mitral apparatus assessment by delayed enhancement...

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 . 6 , N O . 2 , 2 0 1 3

© 2 0 1 3 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 / $ 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 . h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j c m g . 2 0 1 2 . 0 8 . 0 1 6

Mitral Apparatus Assessment byDelayed Enhancement CMRRelative Impact of Infarct Distribution on Mitral Regurgitation

Jason S. Chinitz, MD,* Debbie Chen, BA,* Parag Goyal, MD,* Sean Wilson, MD,*Fahmida Islam, BA,* Thanh Nguyen, PHD,† Yi Wang, PHD,† Sandra Hurtado-Rua, PHD,‡Lauren Simprini, MD,§ Matthew Cham, MD,� Robert A. Levine, MD,¶Richard B. Devereux, MD,* Jonathan W. Weinsaft, MD*†

New York, New York; and Boston, Massachusetts

O B J E C T I V E S This study sought to assess patterns and functional consequences of mitral

apparatus infarction after acute myocardial infarction (AMI).

B A C K G R O U N D The mitral apparatus contains 2 myocardial components: papillary muscles and the

adjacent left ventricular (LV) wall. Delayed-enhancement cardiac magnetic resonance (DE-CMR) enables in vivo

study of inter-relationships and potential contributions of LV wall and papillary muscle infarction (PMI) to mitral

regurgitation (MR).

M E T H O D S Multimodality imaging was performed: CMR was used to assess mitral geometry and

infarct pattern, including 3D DE-CMR for PMI. Echocardiography was used to measure MR. Imaging

occurred 27 � 8 days after AMI (CMR, echocardiography within 1 day).

R E S U L T S A total of 153 patients with first AMI were studied; PMI was present in 30% (n � 46 [72%

posteromedial, 39% anterolateral]). When stratified by angiographic culprit vessel, PMI occurred in 65% of

patients with left circumflex, 48% with right coronary, and only 14% of patients with left anterior descending

infarctions (p �0.001). Patients with PMI had more advanced remodeling as measured by LV size and mitral

annular diameter (p �0.05). Increased extent of PMI was accompanied by a stepwise increase in mean infarct

transmurality within regional LV segments underlying each papillary muscle (p �0.001). Prevalence of lateral

wall infarction was 3-fold higher among patients with PMI compared to patients without PMI (65% vs. 22%,

p �0.001). Infarct distribution also impacted MR, with greater MR among patients with lateral wall infarction

(p � 0.002). Conversely, MR severity did not differ on the basis of presence (p � 0.19) or extent (p � 0.12)

of PMI, or by angiographic culprit vessel. In multivariable analysis, lateral wall infarct size (odds ratio 1.20/%

LV myocardium [95% confidence interval: 1.05 to 1.39], p � 0.01) was independently associated with

substantial (moderate or greater) MR even after controlling for mitral annular (odds ratio 1.22/mm [1.04 to

1.43], p � 0.01), and LV end-diastolic diameter (odds ratio 1.11/mm [0.99 to 1.23], p � 0.056).

C O N C L U S I O N S Papillary muscle infarction is common after AMI, affecting nearly one-third of

patients. Extent of PMI parallels adjacent LV wall injury, with lateral infarction—rather than PMI—

associated with increased severity of post-AMI MR. (J Am Coll Cardiol Img 2013;6:220–34) © 2013 by

the American College of Cardiology Foundation

From the *Department of Medicine, Greenberg Cardiology Division, Weill Cornell Medical College, New York, New York;†Department of Radiology, Weill Cornell Medical College, New York, New York; ‡Department of Biostatistics, Weill CornellMedical College, New York, New York; §Cardiology Division, Memorial Sloan Kettering Cancer Center, New York, New York;�Radiology, Mount Sinai Medical Center, New York, New York; and the ¶Massachusetts General Hospital, Harvard MedicalSchool, Boston, Massachusetts. This work was supported by Lantheus Medical Imaging, and a Doris Duke Clinical ScientistDevelopment Award (K23 HL102249-01) to Dr. Weinsaft. The authors have reported they have no relationships relevant tothe contents of this paper to disclose.

Manuscript received January 11, 2012; revised manuscript received August 17, 2012, accepted August 20, 2012.

http://dx.doi.org/10.1016/j.jcmg.2012.08.016

Murcocsd

terSMdgwae

idfolzbtdi

1sotipsutiDotDrimg

3-dimensional

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 . 6 , N O . 2 , 2 0 1 3

F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.

Mitral Apparatus Assessment by DE-CMR

221

itral regurgitation (MR) is a seriousconsequence of acute myocardial in-farction (AMI) that confers risk foradverse outcomes, including heart fail-

re and death (1–3). Despite advances in coronaryeperfusion, MR remains common after AMI, oc-urring in as many as 50% of patients (1,4). More-ver, MR can beget MR, as regurgitation itselfontributes to chamber dilation, hypertrophy, andystolic dysfunction (5–7). This cascade can furtheristort both chamber and valve geometry, resulting

See page 235

in progressive MR. While mitral repair/replacement caneffectively treat severe MR and medical therapiescan prevent adverse remodeling, clinical outcomescan be compromised if therapy is delayed (8–11).Consistent with these clinical observations, animalstudies have demonstrated that early treatment ofeven moderate MR can reduce left ventricular (LV)volumes, attenuate adverse remodeling, and preventcontractile dysfunction (6). Thus, identification ofindexes that predict development and progressionof post-AMI MR is important for implementationof effective therapeutic interventions.

The mitral valve apparatus includes 2 myocardialcomponents: papillary muscles and the adjacent LVwall (12). Both are believed to contribute to mitralvalve integrity through direct contractile effects tomaintain valve closure as well as secondary effectson valve geometry (12,13). The link between pap-illary muscle infarction (PMI) and severe MR iswell established in the context of papillary rupture(14–16). It is also possible that lesser degrees ofpapillary necrosis can produce lesser MR that in-creases with time. However, PMI can occur with-out rupture, a phenomenon originally reported inexperimental as well as autopsy studies (17–19), andwas recently shown in vivo using delayed-enhancement(DE) cardiac magnetic resonance (CMR), which enableshigh-resolution imaging of both PMI and LV wallinfarcts in a manner that closely agrees with pathology-evidenced myocyte necrosis (20–22). Both PMI and LVinfarct distribution have been studied as isolated predic-tors of MR. However, results have varied (5,23–26),possibly because their inter-relationship and potentialindependent contributions to MR have not been exam-ined.

This study examined patterns and consequencesof mitral apparatus infarction after AMI. To thisend, CMR was performed to assess cardiac func-
tion, geometry, and infarct pattern, including high- u
resolution 3-dimensional (3D) DE-CMR for PMI.Transthoracic echocardiography (echo) was per-formed for dedicated assessment of MR. Theaims were to determine 1) clinical and imagingpredictors of post-AMI PMI; and 2) relativeimpact of PMI and adjacent LV wall infarctionon early post-AMI MR.

M E T H O D S

Population. The population consisted of consecu-ive patients with first ST-segment elevation MInrolled in a prospective registry of post-AMIemodeling (clinical trial #NCT00539045) betweeneptember 2006 and August 2011 at Weill Cornelledical College. Patients with intrinsic mitral

ysfunction (prolapse, rheumatic, prior sur-ery) were excluded. Imaging was performedithin a pre-specified time limit of 6 weeks

fter AMI. All patients underwent CMR andcho within 1 day.

