- Email: [email protected]
Clinical and Angiographic Predictors of ST-Segment Recovery After Primary Percutaneous Coronary Intervention
Clinical and Angiographic Predictors of ST-Segment Recovery After Primary Percutaneous Coronary Intervention Niels J. W. Verouden, MD, Joost D. E. Haeck, MD, Wichert J. Kuijt, MD, Martijn Meuwissen, MD, PhD, Karel T. Koch, MD, PhD, José P. S. Henriques, MD, PhD, Jan Baan, MD, PhD, Marije M. Vis, MD, Jan J. Piek, MD, PhD, Jan G. P. Tijssen, PhD, and Robbert J. de Winter, MD, PhD* Important determinants of incomplete ST-segment recovery in patients undergoing primary percutaneous coronary intervention (PCI) for ST-segment elevation myocardial infarction (STEMI) have been incompletely characterized. Early risk stratification could identify patients with STEMI and incomplete ST-segment recovery who may benefit from adjunctive therapy. For the present study, we analyzed 12-lead electrocardiograms from 2,124 patients with STEMI who underwent primary PCI at our institution from 2000 to 2007. ST-segment recovery was defined as percent change in cumulative ST-segment deviation between preprocedural and immediately postprocedural electrocardiograms and categorized as incomplete when <50%. A total of 1,032 patients (49%) had incomplete ST-segment recovery. After multivariable adjustment, age >60 years (adjusted odds ratio [OR] 1.28, 95% confidence interval [CI] 1.06 to 1.54, p ⴝ 0.011), diabetes mellitus (OR 1.36, 95% CI 1.02 to 1.82, p ⴝ 0.034), left anterior descending coronary artery–related STEMI (OR 1.92, 95% CI 1.61 to 2.30, p<0.001), and multivessel disease (OR 1.34, 95% CI 1.10 to 1.63, p ⴝ 0.004) were independent predictors of incomplete ST-segment recovery. Current smoking (OR 0.79, 95% CI 0.65 to 0.95, p ⴝ 0.013) and a preprocedural Thrombolysis In Myocardial Infarction grade <3 flow (OR 0.70, 95% CI 0.53 to 0.93, p ⴝ 0.014) were inversely related to ST-segment recovery. Incomplete ST-segment recovery was a strong predictor of long-term mortality (hazard ratio 2.07, 95% CI 1.59 to 2.69, p <0.001) in addition to identified characteristics that independently predicted incomplete ST-segment recovery. In conclusion, incomplete ST-segment recovery at the end of PCI occurred significantly more often in the presence of an age >60 years, nonsmoking, diabetes mellitus, left anterior descending coronary artery–related STEMI, multivessel disease, and preprocedural Thrombolysis In Myocardial Infarction grade 3 flow. Patients with STEMI and these clinical features are at increased risk of impaired myocardial salvage and are appropriate candidates for adjunctive therapy. © 2010 Elsevier Inc. All rights reserved. (Am J Cardiol 2010;105:1692–1697) In patients with ST-segment elevation myocardial infarction (STEMI), thrombolysis has been replaced by primary percutaneous coronary intervention (PCI) as the preferred treatment strategy.1,2 Nevertheless, primary PCI only facilitates myocardial reperfusion by restoring coronary epicardial flow. Microvascular dysfunction is a common complication after “successful” primary PCI and may be induced by distal embolization, reperfusion injury, or bioactive factors causing vasoconstriction downstream. Microvascular dysfunction is quantified by ST-segment recovery as measured on the 12-lead electrocardiogram (ECG). ST-segment recovery is a strong and independent predictor of adverse cardiac remodeling and increased mortality.3,4 As a result, incomplete ST-segment recovery often is a reason to administer adjunctive therapy to increase suboptimal micro-
Department of Cardiology of the Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands. Manuscript received October 31, 2009; revised manuscript received and accepted January 25, 2010. *Corresponding author: Tel: 31-20-566-9111; fax: 31-20-696-2609. E-mail address: [email protected] (R.J. de Winter). 0002-9149/10/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.01.343
vascular reperfusion. Despite the value of ST-segment recovery as a predictor of outcome,5– 8 little is known about which factors determine ST-segment recovery itself. Knowledge of powerful clinical predictors of incomplete ST-segment recovery could increase early risk stratification and thus improve therapy in patients with STEMI undergoing primary PCI. Therefore, we sought to identify independent predictors of incomplete ST-segment recovery in an unselected cohort of patients with STEMI undergoing primary PCI. Methods Data analyzed in our study were obtained from patients with STEMI who underwent primary PCI at the Academic Medical Center, University of Amsterdam (Amsterdam, The Netherlands), from November 1, 2000, to January 1, 2007. In general, primary PCIs were performed according to current guidelines. Patients with an indication for primary PCI received aspirin (500 mg) and unfractionated heparin (5,000 IU) during transportation to the catheterization laboratory. During 2005, clopidogrel 300 or 600 mg was added www.AJConline.org
Coronary Artery Disease/Predictors of ST-Segment Recovery in STEMI
as pretreatment before primary PCI. Glycoprotein IIb/IIIa inhibitors were not routinely used but administered during the procedure at the discretion of the operator. We obtained information on 1-year vital status from the institutional follow-up database of patients after PCI. Patients are surveyed 1 year after PCI by a mailed, selfadministered questionnaire. Follow-up information was synchronized with computerized, long-term mortality records from the National Death Index and local authorities, which were updated until October 2008. We reviewed outpatients’ files and contacted general practitioners by telephone in case of conflicting or missing data. Follow-up could be obtained in 2,103 of 2,124 patients (99%). From the local electronic database at the catheterization laboratory, we abstracted baseline demographic variables and procedural and angiographic information that had been prospectively collected and entered by specialized nurses and interventional cardiologists concurrently with routine patient care. This information included an operator’s online assessment of anterograde flow using the Thrombolysis In Myocardial Infarction (TIMI) scale,9 extent of coronary artery disease, and timing of re-establishment of anterograde flow through the infarct-related artery. We retrospectively sought to collect 12-lead ECGs recorded immediately before arterial puncture at the catheterization laboratory for all patients with STEMI who underwent primary PCI at our institution. These preprocedural ECGs were compared with 12-lead ECGs recorded at the time of last contrast injection (i.e., immediate postprocedural ECGs), before patients were transferred to the coronary or intensive care unit. ECGs of included patients were analyzed by 1 investigator (NV), unaware of the clinical, angiographic, and outcome data. ST-segment deviation was measured with a handheld caliper and magnifying glass at 80 ms after the J-point in all available leads. ST-segment deviation was measured to the nearest 0.05 mV with the TP segment as the preferred isoelectric baseline. We defined ST-segment recovery as the percent change of summed ST-segment deviations on the 12-lead ECG immediately after PCI compared to the ECG before PCI. The primary outcome of this analysis was the occurrence of incomplete STsegment recovery. We defined ST-segment recovery as incomplete if ⬍50%. All consecutive patients with STEMI who underwent primary PCI at our institution according to our digital database were eligible for inclusion. We excluded patients who underwent primary PCI of the left main coronary artery or a bypass graft because of the distinctive electrocardiographic pattern observed in such cases. Patients from whom a preprocedural or postprocedural ECG could not be retrieved were excluded from this analysis. Furthermore, patients with electrocardiographic recordings containing a complete left bundle branch block, sustained ventricular arrhythmias (among others, an accelerated idioventricular rhythm), a paced rhythm, or severe artifacts that prohibit accurate ST-segment evaluation were excluded. For patients with STEMI who were referred for primary PCI but in whom we observed normalization of ST-segment elevation on arrival at the catheterization laboratory, ST-segment re-
1693
Table 1 Baseline and procedural characteristics Variable Age (years), mean ⫾ SD Men Current smoker* Diabetes mellitus† Hypertension‡ Hypercholesterolemia§ Previous myocardial infarction Previous coronary bypass grafting Previous percutaneous coronary intervention Left anterior descending coronary artery–related myocardial infarction Multivessel disease Preprocedural Thrombolysis In Myocardial Infarction grade 3 flow Stent inserted Glycoprotein IIb/IIIa inhibitor used Postprocedural Thrombolysis In Myocardial Infarction grade 3 flow Intra-aortic balloon counterpulsation used Total ischemic time (minutes) (25th, 75th percentile)储 Peak creatine kinase-MB fraction (g/L) (25th, 75th percentile)¶
Included Patients (n ⫽ 2,124) 61 ⫾ 13 72% 46% 11% 31% 22% 11% 1% 7% 46% 31% 11% 87% 31% 89% 7% 172 (130, 250) 248 (131, 435)
* Defined as patients who smoke ⱖ1 tobacco product per day or smoked in the 30 days before admission. † Defined as a documented history of diabetes mellitus diagnosed and/or treated by a physician. ‡ Defined as a documented history of hypertension diagnosed and/or treated by a physician. § Defined as a documented history of hypercholesterolemia diagnosed and/or treated by a physician. 储 Data available in 1,897 patients. ¶ Data available in 1,272 patients.
