F

I

TGT

BL

S

A

o

s

r

a

t

m

a

F

ItpooanwoCmo

FCCDrBSaC

M

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 . 4 , 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 1 . 0 0 5

R O M P I C T U R E S T O P R A C T I C E P A R A D I G M S

maging Cardiac Resynchronization Therapy

heodore Abraham, MD,* David Kass, MD,* Giovanni Tonti, MD,†ery F. Tomassoni, MD,‡ William T. Abraham, MD,§ Jeroen J. Bax, MD,�homas H. Marwick, MBBS, PHD¶

altimore, Maryland; Campobasso, Italy; Lexington, Kentucky; Columbus, Ohio;eiden, the Netherlands; and Brisbane, Australia

ection Editor: Mani A. Vannan, MBBS

lthough a prognostic benefit has been shown from cardiac resynchronization therapy, questions are

ften directed toward the prediction of symptomatic or functional benefit. Recent multicenter trials have

hown the pitfalls of current mechanical markers of left ventricular synchrony, but these negative trial

esults have not marked the conclusion of efforts to predict outcome. Potential new contributors to the

ssessment of mechanical synchrony include echocardiographic and magnetic resonance techniques for

he assessment of myocardial deformation. Nonsynchrony markers that seem promising include assess-

ent of the location and extent of myocardial scar and imaging of the coronary venous and phrenic nerve

natomy. (J Am Coll Cardiol Img 2009;2:486 –97) © 2009 by the American College of Cardiology

oundation

psnTie

DP

Ts

arylaants

theportis a cical

archthe

n the last 2 years, 2 important multicenterrials involving 172 (1) and nearly 500 (2)atients showed failure of mechanical markersf left ventricular (LV) synchrony to predictutcome after cardiac resynchronization ther-py (CRT). These findings contradicted aumber of single center studies, reviewed else-here (3), suggesting that mechanical markersf LV synchrony could predict outcome afterRT. The negative trial results have notarked the conclusion of efforts to predict

utcome from CRT, and indeed, a less well

rom the *Johns Hopkins Medical Institutions, Baltimore, Mampobasso, Campobasso, Italy; ‡Lexington Cardiology Consultolumbus, Ohio; �Leiden University, Leiden, the Netherlands; andrs. Abraham and Kass are consultants and receive research sup

esearch grants from Siemens. Dr. Tomassoni received honoraria orax has received research grants from GE Healthcare, BMS Medcientific, Medtronic, and Biotronik. Dr. Marwick has received resend Siemens. Supported in part by a Program grant (519823) fromanberra, Australia.

anuscript received November 18, 2008; revised manuscript received

ublicized multicenter study of 161 patients (4)howed that imaging of both synchrony andonsynchrony markers could predict response.he goal of this review is to evaluate the role of

maging (not just of dyssynchrony) to thevaluation of potential candidates for CRT.

efining the Benefits of CRT inopulations and Individuals

he use of CRT began with a series of acutetudies in patients with depressed LV function,

nd; †Catholic University of Sacred Heart of, Lexington, Kentucky; §Ohio State University,¶University of Queensland, Brisbane, Australia.from Boston Scientific. Dr. Tonti has receivedonsultant for Siemens and Biosense Webster. Dr.Imaging, Edwards Lifescience, St. Jude, Bostongrant support from GE Medical Systems, Philips,National Health and Medical Research Council,

December 30, 2008, accepted January 9, 2009.

sptActfhC

useavripcpnpLsa4ehlaarocwoc

6s

witttasatrtmttt

D

Eotateemamc

pftc

ardiography

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 . 4 , 2 0 0 9

A P R I L 2 0 0 9 : 4 8 6 – 9 7

Abraham et al.

Imaging CRT

487

howing improvement in invasive hemodynamicarameters (5) and echocardiographic manifesta-ions of LV dyssynchrony and stroke volume (6).fter encouraging studies of acute and longer-term

linical effects (7–9), several randomized controlledrials involving �4,000 patients established the roleor CRT in heart failure (10–16). These resultsave given a strong mandate for the routine use ofRT in eligible heart failure patients (Table 1).In the context of this evidence base, the need to

se imaging for the identification of mechanicalynchrony might reasonably be questioned. How-ver, the therapeutic expectations of patients indvanced heart failure are not restricted to sur-ival but also functional status, which leads to theather controversial definition as to whether anndividual is a responder to CRT. Clinical endoints (New York Heart Association functionallass, quality of life score, exercise capacity ex-ressed as 6-min walking distance), hemody-amic response, and echocardiographic endoints (improved LV systolic function or reverseV remodeling) have been used to evaluate re-

ponse to CRT. Clinical assessment is subjectivend unreliable given the placebo effect of CRT;0% of patients randomized to no CRT experi-nced a significant reduction in symptoms. Acuteemodynamic response may not be predictive of

ong-term response. Changes in ejection fractionre modest, with most studies reporting an aver-ge improvement of 4% to 5%. Evidence ofeverse remodeling provides an objective meansf assessment of response to CRT—a 15% de-rease in LV end-systolic volume (ESV) indexith CRT has emerged as a consistent parameterf reverse remodeling and predictive of long-termlinical outcomes (17)—this is attained in about