Comprehensive clinical data were collected,ncluding cardiac risk factors, coronary arteryisease history, New York Heart Associationunctional class, and medication regimen. Cor-nary angiograms were reviewed for infarctocation and coronary dominance. Cardiac en-ymes (creatine phosphokinase, myocardialand fraction) were collected for serologic es-imation of infarct size. The study was con-ucted in accordance with the Weill Cornell

nstitutional review board.Image acquisition. CARDIAC MAGNETIC

RESONANCE. CMR was performed using.5-T scanners (General Electric, Wauke-ha, Wisconsin). Examinations consistedf 2 components: 1) cine-CMR for func-ion/morphology; and 2) DE-CMR fornfarct quantification. Cine-CMR waserformed using a steady-state free precession pulseequence. Short-axis images were acquired contig-ously from the level of the mitral valve annulushrough the apex. Long-axis images were acquiredn standard 2-, 3-, and 4-chamber orientations.

E-CMR was performed after intravenous gad-linium infusion (0.2 mmol/kg). In all patients,ailored PMI imaging was attempted using a 3DE-CMR pulse sequence that provides high-

esolution imaging (matrix size 256 � 256; typicaln-plane resolution 1.4 � 1.4mm, slice thickness 5

m) acquired during free-breathing by a navigatorating algorithm (27). When 3D DE-CMR was

A B B

A N D

AMI �

infarc

CI � c

CMR �

reson

DE �

Echo

LAD �

artery

LCX �

LV �

MR �

OR �

PMI �

infarc

RCA �

RF �

2D �

3D �

nsuccessful (16%), imaging was limited to b

R E V I A T I O N S

A C R O N YM S

acute myocardial

tion

onfidence interval

cardiac magnetic

ance

delayed-enhanced

� echocardiography

left anterior descending

left circumflex

left ventricular

mitral regurgitation

odds ratio

papillary muscle

tion

right coronary artery

regurgitant fraction

2-dimensional

reath-

N

iape

tT


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


222

held 2-dimensional (2D) DE-CMR. For both 2Dand 3D DE-CMR, inversion times were adjustedto null viable myocardium and contiguous short-axis images were acquired throughout the LV.

ECHOCARDIOGRAPHY. Transthoracic echo wasperformed by experienced sonographers using com-mercially available equipment (General ElectricVivid-7; Philips ie33) with phased-array transduc-ers. Images were acquired using a pre-specifiedregistry protocol, including dedicated MR assess-ment in accordance with American Society ofEchocardiography consensus guidelines (28). Imag-ing was performed in parasternal, subcostal, andapical 2-, 3-, and 4- chamber orientations. ColorDoppler was used to assess presence and severity ofMR on the basis of jet area, depth, vena contracta,and directionality. Pulsed-wave Doppler includedassessment of MR duration and pulmonary veinflow profile. Two-dimensional imaging was usedfor related parameters, including mitral valve andannular morphology.Image interpretation. CARDIAC MAGNETIC RESO-

ANCE. DE-CMR was used to detect PMI. Inaccordance with established criteria (13), papillarymuscles were defined as discrete myocardial struc-tures within the LV cavity that were typicallycircular in cross-sectional shape and either antero-

Figure 1. Representative Examples of Complete and Partial PM

Typical examples of (A) complete papillary muscle infarction (PMI) anetic resonance. Each example is composed of 2 short-axis imagesoften associated with transmural infarction of the adjacent left ven
infarction. Upper right shows bilateral, complete PMI with transmural in
lateral or posteromedial in location. PMI wasdeemed present if any papillary hyperenhancementwas evident on DE-CMR short-axis images. PMIwas further categorized by location and extent—partial or complete—stratified using a threshold of�50% papillary myocardium (Fig. 1).

Global LV infarct size (% LV myocardium) wasquantified on DE-CMR using the full-width athalf-maximum method (22,29). Reproducibility ofinfarct size quantification, as tested in a subgroupcomprising 15% of the study population (n � 23),yielded small intraobserver (� � 0.2 � 1.5% LVmyocardium [limits of agreement � �2.7% to�3.1%]) and interobserver differences (� � �0.6 �1.3% LV myocardium [limits of agreement ��3.3% to �2.0%]). Left ventricular infarct trans-murality was scored using a 17-segment model, 5points per segment scoring system (0 � no hyper-enhancement; 1 � 1% to 25%; 2 � 26% to 50%;3 � 51% to 75%; 4 � 76% to 100%) (30). Relativenfarct burden within designated anterior, inferior,nd lateral territories (Fig. 2) was calculated as theroportion of segmental scores contained withinach territory, multiplied by global LV infarct size.

The CMR scan was also analyzed for conven-ional indexes of cardiac structure and function.he LV ejection fraction and cavity volumes were

(B) partial PMI detected by delayed-enhancement cardiac mag-in the affected papillary muscle. As shown, complete PMI waslar wall, whereas partial PMI was associated with subendocardial

I

ndwithtricu

farction of the inferior and lateral walls.

nhaivib

ten


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


223

based on end-diastolic and end-systolic endocardialcontours of contiguous short-axis images, withend-diastolic diameter measured in short axis atmid-LV level. Regional LV contractile functionwas graded using a 17-segment, 5-point scoringsystem (0 � normal contraction; 1 � mild hypoki-

esia; 2 � moderate hypokinesia; 3 � severeypokinesia; 4 � akinesia; 5 � dyskinesia). Lefttrial and mitral geometric indexes were measuredn accordance with established methodological con-entions (26,31). Left atrial area was planimeteredn 4-chamber orientation, with volume calculatedased on area and chamber length (8/3� · area2/

length) from the posterior left atrium to the mitralannulus (31). Mitral indexes were measured in3-chamber orientation (26). Mitral annular diame-ter was measured in a linear plane extending fromthe respective junctions of anterior and posteriorvalve leaflets with the atrial wall. Coaptation depthwas defined as the distance between leaflet coap-tation and the mitral annulus. Tenting area en-compassed the area enclosed between the annulusand the mitral valve leaflets. Atrial and mitralannular indexes were measured during ventricular

Anterior

17

13

1

7

10

4

15

14

8 12

9

2

311

6

5

16

La

14

8

9

2

3

1. Basal Anterior

2. Basal Anteroseptal

3. Basal Inferoseptal

4. Basal Inferior

5. Basal Inferolateral

6. Basal Anterolateral

7. Mid Anterior

8. Mid Anteroseptal

9. Mid Inferoseptal

Figure 2. Infarct Distribution

Bullseye plots illustrating segments assigned to each regional infarcanterolateral/inferolateral segments; inferior � inferior/inferoseptalin green in respective bullseye plots), so that total myocardium sub

end systole.

ECHOCARDIOGRAPHY. Echocardiography was thereference for MR, which was quantified on the basisof regurgitant fraction (RF). The RF was calculatedon the basis of differential stroke volume (SV) ascalculated (VTI · �r2) using Doppler and 2D echoindexes acquired at the mitral and aortic valveannuli (RF � [SVmitral � SVaorta ] / SVmitral ·100%). Reproducibility of MR quantification basedon regurgitant fraction (32,33) as well as studycenter expertise for MR assessment (34–36) havebeen previously reported. Severity of MR wasgraded using established cutoffs in accordance withAmerican Society of Echocardiography guidelines(mild, RF �30%; moderate, 30% to 39%;moderate-severe, 40% to 49%; and severe, �50%)(28). Each respective modality (CMR, echo) wasinterpreted by an experienced reader blinded toclinical data and other imaging tests.Statistical methods. Comparisons of continuous vari-ables were made using Student t test (expressed asmean � SD) for 2 group comparisons, with adjunctiveuse of Levene test to assess for approximate equality ofvariance between groups. Analysis of variance wasused for multiple group comparisons. Categorical

al

12

11

6

5

16

Inferior

17

13

1

7

10

4

15

14

8 12

9

2

311

6

5

16

10. Mid Inferior

11. Mid Inferolateral

12. Mid Anterolateral

13. Apical Anterior

14. Apical Septal

15. Apical Inferior

16. Apical Lateral

17. Apex

tegory (anterior � anterior/anteroseptal segments; lateral �

ents). Each category was composed of 5 segments (highlightedded by each was equivalent.