covery analysis after PCI was not appropriate, and therefore these patients were also excluded. We assessed the association between ST-segment recovery, clinical characteristics (age, gender, hypertension, hypercholesterolemia, diabetes mellitus, current smoking, previous MI, coronary artery bypass grafting, or PCI), and procedural characteristics (left anterior descending coronary artery [LAD]–related MI, preprocedural and postprocedural TIMI-graded flow, use of glycoprotein IIb/IIIa inhibitors, and intra-aortic balloon pump insertion) using binary logistic regression. This association was studied in crude and multivariable models. Backward selections were used to select the most parsimonious set of predictive variables and results were expressed as odds ratios (ORs) with 95% confidence intervals (CIs). Total ischemic time and body mass index were not included in multivariable analyses because of a considerable amount of missing values. To construct a ST-segment recovery prediction model, a simple risk score was calculated. We assigned points according to the adjusted OR of each predictive characteristic that would be available after emergency coronary angiography but before PCI. A score was calculated by adding up points corresponding to each patient’s risk factors. Here-
1694
The American Journal of Cardiology (www.AJConline.org)
Table 2 Univariable and multivariable-adjusted associations between patient characteristics and incomplete ST-segment recovery Variable
Age (years) ⬎60 Men Current smoker Diabetes mellitus Hypertension Hypercholesterolemia Previous myocardial infarction Previous coronary bypass grafting Previous percutaneous coronary intervention
Incomplete ST-Segment Recovery
53% (603/1,128) 49% (744/1,531) 43% (421/984) 59% (138/235) 52% (343/658) 47% (219/465) 51% (122/237) 53% (8/15) 47% (69/146)
Univariable Analysis
Multivariable Analysis*
OR
CI
p Value
OR
CI
p Value
1.52 1.00 0.65 1.58 1.23 0.93 1.14 1.21 0.94
1.28–1.80 0.83–1.21 0.55–0.77 1.20–2.08 1.02–1.48 0.75–1.14 0.87–1.49 0.44–3.35 0.68–1.32
⬍0.001 0.99 ⬍0.001 0.001 0.03 0.47 0.35 0.71 0.74
1.28 — 0.79 1.36 — — — — —
1.06–1.54 — 0.65–0.95 1.02–1.82 — — — — —
0.011 — 0.013 0.034 — — — — —
* Backward selection model including patient characteristics and available procedural characteristics: left anterior descending coronary artery–related myocardial infarction, multivessel disease, preprocedural Thrombolysis In Myocardial Infarction grade ⬍3 flow, stent insertion, glycoprotein IIb/IIIa inhibitor use, postprocedural Thrombolysis In Myocardial Infarction grade ⬍3 flow, and intra-aortic balloon counterpulsation used.
Table 3 Univariable and multivariable-adjusted associations between procedural characteristics and incomplete ST-segment recovery Variable
Left anterior descending coronary artery–related myocardial infarction Multivessel disease Stent inserted Glycoprotein IIb/IIIa inhibitor used Intra-aortic balloon counterpulsation used Preprocedural Thrombolysis In Myocardial Infarction grade flow ⬍3 Postprocedural Thrombolysis In Myocardial Infarction grade flow ⬍3
Incomplete ST-Segment Recovery
Univariable Analysis
Multivariable Analysis*
OR
95% CI
p Value
OR
95% CI
p Value
57% (553/971)
1.86
1.57–2.21
⬍0.001
1.92
1.61–2.30
⬍0.001
54% (351/651) 48% (894/1,858) 56% (361/648) 62% (89/144)
1.36 0.86 1.51 1.78
1.13–1.64 0.67–1.11 1.25–1.82 1.26–2.52
0.001 0.25 ⬍0.001 0.001
1.34 — 1.33 —
1.10–1.63 — 1.09–1.62 —
0.004 — 0.004 —
48% (905/1,889)
0.78
0.60–1.03
0.08
0.70
0.53–0.93
0.014
68% (153/226)
2.43
1.81–3.26
⬍0.001
2.27
1.66–3.09
⬍0.001
* Backward selection model with procedural characteristics and patient characteristics as listed in Table 2.
