Canadian Journal of Cardiology 28 (2012) 40–47

Clinical ResearchReperfusion Strategies and Outcomes of ST-SegmentElevation Myocardial Infarction Patients in Canada:

Observations From the Global Registry of AcuteCoronary Events (GRACE) and the Canadian Registry of

Acute Coronary Events (CANRACE)Andrew Czarnecki, MD,a Robert C. Welsh, MD,b Raymond T. Yan, MD,c

J. Paul DeYoung, MD,d Richard Gallo, MD,e Barry Rose, MD,f Francois R. Grondin, MD,g

Jan M. Kornder, MD,h Graham C. Wong, MD,i Keith A. A. Fox, MB, ChB, FRCP,j

Joel M. Gore, MD,k Shaun G. Goodman, MD, MSc,a,c and Andrew T. Yan, MD;a,c on behalf ofthe Global Registry of Acute Coronary Events (GRACE/GRACE2) and Canadian Registry of

Coronary Events (CANRACE) Investigatorsa Terrence Donnelly Heart Centre, Division of Cardiology, St Michael’s Hospital, University of Toronto, Toronto, Ontario, Canada

b Mazankowski Alberta Heart Institute, University of Alberta, Edmonton, Alberta, Canadac Canadian Heart Research Centre, Toronto, Ontario, Canadad Cornwall Community Hospital, Cornwall, Ontario, Canada

e Montréal Heart Institute, Université de Montréal, Montréal, Québec, Canadaf Health Sciences Centre, St John’s, Newfoundland, Canada

g Hôtel-Dieu de Lévis, Lévis, Québec, Canadah Surrey Memorial Hospital, Surrey, British Columbia, Canada

i University of British Columbia, Vancouver, British Columbia, Canadaj Cardiovascular Research, Division of Medical and Radiological Sciences, University of Edinburgh, Edinburgh, Scotland

k Department of Medicine, Division of Cardiovascular Medicine, University of Massachusetts Medical School, Worcester, Massachusetts, USA

ABSTRACTBackground: We examine the clinical characteristics and outcomes ofST-elevation myocardial infarction (STEMI) patients receiving variousreperfusion therapies in 2 contemporary Canadian registries.Methods: Of 4045 STEMI patients, 2024 received reperfusion therapyand had complete data on invasive management. They were stratified

by reperfusion strategy used: primary percutaneous coronary interven-

See page 45 for disclosure information.

0828-282X/$ – see front matter © 2012 Canadian Cardiovascular Society. Publisheddoi:10.1016/j.cjca.2011.09.011

RÉSUMÉIntroduction : Nous avons examiné les caractéristiques et les résul-tats cliniques des patients ayant un infarctus du myocarde (IM) avecsus-décalage du segment ST de 2 registres canadiens contemporainset recevant divers traitements de reperfusion.Méthodes : Des 4 045 patients ayant un IM avec sus-décalage du

segment ST, 2 024 ont reçu un traitement de reperfusion et ont eu des

The optimal therapeutic approach for ST-segment elevationmyocardial infarction (STEMI) has been long debated. Both

Received for publication May 31, 2011. Accepted September 15, 2011.

Corresponding author: Dr Andrew T. Yan or Dr Shaun G. Goodman, St Mi-chael’s Hospital, Division of Cardiology, 30 Bond St., Room 6-030 Queen, Toronto,Ontario M5B 1W8, Canada. Tel.: �1-416-864-5465; fax: �1-416-864-5159.

E-mail: yana@smh.ca

fibrinolysis and primary percutaneous coronary intervention(PCI) are well-established reperfusion strategies with a wealthof literature supporting their efficacy with respect to both mor-bidity and mortality.1 While the superiority of timely primaryPCI has been established and most Canadian interventionalcentres have adopted primary PCI as the standard of care,2

many patients presenting to hospitals without on-site PCI fa-

cilities continue to receive fibrinolysis.

by Elsevier Inc. All rights reserved.

Czarnecki et al.Reperfusion Strategies and Outcomes of STEMI

41

tion (PCI) (n �716); fibrinolysis with rescue PCI (n �177); fibrinolysiswith urgent/elective PCI (n �210); and fibrinolysis without PCI(n �921). Data were collected on clinical and laboratory findings, andoutcomes.Results: Compared with fibrinolytic-treated patients, patients treatedwith primary PCI were younger and had higher Killip class, had longertime to delivery of reperfusion therapy, and utilized more antiplatelettherapy but less heparin, �-blockers and angiotensin-converting en-zyme inhibitors. In-hospital death occurred in 2.7% of patients treatedwith primary PCI, 1.7% fibrinolysis-rescue PCI, 1.0% fibrinolysis-ur-gent/elective PCI, and 4.8% fibrinolysis-alone (P �0.009); the rates ofdeath/reinfarction were 3.9%, 4.0%, 4.3%, and 7.1% (P �0.032),respectively. The rate of shock was highest in the primary PCI group.Rates of heart failure or major bleeding were similar in the 4 groups.In multivariable analysis, no PCI during hospitalization was associatedwith death and reinfarction (adjusted odds ratio � 1.66; 95% confi-dence interval, 1.03-2.70; P �0.04).Conclusions: Clinical features, time to reperfusion, and medicationutilization differed with respect to the reperfusion strategy. While lowrates of re-infarction/death were observed, these complications oc-curred more frequently in those who did not undergo PCI during indexhospitalization.