ter

17

13

1

7

10

4

15

t casegm

variables were compared using Pearson chi-square or,

eKTg

rtss

Association; PMI � papilla


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


224

when �5 outcomes were expected per cell, Fisherxact test. Ordinal variables were compared usingruskal-Wallis 1-way analysis or the Jonckheere-empestra test, with the latter used to test between-

Conventional Imaging Characteristics in Relation to PMI

Overall (n � 153) PM

57 � 12

82% (126)

factors

46% (70)

mia 46% (70)

25% (38)

34% (52)

26% (40)

disease

5% (7)

1% (1)

ss, I/II/III/IV 82% (126)/14% (22)/3% (4)/1% (1) 78% (36)/

tions

97% (148)

inhibitor 94% (144)

5% (7)

7% (10)

100% (153)

95% (145)

26% (39)

ary intervention 75% (114)

nance

tion/geometry

% 52.5 � 11.9

0.8 � 0.6

5.5 � 0.5

ted 0.03 � 0.003 0

me, ml 153.1 � 42.4 1

me index, ml/m2 77.6 � 19.0

e, ml 75.5 � 37.4

e index, ml/m2 38.3 � 18.4

g 131.1 � 32.3 1

ndex, g/m2 66.4 � 14.2

r, cm 3.6 � 0.52 22.1 � 4.9

, ml 79.2 � 27.6

index, ml/m2 40.4 � 14.0

etry

, cm 2.8 � 0.4

, cm 0.6 � 0.2

0.9 � 0.4

% (n). Boldface indicates p values �0.05.bypass graft surgery; EDD � end-diastolic diameter; HMG CoA � 3-hydroxy-3-ry muscle infarction.

roup differences in graded MR severity. Logistic v

egression was used to evaluate multivariable associa-ions between imaging parameters and MR. Two-ided p �0.05 was considered indicative of statisticalignificance. Calculations were performed using SPSS

(n � 46) PMI� (n � 107) p Value

� 13 56 � 12 0.06

% (41) 79% (85) 0.15

% (23) 44% (47) 0.49

% (18) 49% (52) 0.28

% (12) 24% (26) 0.81

% (14) 36% (38) 0.54

% (13) 25% (27) 0.70

% (4) 3% (3) 0.20

% (1) 0% (0) 0.30

(7)/4% (2)/2% (1) 84% (90)/14% (15)/2% (2)/0% (0) 0.34

% (45) 96% (103) 1.00

% (44) 94% (100) 0.73

% (2) 5% (5) 1.00

% (4) 6% (6) 0.49

% (46) 95% (107) —

% (42) 96% (103) 0.24

% (12) 25% (27) 0.91

% (34) 75% (80) 0.91

� 11.8 52.6 � 12.0 0.89

� 0.6 0.8 � 0.6 0.45

� 0.5 5.5 � 0.5 0.02

� 0.003 0.032 � 0.003 0.04

� 39.2 148.7 � 43.2 0.049

� 17.6 76.0 � 19.4 0.11

� 37.2 73.2 � 37.4 0.25

� 18.3 37.4 � 18.5 0.38

� 33.5 128.8 � 31.6 0.17

� 14.7 65.8 � 14.0 0.42

� 0.6 3.6 � 0.5 0.91

� 5.4 22.0 � 4.7 0.87

� 31.2 79.0 � 26.0 0.90

� 15.0 40.7 � 13.6 0.67

� 0.4 2.8 � 0.4 0.046

� 0.2 0.6 � 0.2 0.82

� 0.4 0.9 � 0.4 0.64

ylglutaryl-coenzyme A; MI � myocardial infarction; NYHA � New York Heart

Table 1. Clinical and

I�

Clinical

Age, yrs 60

Male 89

Atherosclerosis risk

Hypertension 50

Hypercholesterole 39

Diabetes mellitus 26

Tobacco use 30

Family history 28

Prior coronary artery

PCI 9

CABG 2

NYHA functional cla 15%

Cardiovascular medica

Beta-blocker 98

HMG CoA-reductase 96

Loop diuretic 4

Nitroglycerin 9

Aspirin 100

Thienopyridines 91

MI treatment strategy

Thrombolysis 26

Percutaneous coron 74

Cardiac magnetic reso

Left ventricular func

Ejection fraction, 52.3

Wall motion score 0.9

EDD, cm 5.7

EDD, height adjus .033

End-diastolic volu 63.4

End-diastolic volu 81.4

End-systolic volum 80.8

End-systolic volum 40.3

Myocardial mass, 36.7

Myocardial mass i 67.8

Left atrial geometry

Left atrial diamete 3.6

Left atrial area, cm 22.2

Left atrial volume 79.6

Left atrial volume 39.7

Mitral annular geom

Annulus diameter 2.9

Coaptation height 0.6

Tenting area, cm2 0.9

Values are mean � SD orCABG � coronary artery meth

ersion 19.0 (SPSS, Chicago, Illinois).

wo

ds(cpRL

tlivpD3wit4

oaa


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


225

R E S U L T S

The population consisted of 153 patients withAMI. CMR was performed 27 � 8 days after AMI(90% �14 days). Echocardiography was performedwithin 1 day of CMR in all patients (95% within 4 h).

Papillary muscle infarction was present in 30%(n � 46) of patients, among whom 72% had postero-medial involvement and 39% had anterolateral papil-lary involvement. Concomitant involvement of bothpapillary muscles was present in 11% (5 of 46) ofaffected patients. Figure 1 provides representativeexamples of PMI identified by DE-CMR.

Table 1 details clinical and conventional imagingcharacteristics of the study population, with strati-fication on the basis of presence or absence of PMI.Patients with PMI were clinically similar to patientswithout PMI. Regarding imaging, results demon-strate that PMI was associated with more advancedremodeling based on LV chamber size and mitralannular diameter (p � 0.05).Infarct parameters. Among the total of 46 patients

ith PMI, nearly half (48%) occurred in the contextf RCA culprit-vessel infarction, with the remain-

Table 2. Infarct Size and Distribution

PMI

Present(n � 46)

Absent(n � 107)

Infarct size

DE-CMR

% LV hyperenhancement 16.0 � 10.9 12.3 � 8.9

Cardiovascular enzymes

Creatine phosphokinase 2,590 � 2,344 2,164 � 1,83

Creatine phosphokinase-MB 243 � 207 199 � 188

Duration of symptoms

Chest pain interval, h 12.4 � 9.4 10.2 � 8.4

Infarct distribution

DE-CMR

Anterior wall 35% (16) 70% (75)

Lateral wall 65% (30) 22% (23)

Inferior wall 72% (33) 45% (48)

Anterior � inferior wall 15% (7) 20% (21)

Lateral � anterior wall 17% (8) 12% (13)

Lateral � inferior wall 50% (23) 11% (12)

Lateral � anterior � inferior wall 9% (4) 4% (4)

Angiography, infarct-related artery

Left anterior descending 28% (13) 72% (77)

Left circumflex 24% (11) 6% (6)

Right coronary 48% (22) 22% (24)

Anatomically dominant artery† 57% (26) 26% (28)

Values are mean � SD or % (n). Boldface indicates p values �0.05. *5 subjects hacoronary artery.
CMR � cardiac magnetic resonance; DE � delayed enhancement; LV � left ventri
er near evenly divided between left anterior de-cending artery (LAD [28%]) and left circumflexLCX [24%]) involvement. When stratified byulprit vessel, PMI occurred in 65% (11 of 17) ofatients with LCX, 48% (22 of 46) of patients withCA, and only 14% (13 of 90) of patients withAD infarcts (p �0.001).Table 2 examines infarct-related parameters in rela-

ion to PMI. As shown, patients with PMI were moreikely to have angiography-evidenced LCX culprit vesselnvolvement (24% vs. 6%, p � 0.001) or RCA culpritessel involvement (48% vs. 22%, p � 0.002) thanatients without PMI. Regarding infarct distribution onE-CMR, prevalence of lateral wall infarction was

-fold higher among patients with, compared to patientsithout, PMI (65% vs. 22%, p � 0.001). Prevalence of

nferior wall infarction was 1.6-fold higher among pa-ients with, compared to patients without, PMI (72% vs.5%, p � 0.003).

