Available online at www.sciencedirect.com

ScienceDirect

Journal of Electrocardiology xx (2015) xxx – xxxwww.jecgonline.com

Electrocardiogram score for the selection of reperfusion strategy in earlylatecomers with ST-segment elevation myocardial infarctionYu-Jiao Zhang, MD, PhD, a Wen Zheng, MD, PhD, a,⁎ Jian Sun, MD, PhD, a

Guo-Li Li, MD, b Bao-Rong Chi, MDa,⁎a Department of Cardiology, the First Hospital of Jilin University, Changchun, China

b Department of Cardiology, the People’s Hospital of Jilin City, Jilin, China

Abstract Objective: The clinical benefit of percutaneous coronary intervention (PCI) is controversial in ST-

⁎ CorrespondingHospital of Jilin Univ

E-mail addresses:

http://dx.doi.org/10.100022-0736/© 2014 El

segment elevation myocardial infarction (STEMI) patients presenting 12–72 hours after symptomonset. Several studies suggested this conflicting result was associated with myocardial area at risk(MaR) of enrolled patients. MaR could be estimated by the electrocardiogram (ECG) score. Ourobjective was to evaluate the benefits of PCI in STEMI latecomers with different MaR.Methods: We constructed a prospective cohort involving 436 patients presenting 12–72 hours afterSTEMI onset and who met an inclusion criteria. 218 underwent PCI and 218 received the optimalmedical therapy (OMT) alone. Individual MaR was quantified by the combined Aldrich ST andSelvester QRS score. The primary endpoint was a composite of cardiovascular death, reinfarction orrevascularization within two years.Results: The 2-year cumulative primary endpoint rate was respectively 9.2% in PCI group and 5.3%in OMT group when MaR b 35% (adjusted hazard ratio for PCI vs. OMT, 1.855; 95% confidenceinterval [CI], 0.617–5.575; P = 0.271), and was 12.8% in PCI group and 23.1% in OMT groupwhen MaR ≥35% (adjusted hazard ratio for PCI vs. OMT, 0.448; 95% CI, 0.228–0.884;P = 0.021).Conclusion: The benefit of PCI for the STEMI latecomers was associated with the MaR. PCI,compared with OMT, could significantly reduce the 2-year primary outcomes in patients withMaR ≥ 35%, but not in ones with MaR b 35%.© 2014 Elsevier Inc. All rights reserved.

Keywords: Electrocardiography/electrocardiogram(s); Myocardial infarction; Myocardial area at risk; Percutaneous coronaryintervention; Selvester QRS score; Aldrich ST score

Introduction

Primary percutaneous coronary intervention (PCI) is therecommended treatment for patients with ST-segment eleva-tion myocardial infarction (STEMI) within 12 hours afterischemic symptom onset [1]. However, the clinical benefit ofPCI vs. optimal medical therapy (OMT) alone is controversialin stable patients presenting over 12 hours timeframe [2–4]. Acomprehensive meta-analysis of 10 trials comparing theefficacy of late PCI vs. medical therapy alone in 3560 patientsrandomized over 12 hours after STEMI indicated thisconflicting result could be associated with enrolled patients’ischemia in the infarct-related artery territory [5].

authors at: Department of Cardiology, the Firstersity, Changchun, China, 130021.zhengw.jlu@gmail.com, chibr@jlu.edu.cn

16/j.jelectrocard.2015.01.004sevier Inc. All rights reserved.

This ischemic myocardial tissue within the vascularterritory that is distal to the culprit lesion of the infarct-related artery is usually called myocardial area at risk (MaR)[6]. MaR includes salvageable and infarcted myocardium.For example, following acute coronary occlusion the MaR isentirely ischemic, but over time the MaR gradually becomesa mix of both ischemic and infarcted myocardium. Theextent of MaR is typically identified and qualified by single-photoemission computed tomography (SPECT). Althoughthe SPECT analysis of MaR is a promising risk-stratificationtool for STEMI, it had very limited clinical applicability dueto high cost and uncommon availability [3,4]. In contrast, astandard 12-lead electrocardiogram (ECG) is inexpensiveand easily applied before reperfusion treatment has com-menced, and several algorithms have been developed toestimate MaR by using the Aldrich ST score and SelvesterQRS score [7–10]. The Aldrich ST score estimates the sizeof ischemic myocardium and Selvester QRS score estimates

2 Y.-J. Zhang et al. / Journal of Electrocardiology xx (2015) xxx–xxx

the size of the infarcted myocardium. As time progressesischemia is replaced by infarction and the primary ECG ST-segment deviation is replaced by QRS complex changes,especially in STEMI latecomers. Thus the combined utiliza-tion of Aldrich and Selvester score would be more stable andaccurate to estimate the MaR of these patients [11].