données complètes sur un traitement effractif. Ils ont été stratifiésselon la stratégie de reperfusion utilisée : l’intervention coronariennepercutanée (ICP) primaire (n � 716); la fibrinolyse avec ICP urgente(n � 177); la fibrinolyse avec ICP urgente ou élective (n � 210); et lafibrinolyse sans ICP (n � 921). Les données ont été recueillies desdécouvertes cliniques et de laboratoire, et des résultats.Résultats : Comparativement aux patients traités par fibrinolyse, les pa-tients traités par une ICP primaire étaient plus jeunes et avaient uneclassification de Killip plus élevée, avaient un délai d’attente plus longavant le traitement de reperfusion et utilisaient plus de traitements anti-plaquettaires, mais moins d’héparine, de �-bloquants et d’inhibiteurs del’enzyme de conversion de l’angiotensine. Il y a eu mortalité à l’hôpitalchez 2,7 % des patients traités par l’ICP primaire, 1,7 % par la fibrinolyseavec ICP urgente, 1,0 % par la fibrinolyse avec ICP urgente ou élective, et4,8 % par la fibrinolyse seule (P � 0,009); les taux de mortalité et derécidive d’infarctus ont été de 3,9 %, 4,0 %, 4,3 % et 7,1 % (P � 0,032),respectivement. Le taux de choc a été plus élevé dans le groupe d’ICPprimaire. Les taux d’insuffisance cardiaque ou de saignement majeur ontété similaires dans les 4 groupes. Dans l’analyse multivariée, aucune ICPdurant l’hospitalisation n’a été associée à la mortalité et à la récidived’infarctus (ratio d’incidence approché ajusté � 1,66; intervalle de confi-ance de 95 %, 1,03-2,70; P � 0,04).Conclusions : Les caractéristiques cliniques, le délai d’attente avantla reperfusion et l’utilisation de la médication ont différé selon lastratégie de reperfusion. Tandis que de faibles taux de récidived’infarctus et de mortalité ont été observés, ces complications se sontprésentées plus fréquemment chez ceux qui n’avaient pas subi d’ICP

Particularly important in the context of limited access toprimary PCI in Canada, a number of studies have suggestedthat very early treatment with fibrinolytics may offer a benefitover primary PCI with respect to the incidence of cardiogenicshock and mortality. The Which Early ST-Elevation (WEST)study randomized Canadian STEMI patients to primary PCIor fibrinolysis with or without mechanical cointervention.3 Al-though no difference was demonstrated in the composite end-point in this pilot study, primary PCI was associated with non-significantly lower rates of reinfarction but increased heartfailure (HF) and cardiogenic shock, as compared with patientstreated with tenecteplase followed by either usual care or earlyPCI. In addition, a subgroup analysis of the Comparison of An-gioplasty and Prehospital Thrombolysis in Acute Myocardial In-farction (CAPTIM) trial suggested that early (� 2 hours) treat-ment with fibrinolytics resulted in lower rates of cardiogenic shockwhen compared with primary PCI with enrollment in the pre-hospital environment.4,5 While these trials were not adequatelypowered to definitively guide routine clinical practice, there is anongoing randomized trial seeking to address this issue.6 Further-more, the use of currently available reperfusion treatment strate-gies and their associated clinical outcomes have not been well de-scribed in Canadian patients in the “real world.”

Accordingly, the objective of this observational study was todescribe the patient characteristics, reperfusion modalities, andclinical outcomes in STEMI patients receiving primary PCI orfibrinolysis in the Global Registry of Acute Coronary Events(GRACE) and Expanded GRACE (GRACE2) and the Cana-dian Registry of Acute Coronary Events (CANRACE).

MethodsThe GRACE, GRACE2, and CANRACE projects have

been previously described.7–9 Between 1999 and 2008, pa-

tients with acute coronary syndromes (ACS) were enrolledfrom 55 hospitals across Canada. The CANRACE registry wasinitiated in 2008 after GRACE and GRACE2 ceased patientenrolment in 2007. Patients were included if they were 18 yearsold or older, and they had a diagnosis of ACS with 1 of thefollowing: electrocardiographic changes consistent with ACS,serial increases in serum biochemical markers of cardiac necro-sis, and/or documented history of coronary artery disease. Pa-tients were excluded if they had an ACS precipitated by seriousnoncardiac comorbidities. Consecutive patient recruitmentwas encouraged at all sites. Standardized data collection formswere used and included patient demographics, clinical charac-teristics, treatment, and outcome data. Forms were completedby designated study coordinators or treating physicians. Localethics board approval was obtained at each participating hos-pital, and study participants provided informed consent. Thepurpose of this observational study was to describe the “realworld” patient characteristics and clinical outcomes of patientswith STEMI in Canada treated with reperfusion therapy.Therefore, patients who did not receive reperfusion were ex-cluded from our analysis.