Table 1. Landmark Trials in CRT

Study (Ref. #) NYHA Functional Class

MIRACLE (16)* III, IV

MUSTIC SR (14) III

MUSTIC AF (15) III

CONTAK CD (18)‡ III to IV

MIRACLE ICD (17)§ III to IV

COMPANION (20) III, IV

CARE HF (19) III, IV

Left ventricular ejection fraction �35% for all trials. *Includes 71 patients enrolledcrossover phase. §Excludes class II patients. �Echo-based criteria for QRS �150CARE-HF � Cardiac Resynchronization-Heart Failure; COMPANION � ComparisCONTAK-Cardiac Defibrillator; CRT � cardiac resynchronization therapy; MIRACLE �Randomized Clinical Evaluation Implantable Cardioverter Defibrillator trial; MUSTIC A

Stimulation in Cardiomyopathies–Sinus Rhythm; NYHA � New York Heart Association; 6 MH

0% of patients, in contrast with a clinical re-ponse rate in around 70%.

The CRT-responder concept should be appliedith caution. First, there is little evidence to justify

ts application to considerations of survival. Al-hough reverse remodeling is an important media-or of improved survival in heart failure, and seemso be associated with the response to CRT (17,18),

9.5% reduction of ESV has a sensitivity andpecificity of only 70% for prediction of mortality,nd survival benefit from myocardial revasculariza-ion has been documented in the absence of reverseemodeling (19). Second, there are problems withrying to make this distinction on the basis ofechanical markers of LV synchrony. Third, al-

hough changes in synchrony has been proposed ashe sine qua non of CRT response (20), the effect ofherapy could be mediated in other ways (21).

efinition of Dyssynchrony

lectrical dyssynchrony. The measurementf electrical dyssynchrony has ranged fromhe simple lumped QRS duration, used inll of the large multicenter trials of biven-ricular pacing therapy, to more complexlectrical activation maps. The latter showarly right heart stimulation that rapidlyoves to the lateral free wall (this delay is

bout 70 ms in the canine heart). Theechanical map shows a slower spread of

ontraction but follows a similar pattern.Electrical dyssynchrony involves the

rimary activation delay typically mani-ested by a widened QRS complex. Theransformation of electrical into mechanical dyssyn-hrony is complex because the latter is not solely

QRS(ms)

Follow-Up(Months) End Points

130 6 NYHA, QoL, 6MHW

150 3 QoL, 6MHW, VO2

200† 3 QoL, 6MHW, VO2

120 6 Composite

130 6 NYHA, QoL, 6MHW, VO2

120 12 Morbidity � mortality

120� 29¶ Morbidity � mortality

-month pilot study. †RV-paced QRS. ‡Includes 248 patients enrolled in 3-month¶Average.f Medical Therapy, Pacing and Defibrillation in Heart Failure; CONTAK CD �ticenter InSync Randomized Clinical Evaluation; MIRACLE ICD � Multicenter InSyncMultisite Stimulation in Cardiomyopathies–Atrial Fibrillation; MUSTIC SR � Multisite

A B B

A N D

2D �

3D �

CMR �

reson

CRT �

therap

CS �

ICE �

echoc

LV �

RV �

TDE �

echoc

�

�

�

�

�

�

�

in 3ms.on oMulF �

W � 6-min hall walking test; QoL � quality of life; VO2 � maximum oxygen

R E V I A T I O N S

A C R O N YM S

2-dimensional

3-dimensional

cardiac magnetic

ance

cardiac resynchronization

y

coronary sinus

intracardiac

ardiography

left ventricle/ventricular

right ventricle/ventricular

tissue Doppler

uptake.

dmecca(mcrsMmnscwlMbvsdTtslmL(gidr

a

erwaTeddcswc

I

EmrtabstaiocenTvids

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 . 4 , 2 0 0 9

A P R I L 2 0 0 9 : 4 8 6 – 9 7

Abraham et al.

Imaging CRT

488

etermined by when the tissue is excited. Rather,echanical events require the cellular process of

xcitation–contraction coupling to generate myo-yte force, and are thus influenced by processes thatontrol calcium cycling, myofilament calcium inter-ctions, regional loading, fibrosis, and other factors22). Indeed, disparities in the timing of regionalechanical function may not be coupled to electri-

al stimulation delay. This might be caused byegional loading differences, fibrosis, and contractiletrength of one part of the wall versus another.

echanical imaging methods detect motion of theuscle but not its activation process, and so cannot

ecessarily delineate mechanisms for apparent dys-ynchrony. This is important because dyssynchronyaused by electrical delay can be targeted by CRT,hereas that caused by regional properties and/or

oading disparities may not be.echanical dyssynchrony. In left bundle branchlock or right ventricular (RV) pacing, septal acti-ation occurs first, and results in pre-stretch of thetill-quiescent lateral wall, which consequently re-uces the peak rate of pressure increase (dP/dtmax).he delayed lateral wall contraction generates sys-

olic forces that are also partly dissipated by re-tretching the now-early-relaxing septal region,owering net cardiac output. Discoordinate papillary

uscle activation can further compromise overallV function by exacerbating mitral regurgitation

23) (Fig. 1). Disparities in wall stiffening thatenerate discoordinate motion are most markedn early systole (isovolumic contraction, loweringP/dtmax), and late systole as one territory enterselaxation ahead of the other (24).

