Previous studies of the role of multivessel PCI in patients with STEMI have come to disparate conclusions. A retrospective, single-center analysis by Corpus et al2 found that 5.1% of patients with STEMI with multivessel coronary artery disease received multivessel interventions during the initial catheterization. Although the in-hospital outcomes were similar between patients treated with multivessel PCI and those who received revascularization of the infarctrelated artery alone, reinfarction, the major cardiac adverse events were greater at 1 year in the group treated with multivessel PCI.2 Other studies comparing the 2 strategies have found either no difference or improved outcomes in patients treated with multivessel PCI.3,6,10 –14 Finally, a small randomized study comparing treatment of multivessel coronary artery disease with revascularization of the infarctrelated artery alone in patients with STEMI found the incidence of major adverse cardiac events at 1 year to be similar in the 2 groups. However, patients treated with single-vessel PCI were significantly more likely to undergo repeat revascularization at 1 year.15 That study was likely underpowered to detect differences in other important clinical outcomes. Several limitations of our analysis should be considered. First, our study was not a randomized trial, and unmeasured confounders, such as the reason multivessel PCI was performed instead of a staged procedure, could have affected the clinical outcomes seen in our analysis. Second, our data set contained a proportion of patients with late presentations after symptom onset. In addition, the limitations of our data set prevented the analysis of staged procedures or a detailed description of the cause of death. The outcomes were limited to those that occurred during hospitalization; therefore, the long-term outcomes associated with multivessel PCI

could not be evaluated. Finally, we were unable to reliably identify the infarct-related artery, which might have affected our results, given that the risk of intervention on the noninfarct-related artery could depend on the involvement of the left anterior descending artery.16 A rigorous attempt to identify the infarct-related artery was undertaken that accounted for both pre-PCI stenosis and thrombolysis in myocardial infarction flow. With the exception of patients with chronic total occlusions, our method likely reliably identified the vast majority of infarct-related arteries. Patients in cardiogenic shock because of acute myocardial have high mortality and remain challenging to treat. In the randomized trial, should we emergently revascularize occluded coronaries for cardiogenic shock (SHOCK), the 30-day mortality rate was 46.7% among patients undergoing early revascularization and 56.0% among patients assigned to initial medical stabilization—a difference that was not signficantly different statistically.17 At 6 months of follow-up, however, a statistically significant 12.8% absolute reduction in mortality was found among patients assigned to early revascularization, suggesting that establishing coronary flow early in the course of shock is beneficial. In the SHOCK trial, the 30-day mortality rates were significantly lower among patients who had undergone successful PCI compared with those who had undergone unsuccessful PCI.17 Although revascularization of the noninfarct-related artery improves blood flow to the myocardium, the present study has provided data supporting a strategy of revascularization of the infarct-related artery alone instead of a strategy of multivessel PCI. Although we could not prove the mechanism by which revascularization of the noninfarct-related artery is associated with mortality using our data, possible explanations of the increased risk include an increase in procedure-related complications such as

512

The American Journal of Cardiology (www.AJConline.org)

