Combined Assessment of Myocardial Perfusion and Late Gadolinium Enhancement in Patients After Percutaneous Coronary Intervention or Bypass Grafts: A Multicenter Study of an Integrated

J A C C : C A R D I O V A S C U L A R I M A G I N G V O L . 2 , N O . 1 1 , 2 0 0 9

© 2 0 0 9 B Y T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N I S S N 1 9 3 6 - 8 7 8 X / 0 9 / $ 3 6 . 0 0

P U B L I S H E D B Y E L S E V I E R I N C . D O I : 1 0 . 1 0 1 6 / j . j c m g . 2 0 0 9 . 0 5 . 0 1 1

Combined Assessment of Myocardial Perfusionand Late Gadolinium Enhancement in PatientsAfter Percutaneous Coronary Interventionor Bypass GraftsA Multicenter Study of an Integrated Cardiovascular MagneticResonance Protocol

Peter Bernhardt, MD,*† Jochen Spiess, MD,* Benny Levenson, MD,‡ Günter Pilz, MD,†Berthold Höfling, MD,† Vinzenz Hombach, MD,* Oliver Strohm, MD§

Ulm, Hausham, and Berlin, Germany; and Calgary, Canada

O B J E C T I V E S We sought to assess the accuracy of an integrated cardiac magnetic resonance (CMR)

protocol for the diagnosis of relevant coronary artery or bypass graft stenosis in patients with suspected

coronary artery disease (CAD) or with previously performed percutaneous coronary intervention (PCI) or

coronary bypass graft surgery (CABG).

B A C K G R O U N D CMR is suitable for diagnosing inducible myocardial ischemia in patients with suspected

CAD and has been proven to be a helpful diagnostic tool for decision of further treatment. However, little is known

about its diagnostic accuracy in patients with known CAD who previously were treated by PCI or CABG.

M E T H O D S A total of 477 patients with suspected CAD, 236 with previous PCI, and 110 after CABG

referred for coronary X-ray angiography (CXA) underwent an integrated CMR examination before CXA.

Myocardial ischemia was assessed using first-pass perfusion after vasodilator stress with adenosine (140

�g/kg/min for 3 min) using gadolinium-based contrast agents (0.1 mmol/kg). Late gadolinium

enhancement (LGE) was assessed 10 min after a second contrast bolus.

R E S U L T S CXA demonstrated a relevant coronary vessel stenosis (�70% luminal reduction) in 313 (38%)

patients using quantitative coronary analysis. The combination of CMR perfusion and LGE assessment for

detecting a relevant coronary stenosis in patients with suspected CAD yielded sensitivity and specificity of

0.94 and 0.87, in PCI patients 0.91 and 0.90, and in CABG patients 0.79 and 0.77, respectively.

C O N C L U S I O N S A combined CMR protocol for the assessment of myocardial perfusion and LGE is

feasible for the detection of relevant coronary vessel stenosis even in patients who previously were

treated by PCI or CAG in a routine clinical setting. However, diagnostic accuracy is reduced in patients

with CABG. This could be due to different flow and perfusion kinetic. Further studies are needed to

optimize the clinical protocols especially in post-surgical patients. (J Am Coll Cardiol Img 2009;2:

1292–1300) © 2009 by the American College of Cardiology Foundation

From the *Department of Internal Medicine II, Cardiovascular MRI Unit, University of Ulm, Ulm, Germany; †CMR-Center at theHospital Agatharied, Academic Teaching Hospital of the University of Munich, Hausham, Germany; ‡Cardiac Praxis, CatheterizationLaboratory at the St.-Gertrauden-Hospital, Berlin, Germany; and the §Stephenson Cardiovascular MR Centre, University of Calgary,Calgary, Canada. Dr. Strohm is a shareholder in and consultant for Circle Cardiovascular Imaging Inc., Calgary, Canada.

Manuscript received February 1, 2009; revised manuscript received April 24, 2009, accepted May 21, 2009.

D(piCbmascghtmcp

bTtsC

aaiatp

M

FwrcvtdwpcC1necS4H

Scpi

itc(53te

rbtSiaemtmbpsqp54nh(secs

(qavt1schmaCr

intervention

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 2 , N O . 1 1 , 2 0 0 9

N O V E M B E R 2 0 0 9 : 1 2 9 2 – 1 3 0 0

Bernhardt et al.