Mechanical dyssynchrony has been measured byrange of imaging modalities, with the largest

�

�

�

�

�

�

� Electrical Inhomog

Mechanical Ineffic-Early septal motion-Late lateral motion

LV systolic/diastolic LV enlargementMital regurgitation

Reduced CoronaryFlow Reserve

InappropriaNeurohormonal R

Figure 1. Contributors to Electrical and Mechanical Dyssynchron

The contributors to mechanical delays include left ventricular (LV) r

coronary flow reserve, and neurohormonal changes. LVH � left ventricu

vidence base being accumulated with echocardiog-aphy (3). Respectively, 30% and 40% of patientsho fulfill eligibility criteria for CRT do not showsymptomatic response or reverse remodeling (18).hus, electrical dyssynchrony alone, as detected by

lectrocardiography, may not be the optimal pre-ictor of CRT response. The concept of disconcor-ance between electrical and mechanical dyssyn-hrony was supported by an experimental study thathowed improvement in ventricular hemodynamicsith CRT despite no change in electrical dyssyn-

hrony (25).

maging Approaches to Mechanical Dyssynchrony

chocardiography. Standard echocardiographicarkers of LV mechanical synchrony have been

ecently reviewed (3). The primary source of varia-ion between these markers relates to how timing isssessed: the simplest are maximal time delaysetween early and late contracting or electricallytimulated regions, total number of regional areashat show delay, and variance in timing of motionround the heart. However, this can miss criticalnformation regarding the geographic distributionf delays that ultimately generate functional dyssyn-hrony, and conversely, hearts with scattered het-rogeneity of activation or contraction may notecessarily have much functional dyssynchrony.he alternative to these segmental approaches are

arious approaches to index dyssynchrony, includ-ng vector mapping that amplifies the index if theelayed regions are clustered together, as well as theo-called CURE (circumferential uniformity ratio

�

�

�

�ity

y

unction Inappropriate Remodeling-Structural (LVH, LV enlargement)-Contractile (proteins, calcium)-Electrical (conduction)

onses

delling, LV systolic and diastolic dysfunction, mitral regurgitation,

ene

ienc

dysf

teesp

y

emo

lar hypertrophy.

eamenoMttcds

fnopmiTt((cnddta

tt

mr(hvdrbaf

3dstdtcc3wrvrCtatmwll

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 . 4 , 2 0 0 9

A P R I L 2 0 0 9 : 4 8 6 – 9 7

Abraham et al.

Imaging CRT

489

stimate) index (26). These approaches can bepplied to virtually any way of assessing electrical orechanical dyssynchrony, and have the potential to

nhance its specificity for therapeutic responsive-ess. The second important variation relates to therientation of the dyssynchrony measurements.ost tissue Doppler echocardiography (TDE)

echniques provide data in the longitudinal orien-ation of the myocardium. However, most fibers areircumferentially oriented, and deformation in thisimension seems to provide a more reliable andtronger signal.

Despite a large evidence base with TDE, thisorm of analysis poses several problems. Signaloise is common, and variability between observers,ften caused by the presence of multiple systoliceaks, seems to be a major problem (2). Second, theeasurement of TDE time delays is time consum-

ng, especially if multiple segments are interrogated.he development of parametric imaging techniques

hat color-code the segments based on time delay27) may reduce spatial variation as well as save timeFig. 2). Third, TDE examines the timing ofontraction in a longitudinal direction, which mayot be the optimal orientation for discerning bothyssynchrony and the impact of CRT (28). Recentata obtained using angle-corrected TDE showhat radial velocities can also detect dyssynchronynd predict acute response to CRT (29).

Tissue velocity measures tissue motion relative tohe transducer and is therefore susceptible to cardiacranslational motion and tethering artifacts. Strain

Figure 2. Parametric Display of Tissue Velocity

In this parametric display (often known by its proprietary name, tiss

wall is evidenced by red coloration of the lateral and yellow coloration

easurements are based on movement of tissueelative to its neighbor and are therefore site specific30). Unfortunately, TDE-based strain analysisas not been shown to be superior to tissueelocity (31), likely reflecting the angle depen-ence of TDE analysis as well as signal-noiseelated problems. In contrast, speckle-tracking-ased measurements of radial, circumferential,nd longitudinal strain (32,33) seem to be botheasible and reliable (Fig. 3).

Quantitative analysis of dyssynchrony with-dimensional (3D) echocardiography requires bor-er detection in a number of 2-dimensional (2D)lices, and creation of a 3D model from whichime-volume data are obtained (34). The systolicyssynchrony index, derived from the dispersion ofime to minimum regional volume (Fig. 4), de-reases after CRT. Unfortunately, the ability toompare timing of all myocardial segments with aD echocardiography approach is probably out-eighed by limitations in edge detection and frame

ates. This technique, speckle strain, and contrastariability imaging (35) are all the focus of currentesearch.ardiac magnetic resonance (CMR). Potential advan-ages of CMR-based techniques for dyssynchronyssessment include high reproducibility, high spa-ial resolution, and the ability to obtain 3D infor-ation including circumferential mechanics (36),hich may be superior to that obtained in the

ongitudinal direction (28). Myocardial tagging al-ows assessment of radial, longitudinal, and circum-

synchronization imaging), delayed activation of the posterolateral

ue of the posterior walls.

f(edpC

ifdistvm

msv(ua(esfacrrt

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 . 4 , 2 0 0 9

A P R I L 2 0 0 9 : 4 8 6 – 9 7

Abraham et al.