This study aimed to evaluate whether the combined ECGscore could be applied (1) to quantify the MaR of STEMIlatecomers, and (2) to evaluate the clinical benefit of PCI vs.OMT for these patients with different stratification of MaR. Itcould provide an easy and quick clinical assessment for theselection of reperfusion strategy.

Fig. 1. Flowchart of study population. STEMI indicates ST-segment elevationmyocardial infarction; SPECT, single-photon emission computed tomography.Propensity score matching was used to ensure similarity of patients in twogroups, matching each PCI case to one medical therapy case.

Methods

All STEMI patients at the First Hospital of Jilin Universitywere recorded in a prospective cohort. Data elements includeddemographic, clinical, angiographic/procedural, and follow-upvariables. Each patient had an outpatient visit at 1 month andseveral follow-up phone calls using a standardized question-naire at 6 months, 1 year, and then annually by trainedpersonnel to document long-term outcomes. In each contact,details of any readmission during that time period and/ormortality information were collected, including the date, theplace, and if the reason for readmission or death wascardiovascular or non-cardiovascular.

Study population

Patients (1) who were over 18 years-old, (2) hospitalizedwith a definitive diagnosis of new-onset STEMI throughJanuary, 2010 to January, 2012 and (3) presented to ourdepartment 12–72 hours after symptom onset were included inthis study. This hospitalization was defined as the indexhospitalization. The exclusion criteria included patients with (1)cardiogenic shock, electrical instability and severe congestiveheart failure (New York Heart Association III or IV) onadmission; (2) electrocardiogram presenting with complete leftor right bundle branch block,Wolff-Parkinson-White syndromeand left ventricular hypertrophy; (3) receiving coronary arterybypass grafting, and (4) in-hospital mortality during the indexhospitalization. In addition, (5) patients without angiographic orcomplete clinical data were also excluded. Then, patients weredivided into two groups according to their treatments: receivingPCI and OMT (PCI group) or OMT alone (OMT group). Wecalculated a propensity score for each patient and matched eachPCI case to one OMT case (Fig. 1). The study was approved bythe Ethics Committee of the First Hospital of Jilin University.

Treatments

All patients received optimal medical therapy, includingaspirin, anticoagulation if indicated, angiotensin-converting-enzyme inhibition (ACEI)/angiotensin receptor blocker(ARB), beta-blockade, and lipid-lowering therapy/plaquestabilization, unless contraindicated. Patients were assignedto perform coronary angiography/PCI within a few hoursafter decision if the condition was permitted, and eachpatient’s angiography record was collected.

ECG acquisition and evaluation

The standard 12-lead ECGs on admission from each patientwere collected within 10 minutes and transferred to a personalcomputer for post-processing. Standard ECG amplitude andduration measurements were performed automatically usingthe algorithm provided with the ECG Research WorkstationSoftware (GE Healthcare, Milwaukee, WI, USA). The resultsby computer measurements were then compared with thoseby manual measurements which were used to increasereproducibility and decrease the inter observer variability.

The Aldrich score in present study was based on the ruleset recommended by Bacharova et al. [12]. This scoringsystem was calculated to estimate the magnitude of STsegment elevation, which was measured at the J point in allrecordings. ST elevation was defined as an elevation of more

Fig. 2. lectrocardiogram scoring examples for estimating MaR. The 12-lead ECGs and algorithm for estimating myocardial area at risk (MaR) were shown in panel A and panel B. Panel A showed a 72 years old female with ECGchange of acute anterior myocardial infarction caused by the occlusion of left anterior descending artery. Panel B showed a 59 years old male with ECGchanges of acute inferior myocardial infarction caused by the occlusion of rightcorona artery.

3Y.-J.

Zhang

etal.

/Journal

ofElectrocardiology

xx(2015)

xxx–xxx

Esry

Table 1Baseline characteristics (N = 436).