bleeding or renal failure, distal embolization associated with PCI, or the loss of collateral flow to other coronary territories.18 –20 Our study provides the largest and most contemporary multicenter examination of the incidence of multivessel PCI as treatment of STEMI in patients with multivessel coronary artery disease. We found that the use of multivessel PCI remains prevalent; however, we failed to demonstrate any short-term benefits of multivessel intervention during primary PCI. In addition, multivessel PCI performed in patients with cardiogenic shock was associated with increased odds of mortality, even after adjusting for confounding variables. The present study differs from previous studies, both observational and randomized, that have found no increased risk with multivessel PCI. Our results suggest that multivessel intervention in patients not in shock undergoing primary PCI does not improve the in-hospital outcomes. Appendix A The following candidate variables were considered but were not identified as predictors for multivessel PCI in the overall population: Patient characteristics: male gender, diabetes, renal failure, previous myocardial infarction, previous congestive heart failure, previous PCI, peripheral vascular disease, chronic lung disease, and New York Heart Association classification. Hospital level characteristics: hospital type (private/community vs university or government) and hospital location (urban, suburban vs rural). Appendix B: Covariates in Mortality Models The following were covariates in both models: male gender, white race, age (linear spline with knots at 65 and 80 years), diabetes, government insurance, ejection fraction (linear spline with knot at 30%), hypercholesterolemia, hypertension, glomerular filtration rate/dialysis (linear spline with dialysis or glomerular filtration rate ⱕ30 patients set to glomerular filtration rate ⫽ 30 and knot at 60), cerebrovascular disease, chronic lung disease, peripheral vascular disease, family history of coronary artery disease, congestive heart failure, previous coronary artery bypass grafting, previous PCI, previous myocardial infarction, previous congestive heart failure, and status (salvage, emergent, and urgent/ elective [referent]). The following were additional covariates in the mortality model among the cardiogenic shock population: body mass index (linear spline with knots at 18.5 and 35 kg/m2); interval from symptom onset to admission (6 to 24 hours, 2 to 7 days, silent myocardial infarction [no interval], and ⱕ6 hours [referent]); preoperative intra-aortic balloon pump; stenosis percentage for proximal left anterior descending, mid-left anterior descending, and circumflex arteries of 100%, 70% to 99%, and ⬍70% (referent); left main stenosis of ⱖ50%; and right coronary artery stenosis of ⱖ70%. The following were additional covariates in the mortality model among the noncardiogenic shock population: body mass index; interval from symptom onset to admission (6 to

24 hours, 2 to 7 days, silent myocardial infarction [no interval], and ⱕ6 hours [referent]); stenosis percentage for proximal left anterior descending, mid-left anterior descending, right coronary, and circumflex arteries of 100%, 70% to 99%, and ⬍70% (referent); and all 2-way interactions between the stenosis percent terms. 1. Roe MT, Cura FA, Joski PS, Garcia E, Guetta V, Kereiakes DJ, Zijlstra F, Brodie BR, Grines CL, Ellis SG. Initial experience with multivessel percutaneous coronary intervention during mechanical reperfusion for acute myocardial infarction. Am J Cardiol 2001;88: 170 –173. 2. Corpus RA, House JA, Marso SP, Grantham JA, Huber KC Jr, Laster SB, Johnson WL, Daniels WC, Barth CW, Giorgi LV, Rutherford BD. Multivessel percutaneous coronary intervention in patients with multivessel disease and acute myocardial infarction. Am Heart J 2004; 148:493–500. 3. Chen LY, Lennon RJ, Grantham JA, Berger PB, Mathew V, Singh M, Holmes DR Jr, Rihal CS. In-hospital and long-term outcomes of multivessel percutaneous coronary revascularization after acute myocardial infarction. Am J Cardiol 2005;95:349 –354. 4. Batchelor WB, Anstrom KJ, Muhlbaier LH, Grosswald R, Weintraub WS, O’Neill WW, Peterson ED. Contemporary outcome trends in the elderly undergoing percutaneous coronary interventions: results in 7,472 octogenarians. National Cardiovascular Network Collaboration. J Am Coll Cardiol 2000;36:723–730. 5. Singh M, Rihal CS, Gersh BJ, Lennon RJ, Prasad A, Sorajja P, Gullerud RE, Holmes DR Jr. Twenty-five-year trends in in-hospital and long-term outcome after percutaneous coronary intervention: a single-institution experience. Circulation 2007;115:2835–2841. 6. Varani E, Balducelli M, Aquilina M, Vecchi G, Hussien MN, Frassineti V, Maresta A. Single or multivessel percutaneous coronary intervention in ST-elevation myocardial infarction patients. Catheter Cardiovasc Interv 2008;72:927–933. 7. Weintraub WS, McKay CR, Riner RN, Ellis SG, Frommer PL, Carmichael DB, Hammermeister KE, Effros MN, Bost JE, Bodycombe DP; American College of Cardiology Database Committee. The American College of Cardiology National Database: progress and challenges. J Am Coll Cardiol 1997;29:459 – 465. 8. Brindis RG, Fitzgerald S, Anderson HV, Shaw RE, Weintraub WS, Williams JF. The American College of Cardiology-National Cardiovascular Data Registry (ACC-NCDR): building a national clinical data repository. J Am Coll Cardiol 2001;37:2240 –2245. 9. Antman EM, Anbe DT, Armstrong PW, Bates ER, Green LA, Hand M, Hochman JS, Krumholz HM, Kushner FG, Lamas GA, Mullany CJ, Ornato JP, Pearle DL, Sloan MA, Smith SC Jr, Alpert JS, Anderson JL, Faxon DP, Fuster V, Gibbons RJ, Gregoratos G, Halperin JL, Hiratzka LF, Hunt SA, Jacobs AK. ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction— executive summary: a report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1999 Guidelines for the Management of Patients With Acute Myocardial Infarction). Circulation 2004;110: 588 – 636. 10. Xu F, Chen YG, Li JF, Li GS, Ji QS, Lv RJ, Li RJ, Sun Y, Zhang W, Li L, Zhang Y. Multivessel percutaneous coronary intervention in Chinese patients with acute myocardial infarction and simple nonculprit arteries. Am J Med Sci 2007;333:376 –380. 11. Katayama N, Horiuchi K, Nakao K, Kasanuki H, Honda T. [Does percutaneous coronary intervention in non-culprit vessels improve the prognosis of acute myocardial infarction complicated by pump failure?] J Cardiol 2005;46:1– 8. 12. Qarawani D, Nahir M, Abboud M, Hazanov Y, Hasin Y. Culprit only versus complete coronary revascularization during primary PCI. Int J Cardiol 2007;123:288 –292. 13. Khattab AA, Abdel-Wahab M, Rother C, Liska B, Toelg R, Kassner G, Geist V, Richardt G. Multi-vessel stenting during primary percutaneous coronary intervention for acute myocardial infarction: a singlecenter experience. Clin Res Cardiol 2008;97:32–38. 14. Kalarus Z, Lenarczyk R, Kowalczyk J, Kowalski O, Gasior M, Was T, Zebik T, Krupa H, Chodor P, Polonski L, Zembala M. Importance of