Regarding infarct size, Table 2 demonstrates thatverall % LV infarction by DE-CMR was largermong patients with PMI (p � 0.03), with parallellbeit nonsignificant relationships between PMI

Posteromedial PMI Antero

p ValuePresent(n � 33)*

Absent(n � 120) p Value

Present(n � 18)* (

0.03 15.4 � 10.8 12.9 � 9.2 0.19 20.2 � 13.3 1

0.25 2,351 � 1,782 2,271 � 2,057 0.85 2,962 � 3,118 2,2

0.30 256 � 206 199 � 189 0.21 191 � 200 2

0.17 12.4 � 9.6 10.4 � 8.5 0.29 13.7 � 9.6 1

<0.001 12% (4) 73% (87) <0.001 78% (14)

<0.001 73% (24) 24% (29) <0.001 61% (11)

0.003 91% (30) 43% (51) <0.001 44% (8)

0.52 12% (4) 20% (24) 0.30 28% (5)

0.39 6% (2) 16% (19) 0.25 44% (8)

<0.001 64% (21) 12% (14) <0.001 39% (7)

0.24 6% (2) 5% (6) 0.68 22% (4)

<0.001 0% (0) 75% (90) <0.001 72% (13)

0.001 33% (11) 5% (6) <0.001 22% (4)

0.002 67% (22) 20% (24) <0.001 6% (1)

<0.001 79% (26) 23% (28) <0.001 6% (1)

ncomitant posteromedial and anterolateral papillary muscle infarction (PMI). †Eith

lateral PMI

Absentn � 135) p Value

2.5 � 8.7 0.03

6 01 � 1,794 0.36

16 � 195 0.70

0.5 � 8.6 0.15

57% (77) 0.09

31% (42) 0.01

54% (73) 0.44

17% (23) 0.33

10% (13) 0.001

21% (28) 0.13

3% (4) 0.007

57% (77) 0.22

10% (3) 0.12

33% (45) 0.02

39% (53) 0.005

d co er left circumflex or right

cular; MB � myocardial band.

i(teai

cc3ociswP(mi

taFtca

mtcswio(g

o1aaa


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


226

and the enzymatic and clinical indexes of peakcreatine phosphokinase and chest pain duration(both p � NS). However, the relation betweenPMI and infarct size varied according to culpritvessel involvement. Among patients with LADinfarcts, LV infarct size (% myocardium) was sim-lar between patients with and without PMI18 � 11% vs. 15 � 9%, p � 0.32). In contradistinc-ion, Figure 3A illustrates that among patients withither RCA or LCX culprit involvement, PMI wasccompanied by a more than 2-fold increase in LVnfarct size (p � 0.001 and p � 0.056, respectively).

Although both RCA and LCX infarction in-reased relative likelihood for PMI, the impact oforonary dominance varied by culprit vessel. FigureB first stratifies both RCA and LCX infarcts basedn PMI and then stratifies patients with PMI byoronary dominance; among patients with RCAnfarcts, PMI exclusively occurred (100%) in theetting of right or codominant coronary anatomy,hereas less than one-half (36%) of patients withMI and LCX infarcts were left or codominant

p � 0.0001). All RCA infarcts with PMI had postero-edial involvement (n � 22); 50% (11 of 22) had

nfarcts of the adjacent mid inferolateral wall.

LV I

nfa

rct

Siz

e (%

)

Right Coronary Artery

PapillaryMuscleInfarction

CoronaryDominance

p = 0.001

A

B

PMI -0

10

20

30

40

50

PMI +

46

52%24/46

–

0%0/22

Non-Dominant

100%22/22

Domina

48%22/46

+

Figure 3. Infarct Size and Coronary Anatomy

(A) Infarct size (mean � SD) stratified by papillary muscle infarction(right) left circumflex (LCX) culprit vessels. (B) Stratification of RCA
nance pattern (lower row). LV � left ventricular.
PMI extent and LV infarct transmurality. In more thanhree-fourths of cases (76%), PMI was partial,s defined by �50% papillary hyperenhancement.igure 4 compares PMI type by anatomic distribu-

ion, demonstrating that PMI type (partial vs.omplete) did not differ between the anterolateralnd posteromedial papillary muscles (p � 0.50).

Figure 5 relates PMI type to both infarct trans-urality and severity of contractile dysfunction in

he adjoining LV wall. As shown (Fig. 5A), in-reased extent of PMI was accompanied by atepwise increase in mean infarct transmuralityithin adjacent LV segments underlying each pap-

llary muscle (p � 0.001). Similar results werebtained (Fig. 5B) when contractile dysfunctionsegmental wall motion score) was used as a surro-ate marker for injury (p � 0.001).

Mitral regurgitation. Among the total study popula-tion, echo-evidenced MR was mild or less in mostcases, with greater severity present in 14% (n � 22)f patients (moderate, n � 7; moderate-severe, n �2; severe, n � 3). Table 3 reports univariatenalyses examining CMR, biomarker, and x-rayngiography parameters in relation to presence orbsence of substantial (moderate or greater) MR.

< 0.001

Left Circumflex Artery

p = 0.06

PMI - PMI +

17

65%11/17

+

36%4/11

Dominant

64%7/11

Non-Dominant

35%6/17

–

I) among patients with (left) right coronary artery (RCA) andLCX infarcts by presence of PMI (upper row) and coronary domi-

p

nt

(PMand


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


227

As shown, patients with moderate or greater MRhad worse LV dysfunction and more advanced LV,left atrial, and mitral annular remodeling (p �0.001) than patients with lesser or absent MR.Regarding infarct pattern, Table 3 also demon-strates that patients with moderate or greater MRhad greater prevalence of lateral wall infarction on

65%33/51

PMI

Posteromedial

27%9/33

Complete

73%24/33

Partial

PMIDistribution

PMIType

Figure 4. PMI Location and Type

Stratification of papillary muscle infarction (PMI) by location (upperamong 46 patients (n � 5 with bilateral PMI).