CMR in PCI and CABG Patients

1293

ecision of necessity of reangiography andintervention in patients with known coro-nary artery disease (CAD) and previous

percutaneous coronary artery interventionPCI) or coronary artery bypass graft (CABG)artly depends on the detection of myocardialschemia and sometimes remains challenging (1).ardiac magnetic resonance (CMR) imaging haseen shown to be a safe and feasible imagingethod (2) to assess myocardial ischemia with high

ccuracy using vasodilator and/or positive inotropictressors as adenosine and dobutamine (3–9). Myo-ardial viability can be assessed noninvasively by lateadolinium enhancement (LGE) techniques withigh accuracy (10–12). Recently, it has been shownhat the combination of perfusion and LGE assess-ent increases accuracy of detecting a relevant

oronary stenosis in patients with suspected, but notreviously documented CAD (13,14).However, most of the published studies have

een performed in patients with suspected CAD.here is little knowledge about the ability of CMR

o detect relevant coronary vessel stenosis or occlu-ion in patients who previously underwent PCI orABG.The primary aim of our study was to assess the

bility of a combined CMR protocol includingdenosine stress perfusion and LGE imaging todentify patients with previous PCI or CABG with

relevant coronary vessel stenosis and to comparehe diagnostic performance to patients withoutrevious PCI or CABG.

E T H O D S

rom May 2004 to October 2007, 823 patientsith a clinical indication for X-ray coronary angiog-

aphy (CXA) were included into the trial in threeenters. Exclusion criteria were contraindication forasodilatory stress-CMR including unstable symp-oms, implanted devices, severe obstructive lungisease, and severe valvular disease. All patientsere taken off vasodilator drugs and caffeine 24 hrior to CMR, and all patients gave informedonsent.MR imaging protocol. Imaging was performed using.5-T whole-body clinical magnetic resonance scan-ers (Intera, Philips Medical Systems, Best, the Neth-rlands, using a cardiac 5-element phased-array re-eiver coil [University of Ulm]; CV/i, GE Medicalystem, Milwaukee, Wisconsin, using a cardiac-element phased-array receiver coil [St.-Gertrauden-
ospital, Berlin], and Signa Excite, GE Medical s
ystem, using a cardiac 8-element phased-array re-eiver coil [Hospital Agatharied]). Heart rate, bloodressure, and oxygen saturation were monitored non-nvasively during the examination.

Functional imaging of the left ventricle was donen 3 long-axis (2-, 3-, and 4-chamber view) orien-ation and in contiguous short-axis orientation toover the left ventricle from the base to the apexsteady-state free precession, repetition time [TR]:.1 ms, echo time [TE]: 2.2 ms, field of view: 32 to4 � 32 to 34 cm, matrix: 256 � 192, slicehickness: 8 mm, no interslice gap; acquisition innd-expirational hold).

After 3 min of adenosine infusion at a constantate of 140 �g/kg/min, 0.1 mmol/kg gadolinium-ased contrast agent (Omniscan, GE Medical Sys-ems, Munich, Germany, or Magnevist, Bayerchering Pharma, Berlin, Germany) was injected

ntravenously during a first-pass perfusion sequences previously described (2,3,15). A hybridcho-planar pulse sequence (TR: 6.4 to 16s, TE: 1.2 to 1.4 ms, flip angle: 25°, slice

hickness: 10 mm, voxel size: 2.8 � 2.8m, 3 to 5 slices acquired at every heart

eat) was used on the GE scanners asreviously described (4). On the Philipscanner, a balanced fast-field echo se-uence (TR: 2.6 ms, TE: 1.3 ms, saturatere-pulse with 100-ms delay, flip angle:0°, 40 dynamics, turbo field echo factor:8, voxel size: 2.8 � 2.9 mm, slice thick-ess: 10 mm, 3 to 4 slices acquired at everyeart beat) was used as previously described5). Perfusion images were acquired inhort-axis orientation using a variable gap to cover thentire left ventricle from base to apex. Breathingommands were trained with the patients prior to thetudy to minimize breathing-related artifacts.

Ten minutes after a second contrast agent bolus0.1 mmol/kg), inversion-recovery gradient-echo se-uences (LGE) were acquired in the same orientations the perfusion sequence (TR: 7.1 ms, TE: 3.2 ms,oxel size: 1.2 � 1.2 mm, slice thickness: 10 mm onhe GE scanners; TR: 5.6 ms, TE: 2.1 ms, voxel size:.2 � 1.2 mm, slice thickness: 10 mm on the Philipscanner). Additionally, a 3D LGE sequence with fulloverage of the left ventricle was performed in breathold (TR: 7.1 ms, TE: 3.2 ms, voxel size: 1.2 � 2.2m, slice thickness: 8 mm). TI was individually

djusted for complete nulling of the myocardium.MR analysis. All CMR studies were evaluated by 1eader of each center locally (J.S., G.P., O.S.) and a

A B B

A N D

CABG

graft

CAD �

CMR �

reson

CXA �

angio

LGE �

enhan

PCI �

artery

econd reader for all centers (P.B.). Investi

R E V I A T I O N S

A C R O N YM S

� coronary artery bypass

coronary artery disease

cardiac magnetic

ance

coronary X-ray

graphy

late gadolinium

cement

percutaneous coronary

gators

ws

u

ptudv

sdhwddt(p

pOKcstLrmtaadw

G �

dark rim artifact.


N O V E M B E R 2 0 0 9 : 1 2 9 2 – 1 3 0 0

Bernhardt et al.


1294

ere blinded to quantitative coronary analysis re-ults and to patients’ history and clinical symptoms.