Imaging CRT

490

erential strain by measurements between dark linesmarkers or tags) left in the myocardium by specialncoding pulses (Fig. 5). Recent advances that haverastically reduced analysis time include harmonichase analysis of tagged CMR and strain-encodedMR (37).Three CMR indices have been used for measur-

ng dyssynchrony. Regional variance is determinedrom the variance of strain magnitude in 28 radiallyisplaced segments for each short-axis section (sim-

lar to the indices used in TDE), or the number ofegments with delayed shortening as a percent ofotal regions examined (38) (Fig. 6). The regionalariance vector is based on the product of a radialagnitude vector with a scalar representing time to

Figure 3. Assessment of Synchrony Using Tissue Doppler and S

Panel A shows synchronous septal (yellow) and lateral (blue) wall m(green) and posterolateral walls (red and blue) using longitudinal stissue Doppler echocardiography.

Figure 4. Evaluation of Synchrony Using 3-Dimensional Echocar

Uniform times to minimum volume indicate synchrony (A). The dys

minimum volume (B).

aximal shortening (Fig. 6B). The resulting vectorum will only have significant magnitude if delayedersus early regions are geographically clusteredFig. 6A, upper panel) (28). Lastly, regional strainniformity is based on regional strain differences atgiven moment in time. Time plots of strain

shortening/stretching) are generated at each of 28venly distributed segments around a short-axislice (28). If segments shorten simultaneously (per-ectly synchronous contraction), the plot appears asstraight line, whereas regionally clustered dyssyn-

hrony generates an undulating plot. The relativeatio of first/zero-order magnitudes derived by Fou-ier analysis (Fig. 6C) generates an index known ashe circumferential uniformity ratio estimate (28).

le Strain

on with tissue Doppler. Panel B shows delay between the septalkle strain, confirmed also with circumferential strain (C). TDE �

graphy

hronous left ventricle is characterized by variation in times to

peck

otipec

dio

sync

lad

siNi

s in

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 . 4 , 2 0 0 9

A P R I L 2 0 0 9 : 4 8 6 – 9 7

Abraham et al.

Imaging CRT

491

Potential disadvantages of CMR include cost,ong imaging time, availability, complex and longnalysis times, and incompatibility with implantedevices. These technical problems will likely be

Figure 5. Use of Tagged Cardiac Magnetic Resonance for the As

The progressive deformation of the grid (A) allows measurement ofment (B). The parametric display (C) shows the time course of contsynchronous (lower row). RV � right ventricle; other abbreviation a

Figure 6. Cardiac Magnetic Resonance Techniques for Strain Me

The regional variance of strain (A) cannot differentiate identical varcontraction clustered in 1 portion of the left ventricular wall (A, topformer displays dyssynchrony. The regional variance vector of princrepresenting time at maximal shortening or instantaneous magnituratio of first/zero-order magnitudes derived by Fourier analysis. The

sus the other so this plot appears sinusoidal. Hearts with more variabili

olved, and CMR may become safe in patients withmplanted devices.uclear medicine techniques. When phase imaging

s performed on perfusion scan data, segmental

sment of Synchrony

time course of deformation in the principal axes of each seg-ion, which can be shown to be synchronous (upper row) or dys-Figure 1.

rement

e of time to peak contraction between segments with delayedersus dispersion of delay through the heart (A, bottom); only thestrain (B) is based on the product of unit vectors with a scalarf shortening. Regional strain uniformity (C) provides a relativert with clustered regions (A, top) shows delays in 1 territory ver-

ses

theract

asu

ianc), vipalde ohea

ty (A, bottom) yield a higher frequency waveform.

catertis

T

Sddaip

dprneicida

ai

ong

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 . 4 , 2 0 0 9

A P R I L 2 0 0 9 : 4 8 6 – 9 7

Abraham et al.

Imaging CRT

492

ounts are proportionate to thickening. Althoughcquisitions are obtained at 8 frames/s, Fourierransformation of the data is believed to provide anffective temporal resolution of 75 frames/s. Theesults of nuclear imaging seem to correspond withhose obtained with TDE (39), although the lim-ted published data on CRT effectiveness indicate aensitivity and specificity �80% (40).

he Role of Imaging in the EP Laboratory

tandard imaging approaches offer limited valueuring CRT implantation. Transthoracic echocar-iography may disturb the sterile field, and poorcoustic windows are likely when optimal position-ng is not possible. Transesophageal echocardiogra-hy requires a second operator.Intracardiac echocardiography (ICE) is useful

uring implantation of biventricular devices, allowsrocedural sterility (the catheter is positioned in theight heart via a subclavian venous sheath), elimi-ates the need for a second operator, and providesxcellent target resolution because of its close prox-mity (Fig. 7). Intraoperative ICE can facilitateannulation of the coronary sinus (CS) by identify-ng the ostium, quantify the degree of mechanicalyssynchrony at the time of the procedure, and

Figure 7. ICE

These typical intracardiac echocardiography (ICE) windows from cattative B-mode images in 2 different patients for short-axis (A) and l

ssess the acute response to CRT.

3D rotational angiography is an alternative thatllows 3D chamber reconstruction at the time ofntervention (Fig. 8) (41). This technique can serve

r positions at the right ventricular base and apex show represen--axis (B) imaging. L � lateral wall, S � septum.