Patient stratification

Of 14,432 patients enrolled, 4045 (28%) had a final diag-nosis of STEMI; patients who did not receive acute reperfusiontherapy (n � 1267) were excluded. Of the 2778 patients whoreceived reperfusion therapy, 2024 had complete data on inva-sive management and were included in this analysis. We strat-ified the study population in accordance with the acute reper-fusion strategy and in the fibrinolysis population by subsequentin-hospital PCI: (1) primary PCI; (2) fibrinolysis with rescuePCI; (3) fibrinolysis with urgent/elective PCI; and (4) fibrino-

durant l’hospitalisation.

lysis without PCI during index hospitalization. Rescue PCI was

42 Canadian Journal of CardiologyVolume 28 2012

defined as PCI after failed fibrinolysis which refers to ongoing/recurrent ischemic discomfort and/or lack of ST-segment ele-vation resolution or recurrent ST elevation. In contrast, ur-gent/elective PCI included: early PCI for cardiogenic shock,PCI for unstable angina, PCI for post-acute myocardial infarc-tion (AMI) ischemia, facilitated PCI (immediate PCI follow-ing successful fibrinolysis, or in conjunction with fibrinolysis),non-emergent adjunctive PCI of non-culprit lesion, and non-emergent elective PCI of suspected culprit lesion. All PCI oc-curred during the index admission. Baseline characteristics,surrogate markers of infarct size, and in-hospital clinical eventswere described. Clinical outcomes in this study included rein-farction, HF, cardiogenic shock, and in-hospital mortality.Rates of major bleeding and sustained ventricular tachycardiaare also presented.

Statistical analysis

Data were analyzed with respect to therapeutic strategy.Continuous variables are reported as means with standarddeviations or medians with interquartile ranges (where ap-propriate), and compared using the Kruskal-Wallis test (orMann-Whitney U test). Categorical variables are presented aspercentages, and compared using Pearson �2 (or Fisher’s exacttest, where appropriate).

Multivariable models were developed for the combinedendpoints including death and reinfarction; death, HF, andshock; as well as death, reinfarction, HF, and shock. The inde-pendent variable was the reperfusion strategy applied (primaryPCI vs all patients treated with fibrinolysis; primary PCI vsfibrinolysis with rescue PCI, or fibrinolysis with urgent/elective

Table 1. Baseline characteristics of the study patients

Primary PCI(n � 716)

Overall(n � 1308)

Re(n

Age 60 � 13 62 � 12Female gender 23.5 26.5Diabetes mellitus 16.4 17.4Smoker 39.9 41.3Hypertension 42.7 45.9Hyperlipidemia 41.1 41.1Peripheral artery disease 2.8 3.7TIA/stroke 2.4 3.8History of bleeding 1.8 1.5Angina 13.3 19.8History of heart failure 1.7 3.2Prior MI 12.2 17.8Prior PCI 9.0 8.1Prior coronary bypass surgery 1.4 2.7Heart rate (Bpm) 77 � 20 75 � 20Systolic BP (mm Hg) 137 � 30 142 � 29 1Diastolic BP (mm Hg) 82 � 19 83 � 19Killip Class

Killip I 88.6 90.6Killip II 7.3 7.3Killip III/IV 4.0 2.1

Cardiac arrest 5.2 3.3Creatinine (�mol/L) 95 � 30 97 � 31GRACE risk score 134 � 33 134 � 35 1

Data are presented as percentages or mean � SD.BP, blood pressure; Bpm, beats per minute; GRACE, Global Registry

intervention; TIA, transient ischemic attack.*Four-group (primary PCI and each individual fibrinolysis group) compa

† Primary PCI vs overall fibrinolysis.

PCI, or fibrinolysis with no PCI), adjusted for the GRACE riskscore, which is a validated predictor of in-hospital mortali-ty,10–12 as well as diabetes and hypertension. Adequate modeldiscrimination and calibration were achieved. Statistical signif-icance was established with a 2-sided P value of � 0.05. SPSSversion 15.0 (SPSS Inc, Chicago, IL) statistical software wasused for analysis.

ResultsOf the 2024 patients included in this study, 716 underwent

primary PCI, and 1308 received fibrinolysis; of these patients177 had fibrinolysis with rescue PCI, 210 had fibrinolysis withurgent/elective PCI, and 921 had fibrinolysis without in-hos-pital PCI.

Baseline characteristics

The baseline demographic and clinical characteristics of thepatients in each group were generally similar (Tables 1 and 2).However, patients treated with fibrinolytic only (ie, no in-hos-pital PCI) were older, and more likely had a history of heartdisease including angina, previous myocardial infarction, andHF. The primary PCI and fibrinolysis-only groups tended tohave higher GRACE risk scores. Medication utilization in thefirst 24 hours was notable for significantly more use of clopi-dogrel and glycoprotein IIb/IIIa inhibitors in the primary PCIpatients, while fibrinolytic-treated patients more frequently re-ceived heparin. �-blocker use was more common in the fi-brinolytic-treated group compared with the primary PCIgroup.

Fibrinolysis

P* P†I)

Urgent/elective PCI(n � 210)

No PCI(n � 921)

60 � 12 63 � 13 � 0.001 0.03423.3 27.3 0.29 0.1213.4 17.9 0.31 0.7039.5 40.6 0.42 0.5042.3 47.2 0.25 0.1541.3 41.9 0.60 0.98

2.9 4.2 0.36 0.263.4 4.3 0.11 0.151.9 1.6 0.35 0.35

15.4 22.1 � 0.001 � 0.0012.4 3.7 0.041 0.11

15.9 19.4 0.001 0.00210.6 8.0 0.36 0.511.0 3.2 0.058 0.15

73 � 20 76 � 20 0.001 0.029142 � 28 143 � 29 0.003 0.00184 � 17 83 � 19 0.27 0.25

0.21 0.02792.3 90.26.3 7.31.4 2.43.0 3.3 0.23 0.042

92 � 21 99 � 34 0.002 0.024131 � 32 135 � 36 0.013 0.069

Coronary Events; MI, myocardial infarction; PCI, percutaneous coronary

scue PC� 177

58 � 1126.619.946.643.536.4

2.31.70.0

13.61.1

11.95.72.3

73 � 2039 � 3185 � 20

90.38.61.14.6

93 � 2330 � 36

of Acute

risons.