PCI group(N = 218)

OMT group(N = 218)

P value

Demographic CharacteristicAge (means ± SD, y) 61 ± 14 61 ± 13 0.818Female (%) 59 (27.1) 58 (26.6) 0.914

Clinical characteristicCongestive heart failure⁎ (%) 118 (54.1) 100 (45.9) 0.085Hypertension (%) 117 (53.7) 113 (51.8) 0.701Hyperlipidemia (%) 13 (6.0) 19 (8.7) 0.271Diabetes mellitus (%) 52 (23.9) 58 (26.6) 0.222COPD (%) 6 (2.8) 9 (4.1) 0.601Renal failure (%) 7 (3.2) 12 (5.5) 0.174

Angiographic characteristicInfarct-related artery (%) 0.939

LAD 113 (51.8) 114 (52.3)LCX 36 (16.5) 38 (17.4)RCA 69 (31.7) 66 (30.3)

TIMI flow grade ininfarct-related artery (%)

0.412

0 206 (94.5) 211 (96.8)1 6 (2.8) 3 (1.4)2 4 (1.8) 3 (1.4)3 2 (0.9) 1 (0.5)

Collateral grade (means ± SD) 0.31 ± 0.66 0.29 ± 0.63 0.711Multivessel disease (%) 25 (11.5) 39 (17.9) 0.058LVEF (means ± SD) 50.0 ± 13.0 48.6 ± 11.2 0.243Estimated MaR (%LV) 35 ± 10 34 ± 8 0.305

SD, standard deviation; LAD, left anterior descending artery; LCX, leftcircumflex coronary artery; RCA, right coronary artery; TIMI, Thrombolysis inMyocardial Infarction; LVEF, left ventricle ejection fraction.

⁎ New York Heart Association functional class II heart failure.

4 Y.-J. Zhang et al. / Journal of Electrocardiology xx (2015) xxx–xxx

than 0.10 mV. The size of the ischemic component of theMaR (%LV) for anterior/inferior AMI was calculated byusing the following formulas:

Anterior−wall AMI %LVð Þ ¼ 3 1:5 No: of leads with ST↑ð Þ−0:4½ �Inferior−wall AMI %LVð Þ ¼ 3 0:6 ∑ST↑I I ; I I I ; aV Fð Þ þ 2:0½ �

The Selvester score estimated the size of the infarctedmyocardium based on modified 32 points QRS-scoring. Everypoint was developed to represent infarction of 3%of the total leftventricular mass (%LV) [7–9]. The electronic GE HealthcareQRS waveform amplitude and duration measurements wererecorded by two separated investigators.

The combined Aldrich and Selvester score, that was totalMaR (%LV), was determined by adding the Aldrich ST andthe Selvester QRS score (Fig. 2) [9,10].

MaR measured by SPECT

A subgroup of 32 patients received baseline viabilityscanning with a SPECT and 99mTc-Sestamibi to assess MaRbefore the procedure.

Angiographic evaluation

Coronary angiography data were assessed off-line in theAngiographic Core Laboratory by personnel unaware of studyallocation. Classification of anterograde coronary flow in theinfarct-related arterywas performed according to Thrombolysisin Myocardial Infarction (TIMI) classification. Angiographiccollateral vessels were scored using the grading systemproposed by Rentrop et al. [13]: 0 = none; 1 = visualizationof branches 0 ~ 1; 2 = branches plus a portion of the epicardialvessel; 3 = visualization of the entire vessel.

Outcomes definition

The primary outcome was major adverse cardiovascularevents (MACE), which was a composite of cardiovasculardeath, reinfarction and revascularization within two years.Cardiovascular death was defined as in or out of hospitalmortality due to ischemic heart disease or sudden cardiac arrest.The definition of reinfarction required two of the followingthree criteria: the persistence of symptoms for 30 or moreminutes, electrocardiographic changes, and elevated cardiacmarkers. The definition of revascularization referred to patientsreceiving revascularization procedures (PCI/CABG, but notincluding staged/assigned procedure) due to recurrent coronarystenosis or persistent chest pain after discharge from the index-hospitalization. Secondary endpoints included the separatecomponents of the primary endpoints within two years.

Statistical analysis

Descriptive data were reported as mean ± standard devia-tion (SD) or frequencies expressed as percentages. Pearsonχ2

and Student t testswere used for comparison among categoricaland continuous variables, respectively. The Spearman'scorrelation coefficient was used to measure the correlationbetween MaR estimated by ECG score vs. by SPECT.