Functional images were analyzed for left ventric-lar end-diastolic and end-systolic volumes by

Figure 1. CMR Interpretation Algorithms for the Diagnosis of Co

Cardiac magnetic resonance (CMR) interpretation algorithms for theenhancement consistent with prior myocardial infarction or presencence of coronary artery or bypass graft stenosis. (B) Perfusion deficing to the diagnosis of coronary artery or bypass graft stenosis. CAB

efect and Absent LGE Perfusion Defect With LGE

ion (Ia) showing perfusion defect of the inferior and inferoseptal) in the absence of late gadolinium enhancement (LGE) (Ib). Perfu-l wall (IIa, arrows) during adenosine with corresponding presence ofubendocardial low signal intensity in the left anterior descending andrfusion areas was not regarded as significant perfusion deficit, but as

e

lanimetry using the multislice approach. Func-ional images were analyzed to obtain left ventric-lar functional parameters as left ventricular end-iastolic and end-systolic volume, and leftentricular ejection fraction.

First-pass perfusion and LGE images were as-essed visually. A relevant perfusion deficit wasefined as a subendocardially beginning hypoen-ancement in at least 2 adjacent segments or slicesith no attributes for imaging artifacts as previouslyescribed (16). Diffuse subendocardial perfusionefects not assignable to a specific coronary supplyerritory were defined as unspecific; dark rim artifact17) was not regarded as perfusion deficit usingreviously described criteria (18,19).Analysis of first-pass perfusion and LGE was

erformed using 2 different algorithms (Fig. 1).ne algorithm was similar to the one described bylem et al. (13). The second algorithm aims to

onsider chronic myocardial infarction in territoriesupplied by previously revascularized coronary ar-eries. In this case, we hypothesized that presence ofGE was not considered as presence of clinically

elevant stenosis of the supplying artery. Relevantyocardial ischemia was defined as mismatch be-

ween hypoenhancement on first-pass perfusionnd enhancement on LGE sequences and classifieds reversible ischemia. A match between first-passeficit and LGE was considered as chronic infarctionith no additional reversible ischemia. Figure 2 shows

ary Artery or CABG Stenosis

gnosis of coronary artery or bypass graft stenosis. (A) Positive latef perfusion deficit during adenosine stress is interpreted as pres-thout corresponding late enhancement is the only condition lead-coronary artery bypass graft.

ron

diae oit wi

Figure 2. Perfusion D

Adenosine-stress perfusmyocardial wall (arrowssion deficit of the lateraLGE (IIb, arrows). The sright coronary artery pe

xamples of stress perfusion as well as of LGE.

susfi

dbCabiiipBp�

asnSfwCdgotppo0

R

Aicccmppfw

((dbu

edi

pAn

fpfmrc


N O V E M B E R 2 0 0 9 : 1 2 9 2 – 1 3 0 0

Bernhardt et al.


1295

Areas of ischemia and areas of LGE were as-igned to the presumably supplying coronary arterysing the 17-seqment model with exclusion ofegment 17 (apex), as this was not visualized on therst-pass perfusion images (20).The CMR data were analyzed for accuracy of

etecting patients with relevant coronary artery orypass graft stenosis as well as detecting.oronary angiography. All coronary catheterizationsnd angiographies were performed as recommendedy the American College of Cardiology and Amer-can Heart Association (1). Blinded analysis ofnvasive data, including degree of luminal narrow-ng of the coronary artery, was performed by inde-endent investigators of the respective centre (J.S.,.L., B.H.). Quantitative coronary analysis waserformed in all patients therefore. A stenosis70% in a coronary artery or in a bypass graft withdiameter �2 mm was considered significant. A

tenotic native vessel with corresponding nonste-otic bypass graft was not considered for ischemia.tatistical analysis. All statistical analysis was per-ormed with commercially available statistic soft-are (StatView 5, SAS Institute, Cary, Northarolina). Data are reported as the mean � stan-ard deviation. Continuous variables betweenroups were compared using the t-test for unpairedbservations. Nominal data were compared usinghe Fisher exact test. Ordinal variables were com-ared using Wilcoxon signed rank test for matchedairs. The McNemar test was used for comparisonf the 2 analysis algorithms. In all cases, a p value �.05 was considered statistically significant.