after, we defined low-, intermediate-, and high-risk groups according to this risk score. Kaplan–Meier estimates according to category of STsegment recovery were determined and compared with the use of the log-rank test. To determine whether the prognostic value of ST-segment recovery was independent of associated patient and procedural characteristics, these variables (age ⬎60 years, smoking, diabetes mellitus, LAD-related MI, multivessel disease, TIMI-graded flow ⬍3 before PCI, use of IIb/IIIa inhibitors, and TIMI-graded flow ⬍3 after PCI) and ST-segment recovery were concurrently entered in a Cox regression model. Results were expressed as hazard ratios with 95% CIs. Normally distributed, continuous variables are expressed as mean ⫾ SD, and other continuous data are expressed as median with interquartile range. All categorical variables are depicted using relative frequency distributions. Characteristics of patients with complete and incomplete ST-segment recovery were compared using the chi-square test for categorical variables and the Student’s t test or Mann– Whitney U test for continuous variables. For all tests,
differences were considered significant if the 2-sided p value was ⬍0.05. All analyses were performed using SPSS 16.0 (SPSS, Inc., Chicago, Illinois). Results A total of 3,185 consecutive patients with STEMI underwent primary PCI at our institution from November 1, 2000, to January 1, 2007. We excluded 67 patients who underwent primary PCI of the left main coronary artery or of a bypass graft. ECGs of 508 patients could not be retrieved. Electrocardiographic exclusion criteria were applicable in 132 of 2,610 patients with a complete set of ECGs. Preprocedural ST-segment normalization had occurred in 354 patients. Thus, 2,124 patients could be included in this analysis. Baseline and procedural characteristics of these patients are presented in Table 1. Incomplete ST-segment recovery occurred in 1,032 of 2,124 patients (49%). The left side of Table 2 displays the univariable relation between occurrence of incomplete ST-segment recovery and various patient characteristics. Incomplete ST-segment
Coronary Artery Disease/Predictors of ST-Segment Recovery in STEMI
recovery was significantly more frequent in older patients and in patients with diabetes mellitus or hypertension, with unadjusted ORs of 1.52 (95% CI 1.28 to 1.80, p ⬍0.001), 1.58 (95% CI 1.20 to 2.08, p ⫽ 0.001), and 1.23 (95% CI 1.02 to 1.48, p ⫽ 0.03), respectively. Patients who were smokers at the time of primary PCI significantly less often had incomplete ST-segment recovery compared to nonsmoking counterparts (unadjusted OR 0.65, 95% CI 0.55 to 0.77, p ⬍0.001). In multivariable logistic regression, age ⬎ 60 years, nonsmoking status, and presence of diabetes mellitus were independently predictive of incomplete STsegment recovery (Table 2, right side). Table 3 presents univariable and multivariable associations between available procedural characteristics and occurrence of incomplete ST-segment recovery. As shown on the left side, incomplete ST-segment recovery was significantly more frequent in patients with LAD-related MI (unadjusted OR 1.86, 95% CI 1.57 to 2.21, p ⬍0.001), in patients with multivessel disease (unadjusted OR 1.36, 95% CI 1.13 to 1.64, p ⫽ 0.001), in patients with postprocedural TIMI-graded flow ⬍3 (unadjusted OR 2.43, 95% CI 1.81 to 3.26, p ⬍0.001), and in patients with a need for intra-aortic balloon pump insertion (unadjusted OR 1.78, 95% CI 1.26 to 2.52, p ⫽ 0.001). Also, incomplete ST-segment recovery was more common in patients who received glycoprotein IIb/IIIa inhibitors during PCI, with an unadjusted OR of 1.51 (95% CI 1.25 to 1.82, p ⬍0.001). Patients with TIMIgraded flow ⬍3 before PCI showed a trend toward less frequent incomplete ST-segment recovery (unadjusted OR 0.78, 95% CI 0.60 to 1.03, p ⫽ 0.08). After multivariable modeling, ORs of LAD-related MI, multivessel disease, glycoprotein IIb/IIIa inhibitor use, and TIMI-graded flow ⬍3 after PCI were sustained, whereas the association between intra-aortic balloon pump insertion and incomplete ST-segment recovery lost significance (Table 3, right side). The association between preprocedural TIMI-graded flow ⬍3 and incomplete ST-segment recovery was more outspoken in multivariable analysis, with an adjusted OR of 0.70 (95% CI 0.53 to 0.93, p ⫽ 0.014). We constructed a patient-specific ST-segment recovery risk score with the 6 predictive variables available before PCI. In accordance with their ORs (as listed in Tables 2 and 3), each characteristic (age ⬎60 years, nonsmoking, diabetes mellitus, multivessel disease, and TIMI-graded flow ⬍3 before PCI) contributed 1 point to the ST-segment recovery risk score, whereas LAD-related MI contributed 3 points. Median score of the study cohort was 3 (interquartile range 1 to 4). The C-statistic of the risk score to predict incomplete ST-segment recovery was 0.61 (95% CI 0.58 to 0.63, p ⬍0.001). By dividing patients into low-risk (0 point to 2 points), intermediate-risk (3 to 4 points), or high-risk (ⱖ5 points) subgroups, we identified 882 patients (41.5%) at low risk, 733 patients (34.5%) at intermediate risk, and 509 patients (24%) at high risk for incomplete ST-segment recovery. The OR for incomplete ST-segment recovery in patients at intermediate risk compared to patients at low risk was 1.81 (95% CI 1.49 to 2.21, p ⬍0.001) and that for patients at high risk was 2.68 (95% CI 2.14 to 3.35, p ⬍0.001). At a median follow-up of 4.1 years (interquartile range 2.7 to 5), 272 patients (85 of 1,078 patients with complete
1695
Figure 1. Cumulative mortality in patients with ⱖ50% versus ⬍50% ST-segment recovery at the end of primary PCI.
ST-segment recovery and 187 of 1,025 patients with incomplete ST-segment recovery) died. Estimated 5-year cumulative mortalities were 9.9% in patients with complete STsegment recovery and 21.4% in patients with incomplete ST-segment recovery (p ⬍0.001; Figure 1). Table 4 presents univariable and multivariable-adjusted relations between long-term mortality, on the 1 hand, and ST-segment recovery and its determinants, on the other. Incomplete ST-segment recovery was a strong univariable prognosticator (unadjusted hazard ratio 2.54, 95% CI 1.96 to 3.28, p ⬍0.001). Moreover, incomplete ST-segment recovery was a strong, independent predictor of long-term mortality in multivariable Cox regression analysis, with an adjusted hazard ratio of 2.07 (95% CI 1.59 to 2.69, p ⬍0.001). Discussion In this cohort of patients with primary PCI, we found that the following characteristics, available before PCI, were independent predictors of incomplete ST-segment recovery: age, smoking status, presence of diabetes mellitus, LADrelated MI, presence of multivessel disease, and preprocedural TIMI flow. By applying a simplified risk score, we could calculate a patient’s a priori risk of incomplete STsegment recovery. Patients in the high-risk category had a 2.7-fold higher risk for incomplete ST-segment recovery (62%) than patients in the low-risk category (38%). Furthermore, incomplete ST-segment recovery was a powerful predictor of long-term mortality in addition to the abovementioned characteristics that independently predicted incomplete ST-segment recovery. First, age being an independent predictor of incomplete ST-segment recovery seems reasonable when we acknowledge the time-dependent nature of arteriolosclerosis. The multifactorial cause of microvascular dysfunction (distal embolization, microvascular constriction through release of bioactive factors, vascular or myocyte edema, etc.) may
1696
The American Journal of Cardiology (www.AJConline.org)
Table 4 Result of unadjusted and multivariable-adjusted long-term mortality analyses for ST-segment recovery Variable
Incomplete ST-segment recovery Age ⬎60 years Current smoker Diabetes mellitus Left anterior descending coronary artery–related myocardial infarction Multivessel disease Preprocedural Thrombolysis In Myocardial Infarction grade ⬍3 flow Glycoprotein IIb/IIIa inhibitors used Postprocedural Thrombolysis In Myocardial Infarction grade ⬍3 flow
Unadjusted
Multivariable Adjusted*
HR
95% CI
p Value
HR
95% CI
p Value
2.54 3.63 0.51 1.76 1.23 1.83 1.03 0.96 2.47
1.96–3.28 2.71–4.86 0.41–0.67 1.29–2.41 0.97–1.57 1.44–2.33 0.71–1.51 0.74–1.25 1.84–3.30
⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.08 ⬍0.001 0.87 0.78 ⬍0.001
2.07 2.79 0.84 1.29 1.18 1.45 0.92 0.76 2.07
1.59–2.69 2.05–3.81 0.64–1.10 0.94–1.77 0.93–1.50 1.13–1.86 0.63–1.36 0.58–1.00 1.51–2.83
⬍0.001 ⬍0.001 0.20 0.12 0.19 0.004 0.69 0.046 ⬍0.001
* All listed variables were simultaneously entered in a Cox regression model. HR ⫽ hazard ratio.