A propensity score model was built to eliminate covariatedifferences; 2 cohorts of 1:1 nearest-neighbor-matchedpatients were consequently obtained [14,15]. The propensity

score was calculated using logistic regression covariatesincluding age, gender, comorbidities (congestive heart failure,hypertension, hyperlipidemia and diabetes), angiographic data(infarct-related vessels, TIMI flow grade, collateral grade,multivessel disease) and estimated MaR. After propensityscore matching, all baseline variables were comparable.

Primary and secondary outcomes during follow-up wereanalyzed using the Kaplan–Meier method, and treatmentswere compared by log-rank test. The covariate-adjusted hazardratio was quantified by utilizing a Cox proportional-hazardregression model that included eleven variables of interest:demographic (age and sex), clinical (congestive heart failure,hypertension, diabetes and hyperlipidemia), angiographic(infarct-related artery, TIMI, collateral flow grade and multi-vessel disease) and treatment (PCI vs. OMT) variables.Proportional-hazard assumptions, linearity of continuousvariables and lack of interactions were tested for each model.In addition, the subgroup analysis of admission time wasperformed by the same Cox proportional-hazards regressionmodel. The relationship between the estimated MaR and 2-year incidence of MACE was assessed as a continuousfunction and depicted with a quadratic regression model. Theincidence was calculated with adjustment for age and sex byeach three points of the MaR, fitting generalized estimatingequations with a Poisson link function.

All analyses were performed with SAS, version 9.3 64-bit(SAS Institute, U.S.A.). A 2-tailed P value b0.05was regardedas statistically significant in all calculations.

5Y.-J. Zhang et al. / Journal of Electrocardiology xx (2015) xxx–xxx

Results

Baseline patients characteristics

In all, 436 STEMI patients with average follow-up of2.7 ± 0.6 years were enrolled in this study. Of 218 (50%)were matched into the PCI group and the other 218 (50%)were matched into the OMT group. A detailed descriptionand distribution of baseline characteristics were given inTable 1. The baseline characteristics were all comparablebetween two groups.

Treatments in two groups

In the PCI group, average time from admission to procedurewas 45 hours. 58 patients underwent immediate PCI (theaverage interval from arrival: 1 hour) and 160 patientsunderwent elective PCI (the average interval from arrival: 65hours). The duration from decision (non-randomization) toPCI was less than 6 hours. PCI was performed successfully in205 patients (94%), and 198 had a TIMI flow grade of threeafter the procedure (successful PCI defined as an open arterywith residual stenosis of less than 50% and a TIMI flowantegrade grade of 2 or 3); 200 patients (92%) were placed atleast one stent, and 191 received drug-eluting stents;glycoprotein IIb/IIIa antagonists was administered to 94%patients during or after procedure. 4 patients (2%) receivedthrombolysis before admission.

In the OMT group, average interval from admission toangiography performed was 3.5 days. Only 2 patients (1%)received thrombolysis before admission.

During 2-year follow-up or until the events occurred, theusage rate of aspirin (96% vs. 94%, P = 0.381), beta-blocker (75% vs. 77%, P = 0.653), ACEI or ARB (62% vs.65%, P = 0.486) and statin (65% vs. 57%, P = 0.077) wassimilar in two groups, except for higher usage rate ofclopidogrel (65% vs. 38%, P b 0.001) in the PCI group thanthe OMT group.

Evaluation the correlation between ECG score and SPECT

In the subgroup of 32 patients (21 patients in the PCIgroup) who received SPECT, the MaR estimated by the ECGwas significantly correlated with that measured by SPECT(r =0.64, P b 0.001).

Table 2Primary and secondary outcomes.

Endpoints (2-ycumulative event rate)

MaR b 35%

PCI (N = 109) OMT (N = 114)

MACE⁎ (%) 10 (9.2) 6 (5.3)Cardiovascular death (%) 1 (0.9) 1 (0.9)Reinfarction (%) 1 (0.9) 3 (2.6)Nonfatal reinfarction (%) 1 (0.9) 3 (2.6)Revascularization† (%) 9 (8.3) 5 (5.4)

P values were calculated with the use of the log-rank test for Kaplan–Meier curve⁎ MACE refers to major adverse cardiovascular event including cardiovascu† Revascularization was defined as another procedure (PCI or CABG but not

persistent chest pain after discharge from the index hospitalization.

