Embed Size (px)
Citation preview
Mprscrpvrf
Fn
2
Journal of the American College of Cardiology Vol. 55, No. 22, 2010© 2010 by the American College of Cardiology Foundation ISSN 0735-1097/$36.00P
Prognostic Significance and Determinantsof Myocardial Salvage Assessed byCardiovascular Magnetic Resonance inAcute Reperfused Myocardial Infarction
Ingo Eitel, MD,* Steffen Desch, MD,* Georg Fuernau, MD,* Lysann Hildebrand, BSC,*Matthias Gutberlet, MD,† Gerhard Schuler, MD,* Holger Thiele, MD*
Leipzig, Germany
Objectives The aim of the study was to determine the prognostic significance and determinants of myocardial salvage as-sessed by cardiovascular magnetic resonance (CMR) in reperfused ST-segment elevation myocardial infarction.
Background In acute myocardial infarction, CMR can retrospectively detect the myocardium at risk and the irreversible injury.This allows for quantifying the extent of salvaged myocardium after reperfusion as a potential strong end pointfor clinical trials and outcome.
Methods We analyzed 208 consecutive ST-segment elevation myocardial infarction patients undergoing primary angio-plasty �12 h after symptom onset. T2-weighted and contrast-enhanced CMR was used to calculate the myocar-dial salvage index (MSI). Patients were categorized into 2 groups defined by the median MSI. The primary endpoint of the study was occurrence of major adverse cardiovascular events defined as death, reinfarction, andoccurrence of new congestive heart failure within 6 months after the index event.
Results The median MSI was 48 (interquartile range 27 to 73). Major adverse cardiovascular events were significantlylower in the MSI � median group (2.9% vs. 22.1%, p � 0.001). The stepwise Cox proportional hazards modelrevealed that the MSI was the strongest predictor of major adverse cardiovascular events at 6-month follow-up(p � 0.001). All prognostic clinical (symptom onset to reperfusion), angiographic (Thrombolysis In MyocardialInfarction flow grade before angioplasty), and electrocardiographic (ST-segment resolution) parameters showedsignificant correlations with the MSI (p � 0.001 for all).
Conclusions This study for the first time demonstrates that the MSI assessed by CMR predicts the outcome in acute reperfusedST-segment elevation myocardial infarction. Therefore, MSI assessment has important implications for patient prog-nosis as well as for the design of future trials intended to test new reperfusion therapy efficacy. (Myocardial SalvageAssessed by Cardiovascular Magnetic Resonance—Impact on Outcome; NCT00952224). (J Am Coll Cardiol 2010;55:2470–9) © 2010 by the American College of Cardiology Foundation
ublished by Elsevier Inc. doi:10.1016/j.jacc.2010.01.049
m(rawprt
sa
yocardial salvage is the principal mechanism by whichatients with acute myocardial infarction benefit fromeperfusion therapies (1). To assess the efficacy of reperfu-ion therapy, it is necessary to determine how much myo-ardium is salvaged by measuring the final infarct size inelation to the initial myocardium at risk. The most widelyracticed technique for directly measuring myocardial sal-age currently is single-photon emission computed tomog-aphy (SPECT). This approach has been applied success-ully in trials (2–6) and has confirmed that the degree of
rom the Heart Center, Departments of *Internal Medicine–Cardiology and †Diag-ostic and Interventional Radiology, University of Leipzig, Leipzig, Germany.
CManuscript received September 21, 2009, revised manuscript received December 1,
009, accepted January 13, 2010.
yocardial salvage is an independent predictor of outcome7). However, this approach is limited by its low spatialesolution and the need for injection of the isotope in thecute setting of coronary occlusion, which could interfereith patient care (2,8). Moreover, imaging must be com-leted within 3 h, which is difficult during off-hours and theequirement of 2 separate perfusion studies leads to addi-ional radiation exposure.
See page 2489
Recently, a landmark study showed that the area of high T2ignal in cardiovascular magnetic resonance (CMR) reflects therea at risk in acute reperfused myocardial infarction (9).
linically, Friedrich et al. (10) confirmed that the proportion
obesrsgcae
dmm
M
SaMeccohe
imvCbPPTlAtda((lAozmAgstIaaVg
EmdmaeceCctssatapCd4roCvssw(1ava(3(s1IaFbesrd
Tmiohaid
2471JACC Vol. 55, No. 22, 2010 Eitel et al.June 1, 2010:2470–9 Myocardial Salvage and Prognosis
f myocardial salvage can be assessed retrospectively in humansy comparing T2-weighted (edematous myocardium) and latenhancement CMR images. Furthermore, trials demon-trated that myocardial salvage assessment by CMR is aeproducible tool that identifies and quantifies myocardiumalvage with excellent agreement with SPECT and angio-raphic scores of myocardial salvage (11–13). However, inontrast to the extensive clinical SPECT experience, therere no prognostic myocardial salvage data using CMR as annd point in clinical trials.
The aim of the present CMR study was therefore toetermine the prognostic significance and determinants ofyocardial salvage assessed by CMR in acute reperfusedyocardial infarction.
ethods
tudy population. This prospective trial was conducted atsingle tertiary care center between November 2006 anday 2008. The study protocol was approved by the local
thics committee, and all patients gave written informedonsent. Patients with infarction undergoing primary per-utaneous coronary intervention (PCI) were eligible if thenset of symptoms was less than 12 h before PCI and if theyad ST-segment elevation of at least 0.1 mV in �2xtremity leads or at least 0.2 mV in �2 precordial leads.