Czarnecki et al.Reperfusion Strategies and Outcomes of STEMI

43

The median time (25th-75th percentiles) from symptomonset to reperfusion with primary PCI (first balloon inflation)was 229 (161-370) minutes. Time from symptom onset-to-needle was 143 (85-249) minutes in the fibrinolysis group. Therescue PCI group received fibrinolytic therapy in 124 (80-200)minutes, while the urgent/elective PCI and no PCI groupsreceived fibrinolysis in 145 (81-225) and 145 (90-261) min-utes from symptom onset, respectively (Table 3).

Laboratory measurements of creatine kinase (CK), CK-MB, inotrope use, electrocardiogram analysis, and left ventric-ular function are shown in Table 3.

Clinical outcomes

Cardiovascular events and composite outcomes are listed inTable 4. Death occurred in 2.7% of the primary PCI patients,1.7% of rescue PCI, 1.0% of the urgent/elective PCI, 4.8% ofthe fibrinolysis-alone cohort (P � 0.009), and 3.7% in fibri-nolysis overall (P � 0.2; vs primary PCI). Death or reinfarctionoccurred less frequently with primary PCI, as compared with

Table 2. Medication use in the first 24 hours

Primary PCI(n � 716)

Overall(n � 1308)

R

Aspirin 96.4 95.8Clopidogrel 93.9 53.7Unfractionated heparin 77.9 54.3Enoxaparin 10.1 48.8Any heparin 85.8 96.7Glycoprotein IIb/IIIa inhibitor 56.2 6.7�-Blockers 77.2 83.5Statin 74.3 67.3ACE inhibitor 54.5 60.8Calcium antagonist 8.6 6.6Inotrope use in first 24 h or after 10.3 7.7Inotrope use within first 24 h 9.1 6.3Inotrope use after first 24 h 4.7 3.3

Data are presented as percentages.ACE, angiotensin-converting enzyme; PCI, percutaneous coronary interv*Four-group (primary PCI and each individual fibrinolysis group) compa† Primary PCI vs overall fibrinolysis.

Table 3. Infarct size and left ventricular function in relation to reperfu

Primary PCI(n � 716)

Overall(n � 1308)

Max CK-MB (first 24 h) (n � 567) 108 (34-265) 98 (37-203)Max CK (first 24 h) (n � 1180) 1446 (628-2778) 930 (436-2110)Time to reperfusion therapy (min)

(n � 1726) 229 (161-370) 143 (85-249)New Q wave (n � 2024) 16.6 13.5New Q wave after index event

(n � 1999) 23.3 24.1Left ventricular function (n � 1410)

Normal 36.9 43.3Mild dysfunction 39.6 38.2Moderate dysfunction 18.7 14.4Severe dysfunction 4.8 4.1

Numerical values represent medians with interquartile range or %.CK, creatine kinase; PCI, percutaneous coronary intervention.*Four-group (primary PCI and each individual fibrinolysis group) compa

† Primary PCI vs overall fibrinolysis.

the patients who received fibrinolysis (3.9% and 6.2%, respec-tively, P � 0.03). Primary PCI had a higher incidence of car-diogenic shock as compared with fibrinolysis overall, with sim-ilar rates of HF and sustained ventricular tachycardia. Rates ofHF, ventricular tachycardia, or major bleeding were similar,with respect to the different fibrinolysis groups. There wasmore major bleeding within the primary PCI group comparedwith the overall fibrinolysis group.

In multivariable analysis, not undergoing PCI was indepen-dently associated with in-hospital death and reinfarction ascompared with primary PCI (adjusted odds ratio � 1.66; 95%confidence interval, 1.03-2.70; P � 0.04) (Table 5). Exclusionof patients who died within the first day from the analysis didnot significantly change the results. A Cox regression analysiswas also performed and showed similar results. In addition,higher GRACE risk score and diabetes independently pre-dicted in-hospital death and reinfarction as well as the compos-ite of death, reinfarction, HF, and shock. Analysis of the overallfibrinolysis group revealed that the reperfusion strategy (PCI vs

Fibrinolysis

P* P†CI7)

Urgent/elective PCI(n � 210)

No PCI(n � 921)

97.6 95.1 0.24 0.3871.1 43.8 � 0.001 � 0.00155.3 54.4 � 0.001 � 0.00151.7 47.4 � 0.001 � 0.00197.1 96.7 � 0.001 � 0.0018.9 2.7 � 0.001 � 0.001

80.3 84.8 0.001 � 0.00173.3 65.0 � 0.001 0.00262.9 60.9 0.037 0.005

5.3 7.3 0.12 0.0675.3 8.1 0.12 0.0494.3 6.5 0.075 0.0222.4 3.6 0.39 0.14

rategies

Fibrinolysis

P* P†scue PCI� 177)

Urgent/elective PCI(n � 210)

No PCI(n � 921)

(63-285) 137 (39-205) 82 (35-179) 0.031 0.99(667-2507) 1044 (420-2023) 802 (404-2047) � 0.001 � 0.001

(80-200) 145 (81-225) 145 (90-261) � 0.001 � 0.00112.4 6.7 15.2 0.003 0.056

29.3 29.5 22.0 0.039 0.680.04 0.046

44.5 47.1 41.736.3 35.7 39.617.1 15.9 13.12.1 1.3 5.6

escue P(n � 17

97.281.552.652.596.023.080.172.357.7

4.08.77.52.9

ention.risons.

sion st

Re(n

1521266

124

risons.