Mea

n I

nfa

rct T

ran

smu

rali

ty

A

0

20

40

60

80

100

No PMI

p < 0.001

Partial PMI Complete PMI

Mea

n W

all

Mo

tio

n S

core

B

0

1

2

3

4

No PMI

p < 0.001


Figure 5. PMI in Relation to Left Ventricular Injury

(A) Infarct transmurality and (B) contractile dysfunction stratified byscore (left) and maximum infarct score (right) in left ventricular (LVinferior/inferolateral segments; anterolateral � mid anterior/anterol
of injury to adjacent LV segments, whether assessed by infarct transmu
DE-CMR (p � 0.001), with nonsignificant differ-ences in infarct-related artery distribution poten-tially attributable to variability in lateral wall coro-nary vascular supply. Among patients with infarctsthat encompassed �1 LV region, only the combi-nation of lateral plus anterior wall infarction wasmore common among patients with increased MR,

51

0.50

83%15/18

Partial17%3/18

Complete

35%18/51

Anterolateral

) and type (lower row). Analysis based on total PMI (n � 51)M

axim

um

In

farc

t Tra

nsm

ura

lity

0

20

40

60

80

100

No PMI

p < 0.001


Max

imu

m W

all

Mo

tio

n S

core

0

1

2

3

4

No PMI

p < 0.001


illary muscle infarction (PMI). Bar graphs based on mean infarctgments adjacent to each papillary muscle (posteromedial � midal wall segments). Note parallels between PMI extent and severity

p =

row

pap) seater

rality or contractile dysfunction (all p � 0.001 for trend).

a11

oD

wwP(


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


228

consistent with the fact that patients with moderateor greater MR had larger overall LV infarct size(both p � 0.01). In multivariable analysis, MR wasindependently associated with presence of lateralwall infarction (odds ratio [OR] 3.86 [95% confi-dence interval (CI): 1.40 to 10.69], p � 0.009) evenfter controlling for overall LV infarct size (OR.07/% LV hyperenhancement [95% CI: 1.02 to.12], p � 0.01).Figure 6 stratifies graded MR severity based

n PMI extent and LV infarct distribution on

Table 3. Cardiac Structure, Function, and Infarct Pattern in Rela

MR�

Mitral annular geometry

Annulus diameter, cm 3.1

Coaptation height, cm 0.7

Tenting area, cm2 1.2

Left atrial geometry

Left atrial diameter, cm 4.2

Left atrial area, cm2 25.7

Left atrial volume, ml 98.8

Left atrial volume index, ml/m2 49.6

Left ventricular function/geometry

Ejection fraction, % 42.7

Wall motion score 1.4

End-diastolic diameter, cm 6.0

End-diastolic volume, ml 191.9

End-diastolic volume index, ml/m2 97.1

End-systolic volume, ml 113.8

End-systolic volume index, ml/m2 57.8

Myocardial mass, g 140.1

Myocardial mass index, g/m2 71.1

Left ventricular infarct pattern

Infarct size

% LV hyperenhancement 20.3

Creatine phosphokinase 2730

Creatine phosphokinase-MB 270

Infarct distribution (DE-CMR)

Anterior wall 64

Lateral wall 68

Inferior wall 59

Anterior � inferior wall 27

Lateral � anterior wall 41

Lateral � inferior wall 36

Lateral � anterior � inferior wall 14

PMI (DE-CMR) 41

Infarct-related artery (x-ray angiography)

Left anterior descending 55

Left circumflex 18

Right coronary 27

Values are mean � SD or % (n). Boldface indicates p values �0.05. *StratifiedMR � mitral regurgitation; other abbreviations as in Tables 1 and 2.

E-CMR—corresponding data are detailed in tab- s

ular form below each bar graph. PMI, categorized byextent of papillary muscle necrosis (p � 0.12) or bybinary presence/absence (p � 0.19), was not associ-ated with increased MR severity. However, LV infarctdistribution did impact MR, as evidenced by increasedMR severity in association with presence of lateral wallinfarction (p � 0.002). Of note, lateral wall infarct size

as more than 3-fold larger among patients (n � 5)ith concomitant anterolateral and posteromedialMI compared to patients with absent or isolated PMI

p � 0.001), paralleling a trend toward increased MR

to Mitral Regurgitation*

22) MR� (n � 131) p Value

.4 2.8 � 0.4 <0.001

.2 0.6 � 0.2 0.001

.5 0.8 � 0.3 <0.001

.6 3.6 � 0.5 <0.001

.7 21.6 � 4.5 <0.001

3.7 75.9 � 25.1 <0.001

6.5 38.9 � 13.0 0.001

2.2 54.2 � 11.0 <0.001

.8 0.7 � 0.5 <0.001

.6 5.5 � 0.7 <0.001

6.7 146.6 � 38.1 <0.001

5.7 74.3 � 15.4 <0.001

0.6 69.1 � 30.5 0.001

7.2 35.0 � 14.2 0.001

2.6 129.6 � 32.0 0.16

8.4 65.7 � 13.3 0.096

2.8 12.3 � 8.6 0.009

009 2215 � 1995 0.29

80 205 � 196 0.28

4) 59% (77) 0.67

5) 29% (38) <0.001

3) 52% (68) 0.53

) 17% (22) 0.24

) 9% (12) 0.001

) 21% (27) 0.10

) 4% (5) 0.09

) 28% (37) 0.23

2) 60% (78) 0.66

) 10% (13) 0.21

) 31% (40) 0.76

he basis of moderate or greater mitral regurgitation threshold.

tion

(n �

� 0

� 0

� 0

� 0

� 5

� 3

� 1

� 1

� 0

� 0

� 4

� 2

� 5

� 2

� 3

� 1

� 1

� 2

� 1

% (1

% (1

% (1

% (6

% (9

% (8

% (3

% (9

% (1

% (4

% (6

on t

everity (p � 0.09) in this group.


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


229

Logistic regression analysis was also used to test theassociation of infarct distribution with MR after control-ling for conventional structural indexes. In the univariateanalysis, presence of lateral wall infarction (OR 1.21/%LV myocardium [95% CI: 1.08 to 1.37], p � 0.001) was

Mit

ral

Reg

urg

itat

ion

(pre

vala

nce

)

No PMI0%

50%

100%

Partial PMI

p = 0.12

Mit

ral

Reg

urg

itat

ion

(pre

vala

nce

)

Lateral Wall

0%

50%

100% p = 0.002

Absent Present

InfAbse

Absent Mild

(RF < 30%

Absent 66% (66) 27% (27)

Present 45% (24) 26% (14)

Absent 63% (45) 25% (18)

Present 56% (45) 28% (23)

Absent 60% (37) 27% (17)

Present 58% (53) 26% (24)

Absent 62% (66) 26% (28)

Present 60% (21) 23% (8)

Complete 27% (3) 46% (5)

Mi

Lat

eral

Infe

rio

rA

nte

rio

rLV I

nfa

rct

Dis

trib

uti

on

Pap

illa

ryM

usc

leIn

farc

tio

nA

B

Figure 6. Mitral Regurgitation Severity

Graded severity of echocardiography-quantified mitral regurgitation(bottom) left ventricular (LV) infarct distribution. Dark blue indicatemoderate; and yellow indicates mild. The corresponding table provRF � regurgitant fraction.

significantly associated with substantial (moderate or

greater) MR, whereas PMI was not (OR 1.76 [95% CI:0.69 to 4.46], p � 0.24). Multivariate modeling was usedto test the independent association of lateral wall infarc-tion with MR after controlling for conventional geomet-ric indexes of mitral annular and LV size. As shown in

Mild

Moderate

Moderate - Severe

Severe

Mitral Regurgitation

omplete PMI

or Wall

0.36

Present

Anterior Wall

p = 0.80

Absent Present

Moderate- Moderate Severe Severe

(RF = 30-39%) (RF = 39-49%) (RF 50%)

3% (3) 4% (4) –

8% (4) 15% (8) 6% (3)

4% (3) 7% (5) 1% (1)

5% (4) 9% (7) 3% (2)

3% (2) 8% (5) 2% (1)

6% (5) 8% (7) 2% (2)

5% (5) 7% (7) 1% (1)

6% (2) 6% (2) 6% (2)

– 27% (3) –

l Regurgitation Severity

R) stratified by (top) papillary muscle infarction (PMI) andvere MR; green indicates moderate to severe; orange indicatesa breakdown of MR severity in relation to PMI and LV infarction.