To ensure that CMR findings reflected acute myocardialnjury, patients were not enrolled if they had a previous
yocardial infarction. Further exclusion criteria were pre-ious fibrinolysis and patients with contraindications toMR at study entry such as implanted pacemakers, defi-rillators, claustrophobia, or metallic intracranial implants.rimary angioplasty and subsequent treatment. PrimaryCI was performed according to standard clinical practice.he decision to use of bare-metal or drug-eluting stents was
eft to the discretion of the interventional cardiologist.dditional use of intra-aortic balloon counterpulsation or
hrombectomy was performed, depending on the hemo-ynamic instability and thrombus in the infarct-relatedrtery. All patients received 500 mg of aspirin and heparin60 U/kg body weight) intravenously before PCI. Clopidogrel600 mg orally during PCI, if not administered before, fol-owed by 75 mg/day for at least 12 months) was mandatory.spirin was given indefinitely at a dose of 100 mg/day. The usef glycoprotein IIb/IIIa inhibitors, angiotensin-converting en-yme inhibitors, beta-blockers, and statins was strongly recom-ended according to guidelines (14).ngiographic analysis. Coronary angiography of the tar-et lesion was performed before and after PCI with theame projections. The angiographic projections used werehose that allowed optimal evaluation of the Thrombolysisn Myocardial Infarction (TIMI) flow of the infarct-relatedrtery or the myocardium supplied by it. Angiographicnalysis included initial and final flow of the culprit vessel.isual assessments were performed offline in the angio-
raphic core laboratory by 2 blinded observers. rlectrocardiographic and enzy-atic analysis. For electrocar-
iographic interpretation, the cu-ulative ST-segment resolution
pproximately 90 min after PCI,xpressed as the percentage, wasalculated by 2 blinded observ-rs as described previously (15).ategorization was performed in
omplete (�70%), partial (�70%o 30%), and no (�30%) ST-egment resolution (15). Plasmaamples for creatine kinase (CK)nd the CK-myocardial band frac-ion were collected on admissionnd subsequently during the hos-italization every 6 h for 2 days.MR. Myocardial salvage was
etermined by CMR on days 1 toafter the index event. The area at
isk, infarct size, and microvascularbstruction (MO) were acquired on a 1.5-T scanner (InteraV, Philips Medical Systems, Best, the Netherlands). Left
entricular (LV) function was assessed by a standard steady-tate free precession technique. For area at risk determination,hort-axis slices covering the whole ventricle using a T2-eighted triple inversion recovery breath-hold pulse sequence
repetition time 2� R-R interval; echo time 80 ms; flip angle80°; voxel size 0.71 � 0.71 � 8.0 mm) were obtained usingbody coil. Late enhancement images covering the whole
entricle were acquired approximately 15 min after intravenousdministration of 0.2 mmol/kg body weight of gadobutrolGadovist, Bayer Schering Pharma, Berlin, Germany). A-dimensional inversion recovery turbo gradient echo sequencerepetition time 2.8 ms; echo time 1.1 ms; flip angle 15°; typicalpatial resolution 1.8 � 1.8 � 5 mm; 2 stacks; pre-pulse delay80 to 280 ms) was used for image acquisition.mage analysis. Offline image analysis was performed onn independent workstation with dedicated software (View-orum release 5.2, Philips Medical Systems) by fullylinded observers. Infarct size, area at risk, and MO werexpressed as percentages of the LV volume, given by theum of the volume of edema, late enhancement, and MOegions for all slices divided by the sum of the LV myocar-ial cross-sectional volumes (%LV).The area of abnormal signal intensity was measured in the
2-weighted images and in the corresponding late enhance-ent images by manual delineation in each of the short-axis
mages. A central core of hypointense signal within the areaf increased T2 signal intensity, which is deemed to beemorrhagic infarction, was included in the area at riskssessment (16). Care was taken to exclude increased signalntensity from the blood pool adjacent to the endocardiumue to slow flow. By consensus, a myocardial region was
Abbreviationsand Acronyms
CK � creatine kinase
CMR � cardiovascularmagnetic resonance
LV � left ventricle/ventricular
MACE � major adversecardiovascular events
MO � microvascularobstruction
MSI � myocardial salvageindex
PCI � percutaneouscoronary intervention
SPECT � single-photonemission computedtomography
TIMI � Thrombolysis InMyocardial Infarction
egarded as affected if at least 10 ad
jacent myocardial pixels
rfpsTp
laaCw(
ainfdnmwamaHhwcSMEpnt
2472 Eitel et al. JACC Vol. 55, No. 22, 2010Myocardial Salvage and Prognosis June 1, 2010:2470–9
evealed a signal intensity of �2 SDs of remote myocardiumor edema and �5 SDs in late enhancement images (17). Inatients with MO, these dark areas were included for infarctize analysis and the area of MO was assessed separately.he following parameters were calculated as describedreviously (10,18):
a. Area at risk � volume edema/volume LV massb. Percentage of infarct size � volume infarct/volume
LV massc. Percentage of MO � volume MO/volume LV massd. Myocardial salvage � area at risk minus infarct sizee. Myocardial salvage index (MSI) � area at risk minus
infarct size/area at risk.