Table 1. Patients’ Characteristics and Hemodynamics Data andof a Coronary Artery or Bypass Graft

All Pa(N �

Age (yrs) 64.3 �

Gender (male) 629 (7

Diabetes mellitus 176 (2

Hypertension 478 (5

Hypercholesterolemia 334 (4

Smoking 182 (2

Segments of LGE 3.2 �

Transmural LGE 153 (1

Left ventricular end-diastolic volume (ml) 173 �

Left ventricular end-systolic volume (ml) 88 �

Left ventricular ejection fraction (%) 49 �

Segments with perfusion deficits 1.5 �

*p � 0.0001. †p � 0.01.
LGE � late gadolinium enhancement.
E S U L T S

total of 876 patients were screened for inclusionnto the study. Fifty-three patients were not in-luded due to claustrophobia (n � 27), contraindi-ation for adenosine (n � 21) or refusal to giveonsent (n � 5). Thus, 823 patients (629 [76%]ale) formed the study population. 477 (58%)

atients with suspected CAD, 236 (29%) withrevious PCI, and 110 (13%) with previous CABGormed the different study groups. CMR studiesere performed 8 � 5 days before CXA.Clinical indication for CXA was relevant angina

74.4%), positive stress test (other than CMR)31%), and/or other abnormalities (8.1%), as, e.g.,yspnea, considered as angina equivalent, new leftundle brunch block, or new nonsustained ventric-lar tachycardia.Mean time between PCI or CABG and CMR

xamination was 314 � 231 days and 423 � 275ays, respectively. Patients’ characteristics are givenn Table 1.

No major or minor complications occurred; allatients could complete the entire protocol. TransientV-block was observed in 81 (9%) patients, but didot require specific treatment or cessation of the study.Quality of perfusion images was sufficient for

urther analysis in all patients. In 138 (16.8%)atients, noncritical artifacts could be seen on per-usion imaging. These artifacts were graded to beotion artifacts due to breathing in 51 cases, dark

im artifacts in 49 cases, and ghosting artifacts in 38ases. A total of 185 (0.9%) segments had to be

parison of Patients With Angiographically Diagnosed Stenosis

ts)

Stenosis(n � 313)

No Stenosis(n � 510)

.3 66.4 � 13.2 63.7 � 11.4

) 253 (80.8%) 376 (73.7%)

) 72 (23.0%) 104 (20.4%)

) 192 (61.3%) 286 (56.1%)

) 128 (40.9%) 206 (40.4%)

) 59 (18.8%) 123 (24.1%)

4.3 � 3.2* 2.3 � 1.8*

) 73 (23.3%)† 80 (15.7%)†

187 � 57 169 � 54

94 � 47 83 � 28

47 � 20 52 � 18

3.4 � 2.1* 0.4 � 1.1*

Com

tien823

12

6.4%

1.4%

8.1%

0.6%

2.1%

2.8

8.6%

63

32

19

2.1

eHiicCtdpn

p4agP

Laa

cpC

ss0oCwa

0


N O V E M B E R 2 0 0 9 : 1 2 9 2 – 1 3 0 0

Bernhardt et al.


1296

xcluded from visual analysis due to severe artifacts.owever, no study had to be excluded due to

nsufficient image quality. In 23 (2.8%) patients,mage quality of the LGE sequences was insuffi-ient for analysis.MR and diagnostic performance. Volumetric data ofhe left ventricle are given in Table 1. Patient groupsid not differ significantly for ventricle functionalarameters. Patients with and without coronary ste-osis did not differ for these parameters, either.Perfusion deficits were detected in 417 (50.7%)

atients in the first-pass perfusion. Two hundred of77 (41.9%) patients with perfusion deficits duringdenosine-stress perfusion were found in the patientroup with suspected CAD, 129 of 236 (54.7%) in the

Table 2. Diagnostic Accuracy of Both CMR Algorithms in All Pat

Sensitivity

Suspected CAD patients

Algorithm A

All patients 0.93

1-vessel disease 0.91

Multiple-vessel disease 0.95

Algorithm B

All patients 0.81



Algorithm A

RCA 0.92

LAD 0.87

LCX 0.80

Algorithm B

RCA 0.82

LAD 0.79

LCX 0.78

PCI patients

Algorithm A

All patients 0.88



Algorithm B

All patients 0.88



Algorithm A

RCA 0.87

LAD 0.80

LCX 0.76

Algorithm B

RCA 0.87

LAD 0.80

LCX 0.76

CI and 88 of 110 (80.0%) in the CABG group. p

GE could be visualized in 312 (37.9%) patients withtransmural extent in 152 patients; 262 of these hadhistory of myocardial infarction.CXA revealed relevant coronary artery stenosis or

hronic occlusion in 173 (36%) patients with sus-ected CAD, in 69 (29%) PCI, and in 71 (65%)ABG patients.Algorithm A performed best in patients with

uspected CAD and yielded a sensitivity of 0.93, apecificity of 0.87, a positive predictive value of.81, a negative predictive value of 0.96, and anverall accuracy of 0.89. In patients with PCI andABG, we observed better CMR interpretation,ith algorithm B resulting in a sensitivity of 0.88