Figure 8. 3-Dimensional Rotational Angiography

This representative image from an electrocardiograph gated3-dimensional rotational angiography show the coronary sinus

hete

and venous vasculature.

aaae4L

btnIsapsi(

odtttroa

P

EtnreoetiLiaumisnrit

ia(

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 . 4 , 2 0 0 9

A P R I L 2 0 0 9 : 4 8 6 – 9 7

Abraham et al.

Imaging CRT

493

s a helpful tool during the implant by providing: 1)ccurate 3D CS reconstruction; 2) multiangle visu-lization of the entire CS; 3) an endocardial view tovaluate branch take-off and valvular anatomy; and) accurate final LV lead tip position in relation toV anatomy.Nonresponse to CRT may be multifactorial,

ut incorrect positioning of the LV lead appearso be important; a lateral LV lead position mayot be optimal for all patients. IntraoperativeCE can quantify the degree of mechanical dys-ynchrony present at the time of the procedure,nd with the use of additional echocardiographicarameters such as tissue Doppler velocity ortrain measurements, can provide real-time phys-ologic information regarding CRT response (42)Fig. 9).

Anatomical visualization and 3D reconstructionf cardiac chambers during implantation of CRTevices may be useful for facilitating lead localiza-ion and assess response. As technology continueso advance, cardiac imaging during EP interven-ions will contribute to better procedural success,educe complications, and aid in the understandingf the interaction between electrical activation andnatomical substrates.

Figure 9. Use of On-Line Vector Velocity Imaging Analysis to Ex

In each state, velocity vectors are graphically represented in time aplot of systole (red) and diastole (blue) is seen below this, and B-mcurve plotted for each (bottom). The left ventricular (LV) dyssynchrand a lack of clear separation of contracting and relaxing elementsthe LV lead in the anterior interventricular vein results to a significa

in a posterolateral vein (C), dyssynchrony does not seem to be improve

ost-Implantation Assessment in the CRT Patient

chocardiography is a useful tool for the evaluation ofherapeutic success during follow-up. Patients who doot show LV resynchronization do not show reverseemodeling after CRT (20). However, reverse remod-ling needs time to occur, and more immediate effectsf successful CRT include an improvement in LVjection fraction and a reduction in mitral regurgita-ion, related to both increased closing force andmprovement in interpapillary muscle synchrony (43).ate follow-up studies can be expected to show

mproved LV ejection fraction and reduced LVESVnd LV end-diastolic volume, however, follow-up issually at 6 months post-implantation and improve-ent may take up to 12 months, particularly in

schemic cardiomyopathy. Reverse remodeling is as-ociated with a reduction in LV annular size andormalization of LV geometry, resulting in a furthereduction in mitral regurgitation (44). Other changesnclude a reduction in RV size and decreases inricuspid regurgitation and pulmonary artery pressure.

Echocardiographic atrioventricular optimizationmproves LV dP/dtmax and stroke volume acutely,nd was performed in many effectiveness studies45). The iterative method uses pulsed-wave Dopp-

ne Synchrony in Short-Axis Images in Native and Paced States

agnitude in yellow/red in the upper right, a parametric M-modeimage is broken into 6 segments automatically and a strain(A) is evidenced by nonalignment of strain and velocity vectorsthe parametric M-mode plot. The biventricular (BiV) pacing witheduction of dyssynchrony (B). During BiV pacing with the LV lead

ami

nd modeonyonnt r

d.

lrvtp

PC

Cd

sfttavspeDc

tion

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 . 4 , 2 0 0 9

A P R I L 2 0 0 9 : 4 8 6 – 9 7

Abraham et al.

Imaging CRT

494

er imaging of mitral inflow (Fig. 10). Contempo-ary CRT devices permit programming of theentricular-ventricular (VV) interval, but the con-ribution of VV optimization to improved systolicerformance is controversial (45) (Fig. 11).

atient Selection for CRT:linical Versus Imaging Criteria

urrent criteria for CRT implantation includerug-refractory heart failure (New York Heart As-

Figure 10. The Iterative Method for Atrioventricular Optimizatio

The sequence starts at a long atrioventricular (AV) delay (180 ms, shing by 20-ms increments, until A-wave truncation appears (80 ms).A-wave truncation disappears and maximum E- and A-wave separa

Figure 11. LV Stroke Volume During VV Optimization

Using 20-ms increments (starting at the left ventricle [LV] activated 80

before the LV), the optimal ventricular-ventricular (VV) delay is the delay as

ociation functional class III to IV), LV ejectionraction �35%, and a wide QRS complex (46). Inhe large CRT trials, Mollema et al. (47) showedhat QRS duration alone yielded only a sensitivitynd specificity of 53%, with an optimized cutoffalue of 163 ms. Conversely, echocardiographictudies have shown that approximately 30% ofatients with a wide QRS complex do not havevidence of LV dyssynchrony (as assessed by tissueoppler imaging). As noted above, although single-

enter studies have shown a high accuracy of a variety

er than intrinsic PR interval to ensure capture) and then shorten-atrioventricular delay is lengthened by 10-ms increments until

is provided.

before the right ventricle [RV], and ending at the RV activated 80 ms

n

ortThen

ms

sociated with the largest LV outflow tract velocity time integral.

ospdotesm

iamCdopdru

btdRitivsqitooci

C

Pgdz

ral

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 . 4 , 2 0 0 9

A P R I L 2 0 0 9 : 4 8 6 – 9 7

Abraham et al.