The CMR core laboratory has excellent reproducibility andow interobserver and intraobserver variability for infarct sizend MSI assessment (13,19). For MSI assessment, the biasnd limits of agreement are �0.3 � 5.0 (13).linical end points. The primary end point of this studyas the occurrence of major adverse cardiovascular events
MACEs), defined as a composite of death, reinfarction,
Figure 1 Study Flow Diagram
CMR � cardiac magnetic resonance; MSI � myocardial salvage index; STEMI � S
Cardiovascular Magnetic Resonance ResultsTable 1 Cardiovascular Magnetic Resonanc
Variable MSI < Me
Edema (%LV) 35.2 (2
Infarct size (%LV) 25.7 (1
Late microvascular obstruction (%LV) 0.9 (0
Left ventricular ejection fraction (%) 46.5 (3
Left ventricular end-diastolic volume (ml) 148.7 (1
Left ventricular end-systolic volume (ml) 77.3 (5
Data are presented as median and interquartile range.LV � left ventricle; MSI � myocardial salvage index.
nd new congestive heart failure within 6 months after thendex event. In a secondary analysis, the individual compo-ents of the primary end point were analyzed. Post-hospitalollow-up included 1 outpatient visit at 6 months. Theiagnosis of reinfarction was based on clinical symptoms,ew ST-segment changes, and increase in the CK-yocardial band level above the reference limit in patientsith normalized values after the index event or if there was
n increase of at least 50% from the last non-normalizedeasurement. New congestive heart failure was defined as
ny congestive heart failure (rales, dyspnea, New Yorkeart Association functional class III to IV) occurring �24after the index event. To avoid double counting of patientsith �1 event, each patient contributed only once to the
omposite MACE end point.tatistical analysis. Patients were grouped by the medianSI into a � median MSI and a � median MSI group.
ach categorical variable is expressed as the number andercentage of patients. Most continuous variables had non-ormal distribution and are therefore presented as mediansogether with the interquartile range.
ent elevation myocardial infarction.
ults
� 104) MSI > Median (n � 104) p Value
3.1) 37.3 (29.1–47.3) 0.34
2.6) 10.3 (5.4–14.3) �0.001
) 0.3 (0.0–0.9) 0.001
3.9) 55.1 (45.8–62.0) �0.001
167.3) 131.7 (111.2–154.7) 0.005
6.6) 58.3 (43.8–75.5) �0.001
T-segm
e Res
dian (n
9.5–4
9.4–3
.4–1.9
8.3–5
25.0–
9.3–9
etbtP
wChdb
sara
mlvid9
M
DCI � p
2473JACC Vol. 55, No. 22, 2010 Eitel et al.June 1, 2010:2470–9 Myocardial Salvage and Prognosis
Differences between groups were assessed by the Fisherxact or the chi-square test for categorical variables and byhe Student t test for continuous data with normal distri-ution. Otherwise, the nonparametric Wilcoxon rank-sumest was used. Correlation analyses were done by usingearson or Spearman tests, as indicated.Univariable and multivariable linear regression analyses
ere performed to characterize predictors of the MSI.ategorical variables included sex, smoking, hypertension,ypercholesterolemia, diabetes, infarct location, number ofiseased vessels, Killip class on admission, and TIMI flow
ain Patient CharacteristicsTable 2 Main Patient Characteristics
Variable
All Patients MSI48 (28–73)(n � 208)
Age (yrs) 66 (55–74)
Male sex 143 (70)
Cardiovascular risk factors
Current smoking 85 (41)
Hypertension 139 (67)
Hypercholesterolemia 71 (34)
Diabetes mellitus 52 (25)
Anterior myocardial infarction 98 (47)
Diseased vessels
1 115 (55)
2 52 (25)
3 41 (20)
Killip class on admission
1 156 (75)
2 42 (20)
3 7 (3)
4 3 (1)
Reperfusion times (min)
Symptom onset to reperfusion 227 (153–374)
Door to balloon 27 (20–35)
TIMI flow grade before PCI
0 118 (57)
1 14 (7)
2 24 (12)
3 52 (25)
TIMI flow grade after PCI
0 3 (1)
1 9 (4)
2 12 (6)
3 184 (88)
Peak creatine kinase (�mol/l/s) 25 (12–45)
ST-segment resolution (%) 69 (46–100)
Concomitant medications
Beta-blockers 203 (98)
ACE inhibitors/AT I antagonist 203 (98)
Aspirin 208 (100)
Clopidogrel 208 (100)
Statins 206 (99)
Aldosterone antagonist 7 (3)
Glycoprotein IIb/IIIa inhibitor 196 (94)
ata are presented as median (interquartile range) or n (%).ACE � angiotensin-converting enzyme; AT I � angiotensin I; MSI � myocardial salvage index; P
efore and after PCI. Continuous variables included age, a
ymptom onset-to-reperfusion time, door-to-balloon timend ST-segment resolution 90 min after PCI. Multivariableegression was performed using only variables with a prob-bility value �0.05 in univariable regression analyses.