Comparison of MACE outcomes between PCI group andOMT group

In all 436 patients, the average MaR was 35%, which wassimilar in two groups (35% vs. 34%, P = 0.305). When 35%was selected as the cut-off, Kaplan–Meier analysis showed 2-year cumulative MACE incidence was similar (P = 0.278)between two groups in patients with MaR b 35%, while PCIgroup had significant lower incidence (P = 0.043) in patientswith MaR ≥ 35% (Table 2 and Fig. 3). The covariates-adjusted hazard ratio (HR) for PCI vs. OMT was 1.855 (95%confidence intervals: 0.617–5.575, P = 0.271) in the patientswith MaR b 35%, and the adjusted HR for PCI vs. OMT inpatients with MaR ≥ 35% was 0.448 (95% confidenceintervals: 0.228–0.884, P = 0.021). Furthermore, the relation-ship between the estimated MaR and the incidence of 2-yearMACE was depicted in the Fig. 4. Two curves representing forthe PCI group and the OMT group respectively crossed around35%. These two curves accompaniedwith each other before theintersection, and separated after the intersection with signifi-cantly ascending in OMT group.

On the basis of the study definition, one reinfarction andfour revascularization events occurred due to in-stent throm-bus/restenosis in PCI group with MaR b 35%, and onereinfarction and five revascularization events occurred due toin-stent thrombus/restenosis in PCI group with MaR ≥ 35%.In secondary outcomes, the cumulative events of PCI group,compared with OMT group, showed more revascularization (9(8.3%) vs. 5(5.4%), P = 0.246) when MaR b 35%, fewercardiovascular death (2(1.8%) vs. 7(6.7%), P = 0.078) andlower reinfarction (4(3.7%) vs. 11(10.8%), P = 0.045) whenMaR ≥ 35% (Table 2). As the “early latecomers” weredefined as 12–72 hours after onset of symptoms, it was along timespan.A subgroup analysis thereforewas performed inorder to reduce the impact of admission time. The resultsshowed no significant difference of benefits in differentsubgroup of admission time (Fig. 5).

Discussion

In this study, we evaluated the efficacy of PCI for stablepatients presenting 12–72 hours after STEMI onset withdifferentMaR.Our results showed that STEMI latecomerswithsevere ischemia (MaR ≥ 35%) could benefit from the patency

MaR ≥ 35%

P value PCI (N = 109) OMT (N = 104) P value

0.278 14 (12.8) 24 (23.1) 0.0430.972 2 (1.8) 7 (6.7) 0.0780.337 4 (3.7) 11 (10.8) 0.0450.337 4 (3.7) 10 (9.9) 0.0720.246 13 (11.9) 13 (12.8) 0.771

s through one year of follow-up.lar death, reinfarction or revascularization.include staged-procedure) performed due to recurrent coronary stenosis o

r

Fig. 3. Kaplan–Meier curves for the primary endpoint. The primary endpoint was the composite occurrence of cardiovascular death, reinfarction, orrevascularization. Kaplan–Meier estimates of the cumulative event rates in the PCI group and the optimal medical therapy (OMT) group, respectively, were11.0% and 13.8% for all enrolled patients (panel A), 9.2% and 5.3% for patients with MaR b 35% (panel B), and 12.8% and 23.1% for patients withMaR ≥ 35% (panel C). The P value was calculated with the use of the log-rank test.

6 Y.-J. Zhang et al. / Journal of Electrocardiology xx (2015) xxx–xxx

of infarct-related artery compared withmedical therapy only; inconstrast, patientswithmoderate ischemia (MaR b 35%) couldnot get more benefit from the patency of infarct-related arterythan medical therapy only. This study supports the hypothesisthat the benefit of delayed PCI is related to MaR, providing aneasy and quick clinical assessment for the selection ofreperfusion strategy for early latecomers with STEMI.

Fig. 4. Relationship between estimatedMaR and primary endpoint. The scatterplots and fixed quadratic regression curves representing for the PCI group andthe OMT group respectively were depicted for fitting the relationship betweenthe MaR estimated by the combined Aldrich and Selvester score and theadjusted incidence of primary endpoints (cardiovascular death, reinfarction, orrevascularization). The incidence was calculated with adjustment for age andsex by each three points of the MaR, fitting generalized estimating equationswith a Poisson link function. Dotted lines represent 95% confidence intervals ofpredicted incidence. Two curves representing for the PCI group and the OMTgroup respectively crossed around 35%. These two curves accompanied witheach other before the intersection, and separated after the intersection withsignificantly ascending in OMTgroup. R square in PCI groupwas 0.876 and Rsquare in OMT group was 0.834.