likely be negatively influenced by aging. Second, the inverse relation between smoking status and occurrence of incomplete ST-segment recovery exemplifies the “smoker’s paradox,” which implies a favorable outcome of smokers with acute MI or heart failure.10,11 In this study we confirm the significant association between smoking status and STsegment recovery as previously reported in a small patient sample.12 Third, the independent association between presence of diabetes mellitus and occurrence of incomplete ST-segment recovery is in accordance with an increased risk of cardiovascular nonfatal and fatal events in diabetic patients.13,14 In addition, patients with diabetes who have MI have a significantly increased total ischemic time and continue to have a poorer prognosis compared to their nondiabetic counterparts.15 Fourth, patients with LAD-related MI had an adjusted two-fold increased risk of incomplete ST-segment recovery. Differences in ST-segment recovery after primary PCI between patients with anterior and nonanterior STEMI were previously reported in subgroup analyses from the Assessment of Pexelizumab in Acute Myocardial Infarction (APEX AMI) trial.6 A larger area at risk and a subsequent higher risk of microvascular dysfunction and tissue injury may largely account for this phenomenon. Fifth, multivessel disease as an independent characteristic can be explained by the assumption that more extensive atherosclerosis of the epicardial coronary arteries is associated with more outspoken dysfunction of the myocardial microvasculature, resulting in more frequent incomplete ST-segment recovery. The association between presence of multivessel disease and incomplete ST-segment recovery seems in agreement with the reported adverse prognosis in this patient subgroup.16 Interestingly, normal TIMIgraded flow before primary PCI was the sixth characteristic independently predictive of incomplete ST-segment recovery. Because patients with preprocedural ST-segment normalization were excluded from this study, persistent ST-segment elevation in the presence of TIMI grade 3 flow through the infarct-related coronary artery before primary PCI signals persistent microvascular dysfunction not responsive to epicardial reperfusion in these patients. Patients at high risk for microvascular dysfunction can be identified with our ST-segment recovery risk score. These patients are a target population for adjunctive therapy. Preprocedural risk stratification with subsequent deci-
sion making about adjunctive treatment enhances management of patients with STEMI. For example, patients at high risk could especially benefit from glycoprotein IIb/IIIa inhibitors in addition to aspirin, clopidogrel, and unfractionated heparin.17 In contrast, patients at low risk could benefit from a more restrictive antiplatelet regimen, which decreases the incidence of clinically relevant complications.18 –20 Nevertheless, increasing economic pressures have intensified the need for appropriate triage and clinical resource use. Our ST-segment recovery risk score is a tool increasing a clinician’s ability to rapidly and accurately assess risk and thus could be of substantial importance.6 An important limitation of our study is the retrospective collection of ECGs. From our source population, we excluded 16% of patients because of missing electrocardiographic recordings, which is similar to other observational studies with unselected patient samples.5,21 We could not trace patients with STEMI who developed left bundle branch block or ventricular arrhythmias, needed external pacing during PCI, or died during PCI. This may have introduced an outcome assessment bias, although we believe this is marginal. In addition, 1 investigator conducted all ST-segment recovery measurements. Nevertheless, other studies consistently have shown a low variability between observers,8,22 which reflects the straightforwardness of ST-segment analysis. Furthermore, this investigator was not aware of the baseline characteristics and subsequent outcome. Moreover, importation, processing, and analysis of electrocardiographic data were digital.
Acknowledgment: We gratefully acknowledge the technical and nursing staffs of the Cardiac Catheterization Laboratory of the Academic Medical Center, University of Amsterdam, for their skilled assistance and Kurdo Barwari, MSc, for his help in data collection. 1. Grines CL, Browne KF, Marco J, Rothbaum D, Stone GW, O’Keefe J, Overlie P, Donohue B, Chelliah N, Timmis GC. A comparison of immediate angioplasty with thrombolytic therapy for acute myocardial infarction. The Primary Angioplasty in Myocardial Infarction Study Group. N Engl J Med 1993;328:673– 679.
Coronary Artery Disease/Predictors of ST-Segment Recovery in STEMI 2. Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review of 23 randomised trials. Lancet 2003;361:13–20. 3. Gerber BL, Rochitte CE, Melin JA, McVeigh ER, Bluemke DA, Wu KC, Becker LC, Lima JA. Microvascular obstruction and left ventricular remodeling early after acute myocardial infarction. Circulation 2000;101:2734 –2741. 4. Nijveldt R, Beek AM, Hirsch A, Stoel MG, Hofman MB, Umans VA, Algra PR, Twisk JW, van Rossum AC. Functional recovery after acute myocardial infarction: comparison between angiography, electrocardiography, and cardiovascular magnetic resonance measures of microvascular injury. J Am Coll Cardiol 2008;52:181–189. 5. Brodie BR, Stuckey TD, Hansen C, VerSteeg DS, Muncy DB, Moore S, Gupta N, Downey WE. Relation between electrocardiographic STsegment resolution and early and late outcomes after primary percutaneous coronary intervention for acute myocardial infarction. Am J Cardiol 2005;95:343–348. 6. Buller CE, Fu Y, Mahaffey KW, Todaro TG, Adams P, Westerhout CM, White HD, van ’t Hof AW, Van de Werf FJ, Wagner GS, Granger CB, Armstrong PW. ST-segment recovery and outcome after primary percutaneous coronary intervention for ST-elevation myocardial infarction: insights from the Assessment of Pexelizumab in Acute Myocardial Infarction (APEX-AMI) trial. Circulation 2008;118:1335– 1346. 7. McLaughlin MG, Stone GW, Aymong E, Gardner G, Mehran R, Lansky AJ, Grines CL, Tcheng JE, Cox DA, Stuckey T, Garcia E, Guagliumi G, Turco M, Josephson ME, Zimetbaum P. Prognostic utility of comparative methods for assessment of ST-segment resolution after primary angioplasty for acute myocardial infarction: the Controlled Abciximab and Device Investigation to Lower Late Angioplasty Complications (CADILLAC) trial. J Am Coll Cardiol 2004;44: 1215–1223. 8. Van ’t Hof AW, Liem A, de Boer MJ, Zijlstra F. Clinical value of 12-lead electrocardiogram after successful reperfusion therapy for acute myocardial infarction. Zwolle Myocardial infarction Study Group. Lancet 1997;350:615– 619. 9. TIMI Study Group. The Thrombolysis In Myocardial Infarction (TIMI) trial. Phase I findings. N Engl J Med 1985;312:932–936. 10. Barbash GI, White HD, Modan M, Diaz R, Hampton JR, Heikkila J, Kristinsson A, Moulopoulos S, Paolasso EA, Van der Werf T. Significance of smoking in patients receiving thrombolytic therapy for acute myocardial infarction. Experience gleaned from the International Tissue Plasminogen Activator/Streptokinase Mortality Trial. Circulation 1993;87:53–58. 11. Fonarow GC, Abraham WT, Albert NM, Stough WG, Gheorghiade M, Greenberg BH, O’Connor CM, Nunez E, Yancy CW, Young JB. A smoker’s paradox in patients hospitalized for heart failure: findings from OPTIMIZE-HF. Eur Heart J 2008;29:1983–1991.