44 Canadian Journal of CardiologyVolume 28 2012

fibrinolysis) was not associated with HF, death/reinfarction,and death/HF/shock (Table 5). GRACE risk score, the pres-ence of diabetes, and hypertension were independently associ-ated with HF.

DiscussionIn this observational study, patients presenting to hospitals

with STEMI treated with primary PCI were younger, withhigher Killip class, and markers of infarct size. The fibrinolytic-treated group received less antiplatelet drugs and inotropes, butmore heparin, �-blockers and angiotensin-converting enzymeinhibitors. Irrespective of the initial reperfusion strategy, pa-tients had low rates of morbidity and mortality. Those treatedwith primary PCI had the lowest rates of death and reinfarction(driven primarily by less reinfarction), while having more car-diogenic shock and bleeding episodes. The lack of PCI during

Table 4. In-hospital outcomes in relation to reperfusion strategies

Primary PCI(n � 716)

Overall(n � 1308)

R

Length of stay‡ 4 � 3.6 5 � 3.8In-hospital death 2.7 3.7In-hospital death/re-MI 3.9 6.2Cardiogenic shock 4.5 2.7HF/pulmonary edema 11.7 10.2Sustained ventricular tachycardia 4.9 5.7Major bleeding 3.7 1.8

Data are percentages.HF, heart failure; MI, myocardial infarction; PCI, percutaneous coronary*Represents 4-group (primary PCI and each individual fibrinolysis group† Primary PCI vs overall fibrinolysis.‡ Mean � SD.

Table 5. Multivariable analysis

Odds ratio 95% CI P

In-hospital death/re-MIPrimary PCI ReferenceFibrinolysis (overall) 1.44 0.90-2.29 0.13Rescue PCI 0.70 0.25-1.95 0.49Urgent/elective PCI 0.99 0.41-2.37 0.98No PCI 1.66 1.03-2.70 0.039

In-hospital death/re-MI/HF/shockPrimary PCI ReferenceRescue PCI 1.29 0.77-2.15 0.34Urgent/elective PCI 1.03 0.62-1.72 0.91No PCI 0.97 0.70-1.33 0.86

HF/shockPrimary PCI ReferenceRescue PCI 1.30 0.76-2.23 0.34Urgent/elective PCI 0.87 0.50-1.52 0.62No PCI 0.77 0.54-1.07 0.12

In-hospital death/HF/shockPrimary PCI ReferenceFibrinolysis (overall) 0.95 0.70-1.30 0.75

HFPrimary PCI ReferenceFibrinolysis (overall) 0.84 0.61-1.15 0.28

Model adjusted for GRACE risk score, hypertension and diabetes mellitus.CI, confidence interval; GRACE, Global Registry of Acute Coronary

Events; HF, heart failure; MI, myocardial infarction; PCI, percutaneous cor-

onary intervention.

hospital admission was a predictor of higher mortality and re-infarction.

Both fibrinolytic therapy and primary PCI are effective rep-erfusion strategies which reduce morbidity and mortality.1 Anumber of studies have demonstrated a mortality benefit toprimary PCI (if it can be performed in a timely manner inexperienced centres) over fibrinolytic therapy13 and thus, this isthe preferred strategy in the most recent American College ofCardiology and American Heart Association guidelines on themanagement of STEMI.14 However, the results of this studyare consistent with previous registry data suggesting that thesurvival advantage of primary PCI is not necessarily reproduc-ible in the real world.1,15–17 The WEST study revealed a ten-dency toward increased rates of HF and shock in patients ran-domized to primary PCI when compared with pre-hospitalfibrinolysis.3 This effect was previously seen in the CAPTIMstudy where patients treated with pre-hospital fibrinolysis lessthan 2 hours from symptom onset, had significantly lower ratesof cardiogenic shock and a trend toward lower mortality.4,5 Inaddition, the French Unité de Soins Intensifs Coronaires(USIC) 2000 registry showed a clear survival benefit to pre-hospital fibrinolysis largely driven by timely delivery of ther-apy, with no differences in the development of cardiogenicshock or HF.18 The Primary Angioplasty in Patients Trans-ferred From General Community Hospitals to SpecializedPTCA Units With or Without Emergency Thrombolysis(PRAGUE)-2 trial demonstrated a survival advantage with pri-mary PCI when presentation was delayed by more than 3hours;19 in contrast, there were similar survival rates for fi-brinolytic-treated and primary PCI patients treated within 3hours of symptom onset.