nd 0.73, a specificity of 0.90 and 0.77, a positive

Groups and Subanalysis for 1- or Multiple-Vessel Disease

cificity PPV NPV Accuracy

.87 0.81 0.96 0.89

.87 0.63 0.97 0.88

.87 0.71 0.98 0.89

.87 0.78 0.89 0.85

.87 0.60 0.95 0.86

.87 0.68 0.94 0.86

.87 0.67 0.98 0.88

.86 0.68 0.95 0.87

.85 0.60 0.95 0.84

.87 0.63 0.87 0.84

.86 0.63 0.93 0.84

.85 0.56 0.94 0.84

.79 0.71 0.94 0.82

.79 0.46 0.97 0.80

.79 0.50 0.97 0.81

.90 0.79 0.95 0.91

.90 0.62 0.97 0.90

.90 0.69 0.97 0.90

.78 0.49 0.96 0.80

.79 0.45 0.95 0.79

.78 0.39 0.95 0.78

.91 0.71 0.94 0.90

.90 0.63 0.94 0.89

.90 0.58 0.88 0.88

ient

Spe

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

redictive value of 0.79 and 0.85, a negative pre-

da

yfFbanvi

s(msafwss

i

f(fwm

D

Tgaipoo

(aaat“

PV �


N O V E M B E R 2 0 0 9 : 1 2 9 2 – 1 3 0 0

Bernhardt et al.


1297

ictive value of 0.95 and 0.61, and an overallccuracy of 0.91 and 0.75, respectively.

Diagnostic performance of CMR with both anal-sis algorithms for all patient groups and subanalysisor 1- or multiple-vessel disease is given in Table 2.urthermore, diagnostic performance subanalysis ofoth CMR algorithms in all patient groups for the leftnterior descending, left circumflex, and right coro-ary artery perfusion territories as well as for arterial,enous, and mixed bypass grafts in the CABG groups provided in Table 2.

Patients with coronary artery or bypass graft steno-is �70% had significantly more segments with LGEp � 0.0001) and had significantly more often trans-ural extent than patients without coronary artery

tenosis (p � 0.0074). There were no differences inge, sex, cardiovascular risk factors, or left ventricularunctional parameters between patients with andithout angiographically detected stenosis. Compari-

on of patients with and without coronary arterytenosis as detected by CXA is given in Table 3.

In 3 of 9 (33%) patients with false positive CMR

Table 2 Continued

Sensitivity

CABG patients

Algorithm A

All patients 0.73



Algorithm B

All patients 0.73



Algorithm A

RCA 0.76

LAD 0.74

LCX 0.70

Algorithm B

RCA 0.76

LAD 0.74

LCX 0.70

Algorithm A

Arterial graft 0.75

Venous graft 0.74

Mixed graft 0.68

Algorithm B

Arterial graft 0.75

Venous graft 0.74

Mixed graft 0.68

p � 0.01 for comparison of Algorithms A and B. Description of Algorithms A aCABG � coronary artery bypass graft; CAD � coronary artery d

NPV � negative predictive value; PCI � percutaneous coronary intervention; P

nterpretation, the perfusion deficit observed af- a

ected only basal myocardial segments. In 8 of 1942%) patients with false negative CMR results, weound perfusion deficit and LGE match, thus thisas interpreted as nonreversible ischemia in chronicyocardial ischemia.

I S C U S S I O N

he primary finding of our study is that an inte-rated CMR examination using first-pass perfusionnd LGE is a feasible method to detect myocardialschemia and irreversible myocardial injuries inatients with previous coronary artery interventionr CABG with high sensitivity, specificity andverall accuracy.Comparing the amount of inducible ischemia

assessed by vasodilator stress first-pass perfusion)nd the amount of irreversible ischemic damage (asssessed by LGE in the same image orientation)llows predicting a relevant coronary artery stenosishat would benefit from revascularization. Amatch” between the amount of inducible ischemia

cificity PPV NPV Accuracy

.62 0.78 0.56 0.69

.62 0.64 0.71 0.67

.62 0.63 0.73 0.67

.77 0.85 0.61 0.75

.77 0.75 0.75 0.75

.77 0.74 0.77 0.75

.62 0.54 0.81 0.67

.63 0.48 0.84 0.66

.63 0.55 0.76 0.61

.77 0.66 0.84 0.76

.76 0.59 0.86 0.75

.78 0.67 0.80 0.75

.63 0.46 0.86 0.67

.63 0.44 0.86 0.66

.60 0.36 0.85 0.62

.77 0.58 0.88 0.77

.78 0.56 0.88 0.76

.75 0.47 0.88 0.73

is provided in Figure 1.e; LAD � left anterior descending; LCX � left circumflex artery;positive predictive value; RCA � right coronary artery.

Spe

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

nd Biseas

nd LGE is considered a completed myocardial

iooica

nrroidaas

nsmaoses7onrehsptp

Cqaavids

cdmcpaptpsdTflc(m

mbpppbs


N O V E M B E R 2 0 0 9 : 1 2 9 2 – 1 3 0 0

Bernhardt et al.


1298

nfarct and thus would not require revascularizationf the supplying vessel. To account for the presencef irreversible injury in segments with inducibleschemia, we developed an analysis algorithm andorrelated the predictive values to invasive coronaryngiography.

Only the visualization of inducible ischemia inon-LGE myocardium would predict a clinicallyelevant coronary vessel stenosis and mandate forevascularization. We therefore compiled our sec-nd analysis algorithm to identify patients withnducible myocardial ischemia and chronic myocar-ial infarction. The better performance of thislgorithm strengthens the thesis about necessity ofn analysis and interpretation algorithm that con-iders chronic myocardial infarction.