C

eri

p =

nt

)

tra

(Ms seides

Table 4, lateral wall infarct size (OR 1.20/% LV myo-

w1ttdwaai00tv


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


230

cardium [95% CI: 1.05 to 1.39], p � 0.01) and mitralannular diameter (OR 1.22/mm [95% CI: 1.05 to 1.45],p � 0.009) were independently associated with MR,

ith a similar trend for LV chamber diameter (OR.11/mm [95% CI: 0.99 to 1.23], p � 0.056). Substitu-ion of lateral wall motion score (OR 1.25 [95% CI: 1.08o 1.45], p � 0.003) in the model (Table 4) alsoemonstrated an independent association between lateralall dysfunction and MR. In contradistinction, use of

ngiography data did not demonstrate an independentssociation for LCX (Table 4) culprit involvement. Sim-lar results were obtained when RCA (OR 0.68 [95% CI:.21 to 2.18], p � 0.52) or LAD (OR 0.90 [95% CI:.32 to 2.51], p � 0.83) culprit involvement was substi-uted in the model, consistent with overlap in coronaryascular supply to the lateral wall.

D I S C U S S I O N

This study provides new insights regarding preva-lence and physiologic consequences of infarctionwithin 2 components of the mitral valve apparatus:

Table 4. Structural Correlates of Mitral Regurgitation

Lateral W

Univariabl

Odds Ratio (95%

Lateral wall infarct size, % myocardium 1.21 (1.08–1.37)

Mitral annular diameter, mm 1.29 (1.13–1.48)

LV end-diastolic diameter, mm 1.20 (1.09–1.33)

PMI, complete or partial 1.76 (0.69–4.46)

Complete PMI 2.43 (0.59–9.97)

Partial PMI 1.57 (0.59–4.21)

Lateral Wa

Univariable R

Odds Ratio (95% CI)

Lateral wall contractile dysfunction(wall motion score)

1.27 (1.13–1.43)



Infarct-Re

Univariable Reg

Odds Ratio (95% CI)

Left circumflex infarction 2.02 (0.59–6.87)



Boldface indicates p values �0.05.CI � confidence interval; other abbreviations as in Tables 1 and 2.

papillary muscles and the adjacent LV wall. Key

findings are as follows. 1) PMI is common in thecurrent era, affecting nearly one-third (30%) ofpatients, with relative likelihood greater in thecontext of LCX or dominant RCA infarcts. 2) PMItypically occurs in the context of increased LVinfarct size, as evidenced by a greater than 2-foldincrease in infarct size among affected patients withLCX or RCA infarcts. 3) Lateral wall infarction,rather than presence or extent of PMI, is associatedwith increased MR as measured early after AMI. Inmultivariate analysis, post-AMI MR is associatedwith lateral wall infarct size even after controllingfor both mitral annular and LV chamber dilation.

Our findings provide new insight into the rela-tion between PMI and MR. Whereas prior studieshave reported similar PMI prevalence, results havebeen discordant regarding PMI as a cause of MR.For example, whereas Tanimoto et al. (25) foundno association between PMI and MR, Okayama etal. (26) reported that the 2 were linked. Interest-ingly, in the latter study, infarction of both papillary

nfarction

gressionMultivariable Regression

Model �2 � 29.37, p <0.001

p Value Odds Ratio (95% CI) p Value

0.001 1.20 (1.05–1.39) 0.01

<0.001 1.22 (1.05–1.45) 0.009

<0.001 1.11 (0.99–1.23) 0.056

0.24

0.22

0.37

ntractility

essionMultivariable Regression

Model �2 � 32.23, p <0.001


<0.001 1.25 (1.08–1.45) 0.003

<0.001 1.24 (1.05–1.45) 0.009

<0.001 1.07 (0.95–1.21) 0.25

d Artery

sionMultivariable Regression

Model �2 � 23.99, p <0.001


0.26 2.26 (0.59–8.72) 0.24

<0.001 1.20 (1.03–1.39) 0.02

<0.001 1.14 (1.03–1.27) 0.01

all I

e Re

CI)

ll Co

egr

late

res

muscles was associated with increased MR whereas


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


231

single PMI was not. Our results suggest that LVinfarct pattern may explain these variable findings.Among our population, only lateral wall infarctionwas associated with increased MR, and the associ-ation between lateral wall infarct size and MRremained significant even after controlling for PMIas well as conventional indexes of LV and mitralgeometry. PMI varied in both distribution andextent, with greater magnitude of papillary necrosisassociated with increased infarct transmuralitywithin the adjacent LV wall, suggesting that priorreported associations between bilateral PMI andMR may be attributable to increased lateral wallinfarct size. Consistent with our results, an associ-ation between lateral wall infarction and MR wasreported among patients with advanced LV dys-function (LV ejection fraction �40%) (23), al-though this prior study did not assess PMI, raisingquestions as to underlying reasons for MR. To thebest of our knowledge, no prior study has examinedthe inter-relationship of PMI and LV infarct dis-tribution on post-AMI MR.

Our finding that PMI has a negligible impact onMR is consistent with prior animal studies thathave examined papillary physiology. Using a caninemodel in which coronary flow was reduced in agraded fashion, Matsuzaki et al. (37) found thatpapillary dysfunction alone was insufficient to causeMR, which was instead associated with LV free(lateral) wall hypokinesis and chamber dilation. Theimportance of lateral wall dysfunction was alsodemonstrated by Messas et al. (38), who performedtargeted LCX occlusion in a sheep model and foundthat LV inferobasal ischemia produced MR,whereas the addition of papillary ischemia paradox-ically decreased MR by reducing leaflet tetheringand improving valve coaptation. Regarding mitralgeometry, animal studies by Kaul et al. (39) foundthat MR was associated with incomplete mitralvalve closure rather than papillary ischemia. Clinicalstudies have also supported this concept. Uemura etal. (40), studying 40 patients with prior MI, re-ported that papillary contractile dysfunction wasassociated with induction of leaflet tethering andMR attenuation, consistent with our multivariableanalysis demonstrating that extent of PMI was notassociated with increased MR. Taken together,these prior data support our observation that lateralwall injury and impaired valve coaptation, ratherthan papillary dysfunction, are key determinants ofpost-AMI MR.

It is important to recognize that animal studies
that have assessed LV injury patterns in vivo have
typically relied on LV or papillary contractile dys-function as surrogates for infarction (37–39), withquestions remaining as to the direct impact ofmyocyte necrosis on mitral valve function. Otherstudies, which have assessed infarct size directly,have demonstrated the importance of lateral wallinfarction using post-mortem analysis, raising un-certainty as to the impact of intrinsic differencesbetween in vivo assessment of mitral geometry/MRand ex-vivo infarct quantification (18,41). To ad-dress these issues in a clinical setting, our studyemployed multimodality imaging for in vivo assess-ment of infarct size, mitral valve geometry, andMR. The CMR and echo were acquired using atailored imaging protocol, with both tests per-formed within a 1-day interval. Results, obtained ina broad cohort of post-AMI patients, offer clinicalevidence that both impaired mitral valve geometry(annular diameter) and lateral wall infarct size (%hyperenhancement) are independently associatedwith post-AMI MR, with a lesser, albeit signifi-cant, magnitude of association for LV chambergeometry as measured by end-diastolic diameter.