Our data describe real-world therapeutic and outcome dis-parities seen in patients presenting with STEMI. The increasedrates of cardiogenic shock seen with primary PCI may be ex-plained by 2 factors. First, there is considerable evidence thatdelays in establishing reperfusion are strongly associated withpoor outcomes.20 In our study, the difference in median timeto delivery of reperfusion therapy was 86 minutes longer forpatients treated with primary PCI. Increased rates of cardio-genic shock in the CAPTIM trial were largely explained by theinitiation of shock during the 1-hour delay from the time ofrandomization, to admission and primary PCI. Shock did not

Fibrinolysis

P* P†CI7)

Urgent/elective PCI(n � 210)

No PCI(n � 921)

6 5 � 4.7 5 � 4.8 � 0.001 � 0.0011.0 4.8 0.009 0.204.3 7.1 0.032 0.0311.0 2.9 0.058 0.0379.2 9.8 0.28 0.336.7 5.3 0.72 0.541.9 1.7 0.10 0.016

ntion.risons.

escue P(n � 17

4 � 3.1.74.03.4

13.86.02.3

interve) compa

occur in patients treated with pre-hospital fibrinolysis. Sec-

Czarnecki et al.Reperfusion Strategies and Outcomes of STEMI

45

ondly, selection bias may have had a significant impact on ourobservational data. High-risk patients and certainly those withcardiogenic shock have a significant survival advantage withprimary PCI.21,22 It is likely that patients with early high-riskfeatures or those with early signs of shock, were preferentiallymanaged with early PCI. This is consistent with the trendstoward lower ejection fractions and higher CK levels observedin our cohort.

Patients in whom PCI was not performed during indexhospitalization had the highest event rates, and no PCI duringadmission for STEMI was an independent predictor of deathand reinfarction. Mortality in STEMI is closely linked to re-establishment of Thrombolysis in Myocardial Infarction(TIMI) 3 flow, which occurs in only 50%-60% of patientsgiven fibrinolysis.23–25 A recent meta-analysis26 which includesGrupo de Análisis de la Cardiopatía Isquémica Aguda (GRA-CIA)-1,27 Trial of Routine Angioplasty and Stenting After Fi-brinolysis to Enhance Reperfusion in Acute Myocardial Infarc-tion (TRANSFER-AMI),28 and the Norwegian Study onDistrict Treatment of ST-Elevation Myocardial Infarction(NORDISTEMI),29 concluded that routine early PCI wasbeneficial for all STEMI patients. They demonstrated a clearreduction in death/reinfarction, reinfarction, and recurrentischemia at 30 days. These findings are consistent with ourdata, suggesting that 1 of the key benefits of PCI is related to areduction in reinfarction. Our observational data are also inaccord with the findings previously reported in the FrenchRegistry of Acute ST-Elevation and Non-ST–Elevation Myo-cardial Infarction (FAST-MI) registry,30 suggesting an associ-ation between angiography (with revascularization where ap-propriate) after fibrinolytic therapy and better outcomes in thereal world.

Previously published data by the Canadian CardiovascularOutcomes Research Team (CCORT) group reported a 12.3%in-hospital mortality for patients with AMI31 in the setting ofsimilar rates of reperfusion.32 The GRACE, GRACE2, andCANRACE registries are not population-based. The hospitalsin these registries were not chosen at random and they includedmany academic institutions, as well as community hospitals,which may not represent the practice patterns of all hospitals(especially rural). Furthermore, the need for informed consentimplies that patients who died shortly after presentation mightnot have been enrolled. Accordingly, our outcomes may not bedirectly comparable with those found by the CCORT group.Indeed, the CCORT study did show significant interregionalvariability in management and outcomes. Furthermore, hospi-tals which chose to participate in these registries, may have hadparticular research interests (evidenced by their participation)including those in quality improvement, which may be re-flected in our improved outcomes. The CCORT group alsopublished data on all AMIs and their data reflect the years 1997to 2000. Our data addressed a much more specific cohort in amore contemporary therapeutic environment. STEMI care hasimproved over the past decade and our data are likely reflectiveof that.

Our study has certain limitations. This is an observationalstudy, and because of the non-randomized nature of our com-parisons and potential unmeasured confounding variables,causality cannot be inferred. Furthermore, due to the need forinformed consent, early deaths might have been excluded.

However, it should be emphasized that the main purpose of

this study is to examine the clinical characteristics and in-hos-pital outcomes of patients treated with various reperfusionstrategies in the “real world.” Data collected in the GRACE/GRACE2 and CANRACE were collected over time from 1999to 2008. There have been significant temporal improvementsin STEMI care, which are beyond the scope of this study; forexample, routine early administration of clopidogrel has in-creased over time. Furthermore, routine PCI after fibrinolysishas shown significant benefits and has become the standard ofcare. Over and above improvements in the medical manage-ment of STEMI care, the systems behind the administration ofcare have changed as well.15 Primary PCI has become the pre-ferred strategy at most interventional centres in Canada, andaccess to timely reperfusion via well-organized protocols hasbeen the focus of quality improvement initiatives. Our datamay serve as a useful benchmark for future comparisons.

In conclusion, this study demonstrates that primary PCIand fibrinolysis with either rescue or urgent/elective PCI isassociated with similar clinical outcomes in the real world. Pa-tients treated with primary PCI trended toward having lowerejection fractions and higher biomarkers and Killip class, whilerequiring more inotropic support and utilizing more antiplate-let therapy. Fibrinolytic-treated patients were more likely tohave previous myocardial infarction, while the utilization ofheparin, �-blockers and angiotensin-converting enzyme inhib-itors was more common. Furthermore, our observational datamay support the use of routine angiography for patients withSTEMI.