Imaging CRT

495

f parameters to predict CRT response, reliable mea-urement of LV dyssynchrony remains a work inrogress. None of these different techniques wasefinitively superior in the PROSPECT (Predictorsf Response to CRT) trial (2). Indeed, the message ofhis trial was that the reproducibility for these param-ters was limited, and that both the sensitivity and thepecificity of the echocardiographic techniques wereodest.How should these observations be incorporated

nto decision making? Patients with heart failurend a wide left bundle branch block but noechanical dyssynchrony should not be deniedRT (3), although it is reasonable to have aialog with the patient regarding the possibilityf symptomatic or functional nonresponse. In theatient with congestive heart failure, systolicysfunction, mechanical dyssynchrony and a nar-ow QRS, trial data do not currently justify these of CRT (1).Perhaps the focus on mechanical synchrony has

een a distraction from more important contribu-ions of imaging: the detection of scar, correspon-ence of pacing site and maximum VV delay, andV and LV status. The presence of scar tissue

n the target zone of the LV lead may limit responseo CRT (48), and the total scar burden is alsomportant (18). Contrast-enhanced CMR may pro-ide delineation of scar tissue with the highestpatial resolution (Fig. 12). The LV lead is fre-uently not positioned in the site of latest mechan-cal activation, and patients with this lead posi-ion show a worse outcome (49–51). Provisionf information on venous anatomy (Fig. 13) (52)r the cardiophrenic bundle (e.g., with multisliceomputed tomography) is important before CRT

Figure 12. Contrast-Enhanced Cardiac Magnetic Resonance to Id

These examples of lateral scar show nontransmural (A) and transmu

mplantation.

onclusions

roblems with the reproducibility of the currenteneration of imaging techniques has led to aegree of nihilism about imaging in resynchroni-ation therapy. In fact, imaging has important

ify the Location and Thickness of Myocardial Scar

(B) scar thickness.

Figure 13. Use of Computed Tomographic Coronary VenographLead Localization

This electrocardiograph-gated image shows the relationship of theonary artery (RCA) and left circumflex artery (CX) to the coronary sigreat cardiac vein (GCV) and superior cardiac vein (SCV), posterior itricular vein (PIV) and left marginal veins (LMV), and posterior veinleft ventricle (PVLV) and side branches (��). Reprinted with permissi

ent

y to Plan

right cor-nus (CS),nterven-of theon from

Van de Veire et al. (52).

cadtnp

wm

R

MP

R

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 . 4 , 2 0 0 9

A P R I L 2 0 0 9 : 4 8 6 – 9 7

Abraham et al.

Imaging CRT

496

ontributions such as the characterization of LVnd RV function and the assessment of myocar-ial viability and atrioventricular delay. Robustechniques of high temporal resolution areeeded for the assessment of LV synchrony. Atresent, multiple imaging modalities contribute

1999;99:1567–73. Heart 2007;93:167–

ith this useful but not uniformly effective treat-ent modality.

eprint requests and correspondence: Dr. Thomas H.arwick, University of Queensland School of Medicine,

rincess Alexandra Hospital, Ipswich Road, Brisbane,

components of the decision matrix to proceed Qld 4102, Australia. E-mail: t.marwick@uq.edu.au.

1

2

2

2

2

2

2

2

2

E F E R E N C E S

1. Beshai JF, Grimm RA, Nagueh SF, etal. Cardiac-resynchronization therapy inheart failure with narrow QRS com-plexes. N Engl J Med 2007;357:2461–71.

2. Chung ES, Leon AR, Tavazzi L, etal. Results of the Predictors of Re-sponse to CRT (PROSPECT) trial.Circulation 2008;117:2608–16.

3. Gorcsan J III, Abraham T, AglerDA, et al. Echocardiography for car-diac resynchronization therapy: rec-ommendations for performance andreporting—a report from the Amer-ican Society of EchocardiographyDyssynchrony Writing Group en-dorsed by the Heart Rhythm Soci-ety. J Am Soc Echocardiogr 2008;21:191–213.

4. Parsai C, Bijnens B, Sutherland GR,et al. Toward understanding responseto cardiac resynchronization therapy:left ventricular dyssynchrony is onlyone of multiple mechanisms. EurHeart J 2008 Nov 11 [E-pub ahead ofprint].

5. Saxon LA, Kerwin WF, CahalanMK, et al. Acute effects of intraop-erative multisite ventricular pacingon left ventricular function and ac-tivation/contraction sequence in pa-tients with depressed ventricularfunction. J Cardiovasc Electro-physiol 1998;9:13–21.

6. Auricchio A, Stellbrink C, Block M,et al., for the Pacing Therapies forCongestive Heart Failure StudyGroup and the Guidant CongestiveHeart Failure Research Group. Effectof pacing chamber and atrioventriculardelay on acute systolic function ofpaced patients with congestive heartfailure. Circulation 1999;99:2993–3001.

7. Leclercq C, Cazeau S, Le Breton H,et al. Acute hemodynamic effects ofbiventricular DDD pacing in patientswith end-stage heart failure. J AmColl Cardiol 1998;32:1825–31.