As DE-CMR provides a highly sensitive tool forassessment of infarct morphology, an additionalstudy goal was to examine the relationship betweeninfarct burden within the papillary muscles andadjacent LV wall. Prior DE-CMR studies, whichhave also reported increased prevalence of PMI inassociation with RCA and LCX culprit vesselinvolvement, have categorized PMI and LV infarctdistribution in a binary fashion (25,26). Our studyexamined graded extent of both papillary and LVwall infarction. Results demonstrate that papillarynecrosis occurs in parallel with infarction of theadjacent LV wall, as evidenced by increased LVinfarct size in association with greater extent ofPMI. Rather than the notion of papillary muscles asend-organ structures particularly susceptible to jeop-ardized arterial supply, our results support the conceptthat papillary muscles and LV myocardium are simi-larly vulnerable to impaired perfusion. Taken togetherwith our findings linking lateral wall infarct size toMR, these data emphasize the importance of promptcoronary reperfusion towards the common goals ofpreserving LV viability, preventing adverse remodel-ing, and minimizing MR after AMI.

In addition to CMR and echo, our study in-cluded analysis of x-ray angiography for assessmentof mitral apparatus perfusion patterns. Results dem-onstrated that right and left coronary dominancebore a variable impact on PMI. Among patients
with RCA infarcts, all with PMI were right (or co-)


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


232

dominant whereas for patients with LCX infarcts,less than half (36%) with PMI were left (or co-)dominant (p � 0.001). Additionally, we observed anearly 2-fold higher prevalence of posteromedial(72%) compared to anterolateral (39%) PMI. Thisdifference may be explained by the fact that theposteromedial papillary muscle is typically suppliedby a single coronary artery, whereas the anterolat-eral papillary muscle is more frequently concomi-tantly perfused by 2 coronary arteries (42). Regard-ing MR, infarct-related culprit vessel was notindependently associated with valvular regurgita-tion, consistent with the concept of variable arterialsupply to the lateral wall.Study limitations. Several limitations should be rec-ognized. First, despite inclusion of a broad cohort ofpost-AMI patients, relatively few (14%) had sub-stantial (moderate or greater) MR, consistent withprior population based studies that have examinedMR after AMI (43,44). This limited our ability toexamine the independent contributions of all struc-tural variables in relation to MR, and may haveresulted in some degree of overfitting in our mul-tivariate models, which tested infarct distribution inrelation to the 2 established indexes (mitral annular,left ventricular size) conventionally associated withMR. Additionally, as very few affected patients hadsevere MR, structural variables examined in thisstudy could not be correlated with graded severity ofvalvular regurgitation. Although MR severity wasmeasured based on regurgitant fraction, it isimportant to recognize that other well-validatedmethods for MR assessment, including effectiveregurgitant orifice area, were not tested. Finally,as MR was assessed within 6 weeks (27 � 8 days)of AMI, current findings may not apply tolonger-term severity of post-AMI MR. Indeed,MR can dynamically change over time and be

lation 1997;96:827–33. Popovic AD. Early

therapy. Additional research is necessary to ex-amine the independent utility of mitral apparatusinfarct pattern—including PMI—for predictinglong-term MR severity.

C O N C L U S I O N S

In summary, this study demonstrates the inter-relationship between papillary muscle and regionalLV infarction and identifies lateral wall infarction(rather than PMI) as an independent marker forincreased MR as measured early after AMI. This isthe first in vivo study to directly link regional LVinfarct size and, by extension, infarct transmuralityto MR. These findings are relevant to the design oftherapies to treat MR, and to the tailoring of theirapplication to individual patients. Applied clini-cally, findings suggest that LV plication therapies(45) might be considered for CAD patients withsevere MR and transmural infarction, whereas cor-onary reperfusion might be more beneficial as aprimary strategy for patients with preserved viabilityin the lateral wall. DE-CMR also has the potentialto guide interventional treatments for MR, such astargeted application of papillary repositioning(46,47) based on infarct distribution. More broadly,improved predictive models for MR are clinicallyimportant to better identify and promptly treat at-riskpatients before development of adverse consequencessuch as cardiac chamber remodeling, heart failure, andarrhythmias. Further investigation is needed to testlong-term predictors of post-AMI MR, as well as theutility of DE-CMR guided strategies to prevent ortreat MR and its complications.

Reprint requests and correspondence: Dr. Jonathan W.Weinsaft, Weill Cornell Medical College, Starr-4, 525East 68th Street, New York, New York 10021. E-mail:

influenced by LV remodeling as well as medical [email protected].

R E F E R E N C E S

1. Amigoni M, Meris A, Thune JJ, et al.Mitral regurgitation in myocardial in-farction complicated by heart failure, leftventricular dysfunction, or both: prog-nostic significance and relation to ven-tricular size and function. Eur Heart J2007;28:326–33.

2. Lamas GA, Mitchell GF, Flaker GC,et al, for the Survival and VentricularEnlargement Investigators. Clinicalsignificance of mitral regurgitation af-ter acute myocardial infarction. Circu-

3. Grigioni F, Enriquez-Sarano M, ZehrKJ, Bailey KR, Tajik AJ. Ischemic mi-tral regurgitation: long-term outcomeand prognostic implications with quan-titative Doppler assessment. Circulation2001;103:1759–64.

4. Ennezat PV, Darchis J, Lamblin N, etal. Left ventricular remodeling is as-sociated with the severity of mitralregurgitation after inaugural anteriormyocardial infarction—optimal tim-ing for echocardiographic imaging.Am Heart J 2008;155:959–65.

5. Neskovic AN, Marinkovic J, Bojic M,
predictors of mi-
tral regurgitation after acute myocar-dial infarction (abstr). Am J Cardiol1999;84:329–32.

6. Beeri R, Yosefy C, Guerrero JL, et al.Early repair of moderate ischemic mi-tral regurgitation reverses left ventric-ular remodeling: a functional and mo-lecular study. Circulation 2007;116Suppl:I288–93.

7. Beeri R, Yosefy C, Guerrero JL, etal. Mitral regurgitation augmentspost-myocardial infarction remodel-ing failure of hypertrophic compen-sation. J Am Coll Cardiol 2008;51:
476 – 86.
mailto:[email protected]

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


233

8. Stevens LM, Basmadjian AJ, BouchardD, et al. Late echocardiographic andclinical outcomes after mitral valve re-pair for degenerative disease. J CardSurg 2010;25:9–15.

9. Kongsaerepong V, Shiota M, GillinovAM, et al. Echocardiographic predic-tors of successful versus unsuccessfulmitral valve repair in ischemic mitralregurgitation. Am J Cardiol 2006;98:504–8.

10. Go AS, Hylek EM, Phillips KA, et al.Prevalence of diagnosed atrial fibrilla-tion in adults: national implicationsfor rhythm management and strokeprevention. The Anticoagulation andRisk Factors in Atrial Fibrillation(ATRIA) study. JAMA 2001;285:2370–5.

11. Wolf PA, Mitchell JB, Baker CS,Kannel WB, D’Agostino RB. Impactof atrial fibrillation on mortality,stroke, and medical costs. Arch InternMed 1998;158:229–34.

12. Perloff JK, Roberts WC. The mitralapparatus. Functional anatomy of mi-tral regurgitation. Circulation 1972;46:227–39.

13. Roberts WC, Cohen LS. Left ventric-ular papillary muscles. Description ofthe normal and a survey of conditionscausing them to be abnormal. Circu-lation 1972;46:138–54.

14. Vlodaver Z, Edwards JE. Rupture ofventricular septum or papillary musclecomplicating myocardial infarction.Circulation 1977;55:815–22.

15. Wei JY, Hutchins GM, Bulkley BH.Papillary muscle rupture in fatal acutemyocardial infarction: a potentiallytreatable form of cardiogenic shock.Ann Intern Med 1979;90:149–52.

16. Nishimura RA, Schaff HV, Shub C,Gersh BJ, Edwards WD, Tajik AJ.Papillary muscle rupture complicatingacute myocardial infarction: analysisof 17 patients. Am J Cardiol 1983;51:373–7.