Assessment of a relevant restenosis or new ste-osis in an artery after intervention has been de-cribed using different diagnostic tools. A recenteta-analysis proved exercise treadmill testing to berather poor diagnostic test to identify the presencef graft stenosis, yielding a sensitivity of 45% and apecificity of 82% (21–24). Dobutamine stresschocardiography with contrast-enhanced perfu-ion imaging has been shown to have a sensitivity of3% and a specificity of 75% (25) for the evaluationf coronary artery stenosis. Nuclear medicine tech-iques have also been evaluated for the detection ofestenosis in patients after PCI. Single-photonmission computed tomography (SPECT) imagingas been shown to have a sensitivity of 93% and apecificity of 77% for detection of restenosis inatients after PCI (23); however, SPECT is knowno suffer from attenuation artifacts (26) and exposes

Table 3. Comparison of Patients With Angiographically Diagno

All Pat(N �

Age (yrs) 64.3 �

Gender (female) 194 (2

Diabetes mellitus 176 (2

Hypertension 478 (5

Hypercholesterolemia 334 (4

Smoking 182 (2

Segments of LGE 3.2 �

Transmural LGE 153 (1

Left ventricular end-diastolic volume (ml) 173 �

Left ventricular end-systolic volume (ml) 88 �

Left ventricular ejection fraction (%) 49 �

Segments with perfusion deficits 1.5 �

*p � 0.0001 versus no stenosis. †p � 0.01 versus no stenosis. Description of AAbbreviation as in Table 1.

atients to unwanted radiation. p

CMR offers the ability to assess different aspects ofAD in a combined protocol; it has been shown touantify ventricular functional parameters (27,28),ssess first-pass perfusion during stress (2–8,13–15),nd allow for noninvasive assessment of myocardialiability (10–15,20,29) in a single study. Recent stud-es have shown that CMR provides a high accuracy inetecting myocardial perfusion defects in patients withuspected CAD (3–6,8,13,14).

We could demonstrate that patients after bypassan be assessed with CMR, but with a loweriagnostic yield. Delayed contrast appearance inyocardial tissue during stress perfusion study

ould have different etiologies, such as post-stenoticressure reduction in the supplying coronary artery,ltered coronary artery anatomy with changed flowroperties, slower flow and/or longer transit time ofhe contrast bolus through bypass grafts. A com-arison between radial artery grafts and longeraphenous venous bypass grafts found significantifferent flow characteristics and flow response (29).hese data strengthen the hypothesis of differentow characteristics in different bypass types thatould possibly cause different myocardial perfusion30) and thus be a limitation in the correct assess-ent of graft patency.This may be a universal limitation for all stressethods after bypass grafts and may mandate a

aseline-stress study immediately after surgery. Sup-orting this theory is a study by Fenchel et al. (31) thatrovided information on the success of interventionalrocedures in 18 patients performing stress-CMR as aaseline after intervention. The lack of a “baseline”tudy could explain the lower specificity of CMR

Stenosis of a Coronary Artery Bypass Graft

ts)

Stenosis(n � 313)

No Stenosis(n � 510)

3 66.4 � 13.2 63.7 � 11.4

) 60 (19.2%) 134 (26.3%)

) 72 (23.0%) 104 (20.4%)

) 192 (61.3%) 286 (56.1%)

) 128 (40.9%) 206 (40.4%)

) 59 (18.8%) 123 (24.1%)

4.3 � 3.2* 2.3 � 1.8

) 73 (23.3%)† 80 (15.7%)

187 � 57 169 � 54

94 � 47 83 � 28

47 � 20 52 � 18

3.4 � 2.1* 0.4 � 1.1

ithms A and B is provided in Figure 1.

sed

ien823

12.

3.6%

1.4%

8.1%

0.6%

2.1%

2.8

8.6%

63

32

19

2.1

lgor

erformance in CABG patients in our study.

ptCvs(snnsctCs

toebicccStptrtm

slof

aoii

C

OsppoipavCgwg

RD8

R


N O V E M B E R 2 0 0 9 : 1 2 9 2 – 1 3 0 0

Bernhardt et al.