For the combined clinical end point, the Kaplan-Meierethod was applied, and differences were assessed by the
og-rank test. Simple Cox regression analysis using the sameariables as defined previously as well as the CMR variablesn Table 1 were used to identify predictors of MACEsuring 6 months. Hazard ratios with their corresponding5% confidence intervals are reported. All variables that
I < Median MSI28 (19–40)(n � 104)
MSI > Median MSI73 (60–82)(n � 104) p Value
67 (55–75) 64 (55–74) 0.40
73 (70) 70 (67) 0.55
43 (41) 42 (40) 0.89
68 (65) 71 (68) 0.66
33 (32) 38 (37) 0.47
26 (25) 26 (25) 1.00
41 (39) 57 (55) 0.03
0.86
58 (56) 57 (55)
27 (26) 25 (24)
19 (18) 22 (21)
0.02
69 (66) 87 (84)
30 (29) 12 (11)
3 (3) 4 (4)
2 (2) 1 (1)
85 (170–425) 186 (126–301) 0.003
25 (19–35) 28 (22–35) 0.07
0.005
69 (66) 49 (47)
7 (7) 7 (7)
12 (12) 12 (12)
16 (15) 36 (35)
0.36
3 (3) 0 (0)
5 (5) 4 (4)
7 (7) 5 (5)
89 (86) 95 (91)
41 (23–61) 16 (7–28) �0.001
63 (35–82) 80 (52–100) 0.007
01 (97) 102 (98) 0.66
01 (97) 102 (98) 0.66
04 (100) 104 (100) 1.00
04 (99) 104 (100) 1.00
03 (99) 103 (99) 100
5 (5) 2 (2) 0.25
97 (93) 99 (95) 0.55
ercutaneous coronary intervention; TIMI � Thrombolysis In Myocardial Infarction.
MS
2
1
1
1
1
1
ppeared to be associated with MACEs at the p � 0.05
lCweMar
Mwwcs2s
R
Op(tmCcqPt(wop
thtsa
ttPcm7AesPpbo9s0
h
2474 Eitel et al. JACC Vol. 55, No. 22, 2010Myocardial Salvage and Prognosis June 1, 2010:2470–9
evel in univariable analysis were then tested in a multipleox regression analysis, based on a stepwise algorithmith the p value set at 0.05 for entering and 0.1 for
xclusion. The extent of late MO was a better predictor ofACEs than the presence of late MO in univariable
nalysis and was therefore included in the multivariable Coxegression model.
For additional comparison of the prognostic value of theSI, infarct size and late MO with regard to MACEs as
ell as mortality, receiver-operator characteristic curvesere generated, and the areas under the curves were
alculated. All statistical tests were performed with SPSSoftware, version 15.0 (SPSS, Inc., Chicago, Illinois). A-tailed p value �0.05 was considered statisticallyignificant.
esults
f 267 eligible ST-segment elevation myocardial infarctionatients, this prospective CMR study included 208 patientsFig. 1). The main reasons for exclusion from the study werehe lack of a CMR examination (n � 23) and previousyocardial infarction (n � 29). Reasons for not undergoingMR are listed in Figure 1. In all remaining 208 patients,
linical outcome data were available and CMR imageuality was suitable to assess myocardial salvage.atient characteristics. Demographic and clinical charac-
eristics are shown in Table 2. The baseline characteristicsage, sex, risk factors) were similar between groups. Patientsith an MSI � median had a significantly higher frequencyf anterior myocardial infarction and lower Killip class atresentation.The MSI values according to hours from symptom onset
o treatment are shown in Figure 2, where the MSI wasighest within the first 2 h but decreased thereafter. Theime from symptom onset to reperfusion was significantlyhorter in the MSI � median group (p � 0.003). There was
Figure 2 Amount of Myocardial Salvage Accordingto Time From Symptom Onset to Reperfusion
PCI � percutaneous coronary intervenion.
significant inverse correlation of symptom duration and
he MSI (r � �0.330, p � 0.001). The door-to-balloonime showed no significant differences between groups.atients with anterior myocardial infarction had a signifi-antly higher MSI compared with patients with nonanterioryocardial infarction (58.6 [interquartile range 33.2 to
3.9] vs. 39.9 [interquartile range 25.4 to 70.8]; p � 0.01).ngiographic analysis, ST-segment resolution, and
nzymatic analysis. Patients with an MSI � median had aignificantly higher frequency of TIMI flow grade 3 beforeCI (p � 0.005). The MSI was significantly higher inatients with residual TIMI flow (TIMI flow grade 2 to 3)efore PCI (p � 0.001) (Fig. 3A). After PCI, the majorityf patients had TIMI flow grade 3 in both groups (86% vs.1%, p � 0.36). TIMI flow before and after PCI correlatedignificantly with the MSI (r � 0.317, p � 0.001 and r �.160, p � 0.03, respectively).Of the 208 patients 104 (50%) had complete, 67 (32%)
ad partial, and 37 (18%) had no ST-segment resolution 90
Figure 3 MSI According to TIMI Flow GradeBefore PCI and ST-Segment Resolution
Box (25th percentile, median, and 75th percentile) and whisker (10th and 90thpercentiles) plots of MSI according to Thrombolysis In Myocardial Infarction(TIMI) flow grade before PCI (A) and ST-segment resolution (B). Abbreviationsas in Figures 1 and 2.