1697
12. Albertal M, Cura F, Escudero AG, Thierer J, Trivi M, Padilla LT, Belardi J. Mechanism involved in the paradoxical effects of active smoking following primary angioplasty: a subanalysis of the protection of distal embolization in high-risk patients with acute myocardial infarction trial. J Cardiovasc Med 2008;9:810 – 812. 13. Danaei G, Lawes CM, Vander HS, Murray CJ, Ezzati M. Global and regional mortality from ischaemic heart disease and stroke attributable to higher-than-optimum blood glucose concentration: comparative risk assessment. Lancet 2006;368:1651–1659. 14. Geiss LS, Herman WH, Smith PJ. Mortality in Non–Insulin-Dependent Diabetes. Diabetes in America, 2nd Ed. Bethesda: National Institute of Diabetes and Digestive and Kidney Diseases, National Institute of Health, 1995:233–258. 15. Donahoe SM, Stewart GC, McCabe CH, Mohanavelu S, Murphy SA, Cannon CP, Antman EM. Diabetes and mortality following acute coronary syndromes. JAMA 2007;298:765–775. 16. Van der Schaaf RJ, Timmer JR, Ottervanger JP, Hoorntje JC, de Boer MJ, Suryapranata H, Zijlstra F, Dambrink JH. Long-term impact of multivessel disease on cause-specific mortality after ST elevation myocardial infarction treated with reperfusion therapy. Heart 2006;92: 1760 –1763. 17. Kandzari DE, Hasselblad V, Tcheng JE, Stone GW, Califf RM, Kastrati A, Neumann FJ, Brener SJ, Montalescot G, Kong DF, Harrington RA. Improved clinical outcomes with abciximab therapy in acute myocardial infarction: a systematic overview of randomized clinical trials. Am Heart J 2004;147:457– 462. 18. Kereiakes DJ, Berkowitz SD, Lincoff AM, Tcheng JE, Wolski K, Achenbach R, Melsheimer R, Anderson K, Califf RM, Topol EJ. Clinical correlates and course of thrombocytopenia during percutaneous coronary intervention in the era of abciximab platelet glycoprotein IIb/IIIa blockade. Am Heart J 2000;140:74 – 80. 19. Montalescot G, Barragan P, Wittenberg O, Ecollan P, Elhadad S, Villain P, Boulenc JM, Morice MC, Maillard L, Pansieri M, Choussat R, Pinton P. Platelet glycoprotein IIb/IIIa inhibition with coronary stenting for acute myocardial infarction. N Engl J Med 2001;344: 1895–1903. 20. Eikelboom JW, Mehta SR, Anand SS, Xie C, Fox KA, Yusuf S. Adverse impact of bleeding on prognosis in patients with acute coronary syndromes. Circulation 2006;114:774 –782. 21. De Luca G, Suryapranata H, Ottervanger JP, Hoorntje JC, Gosselink AT, Dambrink JH, de Boer MJ, van ’t Hof AW. Postprocedural single-lead ST-segment deviation and long-term mortality in patients with ST-segment elevation myocardial infarction treated by primary angioplasty. Heart 2008;94:44 – 47. 22. Claeys MJ, Bosmans J, Veenstra L, Jorens P, De R, Vrints CJ. Determinants and prognostic implications of persistent ST-segment elevation after primary angioplasty for acute myocardial infarction: importance of microvascular reperfusion injury on clinical outcome. Circulation 1999;99:1972–1977.
Department of Cardiology of the Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands. Manuscript received October 31, 2009; revised manuscript received and accepted January 25, 2010. *Corresponding author: Tel: 31-20-566-9111; fax: 31-20-696-2609. E-mail address: [email protected] (R.J. de Winter). 0002-9149/10/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.01.343
vascular reperfusion. Despite the value of ST-segment recovery as a predictor of outcome,5– 8 little is known about which factors determine ST-segment recovery itself. Knowledge of powerful clinical predictors of incomplete ST-segment recovery could increase early risk stratification and thus improve therapy in patients with STEMI undergoing primary PCI. Therefore, we sought to identify independent predictors of incomplete ST-segment recovery in an unselected cohort of patients with STEMI undergoing primary PCI. Methods Data analyzed in our study were obtained from patients with STEMI who underwent primary PCI at the Academic Medical Center, University of Amsterdam (Amsterdam, The Netherlands), from November 1, 2000, to January 1, 2007. In general, primary PCIs were performed according to current guidelines. Patients with an indication for primary PCI received aspirin (500 mg) and unfractionated heparin (5,000 IU) during transportation to the catheterization laboratory. During 2005, clopidogrel 300 or 600 mg was added www.AJConline.org
Coronary Artery Disease/Predictors of ST-Segment Recovery in STEMI
as pretreatment before primary PCI. Glycoprotein IIb/IIIa inhibitors were not routinely used but administered during the procedure at the discretion of the operator. We obtained information on 1-year vital status from the institutional follow-up database of patients after PCI. Patients are surveyed 1 year after PCI by a mailed, selfadministered questionnaire. Follow-up information was synchronized with computerized, long-term mortality records from the National Death Index and local authorities, which were updated until October 2008. We reviewed outpatients’ files and contacted general practitioners by telephone in case of conflicting or missing data. Follow-up could be obtained in 2,103 of 2,124 patients (99%). From the local electronic database at the catheterization laboratory, we abstracted baseline demographic variables and procedural and angiographic information that had been prospectively collected and entered by specialized nurses and interventional cardiologists concurrently with routine patient care. This information included an operator’s online assessment of anterograde flow using the Thrombolysis In Myocardial Infarction (TIMI) scale,9 extent of coronary artery disease, and timing of re-establishment of anterograde flow through the infarct-related artery. We retrospectively sought to collect 12-lead ECGs recorded immediately before arterial puncture at the catheterization laboratory for all patients with STEMI who underwent primary PCI at our institution. These preprocedural ECGs were compared with 12-lead ECGs recorded at the time of last contrast injection (i.e., immediate postprocedural ECGs), before patients were transferred to the coronary or intensive care unit. ECGs of included patients were analyzed by 1 investigator (NV), unaware of the clinical, angiographic, and outcome data. ST-segment deviation was measured with a handheld caliper and magnifying glass at 80 ms after the J-point in all available leads. ST-segment deviation was measured to the nearest 0.05 mV with the TP segment as the preferred isoelectric baseline. We defined ST-segment recovery as the percent change of summed ST-segment deviations on the 12-lead ECG immediately after PCI compared to the ECG before PCI. The primary outcome of this analysis was the occurrence of incomplete STsegment recovery. We defined ST-segment recovery as incomplete if ⬍50%. All consecutive patients with STEMI who underwent primary PCI at our institution according to our digital database were eligible for inclusion. We excluded patients who underwent primary PCI of the left main coronary artery or a bypass graft because of the distinctive electrocardiographic pattern observed in such cases. Patients from whom a preprocedural or postprocedural ECG could not be retrieved were excluded from this analysis. Furthermore, patients with electrocardiographic recordings containing a complete left bundle branch block, sustained ventricular arrhythmias (among others, an accelerated idioventricular rhythm), a paced rhythm, or severe artifacts that prohibit accurate ST-segment evaluation were excluded. For patients with STEMI who were referred for primary PCI but in whom we observed normalization of ST-segment elevation on arrival at the catheterization laboratory, ST-segment re-
1693
Table 1 Baseline and procedural characteristics Variable Age (years), mean ⫾ SD Men Current smoker* Diabetes mellitus† Hypertension‡ Hypercholesterolemia§ Previous myocardial infarction Previous coronary bypass grafting Previous percutaneous coronary intervention Left anterior descending coronary artery–related myocardial infarction Multivessel disease Preprocedural Thrombolysis In Myocardial Infarction grade 3 flow Stent inserted Glycoprotein IIb/IIIa inhibitor used Postprocedural Thrombolysis In Myocardial Infarction grade 3 flow Intra-aortic balloon counterpulsation used Total ischemic time (minutes) (25th, 75th percentile)储 Peak creatine kinase-MB fraction (g/L) (25th, 75th percentile)¶
Included Patients (n ⫽ 2,124) 61 ⫾ 13 72% 46% 11% 31% 22% 11% 1% 7% 46% 31% 11% 87% 31% 89% 7% 172 (130, 250) 248 (131, 435)
* Defined as patients who smoke ⱖ1 tobacco product per day or smoked in the 30 days before admission. † Defined as a documented history of diabetes mellitus diagnosed and/or treated by a physician. ‡ Defined as a documented history of hypertension diagnosed and/or treated by a physician. § Defined as a documented history of hypercholesterolemia diagnosed and/or treated by a physician. 储 Data available in 1,897 patients. ¶ Data available in 1,272 patients.