The ST deviations were highly correlated with MaRmeasured by SPECT only when ischemia was present duringthe initial time following coronary occlusion [16]. And theQRSabnormities have been well-developed to estimate infarct sizeduring the infarction process [13,16–18]. Bacharova L et al.[19] indicated that these changes of electrogenesis manifestingas ST segment deviations and QRS complex changes afterpersisted occlusion of a coronary artery and consequentischemia/infarction were related to slowing of ventricularactivation and changes in morphology and duration of actionpotentials in the affected area. Irene E.G. van Hellemond et al.[9,10] showed the combined Aldrich and Selvester score,estimating the MaR, was moderately correlated with thatmeasured by SPECT in either acute anterior or inferior-wallMI.Our studies similarly observed a significant correlation (r =0.64, P b 0.001) between the combined score and SPECT formeasuring MaR in STEMI latecomers.

The accepted definition of “early latecomers” refers topatients with STEMI presenting 12–72 hours after symptomsonset [20]. In China, there are only 3% STEMI patients whounderwent primary PCI within recommended 12 hours [21],and the overall rates of timely reperfusion therapies did notimproved during the past 10 years [22]. Therefore, there was aconsiderable proportion of “early latecomers” facing thedecision of whether should be performed delayed PCI. Ourstudy showed patients with moderate MaR could not get moreclinical benefit from PCI than OMT alone, and there was anincreased risk of repeat revascularization within the 2-yearfollow-up. Excess revascularization might derive on PCI-related restenosis because 4 of 9 events were related. Themechanism of PCI-related restenosis could be a consequenceof distal atherothrombotic embolization or microvascularplugging in the early stage or in-stent thrombosis in the latestage [3,23]. Although clopidogrel was prescribed at dischargein almost all patients with PCI, the agents were being taken inonly 65% patients when the events occurred. This could beanother reason for in-stent thrombosis. On the other hand,

Fig. 5. Subgroup analysis. Hazard ratios (black spots), 95% CIs (horizontal lines), cumulative 2-year primary outcomes (cardiovascular death, reinfarction, orevascularization) for different subgroups of admission time (12–48 hours and 48 hours–72 hours) were shown.

7Y.-J. Zhang et al. / Journal of Electrocardiology xx (2015) xxx–xxx

there was a high-risk of cardiac function deterioration,cardiogenic shock or even death for patients with severeMaR. PCI could significantly reduce the short-term cardio-vascular events by reducing infarction area and improvementon cardiac function [24,25]. Our result showed lowercumulative incidence of cardiovascular death (PCI vs. OMT:1.8% vs. 6.7%) and reinfarction (3.7% vs. 10.8%) occurred inPCI group when MaR ≥ 35%. Therefore, the estimated MaRis effective to stratify the risk for reperfusion strategy selection.

As the present study was based on a prospectiveobservational cohort, the grouping and the selection ofoperation timing were not randomly assigned compared withrandomized controlled trials. Therefore, there could exist somepotential impact factors leading to selective bias, but weconducted this cohort through propensity score matching andadjusted possible risk factors to reduce these biases. Of note,propensity score matching could result in the loss of cases, andthus we conducted a similar model in the subset of 681patients, and the trend of benefit did not change (adjusted HR:0.42, 95% CI: 0.20–0.94, P = 0.035). So we think it isreasonable to only conserve the patients with comparablebaseline characteristics (propensity score matching) forreducing selective bias given non-randomized sample selec-tion in this study. In addition, only patients alive at dischargefrom the index-hospitalization were enrolled. Although it couldunderestimate the event rates of two groups, this recruitmentwould reduce the unbalanced in-hospital mortality between twogroups due to time-to-treatment bias, especially in stable patients.

Conclusions

In summary, the benefit of PCI for the STEMI latecomerswas associatedwith theMaR. PCI, comparedwith OMT, couldsignificantly reduce the 2-year primary outcomes in patients

r

withMaR ≥ 35%, but not in oneswithMaR b 35%. TheMaRestimated by the combined Aldrich and Selvester score canprovide an easy and quick clinical assessment for the selectionof reperfusion strategy for early latecomers with STEMI.