AcknowledgementsThe authors thank all the study investigators, coordinators,

and patients who participated in the GRACE and CANRACEregistries. A list of participating GRACE/GRACE2 investiga-tors may be found in Goodman SG, et al.7

Funding SourcesGRACE/GRACE2 and CANRACE were sponsored by

Sanofi-Aventis and Bristol-Myers Squibb. Dr Andrew Yan issupported by a New Investigator Award from the Heart andStroke Foundation of Canada.

DisclosuresRobert C. Welsh: honoraria from Astra Zeneca, Eli Lilly,

Boeringher Ingelheim, sanofi-aventis, Johnson and Johnson,Hoffman LaRoche, and Schering Plough; J. Paul DeYoung,speaker and consulting honoraria from sanofi-aventis, Pfizer,Merck, Servier; Richard Gallo: research grant support fromsanofi-aventis and Bristol-Myers Squibb; Barry Rose: speakerand consulting honoraria and research grant support fromsanofi-aventis and Bristol-Myers Squibb; Francois R. Grondin:consulting honoraria and research grant support from sanofi-aventis and Bristol-Myers Squibb; Jan M. Kornder: consultinghonoraria and research grant support from sanofi-aventis andBristol-Myers Squibb; Graham C. Wong: consulting hono-raria and research grant support from sanofi-aventis and Bris-tol-Myers Squibb; Keith A. A. Fox: research grant support andhonoraria from sanofi-aventis, GlaxoSmithKline, Bristol-My-ers Squibb, and Lilly; Joel M. Gore, research grant supportfrom sanofi-aventis, and Bristol-Myers Squibb; Shaun G.

Goodman: speaker and consulting honoraria and research

46 Canadian Journal of CardiologyVolume 28 2012

grant support from sanofi-aventis, Bristol-Myers Squibb, As-traZeneca, Daiichi Sanyko, Eli Lilly, Merck, and The Medi-cines Company; Andrew T. Yan: research grant support, andconsulting honoraria from sanofi-aventis and Bristol-MyersSquibb. Andrew Czarnecki and Raymond T. Yan have no con-flicts of interest to disclose.

The industrial sponsors had no involvement in the studyconception or design; collection, analysis, and interpretation ofdata; in the writing, review, or approval of the manuscript; andin the decision to submit the manuscript for publication.

References

1. Huynh T, Perron S, O’Loughlin J, et al. Comparison of primary percu-taneous coronary intervention and fibrinolytic therapy in ST-segment-elevation myocardial infarction: Bayesian hierarchical meta-analyses ofrandomized controlled trials and observational studies. Circulation 2009;119:3101-9.

2. Ko DT, Donovan LR, Huynh T, et al. A survey of primary percutaneouscoronary intervention for patients with ST segment elevation myocardialinfarction in Canadian hospitals. Can J Cardiol 2008;24:839-43.

3. Armstrong PW, WEST Steering Committee. A comparison of pharma-cologic therapy with/without timely coronary intervention vs. primarypercutaneous intervention early after ST-elevation myocardial infarction:The WEST (which early ST-elevation myocardial infarction therapy)study. Eur Heart J 2006;27:1530-8.

4. Bonnefoy E, Lapostolle F, Leizorovicz A, et al. Primary angioplasty versusprehospital fibrinolysis in acute myocardial infarction:A randomisedstudy. Lancet 2002;360:825-9.

5. Bonnefoy E, Steg PG, Boutitie F, et al. Comparison of primary angioplastyand pre-hospital fibrinolysis in acute myocardial infarction (CAPTIM) trial: a5-year follow-up. Eur Heart J 2009;30:1598-606.

6. Armstrong PW, Gershlick A, Goldstein P, et al. The strategic reperfusionearly after myocardial infarction (STREAM) study. Am Heart J 2010;160:30-35.e1.

7. Goodman SG, Huang W, Yan AT, et al. The expanded global registry ofacute coronary events: baseline characteristics, management practices, andhospital outcomes of patients with acute coronary syndromes. Am Heart J2009;158:193-201.e1-5.

8. GRACE Investigators. Rationale and design of the GRACE (global regis-try of acute coronary events) project: a multinational registry of patientshospitalized with acute coronary syndromes. Am Heart J 2001;141:190-9.

9. Steg PG, Goldberg RJ, Gore JM, et al. Baseline characteristics, manage-ment practices, and in-hospital outcomes of patients hospitalized withacute coronary syndromes in the global registry of acute coronary events(GRACE). Am J Cardiol 2002;90:358-63.

10. Gale CP, Manda SO, Weston CF, Birkhead JS, Batin PD, Hall AS. Eval-uation of risk scores for risk stratification of acute coronary syndromes inthe myocardial infarction national audit project (MINAP) database. Heart2009;95:221-7.

11. Granger CB, Goldberg RJ, Dabbous O, et al. Predictors of hospital mor-tality in the global registry of acute coronary events. Arch Intern Med2003;163:2345-53.

12. Yan AT, Yan RT, Tan M, et al. Risk scores for risk stratification in acutecoronary syndromes: useful but simpler is not necessarily better. Eur

Heart J 2007;28:1072-8.

13. Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenousthrombolytic therapy for acute myocardial infarction: a quantitative re-view of 23 randomised trials. Lancet 2003;361:13-20.

