Ctws3mpFr

TcN

oPS

0d

Noninvasive Imaging of Cardiac Venous Anatomy With 64-SliceMulti-Slice Computed Tomography and Noninvasive Assessment

of Left Ventricular Dyssynchrony by 3-Dimensional TissueSynchronization Imaging in Patients With Heart Failure Scheduled

for Cardiac Resynchronization Therapy

Nico R. Van de Veire, MD, PhDa,b,*, Nina Ajmone Marsan, MDa, Joanne D. Schuijf, MSc, PhDa,Gabe B. Bleeker, MD, PhDa, Maurits C.E.F. Wijffels, MD, PhDa, Lieselotte van Erven, MD, PhDa,

Eduard R. Holman, MD, PhDa, Johan De Sutter, MD, PhDb, Ernst E. van der Wall, MD, PhDa,Martin J. Schalij, MD, PhDa, and Jeroen J. Bax, MD, PhDa

Objectives of this study were to perform a prospective head-to-head comparison betweenmulti-slice computed tomography (MSCT) venography and invasive venography in cardiacresynchronization therapy (CRT) candidates as well as to evaluate the relation between leftventricular (LV) lead position and effect on LV dyssynchrony and immediate response toCRT. Twenty-one consecutive heart failure patients scheduled for CRT implantation wereprospectively enrolled to undergo 64-slice MSCT to visualize the venous system, invasivevenography during device implantation, and tri-plane tissue synchronization imaging (TSI)before and after implantation. Excellent agreement between MSCT and invasive venogra-phy was noted. No significant differences were observed between both techniques regardingvessel diameters. In 12 patients, a match was observed between the area of latest mechan-ical activation (on TSI) and LV lead position. These patients showed a significant decreasein LV dyssynchrony (43 � 7 ms to 11 � 9 ms, p <0.0001) with acute reduction in LVend-systolic volume (188 � 54 ml to 162 � 48 ml, p <0.01) and improvement in LV ejectionfraction (22% � 9% to 34% � 9%, p <0.01). Patients with a mismatch between area oflatest activation and LV lead position remained dyssynchronous without improvement inLV function. In conclusion, visualization of major tributaries of the coronary sinus wascomparable between invasive venography and MSCT venography. Optimal LV leadpositioning in a vein draining the area of latest mechanical activation (determined fromtri-plane TSI) resulted in acute improvement of LV dyssynchrony and systolic functionafter CRT implantation. © 2008 Elsevier Inc. All rights reserved. (Am J Cardiol 2008;

101:1023–1029)

cDotcstvstvpMaLtssf

ardiac resynchronization therapy (CRT) is an attractivereatment for highly symptomatic heart failure patients withide QRS complex and depressed left ventricular (LV)

ystolic function.1 In large randomized clinical trials, up to0% of patients do not respond favorably to CRT.2,3 Twoajor issues need to be addressed during the selection

rocess of potential candidates to improve the success rate.3irst, LV dyssynchrony appears an important predictor ofesponse to CRT.4 The amount of intraventricular dyssyn-

aDepartment of Cardiology, Leiden University Medical Center, Leiden,he Netherlands; and bGhent University, Gent, Belgium. Manuscript re-eived September 11, 2007; revised manuscript received and acceptedovember 18, 2007.

Dr. Bax was supported by research grants from Medtronic, Minneap-lis, Minnesota, Boston Scientific, Natich, Massachusetts, St. Jude, St.aul, Minnesota, GE Health Care, Amersham, UK, and Bristol-Myersquibb Medical Imaging, N Bellerica, Massachusetts.

*Corresponding author: Tel: 32-9-332-51-46.

tE-mail address: nico.vandeveire@ugent.be (N. Van de Veire).