Contributors

Study concept and design: Dr. Bao-Rong Chi, Dr. Yu-Jiao Zhang, Dr. Wen Zheng, Dr. Jian Sun and Dr. Guo-Li Li.

Manuscript writing:Dr. Yu-JiaoZhang andDr.Wen Zheng.Critical revision of the manuscript for important intellectual

content: Dr. Jian Sun, Dr. Bao-Rong Chi and Dr. Guo-Li Li.Dr. Bao-Rong Chi and Dr. Wen Zheng have full access to

all of the data in the study and take responsibility for theintegrity of the data and the accuracy of the data analysis andare the guarantors.

Conflict of interest

No conflict of interest.

Patient consent

Obtained.

Ethics approval

The Ethics Committee of the First Hospital of JilinUniversity.

Acknowledgement

Researchwas supported byResearchFunds of Jilin ProvincialScience &Technology Department (20130206021SF).

8 Y.-J. Zhang et al. / Journal of Electrocardiology xx (2015) xxx–xxx

References

[1] American College of Emergency P, Society for Cardiovascular A,Interventions, et al. 2013 ACCF/AHA guideline for the management ofST-elevation myocardial infarction: a report of the American College ofCardiology Foundation/American Heart Association Task Force onPractice Guidelines. J Am Coll Cardiol 2013;61(4):e78–e140, http://dx.doi.org/10.1016/j.jacc.2012.11.019 [published Online First: Epub Date].

[2] Erne P, SchoenenbergerAW,BurckhardtD, et al. Effects of percutaneouscoronary interventions in silent ischemia after myocardial infarction: theSWISSI II randomized controlled trial. JAMA 2007;297(18):1985–91,http://dx.doi.org/10.1001/jama.297.18.1985 [published Online First:Epub Date].

[3] Hochman JS, Lamas GA, Buller CE, et al. Coronary intervention forpersistent occlusion after myocardial infarction. N Engl J Med2006;355(23):2395–407, http://dx.doi.org/10.1056/NEJMoa066139[published Online First: Epub Date].

[4] Schomig A, Mehilli J, Antoniucci D, et al. Mechanical reperfusion inpatients with acute myocardial infarction presenting more than 12hours from symptom onset: a randomized controlled trial. JAMA2005;293(23):2865–72, http://dx.doi.org/10.1001/jama.293.23.2865[published Online First: Epub Date].

[5] Abbate A, Biondi-Zoccai GG, Appleton DL, et al. Survival and cardiacremodeling benefits in patients undergoing late percutaneous coronaryintervention of the infarct-related artery: evidence from a meta-analysisof randomized controlled trials. J Am Coll Cardiol 2008;51(9):956–64,http://dx.doi.org/10.1016/j.jacc.2007.11.062 [published Online First:Epub Date].

[6] Rochitte CE, Azevedo CF. The myocardial area at risk. Heart 2011,http://dx.doi.org/10.1136/heartjnl-2011-301332 [published OnlineFirst: Epub Date].

[7] Ideker RE, Wagner GS, Ruth WK, et al. Evaluation of a QRS scoringsystem for estimating myocardial infarct size. II. Correlation withquantitative anatomic findings for anterior infarcts. Am J Cardiol1982;49(7):1604–14.

[8] Roark SF, Ideker RE, Wagner GS, et al. Evaluation of a QRS scoringsystem for estimating myocardial infarct size. III. Correlation withquantitative anatomic findings for inferior infarcts. Am J Cardiol1983;51(3):382–9.

[9] van Hellemond IE, Bouwmeester S, Olson CW, et al. Consideration ofQRS complex in addition to ST-segment abnormalities in the estimated"risk region" during acute anterior myocardial infarction. J Electrocardiol2011;44(3):370–6, http://dx.doi.org/10.1016/j.jelectrocard.2011.01.004[published Online First: Epub Date].

[10] van Hellemond IE, Bouwmeester S, Olson CW, et al. Consideration ofQRS complex in addition to ST segment abnormalities in the estimationof the 'risk region' during acute inferior myocardial infarction. JElectrocardiol 2013;46(3):215–20, http://dx.doi.org/10.1016/j.jelectro-card.2013.02.004 [published Online First: Epub Date].

[11] Carlsen EA, Hassell ME, van Hellemond IE, et al. The stability ofmyocardial area at risk estimated electrocardiographically in patients withST elevation myocardial infarction. J Electrocardiol 2014;47(4):540–5,http://dx.doi.org/10.1016/j.jelectrocard.2014.04.016 [published OnlineFirst: Epub Date].