14. Kushner FG, Hand M, Smith SC Jr, et al. 2009 focused updates: ACC/AHA guidelines for the management of patients with ST-elevation myo-cardial infarction (updating the 2004 guideline and 2007 focused update)and ACC/AHA/SCAI guidelines on percutaneous coronary intervention(updating the 2005 guideline and 2007 focused update) a report of theAmerican college of cardiology Foundation/American heart associationtask force on practice guidelines. J Am Coll Cardiol 2009;54:2205-41.

15. Kalla K, Christ G, Karnik R, et al. Implementation of guidelines improvesthe standard of care: the viennese registry on reperfusion strategies inST-elevation myocardial infarction (vienna STEMI registry). Circulation2006;113:2398-405.

16. Tiefenbrunn AJ, Chandra NC, French WJ, Gore JM, Rogers WJ. Clinicalexperience with primary percutaneous transluminal coronary angioplastycompared with alteplase (recombinant tissue-type plasminogen activator)in patients with acute myocardial infarction: a report from the secondnational registry of myocardial infarction (NRMI-2). J Am Coll Cardiol1998;31:1240-5.

17. Danchin N, Vaur L, Genes N, et al. Treatment of acute myocardial in-farction by primary coronary angioplasty or intravenous thrombolysis inthe “real world”: one-year results from a nationwide French survey. Cir-culation 1999;99:2639-44.

18. Danchin N, Blanchard D, Steg PG, et al. Impact of prehospital throm-bolysis for acute myocardial infarction on 1-year outcome: results from theFrench nationwide USIC 2000 registry. Circulation 2004;110:1909-15.

19. Widimsky P, Budesinsky T, Vorac D, et al. Long distance transport forprimary angioplasty vs immediate thrombolysis in acute myocardial in-farction. Final results of the randomized national multicentre trial--PRAGUE-2. Eur Heart J 2003;24:94-104.

20. Boersma E, Maas AC, Deckers JW, Simoons ML. Early thrombolytictreatment in acute myocardial infarction: reappraisal of the golden hour.Lancet 1996;348:771-5.

21. Berger PB, Holmes DR Jr, Stebbins AL, et al. Impact of an aggressiveinvasive catheterization and revascularization strategy on mortality in pa-tients with cardiogenic shock in the global utilization of streptokinase andtissue plasminogen activator for occluded coronary arteries (GUSTO-I)trial: an observational study. Circulation 1997;96:122-7.

22. Hochman JS, Sleeper LA, Webb JG, et al. Early revascularization in acutemyocardial infarction complicated by cardiogenic shock. SHOCK inves-tigators. Should we emergently revascularize occluded coronaries for car-diogenic shock. N Engl J Med 1999;341:625-34.

23. Ross AM, Coyne KS, Moreyra E, et al. Extended mortality benefit of earlypostinfarction reperfusion. GUSTO-I angiographic investigators. Globalutilization of streptokinase and tissue plasminogen activator for occludedcoronary arteries trial. Circulation 1998;97:1549-56.

24. The GUSTO investigators. An international randomized trial comparingfour thrombolytic strategies for acute myocardial infarction. N Engl J Med1993;329:673-82.

25. Anderson JL, Karagounis LA, Califf RM. Metaanalysis of five reportedstudies on the relation of early coronary patency grades with mortality andoutcomes after acute myocardial infarction. Am J Cardiol 1996;78:1-8.

26. Borgia F, Goodman SG, Halvorsen S, et al. Early routine percutaneouscoronary intervention after fibrinolysis vs. standard therapy in ST-seg-ment elevation myocardial infarction: a meta-analysis. Eur Heart J 2010;

31:2156-69.

Czarnecki et al.Reperfusion Strategies and Outcomes of STEMI

47

27. Fernandez-Aviles F, Alonso JJ, Castro-Beiras A, et al. Routine invasivestrategy within 24 hours of thrombolysis versus ischaemia-guided con-servative approach for acute myocardial infarction with ST-segmentelevation (GRACIA-1): a randomised controlled trial. Lancet 2004;364:1045-53.

28. Cantor WJ, Fitchett D, Borgundvaag B, et al. Routine early angioplastyafter fibrinolysis for acute myocardial infarction. N Engl J Med 2009;360:2705-18.

29. Bohmer E, Hoffmann P,Abdelnoor M, Arnesen H, Halvorsen S. Efficacyand safety of immediate angioplasty versus ischemia-guided managementafter thrombolysis in acute myocardial infarction in areas with very long

transfer distances results of the NORDISTEMI (NORwegian study on

DIstrict treatment of ST-elevation myocardial infarction). J Am CollCardiol 2010;55:102-10.

30. Danchin N, Coste P, Ferrieres J, et al. Comparison of thrombolysis fol-lowed by broad use of percutaneous coronary intervention with primarypercutaneous coronary intervention for ST-segment-elevation acute myo-cardial infarction: data from the French registry on acute ST-elevationmyocardial infarction (FAST-MI). Circulation 2008;118:268-76.

31. Tu JV, Austin PC, Filate WA, et al. Outcomes of acute myocardial infarc-tion in Canada. Can J Cardiol 2003;19:893-901.

32. Jackevicius CA, Alter D, Cox J, et al. Acute treatment of myocardial

infarction in Canada 1999-2002. Can J Cardiol 2005;21:145-52.