1299

The presence of significant collateral flow inatients with multivessel CAD could account forhe lower sensitivity in our patient group afterAPB; these patients had a higher rate of multiple-

essel disease and presumably more often developedystem of collateral flow. A study by Becker et al.32) showed that stenoses of the coronary vesselsupplying collateral flow to ischemic regions areecessary for the occurrence of coronary steal phe-omenon. Another recent study detected coronaryteal in only one-third of patients with chronic totaloronary occlusion (33). These findings strengthenhe higher rate of false negative findings in ourABG group and thus could explain the lower

ensitivity of CMR performance.In 33% of our false positive CMR interpreta-

ions, we observed myocardial ischemia affectingnly the basal myocardial segments. A possiblexplanation could be a distal anastomosis of theypass graft, restricting flow to the basal segmentsn case of a proximal stenosis or occlusion. Thus, inase of proximal stenosis or occlusion of the nativeoronary artery, a perfusion of the basal segmentsould be restricted.tudy limitations. The patient groups included inhe study had a high pre-test probability for theresence of CAD, as they all had a clinical indica-ion for invasive coronary angiography. Thus, theesults of our study may not be easily transferrableo the general population or to patients with low or

magnetic resonance: a comparison with9. Kwong RY, Schu

khraj S, et al. Dete

The use of different CMR systems and perfusionequences in the 3 centers could be a possibleimitation to our study. However, we did notbserve significant differences regarding CMR per-ormance between our centers.

Another possible limitation is the presence ofrtifacts in 16.8% patients. However, less than 1%f stress perfusion images and less than 3% of LGEmages had to be excluded from analysis due tonsufficient image quality.

O N C L U S I O N S

ur data show that CMR is suitable for noninva-ive detection of relevant coronary artery stenosis inatients with suspected CAD, including those withrevious PCI or CABG. The combined assessmentf inducible ischemia and reversible ischemia in anntegrated CMR protocol allows predicting theresence of a clinically relevant coronary arterynd/or bypass graft stenosis. In patients with pre-ious CABG, sensitivity, specificity, and accuracy ofMR is reduced in comparison to the other patient

roups. Further studies are warranted to determinehether a quantitative assessment in this patientroup could improve the accuracy of the test.

eprint requests and correspondence: Dr. Peter Bernhardt,epartment of Internal Medicine II, University of Ulm,

9077 Ulm, Germany. E-mail: peter.bernhardt@uniklinik-
edium pre-test probability. ulm.de.
1

1

1

E F E R E N C E S

1. Scanion PJ, Faxon DP, Audet AM, et al.ACC/AHA guidelines for coronary an-giography. A report of the American Col-lege of Cardiology/American Heart Asso-ciation Task Force on practice guidelines(Committee on Coronary Angiography).Developed in collaboration with the Soci-ety for Cardiac Angiography and Inter-ventions. Circulation 1999;99:2345–57.

2. Bernhardt P, Steffens M, Kleinertz K,et al. Safety of adenosine stress magneticresonance imaging using a mobile car-diac magnetic resonance system. J Car-diovasc Magn Reson 2006;8:475–8.

3. Giang TH, Nanz D, Coulden R, et al.Detection of coronary artery disease bymagnetic resonance myocardial perfusionimaging with various contrast mediumdoses: first European multi-centre experi-ence. Eur Heart J 2004;25:1657–65.

4. Schwitter J, Nanz D, Kneifel S, et al.Perfusion in coronary artery disease by

positron emission tomography and cor-onary angiography. Circulation 2001;103:2230–5.

5. Nagel E, Klein C, Paetsch I, et al.Magnetic resonance perfusion measure-ments for the noninvasive detection ofcoronary artery disease. Circulation2003;108:432–7.

6. Ishida N, Sakuma H, Motoyasu M, et al.Noninfarcted myocardium: correlationbetween dynamic first-pass contrast-enhanced myocardial MR imaging andquantitative coronary angiography. Radi-ology 2003;229:209–16.

7. Pennell DJ. Cardiovascular magneticresonance and the role of adenosinepharmacologic stress. Am J Cardiol2004;94 Suppl:26D–32D.

8. Lee DC, Simonetti OP, Harris KR, et al.Magnetic resonance versus radionuclidepharmacological stress perfusion imagingfor flow limiting stenoses of varying sever-ity. Circulation 2004;110:48–65.

ssheim AE, Re-cting acute coro-

nary syndrome in the emergency de-partment with cardiac magnetic reso-nance imaging. Circulation 2003;107:531–7.

0. Klein C, Nekolla SG, Bengel FM,et al. Assessment of myocardial viabil-ity with contrast-enhanced magneticresonance imaging: comparison withpositron emission tomography. Circu-lation 2002;105:162–7.

1. Wagner A, Mahrholdt H, Holly TA,et al. Contrast-enhanced MRI and rou-tine single photon emission computed to-mography (SPECT) perfusion imagingfor detection of subendocardial myocardialinfarcts: an imaging study. Lancet 2003;361:374–9.

2. Christiansen JP, Edwards C, SinclairT, et al. Detection of myocardialscar by contrast-enhanced cardiacmagnetic resonance imaging in pa-tients with troponin-positive chestpain and minimal angiographic cor-onary artery disease. Am J Cardiol
2006;97:768 –71.
mailto:[email protected]

mailto:[email protected]

1

1

1

1

1

1

1

2

2

2

3

3

3

3

Kc


N O V E M B E R 2 0 0 9 : 1 2 9 2 – 1 3 0 0

Bernhardt et al.


1300

3. Klem I, Heitner JF, Shah DJ, et al.Improved detection of coronary arterydisease by stress perfusion cardiovas-cular magnetic resonance with the useof delayed enhancement infarctionimaging. J Am Coll Cardiol 2006;47:1630–8.