mv(s(c(apCCg
papiosea4re
2475JACC Vol. 55, No. 22, 2010 Eitel et al.June 1, 2010:2470–9 Myocardial Salvage and Prognosis
in after PCI. ST-segment resolution as a continuousariable was significantly better in the MSI � median groupp � 0.007) and patients with a MSI � median also had aignificantly lower peak CK and CK-myocardial band levelp � 0.001). MSI was significantly higher in patients withomplete ST-segment resolution after PCI (p � 0.001)Fig. 3B). The extent of ST-segment resolution 90 minfter the primary PCI correlated with the MSI (r � 0.287,� 0.001).MR. The median time between the index event andMR was 3 days (interquartile range 2 to 4) for both
Figure 4 CMR Findings in a Patient With Inferior and Anterior M
(A) T2-weighted image (left) of a patient with an occluded right coronary artery (rigsignal (left, white arrows). Extension into the inferior wall of the right ventricle ispanel) showing edema contours (red, lower left panel) and the corresponding latepapillary (dark blue), and infarcted (red) contours (lower right panel) in a patient
roups. Results of CMR were available in all 208 w
atients. In all patients, a high transmural T2 signalbnormality was observed in the infarct region. In 9 (4%)atients, CMR detected no late enhancement (abortednfarction) (20). Figure 4A is an example of an acutelyccluded right coronary artery with corresponding highignal in the inferior left and right ventricular walls. Thevolution of myocardial edema in a patient with leftnterior descending artery occlusion is shown in FigureB. The localization of late enhancement was in the sameegion of the edema in all cases. The median amount ofdema was 35.5%LV (interquartile range 29.2 to 44.9)
ardial Infarction
rows). The area at risk is visually apparent by a thickened myocardium with high(left, black arrows). (B) Short-axis slices of a T2-weighted image (upper left
ncement image (upper right panel) with endocardial (green), epicardial (yellow),nterior myocardial infarction and an MSI � median. Abbreviations as in Figure 1.
yoc
ht, arvisible
enhawith a
ith no significant differences between groups. The
m1rw
s(omiegPgavfmb(Cc(Ivt�6t(
iMa
at
M6a
iMs
D
Tome
UR
A
2476 Eitel et al. JACC Vol. 55, No. 22, 2010Myocardial Salvage and Prognosis June 1, 2010:2470–9
edian infarct size was 16.3%LV (interquartile range0.0 to 26.4), significantly smaller than myocardium atisk (p � 0.001) (Table 1). The median calculated MSIas 48.3 (interquartile range 27.7 to 73.2).Late MO was identified in 134 (64%) patients with a
ignificantly higher occurrence in the MSI � median group85 vs. 49 patients, p � 0.001). Infarct size and the extentf late MO were also significantly larger in the MSI �edian group. LV ejection fraction was significantly higher
n the MSI � median group, whereas end-diastolic andnd-systolic volume indexes were significantly lower in thisroup at this early stage after the index event (Table 1).redictors of myocardial salvage. In a multivariable re-ression model adjusted for significant variables in univari-ble regression analysis using the MSI as the dependentariable, complete ST-segment resolution (p � 0.001), timerom symptom onset to reperfusion (p � 0.002), anterioryocardial infarction (p � 0.004), and TIMI flow grade �2
efore PCI (p � 0.03) were the strongest predictors of MSITable 3).
linical outcome. At 6-month follow-up, there were 10ardiac deaths (9.6%) in the � median MSI group and 11%) in the MSI � median group (p � 0.003) (Fig. 5A).n the MSI � median group, the patient died of a rightentricular infarction. Nonfatal reinfarctions and conges-ive heart failure occurred significantly more often in the
median MSI group (13 vs. 2 events). Consequently, at-month follow-up, MACEs were significantly lower inhe MSI � median group (2.9% vs. 22.1%, p � 0.001)Fig. 5B).
In addition to the MSI, several established markers ofncreased patient risk were associated with increased
ACEs at 6-month follow-up by simple Cox regression
nivariable and Multivariable Linearegression Analysis for the Prediction of MSITable 3 Univariable and Multivariable LinearRegression Analysis for the Prediction of MSI
Variable
MSI
Univariable Multivariable
� p Value � p Value
Age (yrs) �0.07 0.34 — —
Male sex �0.04 0.58 — —
Hypertension 0.07 0.34 — —
Hypercholesterolemia 0.09 0.16 — —
Diabetes mellitus �0.02 0.75 — —
Anterior myocardial infarction 0.17 0.01 0.22 0.004
Current smoking 0.01 0.83 — —
Number of diseased vessels �0.01 0.98 — —
Killip class on admission 0.09 0.21 — —
Symptom onset to reperfusion (min) �0.30 �0.001 �0.23 0.002
Door-to-balloon time (min) 0.08 0.28 — —
TIMI flow grade �2 before PCI 0.28 �0.001 0.16 0.03
TIMI flow grade 3 after PCI 0.13 0.07 0.07 0.37
Complete ST-segment resolution (�70%) 0.26 �0.001 0.31 �0.001
bbreviations as in Table 2.
nalysis (Table 4). Using stepwise multiple Cox regression
nalysis, only MSI remained an independent predictor ofhe combined clinical end point.
Despite significant predictors in univariable analyses, lateO and LV ejection fraction were unable to predict
-month MACEs in stepwise multivariable Cox regressionnalysis.