covery analysis after PCI was not appropriate, and therefore these patients were also excluded. We assessed the association between ST-segment recovery, clinical characteristics (age, gender, hypertension, hypercholesterolemia, diabetes mellitus, current smoking, previous MI, coronary artery bypass grafting, or PCI), and procedural characteristics (left anterior descending coronary artery [LAD]–related MI, preprocedural and postprocedural TIMI-graded flow, use of glycoprotein IIb/IIIa inhibitors, and intra-aortic balloon pump insertion) using binary logistic regression. This association was studied in crude and multivariable models. Backward selections were used to select the most parsimonious set of predictive variables and results were expressed as odds ratios (ORs) with 95% confidence intervals (CIs). Total ischemic time and body mass index were not included in multivariable analyses because of a considerable amount of missing values. To construct a ST-segment recovery prediction model, a simple risk score was calculated. We assigned points according to the adjusted OR of each predictive characteristic that would be available after emergency coronary angiography but before PCI. A score was calculated by adding up points corresponding to each patient’s risk factors. Here-
1694
The American Journal of Cardiology (www.AJConline.org)
Table 2 Univariable and multivariable-adjusted associations between patient characteristics and incomplete ST-segment recovery Variable
Age (years) ⬎60 Men Current smoker Diabetes mellitus Hypertension Hypercholesterolemia Previous myocardial infarction Previous coronary bypass grafting Previous percutaneous coronary intervention
Incomplete ST-Segment Recovery
53% (603/1,128) 49% (744/1,531) 43% (421/984) 59% (138/235) 52% (343/658) 47% (219/465) 51% (122/237) 53% (8/15) 47% (69/146)
Univariable Analysis
Multivariable Analysis*
OR
CI
p Value
OR
CI
p Value
1.52 1.00 0.65 1.58 1.23 0.93 1.14 1.21 0.94
1.28–1.80 0.83–1.21 0.55–0.77 1.20–2.08 1.02–1.48 0.75–1.14 0.87–1.49 0.44–3.35 0.68–1.32
⬍0.001 0.99 ⬍0.001 0.001 0.03 0.47 0.35 0.71 0.74
1.28 — 0.79 1.36 — — — — —
1.06–1.54 — 0.65–0.95 1.02–1.82 — — — — —
0.011 — 0.013 0.034 — — — — —
* Backward selection model including patient characteristics and available procedural characteristics: left anterior descending coronary artery–related myocardial infarction, multivessel disease, preprocedural Thrombolysis In Myocardial Infarction grade ⬍3 flow, stent insertion, glycoprotein IIb/IIIa inhibitor use, postprocedural Thrombolysis In Myocardial Infarction grade ⬍3 flow, and intra-aortic balloon counterpulsation used.
Table 3 Univariable and multivariable-adjusted associations between procedural characteristics and incomplete ST-segment recovery Variable
Left anterior descending coronary artery–related myocardial infarction Multivessel disease Stent inserted Glycoprotein IIb/IIIa inhibitor used Intra-aortic balloon counterpulsation used Preprocedural Thrombolysis In Myocardial Infarction grade flow ⬍3 Postprocedural Thrombolysis In Myocardial Infarction grade flow ⬍3
Incomplete ST-Segment Recovery
Univariable Analysis
Multivariable Analysis*
OR
95% CI
p Value
OR
95% CI
p Value
57% (553/971)
1.86
1.57–2.21
⬍0.001
1.92
1.61–2.30
⬍0.001
54% (351/651) 48% (894/1,858) 56% (361/648) 62% (89/144)
1.36 0.86 1.51 1.78
1.13–1.64 0.67–1.11 1.25–1.82 1.26–2.52
0.001 0.25 ⬍0.001 0.001
1.34 — 1.33 —
1.10–1.63 — 1.09–1.62 —
0.004 — 0.004 —
48% (905/1,889)
0.78
0.60–1.03
0.08
0.70
0.53–0.93
0.014
68% (153/226)
2.43
1.81–3.26
⬍0.001
2.27
1.66–3.09
⬍0.001
* Backward selection model with procedural characteristics and patient characteristics as listed in Table 2.
after, we defined low-, intermediate-, and high-risk groups according to this risk score. Kaplan–Meier estimates according to category of STsegment recovery were determined and compared with the use of the log-rank test. To determine whether the prognostic value of ST-segment recovery was independent of associated patient and procedural characteristics, these variables (age ⬎60 years, smoking, diabetes mellitus, LAD-related MI, multivessel disease, TIMI-graded flow ⬍3 before PCI, use of IIb/IIIa inhibitors, and TIMI-graded flow ⬍3 after PCI) and ST-segment recovery were concurrently entered in a Cox regression model. Results were expressed as hazard ratios with 95% CIs. Normally distributed, continuous variables are expressed as mean ⫾ SD, and other continuous data are expressed as median with interquartile range. All categorical variables are depicted using relative frequency distributions. Characteristics of patients with complete and incomplete ST-segment recovery were compared using the chi-square test for categorical variables and the Student’s t test or Mann– Whitney U test for continuous variables. For all tests,
differences were considered significant if the 2-sided p value was ⬍0.05. All analyses were performed using SPSS 16.0 (SPSS, Inc., Chicago, Illinois). Results A total of 3,185 consecutive patients with STEMI underwent primary PCI at our institution from November 1, 2000, to January 1, 2007. We excluded 67 patients who underwent primary PCI of the left main coronary artery or of a bypass graft. ECGs of 508 patients could not be retrieved. Electrocardiographic exclusion criteria were applicable in 132 of 2,610 patients with a complete set of ECGs. Preprocedural ST-segment normalization had occurred in 354 patients. Thus, 2,124 patients could be included in this analysis. Baseline and procedural characteristics of these patients are presented in Table 1. Incomplete ST-segment recovery occurred in 1,032 of 2,124 patients (49%). The left side of Table 2 displays the univariable relation between occurrence of incomplete ST-segment recovery and various patient characteristics. Incomplete ST-segment
Coronary Artery Disease/Predictors of ST-Segment Recovery in STEMI
recovery was significantly more frequent in older patients and in patients with diabetes mellitus or hypertension, with unadjusted ORs of 1.52 (95% CI 1.28 to 1.80, p ⬍0.001), 1.58 (95% CI 1.20 to 2.08, p ⫽ 0.001), and 1.23 (95% CI 1.02 to 1.48, p ⫽ 0.03), respectively. Patients who were smokers at the time of primary PCI significantly less often had incomplete ST-segment recovery compared to nonsmoking counterparts (unadjusted OR 0.65, 95% CI 0.55 to 0.77, p ⬍0.001). In multivariable logistic regression, age ⬎ 60 years, nonsmoking status, and presence of diabetes mellitus were independently predictive of incomplete STsegment recovery (Table 2, right side). Table 3 presents univariable and multivariable associations between available procedural characteristics and occurrence of incomplete ST-segment recovery. As shown on the left side, incomplete ST-segment recovery was significantly more frequent in patients with LAD-related MI (unadjusted OR 1.86, 95% CI 1.57 to 2.21, p ⬍0.001), in patients with multivessel disease (unadjusted OR 1.36, 95% CI 1.13 to 1.64, p ⫽ 0.001), in patients with postprocedural TIMI-graded flow ⬍3 (unadjusted OR 2.43, 95% CI 1.81 to 3.26, p ⬍0.001), and in patients with a need for intra-aortic balloon pump insertion (unadjusted OR 1.78, 95% CI 1.26 to 2.52, p ⫽ 0.001). Also, incomplete ST-segment recovery was more common in patients who received glycoprotein IIb/IIIa inhibitors during PCI, with an unadjusted OR of 1.51 (95% CI 1.25 to 1.82, p ⬍0.001). Patients with TIMIgraded flow ⬍3 before PCI showed a trend toward less frequent incomplete ST-segment recovery (unadjusted OR 0.78, 95% CI 0.60 to 1.03, p ⫽ 0.08). After multivariable modeling, ORs of LAD-related MI, multivessel disease, glycoprotein IIb/IIIa inhibitor use, and TIMI-graded flow ⬍3 after PCI were sustained, whereas the association between intra-aortic balloon pump insertion and incomplete ST-segment recovery lost significance (Table 3, right side). The association between preprocedural TIMI-graded flow ⬍3 and incomplete ST-segment recovery was more outspoken in multivariable analysis, with an adjusted OR of 0.70 (95% CI 0.53 to 0.93, p ⫽ 0.014). We constructed a patient-specific ST-segment recovery risk score with the 6 predictive variables available before PCI. In accordance with their ORs (as listed in Tables 2 and 3), each characteristic (age ⬎60 years, nonsmoking, diabetes mellitus, multivessel disease, and TIMI-graded flow ⬍3 before PCI) contributed 1 point to the ST-segment recovery risk score, whereas LAD-related MI contributed 3 points. Median score of the study cohort was 3 (interquartile range 1 to 4). The C-statistic of the risk score to predict incomplete ST-segment recovery was 0.61 (95% CI 0.58 to 0.63, p ⬍0.001). By dividing patients into low-risk (0 point to 2 points), intermediate-risk (3 to 4 points), or high-risk (ⱖ5 points) subgroups, we identified 882 patients (41.5%) at low risk, 733 patients (34.5%) at intermediate risk, and 509 patients (24%) at high risk for incomplete ST-segment recovery. The OR for incomplete ST-segment recovery in patients at intermediate risk compared to patients at low risk was 1.81 (95% CI 1.49 to 2.21, p ⬍0.001) and that for patients at high risk was 2.68 (95% CI 2.14 to 3.35, p ⬍0.001). At a median follow-up of 4.1 years (interquartile range 2.7 to 5), 272 patients (85 of 1,078 patients with complete
1695
Figure 1. Cumulative mortality in patients with ⱖ50% versus ⬍50% ST-segment recovery at the end of primary PCI.