8. Kass DA, Chen CH, Curry C, et al.Improved left ventricular mechanicsfrom acute VDD pacing in patientswith dilated cardiomyopathy and ven-tricular conduction delay. Circulation

9. Gras D, Mabo P, Tang T, et al.Multisite pacing as a supplementaltreatment of congestive heart failure:preliminary results of the MedtronicInc. InSync Study. Pacing Clin Elec-trophysiol 1998;21:2249–55.

10. Cazeau S, Leclercq C, Lavergne T, etal. Effects of multisite biventricularpacing in patients with heart failureand intraventricular conduction delay.N Engl J Med 2001;344:873–80.

11. Leclercq C, Walker S, Linde C, et al.Comparative effects of permanentbiventricular and right-univentricularpacing in heart failure patients withchronic atrial fibrillation. Eur Heart J2002;23:1780–7.

12. Abraham WT, Fisher WG, SmithAL, et al. Cardiac resynchronizationin chronic heart failure. N Engl J Med2002;346:1845–53.

13. Young JB, Abraham WT, Smith AL,et al. Combined cardiac resynchroni-zation and implantable cardioversiondefibrillation in advanced chronicheart failure: the MIRACLE ICDTrial. JAMA 2003;289:2685–94.

14. Higgins SL, Hummel JD, Niazi IK, etal. Cardiac resynchronization therapyfor the treatment of heart failure inpatients with intraventricular conduc-tion delay and malignant ventriculartachyarrhythmias. J Am Coll Cardiol2003;42:1454–9.

15. Cleland JG, Daubert JC, Erdmann E,et al. The effect of cardiac resynchro-nization on morbidity and mortality inheart failure. N Engl J Med 2005;352:1539–49.

16. Bristow MR, Saxon LA, Boehmer J,et al. Cardiac-resynchronization ther-apy with or without an implantabledefibrillator in advanced chronic heartfailure. N Engl J Med 2004;350:2140–50.

17. Yu CM, Bleeker GB, Fung JW, et al.Left ventricular reverse remodelingbut not clinical improvement predictslong-term survival after cardiac resyn-chronization therapy. Circulation2005;112:1580–6.

18. St John Sutton M, Keane MG. Re-verse remodelling in heart failure withcardiac resynchronisation therapy.

71.

9. Samady H, Elefteriades JA, AbbottBG, Mattera JA, McPherson CA,Wackers FJ. Failure to improve leftventricular function after coronaryrevascularization for ischemic car-diomyopathy is not associated withworse outcome. Circulation 1999;100:1298 –304.

0. Bleeker GB, Mollema SA, HolmanER, et al. Left ventricular resynchro-nization is mandatory for response tocardiac resynchronization therapy:analysis in patients with echocardio-graphic evidence of left ventriculardyssynchrony at baseline. Circulation2007;116:1440–8.

1. Tang WH, Mullens W, BorowskiAG, et al. Relation of mechanicaldyssynchrony with underlying cardiacstructure and performance in chronicsystolic heart failure: implications onclinical response to cardiac resynchro-nization. Europace 2008;10:1370–4.

2. Sengupta PP, Tondato F, KhandheriaBK, Belohlavek M, Jahangir A. Elec-tromechanical activation sequence innormal heart. Heart Fail Clin 2008;4:303–14.

3. Auricchio A, Prinzen FW. Update onthe pathophysiological basics of car-diac resynchronization therapy. Eu-ropace 2008;10:797–800.

4. Leclercq C, Kass DA. Retiming thefailing heart: principles and currentclinical status of cardiac resynchroni-zation. J Am Coll Cardiol 2002;39:194–201.

5. Leclercq C, Faris O, Tunin R, et al.Systolic improvement and mechanicalresynchronization does not requireelectrical synchrony in the dilated fail-ing heart with left bundle-branchblock. Circulation 2002;106:1760–3.

6. Byrne MJ, Helm RH, Daya S, et al.Diminished left ventricular dyssyn-chrony and impact of resynchroniza-tion in failing hearts with right versusleft bundle branch block. J Am CollCardiol 2007;50:1484–90.

7. Gorcsan J III, Kanzaki H, Bazaz R,Dohi K, Schwartzman D. Usefulnessof echocardiographic tissue synchro-nization imaging to predict acuteresponse to cardiac resynchroniza-

tion therapy. Am J Cardiol 2004;93:1178 – 81.

2

2

3

3

3

3

3

3

3

4

4

4

4

5

5

5

Ksem

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 . 4 , 2 0 0 9

A P R I L 2 0 0 9 : 4 8 6 – 9 7

Abraham et al.

Imaging CRT

497

8. Helm RH, Leclercq C, Faris OP, etal. Cardiac dyssynchrony analysis us-ing circumferential versus longitudinalstrain: implications for assessing car-diac resynchronization. Circulation2005;111:2760–7.

9. Dohi K, Suffoletto MS, SchwartzmanD, Ganz L, Pinsky MR, Gorcsan JIII. Utility of echocardiographic radialstrain imaging to quantify left ventric-ular dyssynchrony and predict acuteresponse to cardiac resynchronizationtherapy. Am J Cardiol 2005;96:112–6.

0. Marwick TH. Measurement of strainand strain rate by echocardiography:ready for prime time? J Am CollCardiol 2006;47:1313–27.

1. Yu CM, Zhang Q, Chan YS, et al.Tissue Doppler velocity is superior todisplacement and strain mapping inpredicting left ventricular reverse re-modelling response after cardiac re-synchronisation therapy. Heart 2006;92:1452–6.