Receiver-operator characteristic curve analyses furtherllustrated that the MSI is the strongest indicator of
ACEs and especially of mortality compared with infarctize and late MO (Fig. 6).
iscussion
o our best knowledge this the largest prospective CMRutcome study to date and the first to report introducingyocardial salvage as a novel strong CMR outcome param-
ter in acute reperfused myocardial infarction. The major
Figure 5 Mortality and MajorAdverse Cardiovascular Event Rate
Kaplan-Meier curve of the incidence of all-cause mortality (A) and the cumula-tive incidence of death, reinfarction, and new congestive heart failure (B) dur-ing the first 6 months after randomization. MSI � myocardial salvage index.
fih6aoTmMtucsMaprshvcsfitCdiiaocopTrasnMaitq
iti
2477JACC Vol. 55, No. 22, 2010 Eitel et al.June 1, 2010:2470–9 Myocardial Salvage and Prognosis
ndings are as follows: 1) patients with a MSI � medianave a significantly lower mortality and MACE rate at-month follow-up; 2) the strongest predictors of the MSIre complete ST-segment resolution, time from symptomnset to reperfusion, anterior myocardial infarction, andIMI flow grade �2 before PCI; and 3) the amount ofyocardial salvage is the strongest inverse correlate ofACEs and mortality at 6-month follow-up. Thus, quan-
ifying the extent of the salvaged area after revascularizationsing CMR might serve as a novel, strong end point forlinical trials investigating the success of reperfusiontrategies.
yocardial salvage assessment by CMR. The myocardialrea at risk is defined as the myocardial tissue within theerfusion bed distal to the culprit lesion of the infarct-elated artery (21). Recently, it has been shown, that in theetting of reperfused (9) and nonreperfused (22) infarction,igh signal intensity areas in T2-weighted images accuratelyisualize the area at risk. The result of reperfusion therapyan therefore be assessed clinically by calculating myocardialalvage as the difference between myocardium at risk andnal infarct size. Friedrich et al. (10) applied this techniqueo patients and systematically assessed myocardial salvage byMR in 92 reperfused infarction patients. This groupemonstrated that the area at risk identified with T2
maging was consistently transmural and exceeded areas ofrreversible injury defined by late enhancement, resulting inmean myocardial salvage of 16 � 11%. Despite acquiringnly 3 short-axis slices, these data are in line with theurrent study, which calculated a mean myocardial salvagef 18 � 12% from a 3-dimensional volume set, and a recentublished study with a mean salvage of 14 � 10% (11).herefore, myocardial salvage assessment using CMR is a
obust method for quantifying the extent of the salvagedrea after reperfusion. However, further improvements inignal intensity and reduction in artifacts of T2 imaging isecessary.
yocardial salvage as clinical outcome parameter. Theccurate identification of the area at risk in acute myocardialnfarction is crucial to the evaluation of future reperfusionherapies and their impact on myocardial salvage. The
Predictors of Major Adverse Cardiac Events inUnivariable and Stepwise Multivariable Cox RegTable 4 Predictors of Major Adverse CardiacUnivariable and Stepwise Multivaria
Variable
U
Hazard Ratio
Killip class on admission 2.01 (1.30–3
TIMI flow grade after PCI 0.57 (0.36–0
Left ventricular ejection fraction (%) 0.95 (0.92–0
ST-segment resolution (%) 0.99 (0.99–1
Late microvascular obstruction (%LV) 1.10 (1.03–1
Infarct size (%LV) 1.08 (1.05–1
MSI 0.95 (0.93–0
CI � confidence interval; other abbreviations as in Tables 1 and 2.
uantification of myocardial salvage might be a better
ndicator of therapeutic efficacy in myocardial infarctionrials than infarct size. SPECT with a technetium tracernjection before reperfusion has been the most widely
Figure 6 Receiver-Operator Characteristic Curves of thePrognostic Value of the MSI, Infarct Size, and Late MO
Receiver-operator characteristic curves showing the prognostic sensitivity andspecificity of the myocardial salvage index (MSI), infarct size (IS), and latemicrovascular obstruction (MO) with regard to major adverse cardiovascularevents (A) and mortality (B) at 6-month follow-up. AUC � area under the curve.
on Analysisnts inox Regression Analysis
ble Stepwise Multivariable
p Value Hazard Ratio (CI) p Value
0.002 — —
0.01 — —
�0.001 — —
0.04 — —
0.004 — —
�0.001 — —
�0.001 0.93 (0.91–0.96) �0.001
ressiEveble C
nivaria
(CI)
.09)
.88)
.98)
.00)
.17)
.12)
.97)
pbrcim2g45sTaf
alwcesrHpIMuptrcMaotCiswsocsipsspcapCramc
eisfobpaSsspsiasad
C
TapfPifm
RmHi
R
2478 Eitel et al. JACC Vol. 55, No. 22, 2010Myocardial Salvage and Prognosis June 1, 2010:2470–9
racticed technique for assessing the area at risk and haseen successfully used to compare the efficacy of variouseperfusion strategies. (2–6) Furthermore, this approach hasonfirmed that the degree of myocardial salvage is anndependent predictor of outcome (7). However, there are
any drawbacks to its widespread use including: 1) need forscans; 2) need for tracer administration; 3) access to a
amma camera �3 h of the tracer administration;) interfering with patient management in the acute setting;) application of radiation dose of both scans; and 6) lowerpatial resolution, particularly of subendocardial infarcts (8).hus, it is desirable to find methods for assessing area at risk
nd subsequent myocardial salvage that are more clinicallyeasible to evaluate outcomes.