002-9149/08/$ – see front matter © 2008 Elsevier Inc. All rights reserved.oi:10.1016/j.amjcard.2007.11.052

hrony can be reliably assessed noninvasively with tissueoppler echocardiography.5 Second, attention to venous anat-my ensures the area of latest activation can be reachedhrough a suitable vein. Particularly in patients with ischemicardiomyopathy, lack of a suitable branch of the coronaryinus may hamper successful endovascular LV lead implan-ation.6 In 2005, preliminary data suggested that the cardiacenous system could be depicted noninvasively using 16-lice multi-slice computed tomography (MSCT).7 Recently,he feasibility of 64-slice MSCT to evaluate the cardiaceins was confirmed in a retrospective analysis of 100atients.8 However, prospective comparisons betweenSCT and invasive venography for assessment of venous

natomy are lacking. Moreover, information on location ofV dyssynchrony and venous anatomy should be integrated

o determine whether a transvenous or a surgical approachhould be selected for LV lead positioning. In the presenttudy, 21 consecutive patients with heart failure scheduledor CRT underwent 64-slice CT of the cardiac venous sys-

em, invasive venography during device implantation, and

www.AJConline.org

3cLtsueii

M

Ausrr

MOr1AritciswrrlcDpdit

cTucutplnwic

tpvsRIpo

vblua(mwpdTfl

atcLi1cvs

bwgHscCsvat

TPuv

CP

P

L

1024 The American Journal of Cardiology (www.AJConline.org)

-dimensional tissue synchronization imaging (TSI) echo-ardiography before CRT to assess location and extent ofV dyssynchrony. Study aims were (1) to perform a head-

o-head comparison between MSCT venography and inva-ive venography in CRT candidates (to further validate these of MSCT for assessing venous anatomy), and (2) tovaluate the effect of CRT when the LV lead is positionednside or outside a vein draining the area of latest mechan-cal activation.

ethods

total of 21 consecutive patients with heart failure sched-led for CRT, underwent 64-slice MSCT before, and inva-ive venography during CRT implantation. Exclusion crite-ia were atrial fibrillation, documented contrast allergy, andeduced renal function (serum creatinine �140 �mol/L).

Imaging was performed with a 64-detector row Toshibaultislice Aquilion 64 system (Toshiba Medical Systems,tawara, Japan) with a collimation of 64 � 0.5 mm and a

otation time of 0.4 seconds. Between 95 and 140 ml (mean08 � 11 ml) of contrast material (Iomeron 400, Braccoltana Pharma GmbH, Konstanz, Germany) at an injection

ate of 5 ml/min was used with an automatic bolus trigger-ng technique. To optimize the scan for venous visualiza-ion, an additional 2-second delay was applied after theontrast bolus reached the descending aorta before trigger-ng the scan. For optimal filling of the cardiac venousystem, the use of saline chasing was omitted. Tube voltageas 120 kV at 300 mA. A segmental reconstruction algo-

ithm allowed inclusion of patients with a range of heartates. The majority of the patients were already takingow-dose � blockers; additional � blockade was consideredontra-indicated because of the presence of heart failure.uring the MSCT examination, electrocardiography waserformed simultaneously for retrospective gating of theata. An initial data set was reconstructed at 75% of the RRnterval, with a slice thickness of 0.5 mm and a reconstruc-ion interval of 0.3 mm.

Data reconstruction was performed on a Vitrea postpro-essing workstation (Vital Images, Plymouth, Minnesota).hree-dimensional volume-rendered reconstructions weresed to obtain general information on the anatomy of theardiac veins. Thereafter, the course of the veins was eval-ated in three orthogonal planes using multiplanar reformat-ing. The tributaries of the coronary sinus evaluated9 includeosterior interventricular vein, posterior vein of the LV, andeft marginal vein. First-degree side branches were alsooted, if present. The diameter of the coronary sinus ostiumas measured in 2 directions (anteroposterior and supero-

nferior), and the diameters of the tributaries were measuredlose to their origin.

Retrograde venography was performed after implanta-ion of the right atrial and right ventricular leads in all 21atients. The coronary sinus was cannulated from a subcla-ian entry site by an 8F guiding catheter. A balloon occlu-ion catheter (Arrow model 6714; Arrow International Inc.,eading, Pennsylvania, or Medtronic model 6215, Medtronic

nc., Minneapolis, Minnesota) was advanced to the proximalart of the coronary sinus. Three venograms (right anterior

blique 30°, anteroposterior and left anterior oblique 30° o

iews) were then obtained with an acceptable occlusionalloon position. A nonionic contrast agent with low osmo-arity (Iomeron 350; Bracco Altana Pharma GmbH) wassed. Next, the balloon occlusion catheter was withdrawnnd the LV pacing lead was positioned, preferably in thepostero)lateral vein. The choice of LV lead position wasade independently by the implanting cardiologist, whoas blinded to the MSCT and dyssynchrony data. At im-lantation, both the sensing and pacing thresholds (at pulseuration of 0.5 ms) of the LV pacing lead were measured.he final position of the LV pacing lead was assessed byuoroscopy.