Coronary Artery Disease/Multivessel PCI in STEMI complete revascularization in patients with acute myocardial infarction treated with percutaneous coronary intervention. Am Heart J 2007; 153:304 –312. 15. Di Mario C, Mara S, Flavio A, Imad S, Antonio M, Anna P, Emanuela P, Stefano DS, Angelo R, Stefania C, Anna F, Carmelo C, Antonio C, Monzini N, Bonardi MA. Single vs multivessel treatment during primary angioplasty: results of the multicentre randomised HEpacoat for cuLPlprit or multivessel stenting for Acute Myocardial Infarction (HELP AMI) Study. Int J Cardiovasc Intervent 2004;6:128 –133. 16. Shishehbor MH, Topol EJ, Mukherjee D, Hu T, Cohen DJ, Stone GW, McClure R, Roffi M, Moliterno DJ. Outcome of multivessel coronary intervention in the contemporary percutaneous revascularization era. Am J Cardiol 2006;97:1585–1590. 17. Hochman JS, Sleeper LA, Webb JG, Sanborn TA, White HD, Talley JD, Buller CE, Jacobs AK, Slater JN, Col J, McKinlay SM, LeJemtel TH; SHOCK Investigators. Early revascularization in acute myocardial infarction complicated by cardiogenic shock. Should We Emer-

513

gently Revascularize Occluded Coronaries for Cardiogenic Shock. N Engl J Med 1999;341:625– 634. 18. Rao SV, O’Grady K, Pieper KS, Granger CB, Newby LK, Van de Werf F, Mahaffey KW, Califf RM, Harrington RA. Impact of bleeding severity on clinical outcomes among patients with acute coronary syndromes. Am J Cardiol 2005;96:1200 –1206. 19. Fokkema ML, Vlaar PJ, Svilaas T, Vogelzang M, Amo D, Diercks GF, Suurmeijer AJ, Zijlstra F. Incidence and clinical consequences of distal embolization on the coronary angiogram after percutaneous coronary intervention for ST-elevation myocardial infarction. Eur Heart J 2009; 30:908 –915. 20. Hochman JS, Lamas GA, Buller CE, Dzavik V, Reynolds HR, Abramsky SJ, Forman S, Ruzyllo W, Maggioni AP, White H, Sadowski Z, Carvalho AC, Rankin JM, Renkin JP, Steg PG, Mascette AM, Sopko G, Pfisterer ME, Leor J, Fridrich V, Mark DB, Knatterud GL. Coronary intervention for persistent occlusion after myocardial infarction. N Engl J Med 2006;355:2395–2407.