ST-segment recovery and 187 of 1,025 patients with incomplete ST-segment recovery) died. Estimated 5-year cumulative mortalities were 9.9% in patients with complete STsegment recovery and 21.4% in patients with incomplete ST-segment recovery (p ⬍0.001; Figure 1). Table 4 presents univariable and multivariable-adjusted relations between long-term mortality, on the 1 hand, and ST-segment recovery and its determinants, on the other. Incomplete ST-segment recovery was a strong univariable prognosticator (unadjusted hazard ratio 2.54, 95% CI 1.96 to 3.28, p ⬍0.001). Moreover, incomplete ST-segment recovery was a strong, independent predictor of long-term mortality in multivariable Cox regression analysis, with an adjusted hazard ratio of 2.07 (95% CI 1.59 to 2.69, p ⬍0.001). Discussion In this cohort of patients with primary PCI, we found that the following characteristics, available before PCI, were independent predictors of incomplete ST-segment recovery: age, smoking status, presence of diabetes mellitus, LADrelated MI, presence of multivessel disease, and preprocedural TIMI flow. By applying a simplified risk score, we could calculate a patient’s a priori risk of incomplete STsegment recovery. Patients in the high-risk category had a 2.7-fold higher risk for incomplete ST-segment recovery (62%) than patients in the low-risk category (38%). Furthermore, incomplete ST-segment recovery was a powerful predictor of long-term mortality in addition to the abovementioned characteristics that independently predicted incomplete ST-segment recovery. First, age being an independent predictor of incomplete ST-segment recovery seems reasonable when we acknowledge the time-dependent nature of arteriolosclerosis. The multifactorial cause of microvascular dysfunction (distal embolization, microvascular constriction through release of bioactive factors, vascular or myocyte edema, etc.) may
1696
The American Journal of Cardiology (www.AJConline.org)
Table 4 Result of unadjusted and multivariable-adjusted long-term mortality analyses for ST-segment recovery Variable
Incomplete ST-segment recovery Age ⬎60 years Current smoker Diabetes mellitus Left anterior descending coronary artery–related myocardial infarction Multivessel disease Preprocedural Thrombolysis In Myocardial Infarction grade ⬍3 flow Glycoprotein IIb/IIIa inhibitors used Postprocedural Thrombolysis In Myocardial Infarction grade ⬍3 flow
Unadjusted
Multivariable Adjusted*
HR
95% CI
p Value
HR
95% CI
p Value
2.54 3.63 0.51 1.76 1.23 1.83 1.03 0.96 2.47
1.96–3.28 2.71–4.86 0.41–0.67 1.29–2.41 0.97–1.57 1.44–2.33 0.71–1.51 0.74–1.25 1.84–3.30
⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.08 ⬍0.001 0.87 0.78 ⬍0.001
2.07 2.79 0.84 1.29 1.18 1.45 0.92 0.76 2.07
1.59–2.69 2.05–3.81 0.64–1.10 0.94–1.77 0.93–1.50 1.13–1.86 0.63–1.36 0.58–1.00 1.51–2.83
⬍0.001 ⬍0.001 0.20 0.12 0.19 0.004 0.69 0.046 ⬍0.001
* All listed variables were simultaneously entered in a Cox regression model. HR ⫽ hazard ratio.
likely be negatively influenced by aging. Second, the inverse relation between smoking status and occurrence of incomplete ST-segment recovery exemplifies the “smoker’s paradox,” which implies a favorable outcome of smokers with acute MI or heart failure.10,11 In this study we confirm the significant association between smoking status and STsegment recovery as previously reported in a small patient sample.12 Third, the independent association between presence of diabetes mellitus and occurrence of incomplete ST-segment recovery is in accordance with an increased risk of cardiovascular nonfatal and fatal events in diabetic patients.13,14 In addition, patients with diabetes who have MI have a significantly increased total ischemic time and continue to have a poorer prognosis compared to their nondiabetic counterparts.15 Fourth, patients with LAD-related MI had an adjusted two-fold increased risk of incomplete ST-segment recovery. Differences in ST-segment recovery after primary PCI between patients with anterior and nonanterior STEMI were previously reported in subgroup analyses from the Assessment of Pexelizumab in Acute Myocardial Infarction (APEX AMI) trial.6 A larger area at risk and a subsequent higher risk of microvascular dysfunction and tissue injury may largely account for this phenomenon. Fifth, multivessel disease as an independent characteristic can be explained by the assumption that more extensive atherosclerosis of the epicardial coronary arteries is associated with more outspoken dysfunction of the myocardial microvasculature, resulting in more frequent incomplete ST-segment recovery. The association between presence of multivessel disease and incomplete ST-segment recovery seems in agreement with the reported adverse prognosis in this patient subgroup.16 Interestingly, normal TIMIgraded flow before primary PCI was the sixth characteristic independently predictive of incomplete ST-segment recovery. Because patients with preprocedural ST-segment normalization were excluded from this study, persistent ST-segment elevation in the presence of TIMI grade 3 flow through the infarct-related coronary artery before primary PCI signals persistent microvascular dysfunction not responsive to epicardial reperfusion in these patients. Patients at high risk for microvascular dysfunction can be identified with our ST-segment recovery risk score. These patients are a target population for adjunctive therapy. Preprocedural risk stratification with subsequent deci-
sion making about adjunctive treatment enhances management of patients with STEMI. For example, patients at high risk could especially benefit from glycoprotein IIb/IIIa inhibitors in addition to aspirin, clopidogrel, and unfractionated heparin.17 In contrast, patients at low risk could benefit from a more restrictive antiplatelet regimen, which decreases the incidence of clinically relevant complications.18 –20 Nevertheless, increasing economic pressures have intensified the need for appropriate triage and clinical resource use. Our ST-segment recovery risk score is a tool increasing a clinician’s ability to rapidly and accurately assess risk and thus could be of substantial importance.6 An important limitation of our study is the retrospective collection of ECGs. From our source population, we excluded 16% of patients because of missing electrocardiographic recordings, which is similar to other observational studies with unselected patient samples.5,21 We could not trace patients with STEMI who developed left bundle branch block or ventricular arrhythmias, needed external pacing during PCI, or died during PCI. This may have introduced an outcome assessment bias, although we believe this is marginal. In addition, 1 investigator conducted all ST-segment recovery measurements. Nevertheless, other studies consistently have shown a low variability between observers,8,22 which reflects the straightforwardness of ST-segment analysis. Furthermore, this investigator was not aware of the baseline characteristics and subsequent outcome. Moreover, importation, processing, and analysis of electrocardiographic data were digital.