4. Cury RC, Cattani CAM, GabureLAG, et al. Diagnostic performanceof stress perfusion and delayed-enhancement MR imaging in patientswith coronary artery disease. Radiol-ogy 2006;240:39–45.

5. Wöhrle J, Nusser T, Merkle N, et al.Myocardial perfusion reserve in car-diovascular magnetic resonance: cor-relation to coronary microvasculardysfunction. J Cardiovasc Magn Re-son 2006;8:781–7.

6. Bernhardt P, Levenson B, AlbrechtA, Engels T, Strohm O. Detectionof cardiac small vessel disease byadenosine-stress magnetic reso-nance. Int J Cardiol 2007;121:261– 6.

7. Di Bella EVR, Parker DL, SinusasAJ. On the dark rim artifact in dy-namic contrast-enhanced MRI myo-cardial perfusion studies. Magn ResonMed 2005;54:1295–9.

8. Arai AE. Magnetic resonance first-pass myocardial perfusion imaging.Top Magn Reson Imaging 2000;11:383–98.

9. Barkhausen J, Hunold P, Jochims M,Debatin JF. Imaging of myocardialperfusion with magnetic resonance. JMagn Reson Imaging 2004;19:750 –7.

0. Cerqueria MD, Weissman NJ, Dilsi-zian V, et al. Standardized myocardialsegmentation and nomenclature fortomographic imaging of the heart: astatement for healthcare professionals
from the Cardiac Imaging Committee
of the Council on Clinical Cardiologyof the American Heart Association.Circulation 2002;105:539–42.

21. Bengston JR, Mark DB, Honan MB,et al. Detection of restenosis afterelective percutaneous transluminalcoronary angioplasty using the exer-cise treadmill test. Am J Cardiol 1990;65:28–34.

22. Honan MB, Bengston JR, Pryor DB,et al. Exercise treadmill testing is apoor predictor of anatomic resteno-sis after angioplasty for acute myo-cardial infarction. Circulation 1989;80:1585–94.

23. Hecht HS, Shaw RE, Bruce TR,Ryan C, Stertzer SH, Myler RK. Use-fulness of tomographic Thallium-201imaging for detection of restenosisafter percutaneous transluminal coro-nary angioplasty. Am J Cardiol 1990;66:1314–8.

24. Chin AS, Goldmann LE, EisenbergMJ. Stress testing after coronary arterybypass graft surgery: a meta-analysis.Can J Cardiol 2003;19:802–8.

25. Elhendy A, Tsutsui JM, O’Leary EL,Xie F, Majeed F, Porter TR. Evalua-tion of restenosis and extend of coro-nary artery disease in patients withprevious coronary interventions by do-butamine stress real-time myocardialcontrast perfusion imaging. Heart2006;92:1480–3.

26. Hendel RC, Berman DS, Cullom SJ,et al. Multicenter clinical trial to eval-uate the efficacy of correction for pho-ton attenuation and scatter in SPECTmyocardial perfusion imaging. Circu-lation 1999;99:2742–9.

27. Bellenger NG, Burgess MI, Ray SG,et al. Comparison of left ventricularejection fraction and volumes in heartfailure by echocardiography, radionu-
clide ventriculography and cardiovas- p
cular magnetic resonance; are they interchangeable? Eur Heart J 2000;21:1387–96.

8. Beygui F, Furber A, Delépine S, et al.Routine breath-hold gradient echoMRI-derived right ventricular mass,volumes and function: accuracy, re-producibility and coherence study.Int J Cardiovasc Imaging 2004;20:509 –16.

9. Kim RJ, Fieno DS, Parrish TB, et al.Relationship of MRI delayed contrastenhancement to irreversible injury, in-farct age, and contractile function.Circulation 1999;100:1992–2002.

0. Chong WCF, Collins P, Webb CM,et al. Comparison of flow characteris-tics and vascular reactivity of radialartery and long saphenous vein grafts.J Cardiothorac Surg 2006;1:4.

1. Fenchel M, Franow A, Stauder NI,et al. Myocardial perfusion afterangioplasty in patients suspected ofhaving single-vessel coronary arterydisease: improvement detected at rest-stress first-pass perfusion MR imag-ing–initial experience. Radiology2005;237:67–74.

2. Becker LC. Conditions for vasodilator-induced coronary steal in experimentalmyocardial ischemia. Circulation 1978;57:1103–10.

3. Werner GS, Figulla HR. Direct as-sessment of coronary steal and asso-ciated changes of collateral hemody-namics in chronis total coronaryocclusions. Circulation 2002;106:435– 40.

ey Words: adenosine stress yardiac magnetic resonance y
erfusion y CABG y PCI.

Documents

Combined Assessment of Myocardial Perfusion and Late Gadolinium Enhancement in Patients After Percutaneous Coronary Intervention or Bypass Grafts: A Multicenter Study of an Integrated