2. Suffoletto MS, Dohi K, CannessonM, Saba S, Gorcsan J III. Novelspeckle-tracking radial strain fromroutine black-and-white echocardio-graphic images to quantify dyssyn-chrony and predict response to cardiacresynchronization therapy. Circula-tion 2006;113:960–8.

3. Lim P, Buakhamsri A, Popovic ZB, etal. Longitudinal strain delay index byspeckle tracking imaging: a newmarker of response to cardiac resyn-chronization therapy. Circulation2008;118:1130–7.

4. Kapetanakis S, Kearney MT, Siva A,Gall N, Cooklin M, Monaghan MJ.Real-time three-dimensional echocar-diography: a novel technique to quan-tify global left ventricular mechanicaldyssynchrony. Circulation 2005;112:992–1000.

5. Kawaguchi M, Murabayashi T, FeticsBJ, et al. Quantitation of basal dyssyn-chrony and acute resynchronizationfrom left or biventricular pacing bynovel echo-contrast variability imaging.J Am Coll Cardiol 2002;39:2052–8.

6. Lardo AC, Abraham TP, Kass DA.Magnetic resonance imaging assess-

ment of ventricular dyssynchrony: cur-

rent and emerging concepts. J AmColl Cardiol 2005;46:2223–8.

37. Osman NF, Sampath S, Atalar E,Prince JL. Imaging longitudinal car-diac strain on short-axis images usingstrain-encoded MRI. Magn ResonMed 2001;46:324–34.

38. Nelson GS, Curry CW, Wyman BT,et al. Predictors of systolic augmenta-tion from left ventricular preexcitationin patients with dilated cardiomyopa-thy and intraventricular conductiondelay. Circulation 2000;101:2703–9.

39. Henneman MM, Chen J, Ypenburg C,et al. Phase analysis of gated myocardialperfusion single-photon emission com-puted tomography compared with tissueDoppler imaging for the assessment ofleft ventricular dyssynchrony. J Am CollCardiol 2007;49:1708–14.

40. Henneman MM, Chen J, Dibbets-Schneider P, et al. Can LV dyssyn-chrony as assessed with phase analysison gated myocardial perfusion SPECTpredict response to CRT? J Nucl Med2007;48:1104–11.

41. Knackstedt C, Muhlenbruch G,Mischke K, et al. Imaging of thecoronary venous system: validation ofthree-dimensional rotational venousangiography against dual-source com-puted tomography. Cardiovasc Inter-vent Radiol 2008;31:1150–8.

42. Calo L, Lamberti F, Loricchio ML, etal. Intracardiac echocardiography:from electroanatomic correlation toclinical application in interventionalelectrophysiology. Ital Heart J 2002;3:387–98.

43. Ypenburg C, Lancellotti P, Tops LF,et al. Acute effects of initiation andwithdrawal of cardiac resynchroniza-tion therapy on papillary muscle dys-synchrony and mitral regurgitation.J Am Coll Cardiol 2007;50:2071–7.

44. Ypenburg C, Lancellotti P, Tops LF,et al. Mechanism of improvement inmitral regurgitation after cardiac re-synchronization therapy. Eur Heart J2008;29:757–65.

45. Stanton T, Hawkins NM, Hogg KJ,Goodfield NE, Petrie MC, McMur-ray JJ. How should we optimize car-diac resynchronization therapy? Eur

Heart J 2008;29:2458–72. t

6. Vardas PE, Auricchio A, Blanc JJ, etal. Guidelines for cardiac pacing andcardiac resynchronization therapy: thetask force for cardiac pacing and car-diac resynchronization therapy of theEuropean Society of Cardiology. De-veloped in collaboration with the Eu-ropean Heart Rhythm Association.Eur Heart J 2007;28:2256–95.

7. Mollema SA, Bleeker GB, van derWall EE, Schalij MJ, Bax JJ. Useful-ness of QRS duration to predict re-sponse to cardiac resynchronizationtherapy in patients with end-stageheart failure. Am J Cardiol 2007;100:1665–70.

8. Ypenburg C, Schalij MJ, Bleeker GB,et al. Impact of viability and scar tissueon response to cardiac resynchroniza-tion therapy in ischaemic heart failurepatients. Eur Heart J 2007;28:33–41.

9. Becker M, Franke A, Breithardt OA,et al. Impact of left ventricular leadposition on the efficacy of cardiac re-synchronisation therapy: a two-dimensional strain echocardiographystudy. Heart 2007;93:1197–203.

0. Becker M, Hoffmann R, Schmitz F, etal. Relation of optimal lead positioningas defined by three-dimensional echo-cardiography to long-term benefit ofcardiac resynchronization. Am J Cardiol2007;100:1671–6.

1. Murphy RT, Sigurdsson G, Mula-malla S, et al. Tissue synchronizationimaging and optimal left ventricularpacing site in cardiac resynchroniza-tion therapy. Am J Cardiol 2006;97:1615–21.

2. Van de Veire NR, Schuijf JD, DeSutter J, et al. Non-invasive visualiza-tion of the cardiac venous system incoronary artery disease patients using64-slice computed tomography. J AmColl Cardiol 2006;48:1832–8.

ey Words: heart failure yynchrony y LV dysfunction ychocardiography y cardiacagnetic resonance y computed

omography.