17. Hider CF, Taylor DE, Wade JD. Theeffect of papillary muscle damage onatrioventricular valve function in theleft heart. Q J Exp Physiol Cogn MedSci 1965;50:15–22.

18. Mittal AK, Langston M Jr., CohnKE, Selzer A, Kerth WJ. Combinedpapillary muscle and left ventricularwall dysfunction as a cause of mitralregurgitation. An experimental study.Circulation 1971;44:174–80.

19. Tsakiris AG, Rastelli GC, AmorimDde S, Titus JL, Wood EH. Effect ofexperimental papillary muscle damageon mitral valve closure in intact anes-thetized dogs. Mayo Clin Proc 1970;45:275–85.

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

1. Fieno DS, Kim RJ, Chen EL, Lomas-ney JW, Klocke FJ, Judd RM.Contrast-enhanced magnetic reso-nance imaging of myocardium at risk:distinction between reversible and ir-reversible injury throughout infarcthealing. J Am Coll Cardiol 2000;36:1985–91.

2. Amado LC, Gerber BL, Gupta SN, etal. Accurate and objective infarct siz-ing by contrast-enhanced magneticresonance imaging in a canine myo-cardial infarction model. J Am CollCardiol 2004;44:2383–9.

3. Srichai MB, Grimm RA, StillmanAE, et al. Ischemic mitral regurgita-tion: impact of the left ventricle andmitral valve in patients with left ven-tricular systolic dysfunction. AnnThorac Surg 2005;80:170–8.

4. D’Ancona G, Biondo D, Mamone G,et al. Ischemic mitral valve regurgita-tion in patients with depressed ven-tricular function: cardiac geometricaland myocardial perfusion evaluationwith magnetic resonance imaging.Eur J Cardiothorac Surg 2008;34:964–8.

5. Tanimoto T, Imanishi T, Kitabata H,et al. Prevalence and clinical signifi-cance of papillary muscle infarctiondetected by late gadolinium-enhancedmagnetic resonance imaging in pa-tients with ST-segment elevationmyocardial infarction. Circulation2010;122:2281–7.

6. Okayama S, Uemura S, Soeda T, et al.Clinical significance of papillary mus-cle late enhancement detected via car-diac magnetic resonance imaging inpatients with single old myocardialinfarction. Int J Cardiol 2011;146:73–9.

7. Nguyen TD, Spincemaille P, Wein-saft JW, et al. A fast navigator-gated3D sequence for delayed enhancementMRI of the myocardium: comparisonwith breathhold 2D imaging. J MagnReson Imaging 2008;27:802–8.

8. Zoghbi WA, Enriquez-Sarano M,Foster E, et al. Recommendationsfor evaluation of the severity of na-tive valvular regurgitation with two-dimensional and Doppler echocardi-ography. J Am Soc Echocardiogr2003;16:777– 802.

9. Flett AS, Hasleton J, Cook C, et al.Evaluation of techniques for the quan-tification of myocardial scar of differ-ing etiology using cardiac magneticresonance. J Am Coll Cardiol Img2011;4:150–6.

0. Sievers B, Elliott MD, Hurwitz LM, etal. Rapid detection of myocardial infarc-tion by subsecond, free-breathing de-
layed contrast-enhancement cardiovas-
cular magnetic resonance. Circulation2007;115:236–44.

1. Lang RM, Biereg M, Devereux RM,et al. Recommendations for chamberquantification: a report from theAmerican Society of Echocardiogra-phy’s Guidelines and Standards Com-mittee and the Chamber Quantifica-tion Writing Group. J Am SocEchocardiogr 2005;18:1440–63.

2. Enriquez-Sarano M, Bailey KR,Seward JB, Tajik AJ, Krohn MJ, MaysJM. Quantitative Doppler assessmentof valvular regurgitation. Circulation1993;87:841–8.

3. Schwammenthal E, Chen C, BenningF, Block M, Breithardt G, LevineRA. Dynamics of mitral regurgitantflow and orifice area. Physiologic ap-plication of the proximal flow conver-gence method: clinical data and exper-imental testing. Circulation 1994;90:307–22.

4. Jones EC, Devereux RB, RomanMJ, et al. Prevalence and correlates ofmitral regurgitation in a population-based sample (the Strong Heart Study).Am J Cardiol 2001;87:298–304.

5. Kontos J, Papademetriou V, WachtellK, et al. Impact of valvular regurgita-tion on left ventricular geometry andfunction in hypertensive patients withleft ventricular hypertrophy: the LIFEstudy. J Hum Hypertens 2004;18:431–6.

6. Palmieri V, Bella JN, Arnett DK, etal. Associations of aortic and mitralregurgitation with body compositionand myocardial energy expenditure inadults with hypertension: the Hyper-tension Genetic Epidemiology Net-work study. Am Heart J 2003;145:1071–7.

7. Matsuzaki M, Yonezawa F, Toma Y, etal. [Experimental mitral regurgitation inischemia-induced papillary muscle dys-function]. J Cardiol 1988;18 Suppl:121–6, discussion 127.

8. Messas E, Guerrero JL, Handschum-acher MD, et al. Paradoxic decrease inischemic mitral regurgitation withpapillary muscle dysfunction: insightsfrom three-dimensional and contrastechocardiography with strain ratemeasurement. Circulation 2001;104:1952–7.

9. Kaul S, Spotnitz WD, Glasheen WP,Touchstone DA. Mechanism of isch-emic mitral regurgitation. An experi-mental evaluation. Circulation 1991;84:2167–80.

0. Uemura T, Otsuji Y, Nakashiki K, etal. Papillary muscle dysfunction atten-uates ischemic mitral regurgitation inpatients with localized basal inferiorleft ventricular remodeling: insightsfrom tissue Doppler strain imaging.
J Am Coll Cardiol 2005;46:113–9.

4

4

4


F E B R U A R Y 2 0 1 3 : 2 2 0 – 3 4

Chinitz et al.


234

41. Gorman JH III, Gorman RC, Plap-pert T, et al. Infarct size and loca-tion determine development of mi-tral regurgitation in the sheep model.J Thorac Cardiovasc Surg 1998;115:615–22.

42. Voci P, Bilotta F, Caretta Q, Mer-canti C, Marino B. Papillary muscleperfusion pattern. A hypothesis forischemic papillary muscle dysfunction.Circulation 1995;91:1714–8.

43. Pastorius CA, Henry TD, HarrisKM. Long-term outcomes of patientswith mitral regurgitation undergoing
percutaneous coronary intervention.Am J Cardiol 2007;100:1218–23.
4. Uddin AM, Henry TD, Hodges JS,Haq Z, Pedersen WR, Harris KM. Theprognostic role of mitral regurgitationafter primary percutaneous coronary in-tervention for acute ST-elevation myo-cardial infarction. Catheter CardiovascInterv 2011;80:779–86.

5. Liel-Cohen N, Guerrero JL, Otsuji Y,et al. Design of a new surgical ap-proach for ventricular remodeling torelieve ischemic mitral regurgitation:insights from 3-dimensional echocar-diography. Circulation 2000;101:2756–63.

6. Hung J, Guerrero JL, Handschum-
acher MD, Supple G, Sullivan S,Levine RA. Reverse ventricular re-
modeling reduces ischemic mitralregurgitation: echo-guided deviceapplication in the beating heart. Cir-culation 2002;106:2594 – 600.

47. Hung J, Chaput M, Guerrero JL, et al.Persistent reduction of ischemic mitralregurgitation by papillary muscle repo-sitioning: structural stabilization of thepapillary muscle-ventricular wall com-plex. Circulation 2007;116 Suppl:I259–63.

Key Words: mitral regurgitationy myocardial infarction y
papillary muscle infarction.

mitral apparatus assessment by delayed enhancement...

Documents