Our study for the first time clearly confirms that CMR ispromising tool that can quantify the success of revascu-
arization reflected by the amount of salvaged myocardiumith a strong predictive value of subsequent clinical out-
ome. So far, infarct size and MO were the most used CMRnd points in clinical trials because previous studies havehown that these parameters are associated with global andegional functional recovery as well as MACEs (23–25).
owever, these trials had small sample sizes and traditionalrognostic markers and scores have not been investigated.n addition, “soft” clinical end points were included in
ACEs such as revascularization or hospital admission fornstable angina (24). In the present study, only harder endoints were included and clinical, angiographic, and elec-rocardiographic parameters, which are known from earlieresearch to affect prognosis in ST-segment elevation myo-ardial infarction, showed significant correlations with the
SI (15,26,27). This may explain why these parameters aressociated with survival. In particular, the strong correlationf ischemic time with the MSI is remarkable. It emphasizeshat the extent of the salvaged myocardium assessed byMR is in accordance with the wavefront phenomenon,
nitially postulated by Reimer et al. (28). Median myocardialalvage was extremely large when PCI was performedithin the first 2 h of symptom onset, whereas it was twice
maller with an additional 2 to 4 h of reperfusion delay, andur data unequivocally show the “flat” part of the salvageurve thereafter (29). These data thus provide additionalupport for the relationship between myocardial salvage toschemic time and confirm current recommendations foratient transport to tertiary centers for the primary PCI asoon after symptom onset as possible (14,30). However,ome patients showed significant myocardial salvage after arolonged ischemic time. This may depend on severallinical factors, including the extent of collateral circulationnd the presence or absence of a history of ischemicreconditioning (27).linical implications. Prognosis in patients with acute
eperfused myocardial infarction is directly related to themount of myocardial salvage, which strengthens the use ofyocardial salvage as a strong surrogate end point for
linical trials investigating the success of reperfusion strat-
gies. Particularly in myocardial infarction trials, there isncreasing interest in measures that can be used as aurrogate end point, allowing a much smaller sample size,aster trials, and a reduction of costs. CMR might be theptimal method to assess myocardial salvage because it cane done retrospectively with 1 scan, without interfering withatient care in the acute setting, at high spatial resolutionnd without ionizing radiation.tudy limitations. Some limitations need attention. First,ome patients had to be excluded from CMR myocardialalvage assessment. Because the baseline characteristics ofatients undergoing and those not undergoing CMR wereimilar, a potential selection bias is unlikely. Second, thenfluence of several clinical variables that were not availablend might affect myocardial salvage were not examined,uch as collateral blood flow and preconditioning. However,ngina onset and recurrence are generally subjective andifficult to assess.
onclusions
his study clearly demonstrated that myocardial salvagessessed with CMR predicts MACEs and mortality inatients with acute reperfused myocardial infarction. There-ore, the retrospective assessment of myocardial salvage afterCI has important implications for patient prognosis and
mproving clinical outcomes, as well as for the design ofuture trials intended to test the efficacy of reperfusionodalities.
eprint requests and correspondence: Dr. Ingo Eitel, Depart-ent of Internal Medicine/Cardiology, University of Leipzig–eart Center, Strümpellstr. 39, 04289 Leipzig, Germany. E-mail:
EFERENCES
1. Braunwald E. Myocardial reperfusion, limitation of infarct size,reduction of left ventricular dysfunction, and improved survival.Should the paradigm be expanded? Circulation 1989;79:441–4.
2. Gibbons RJ, Christian TF, Hopfenspirger M, Hodge DO, Bailey KR.Myocardium at risk and infarct size after thrombolytic therapy foracute myocardial infarction: implications for the design of randomizedtrials of acute intervention. J Am Coll Cardiol 1994;24:616–23.
3. Gibbons RJ, Miller TD, Christian TF. Infarct size measured by singlephoton emission computed tomographic imaging with (99m)Tc-sestamibi: a measure of the efficacy of therapy in acute myocardialinfarction. Circulation 2000;101:101–8.
4. Kastrati A, Mehilli J, Dirschinger J, et al. Myocardial salvage aftercoronary stenting plus abciximab versus fibrinolysis plus abciximab inpatients with acute myocardial infarction: a randomised trial. Lancet2002;359:920–5.
5. Kastrati A, Mehilli J, Nekolla S, et al. A randomized trial comparingmyocardial salvage achieved by coronary stenting versus balloonangioplasty in patients with acute myocardial infarction consideredineligible for reperfusion therapy. J Am Coll Cardiol 2004;43:734–41.
6. Schomig A, Kastrati A, Dirschinger J, et al. Coronary stenting plusplatelet glycoprotein IIb/IIIa blockade compared with tissue plasmin-ogen activator in acute myocardial infarction. Stent versus Thrombol-
ysis for Occluded Coronary Arteries in Patients with Acute Myocar-dial Infarction Study Investigators. N Engl J Med 2000;343:385–91.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
K
2479JACC Vol. 55, No. 22, 2010 Eitel et al.June 1, 2010:2470–9 Myocardial Salvage and Prognosis
7. Ndrepepa G, Mehilli J, Schwaiger M, et al. Prognostic value ofmyocardial salvage achieved by reperfusion therapy in patients withacute myocardial infarction. J Nucl Med 2004;45:725–9.
8. Wagner A, Mahrholdt H, Holly TA, et al. Contrast-enhanced MRIand routine single photon emission computed tomography (SPECT)perfusion imaging for detection of subendocardial myocardial infarcts:an imaging study. Lancet 2003;361:374–9.
9. Aletras AH, Tilak GS, Natanzon A, et al. Retrospective determinationof the area at risk for reperfused acute myocardial infarction withT2-weighted cardiac magnetic resonance imaging: histopathologicaland displacement encoding with stimulated echoes (DENSE) func-tional validations. Circulation 2006;113:1865–70.