The different image series were stored digitally and an-lyzed by an experienced observer without knowledge ofhe MSCT data, at a separate workstation using commer-ially available software (QCA-CMS version 6.0; Medis,eiden, The Netherlands). Images were first analyzed to

dentify the tributaries of the coronary sinus and presence ofst degree side branches. The proximal vein diameters of theoronary sinus, posterior interventricular vein, posteriorein of the left ventricle, and left marginal vein were mea-ured after catheter-based image calibration.

All patients underwent a transthoracic echocardiogramefore and within 72 hours after CRT implantation. Studiesere performed with a commercially available echocardio-raphic platform (VIVID 7; GE Vingmed Ultrasound,orten, Norway), equipped with a 3V-probe for 3-dimen-

ional acquisition. Patients were scanned in left lateral de-ubitus position, from the apical window in tri-plane modus.are was taken to visualize the true LV apex, allowing

imultaneous acquisition of the apical 4-, 2-, and 3-chamberiews. Color-coded TSI was applied to the tri-plane view tossess longitudinal myocardial regional function. Gain set-ings, filters, and pulse repetition frequency were adjusted to

able 1revalence of the major cardiac veins and number of side branchessing multi-slice computed tomography (MSCT) and invasiveenography

n/Total (%)

MSCT Invasive Venography

oronary sinus 21/21 (100%) 21/21 (100%)osterior interventricular vein 21/21 (100%) 18/21 (86%)0 side branches 14/21 (67%) 12/18 (66%)1 side branch 2/21 (9%) 3/18 (17%)2 side branches 4/21 (19%) 3/18 (17%)3 side branches 0/21 (0%) 0/18 (0%)4 side branches 1/21 (5%) 0/18 (0%)osterior vein of LV 20/21 (95%) 20/21 (95%)0 side branches 15/20 (75%) 12/20 (60%)1 side branch 2/20 (10%) 1/20 (5%)2 side branches 3/20 (15%) 6/20 (30%)3 side branches 0/20 (0%) 0/20 (0%)4 side branches 0/20 (0%) 1/20 (5%)eft marginal vein 12/21 (57%) 13/21 (62%)0 side branches 6/12 (50%) 5/13 (38%)1 side branch 2/12 (17%) 3/13 (23%)2 side branches 3/12 (25%) 4/13 (31%)3 side branches 0/12 (0%) 0/13 (0%)4 side branches 1/12 (8%) 1/13 (8%)

ptimize color saturation. Sector size and depth were opti-

msaV

ittwwvt(i

udt

lQpctp3emplwamla

qti

F3tva planted

Fravta

TEr

M3

L

1025Heart Failure/Venous Anatomy and Dyssynchrony in CRT Candidates

ized for the highest possible frame rate. At least 2 con-ecutive beats were recorded from each view, and the im-ges were digitally stored for offline analysis (EchoPac; GEingmed Ultrasound).During postprocessing, the tri-plane data set was frozen

n end-diastole, and the endocardial border was manuallyraced in the apical 4-, 2- and 3-chamber views, respec-ively. Then, using the same heartbeat, the tri-plane data setas frozen in end-systole, and again the endocardial borderas manually traced in the apical 4-, 2-, and 3-chamberiews. Three-dimensional LV end-diastolic and end-sys-olic volumes were generated automatically by the softwareEchopac), and LV ejection fraction was reported accord-ngly.10

During postprocessing, the tri-plane TSI data set wassed to analyze myocardial velocity curves as previouslyescribed.5 TSI automatically calculates the time from

GCV

LMVPVLV

PIV

igure 1. Cardiac venous anatomy of an 80-year-old man with severe h-dimensional volume-rendered reconstruction from 64-slice MSCT dataributaries of the coronary sinus are visualized using both techniques, specentricle (PVLV), and a large left marginal vein (LMV). Note the prominnd PVLV (lightning flash). The left ventricular lead was successfully im

igure 2. No significant differences are observed between invasive venog-aphy or MSCT in quantitative assessment of the diameters of the CS in thenteroposterior (AP) and superoinferior (SI) direction, the posterior inter-entricular vein (PIV), the posterior vein of the left ventricle (PVLV), andhe left marginal vein (LMV). Correlation coefficients between invasivelynd noninvasively measured diameters are mentioned below the X-axis.