[12] Bacharova L, Mateasik A, Carnicky J, et al. The DipolarElectroCARdioTOpographic (DECARTO)-like method for graph-ic presentation of location and extent of area at risk estimated fromST-segment deviations in patients with acute myocardial infarc-

tion. J Electrocardiol 2009;42(2):172–80, http://dx.doi.org/10.1016/j.jelectrocard.2008.12.005 [published Online First: EpubDate].

[13] Rentrop KP, Cohen M, Blanke H, et al. Changes in collateral channelfilling immediately after controlled coronary artery occlusion by anangioplasty balloon in human subjects. J Am Coll Cardiol 1985;5(3):587–92.

[14] D'Agostino Jr RB. Propensity score methods for bias reduction in thecomparison of a treatment to a non-randomized control group. StatMed 1998;17(19):2265–81.

[15] Austin PC. An introduction to propensity score methods for reducing theeffects of confounding in observational studies. Multivar Behav Res2011;46(3):399–424, http://dx.doi.org/10.1080/00273171.2011.568786[published Online First: Epub Date].

[16] Persson E, Pettersson J, Ringborn M, et al. Comparison of ST-segmentdeviation to scintigraphically quantified myocardial ischemia duringacute coronary occlusion induced by percutaneous transluminal coronaryangioplasty. Am J Cardiol 2006;97(3):295–300, http://dx.doi.org/10.1016/j.amjcard.2005.08.044 [published Online First: Epub Date].

[17] Geerse DA, Wu KC, Gorgels AP, et al. Comparison between contrast-enhancedmagnetic resonance imaging and Selvester QRS scoring systemin estimating changes in infarct size between the acute and chronic phasesof myocardial infarction. AnnNoninvasive Electrocardiol 2009;14(4):360–5,http://dx.doi.org/10.1111/j.1542-474X.2009.00327.x [publishedOnline First:Epub Date].

[18] Loring Z, Chelliah S, Selvester RH, et al. A detailed guide forquantification of myocardial scar with the Selvester QRS score in thepresence of electrocardiogram confounders. J Electrocardiol 2011;44(5):544–54, http://dx.doi.org/10.1016/j.jelectrocard.2011.06.008 [pub-lished Online First: Epub Date].

[19] Bacharova L, Szathmary V, Mateasik A. QRS complex and ST segmentmanifestations of ventricular ischemia: the effect of regional slowing ofventricular activation. J Electrocardiol 2013;46(6):497–504, http://dx.doi.org/10.1016/j.jelectrocard.2013.08.016 [published Online First:Epub Date].

[20] Fontanelli A, Bonanno C. Primary percutaneous coronary interventionin 'early' latecomers with ST-segment elevation acute myocardialinfarction: the role of the infarct-related artery status. J Cardiovasc Med2011;12(1):13–8, http://dx.doi.org/10.2459/JCM.0b013e32834038d8[published Online First: Epub Date].

[21] Huo Y. Current status and development of percutaneous coronaryintervention in China. J Zhejiang Univ Sci B 2010;11(8):631–3, http://dx.doi.org/10.1631/jzus.B1001012 [published Online First: Epub Date].

[22] Li J, Li X, Wang Q, et al. ST-segment elevation myocardial infarctionin China from 2001 to 2011 (the China PEACE-Retrospective AcuteMyocardial Infarction Study): a retrospective analysis of hospital data.Lancet 2014, http://dx.doi.org/10.1016/s0140-6736(14)60921-1 [pub-lished Online First: Epub Date].

[23] Galiuto L, Paraggio L, Liuzzo G, et al. Predicting the no-reflowphenomenon following successful percutaneous coronary intervention.Biomark Med 2010;4(3):403–20, http://dx.doi.org/10.2217/bmm.10.55 [published Online First: Epub Date].

[24] Hochman JS, Sleeper LA, White HD, et al. One-year survival followingearly revascularization for cardiogenic shock. JAMA 2001;285(2):190–2.

[25] Wu AH, Parsons L, Every NR, et al. Hospital outcomes in patientspresenting with congestive heart failure complicating acute myocardialinfarction: a report from the Second National Registry of MyocardialInfarction (NRMI-2). J Am Coll Cardiol 2002;40(8):1389–94.