0. Friedrich MG, Abdel-Aty H, Taylor A, Schulz-Menger J, MessroghliD, Dietz R. The salvaged area at risk in reperfused acute myocardialinfarction as visualized by cardiovascular magnetic resonance. J AmColl Cardiol 2008;51:1581–7.
1. Wright J, Adriaenssens T, Dymarkowski S, Desmet W, Bogaert J.Quantification of myocardial area at risk with T2-weighted CMR:comparison with contrast-enhanced CMR and coronary angiography.J Am Coll Cardiol Img 2009;2:825–31.
2. Carlsson M, Ubachs JF, Hedstrom E, Heiberg E, Jovinge S, ArhedenH. Myocardium at risk after acute infarction in humans on cardiacmagnetic resonance: quantitative assessment during follow-up andvalidation with single-photon emission computed tomography. J AmColl Cardiol Img 2009;2:569–76.
3. Thiele H, Sareban M, Engelhardt H, Eitel I, Gutberlet M, Schuler G.Reproducibility of myocardial salvage in acute myocardial infarction bythe use of contrast-enhanced magnetic resonance imaging (abstr).Circulation 2009;120 Suppl:300.
4. van de Werf F, Bax J, Betriu A, et al. Management of acute myocardialinfarction in patients presenting with persistent ST-segment elevation:the Task Force on the Management of ST-Segment Elevation AcuteMyocardial Infarction of the European Society of Cardiology. EurHeart J 2008;29:2909–45.
5. Schroder R. Prognostic impact of early ST-segment resolution in acuteST-elevation myocardial infarction. Circulation 2004;110:e506–10.
6. Ganame J, Messalli G, Dymarkowski S, et al. Impact of myocardialhaemorrhage on left ventricular function and remodelling in patientswith reperfused acute myocardial infarction. Eur Heart J 2009;30:1440–9.
7. Bondarenko O, Beek AM, Hofman MB, et al. Standardizing thedefinition of hyperenhancement in the quantitative assessment ofinfarct size and myocardial viability using delayed contrast-enhancedCMR. J Cardiovasc Magn Reson 2005;7:481–5.
8. Kastrati A, Mehilli J, Dirschinger J, et al. Myocardial salvage aftercoronary stenting plus abciximab versus fibrinolysis plus abciximab inpatients with acute myocardial infarction: a randomised trial. Lancet
2002;359:920–5. i9. Thiele H, Kappl MJ, Conradi S, Niebauer J, Hambrecht R, Schuler G.Reproducibility of chronic and acute infarct size measurement bydelayed enhancement-magnetic resonance imaging. J Am Coll Cardiol2006;47:1641–5.
0. Eitel I, Desch S, Sareban M, et al. Prognostic significance andmagnetic resonance imaging findings in aborted myocardial infarctionafter primary angioplasty. Am Heart J 2009;158:806–13.
1. Lee JT, Ideker RE, Reimer KA. Myocardial infarct size and locationin relation to the coronary vascular bed at risk in man. Circulation1981;64:526–34.
2. Tilak GS, Hsu LY, Hoyt RF Jr., Arai AE, Aletras AH. In vivoT2-weighted magnetic resonance imaging can accurately determinethe ischemic area at risk for 2-day-old nonreperfused myocardialinfarction. Invest Radiol 2008;43:7–15.
3. Wu KC, Zerhouni EA, Judd RM, et al. Prognostic significance ofmicrovascular obstruction by magnetic resonance imaging in patientswith acute myocardial infarction. Circulation 1998;97:765–72.
4. Hombach V, Grebe O, Merkle N, et al. Sequelae of acute myocardialinfarction regarding cardiac structure and function and their prognos-tic significance as assessed by magnetic resonance imaging. Eur Heart J2005;26:549–57.
5. Nijveldt R, Beek AM, Hirsch A, et al. Functional recovery after acutemyocardial infarction: comparison between angiography, electrocardi-ography, and cardiovascular magnetic resonance measures of micro-vascular injury. J Am Coll Cardiol 2008;52:181–9.
6. Stone GW, Cox D, Garcia E, et al. Normal flow (TIMI-3) beforemechanical reperfusion therapy is an independent determinant ofsurvival in acute myocardial infarction: analysis from the primaryangioplasty in myocardial infarction trials. Circulation 2001;104:636 – 41.
7. Nallamothu BK, Bradley EH, Krumholz HM. Time to treatment inprimary percutaneous coronary intervention. N Engl J Med 2007;357:1631–8.
8. Reimer KA, Lowe JE, Rasmussen MM, Jennings RB. The wave-front phenomenon of ischemic cell death. 1. Myocardial infarct sizevs duration of coronary occlusion in dogs. Circulation 1977;56:786 –94.
9. Gersh BJ, Stone GW, White HD, Holmes DR Jr. Pharmacologicalfacilitation of primary percutaneous coronary intervention for acutemyocardial infarction: is the slope of the curve the shape of the future?JAMA 2005;293:979–86.
0. Francone M, Bucciarelli-Ducci C, Carbone I, et al. Impact of primarycoronary angioplasty delay on myocardial salvage, infarct size, andmicrovascular damage in patients with ST-segment elevation myocar-dial infarction: insight from cardiovascular magnetic resonance. J AmColl Cardiol 2009;54:2145–53.
ey Words: cardiovascular magnetic resonance y magnetic resonance
maging y myocardial infarction y myocardial salvage y prognosis.