Acknowledgment: We gratefully acknowledge the technical and nursing staffs of the Cardiac Catheterization Laboratory of the Academic Medical Center, University of Amsterdam, for their skilled assistance and Kurdo Barwari, MSc, for his help in data collection. 1. Grines CL, Browne KF, Marco J, Rothbaum D, Stone GW, O’Keefe J, Overlie P, Donohue B, Chelliah N, Timmis GC. A comparison of immediate angioplasty with thrombolytic therapy for acute myocardial infarction. The Primary Angioplasty in Myocardial Infarction Study Group. N Engl J Med 1993;328:673– 679.
Coronary Artery Disease/Predictors of ST-Segment Recovery in STEMI 2. Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review of 23 randomised trials. Lancet 2003;361:13–20. 3. Gerber BL, Rochitte CE, Melin JA, McVeigh ER, Bluemke DA, Wu KC, Becker LC, Lima JA. Microvascular obstruction and left ventricular remodeling early after acute myocardial infarction. Circulation 2000;101:2734 –2741. 4. Nijveldt R, Beek AM, Hirsch A, Stoel MG, Hofman MB, Umans VA, Algra PR, Twisk JW, van Rossum AC. Functional recovery after acute myocardial infarction: comparison between angiography, electrocardiography, and cardiovascular magnetic resonance measures of microvascular injury. J Am Coll Cardiol 2008;52:181–189. 5. Brodie BR, Stuckey TD, Hansen C, VerSteeg DS, Muncy DB, Moore S, Gupta N, Downey WE. Relation between electrocardiographic STsegment resolution and early and late outcomes after primary percutaneous coronary intervention for acute myocardial infarction. Am J Cardiol 2005;95:343–348. 6. Buller CE, Fu Y, Mahaffey KW, Todaro TG, Adams P, Westerhout CM, White HD, van ’t Hof AW, Van de Werf FJ, Wagner GS, Granger CB, Armstrong PW. ST-segment recovery and outcome after primary percutaneous coronary intervention for ST-elevation myocardial infarction: insights from the Assessment of Pexelizumab in Acute Myocardial Infarction (APEX-AMI) trial. Circulation 2008;118:1335– 1346. 7. McLaughlin MG, Stone GW, Aymong E, Gardner G, Mehran R, Lansky AJ, Grines CL, Tcheng JE, Cox DA, Stuckey T, Garcia E, Guagliumi G, Turco M, Josephson ME, Zimetbaum P. Prognostic utility of comparative methods for assessment of ST-segment resolution after primary angioplasty for acute myocardial infarction: the Controlled Abciximab and Device Investigation to Lower Late Angioplasty Complications (CADILLAC) trial. J Am Coll Cardiol 2004;44: 1215–1223. 8. Van ’t Hof AW, Liem A, de Boer MJ, Zijlstra F. Clinical value of 12-lead electrocardiogram after successful reperfusion therapy for acute myocardial infarction. Zwolle Myocardial infarction Study Group. Lancet 1997;350:615– 619. 9. TIMI Study Group. The Thrombolysis In Myocardial Infarction (TIMI) trial. Phase I findings. N Engl J Med 1985;312:932–936. 10. Barbash GI, White HD, Modan M, Diaz R, Hampton JR, Heikkila J, Kristinsson A, Moulopoulos S, Paolasso EA, Van der Werf T. Significance of smoking in patients receiving thrombolytic therapy for acute myocardial infarction. Experience gleaned from the International Tissue Plasminogen Activator/Streptokinase Mortality Trial. Circulation 1993;87:53–58. 11. Fonarow GC, Abraham WT, Albert NM, Stough WG, Gheorghiade M, Greenberg BH, O’Connor CM, Nunez E, Yancy CW, Young JB. A smoker’s paradox in patients hospitalized for heart failure: findings from OPTIMIZE-HF. Eur Heart J 2008;29:1983–1991.
1697
12. Albertal M, Cura F, Escudero AG, Thierer J, Trivi M, Padilla LT, Belardi J. Mechanism involved in the paradoxical effects of active smoking following primary angioplasty: a subanalysis of the protection of distal embolization in high-risk patients with acute myocardial infarction trial. J Cardiovasc Med 2008;9:810 – 812. 13. Danaei G, Lawes CM, Vander HS, Murray CJ, Ezzati M. Global and regional mortality from ischaemic heart disease and stroke attributable to higher-than-optimum blood glucose concentration: comparative risk assessment. Lancet 2006;368:1651–1659. 14. Geiss LS, Herman WH, Smith PJ. Mortality in Non–Insulin-Dependent Diabetes. Diabetes in America, 2nd Ed. Bethesda: National Institute of Diabetes and Digestive and Kidney Diseases, National Institute of Health, 1995:233–258. 15. Donahoe SM, Stewart GC, McCabe CH, Mohanavelu S, Murphy SA, Cannon CP, Antman EM. Diabetes and mortality following acute coronary syndromes. JAMA 2007;298:765–775. 16. Van der Schaaf RJ, Timmer JR, Ottervanger JP, Hoorntje JC, de Boer MJ, Suryapranata H, Zijlstra F, Dambrink JH. Long-term impact of multivessel disease on cause-specific mortality after ST elevation myocardial infarction treated with reperfusion therapy. Heart 2006;92: 1760 –1763. 17. Kandzari DE, Hasselblad V, Tcheng JE, Stone GW, Califf RM, Kastrati A, Neumann FJ, Brener SJ, Montalescot G, Kong DF, Harrington RA. Improved clinical outcomes with abciximab therapy in acute myocardial infarction: a systematic overview of randomized clinical trials. Am Heart J 2004;147:457– 462. 18. Kereiakes DJ, Berkowitz SD, Lincoff AM, Tcheng JE, Wolski K, Achenbach R, Melsheimer R, Anderson K, Califf RM, Topol EJ. Clinical correlates and course of thrombocytopenia during percutaneous coronary intervention in the era of abciximab platelet glycoprotein IIb/IIIa blockade. Am Heart J 2000;140:74 – 80. 19. Montalescot G, Barragan P, Wittenberg O, Ecollan P, Elhadad S, Villain P, Boulenc JM, Morice MC, Maillard L, Pansieri M, Choussat R, Pinton P. Platelet glycoprotein IIb/IIIa inhibition with coronary stenting for acute myocardial infarction. N Engl J Med 2001;344: 1895–1903. 20. Eikelboom JW, Mehta SR, Anand SS, Xie C, Fox KA, Yusuf S. Adverse impact of bleeding on prognosis in patients with acute coronary syndromes. Circulation 2006;114:774 –782. 21. De Luca G, Suryapranata H, Ottervanger JP, Hoorntje JC, Gosselink AT, Dambrink JH, de Boer MJ, van ’t Hof AW. Postprocedural single-lead ST-segment deviation and long-term mortality in patients with ST-segment elevation myocardial infarction treated by primary angioplasty. Heart 2008;94:44 – 47. 22. Claeys MJ, Bosmans J, Veenstra L, Jorens P, De R, Vrints CJ. Determinants and prognostic implications of persistent ST-segment elevation after primary angioplasty for acute myocardial infarction: importance of microvascular reperfusion injury on clinical outcome. Circulation 1999;99:1972–1977.