he beginning of the QRS complex to peak systolic ve- 1

ocity in every position in the image with reference to theRS interval. The TSI algorithm detects positive velocityeaks within a specified time interval, and the coloroding ranges from green (earliest), over yellow-orangeo red (latest) within this interval. By using surface map-ing and manual tracing of the endocardial borders, a-dimensional rendered volume was automatically gen-rated, allowing for a spatial representation of electro-echanical activation times. Sample volumes were

laced in the basal and mid segments of the septal,ateral, inferior, anterior, posterior, and anteroseptal LValls to measure the time between onset of QRS complex

nd peak systolic myocardial velocity. From these 12easurements, the LV dyssynchrony parameters calcu-

ated include delay between septum and lateral wall11,12

nd standard deviation of all 12 LV segments.13,14

The severity of mitral regurgitation was graded semi-uantitatively from color-flow Doppler in the conven-ional parasternal long-axis and apical 2- and 4-chambermages. Mitral regurgitation was characterized as mild,

GCV

AIV

IV

PVLV

LMV

lure and history of coronary artery bypass grafting. The left panel is aght panel is a fluoroscopic image from invasive venography. All major, a small posterior interventricular vein (PIV), a posterior vein of the leftbranch of the LMV (block arrow) and the connection between the LMVin the LMV.

able 2chocardiographic observations before and after cardiac

esynchronization therapy (CRT) implantation

Before CRT After CRT p Value

itral regurgitation grade 1.5 � 0.9 1.1 � 0.9 �0.001-dimensional volumetric

parametersLV end-diastolic volume (ml) 239 � 56 239 � 55 NSLV end-systolic volume (ml) 183 � 50 168 � 50 �0.05LV ejection fraction (%) 24 � 9 31 � 10 �0.05V dyssynchrony indexesSeptal-to-lateral delay (ms) 84 � 19 45 � 50 0.001Standard deviation 12 LV

segments (ms)43 � 8 26 � 21 0.001

P

eart fai. The riifically

ent side

� (jet area/left atrial area �10%); moderate, 2� (jet

a(a

ptaifcbpvv

R

Ta(u7

(4ncaiwpusvipdIoOudi2

Fepva

FL(i(

1026 The American Journal of Cardiology (www.AJConline.org)

rea/left atrial area 10% to 20%); moderately severe, 3�jet area/left atrial area 20% to 45%); and severe, 4� (jetrea/left atrial area �45%).15

All analyses were performed with the statistical softwarerogram SPSS 12.0.1 (SPSS, Inc., Chicago, Illinois). Con-inuous data are presented as mean � SD. Categorical datare presented as absolute numbers or percentages. Compar-sons between MSCT and invasive venography were per-ormed using the independent-samples t test and Pearson’sorrelation analysis; comparisons between characteristicsefore and after implantation were performed using theaired-samples t test. Comparisons between categoricalariables were performed with Chi-Square testing. A palue �0.05 was considered significant.

esults

here were 13 (62%) men and 8 (38%) women with a meange of 67 � 11 years; all were in New York Heart AssociationNYHA) class III despite optimized medical therapy (86%sed diuretics, 95% angiotensin-converting enzyme inhibitors,1% � blockers, 48% spironolactone). Patients with ischemic

igure 3. (A) The favorable effects of positioning the left ventricular leaffects in patients showing a mismatch between lead position and aarameters calculated from tri-plane TSI data, specifically, septal-to-lateelocity in 12 LV segments (SD12). The lower graphs depict LV-end sfter CRT implantation.

igure 4. Case example of an 85-year-old patient with idiopathic dilatedBBB-configuration). MSCT shows a large left marginal vein with promin

D, E, F). In the 3-dimensional LV dyssynchrony assessment with TSI (I),s positioned in the left marginal vein (G, H), resulting in synchronous ac

J). In addition, a reduction of LV end-systolic volume from 191 to 153 ml was

12, 57%) and those with nonischemic cardiomyopathy (9,3%) were both included; 6 (29%) had a history of percuta-eous coronary intervention, and 6 (29%) had a history oforonary artery bypass grafting (combined with mitral valvennuloplasty in 2 patients). MSCT was successfully performedn all included patients. Average heart rate during scanningas 65 � 14 beats/min, scan time 21 � 11 seconds, helicalitch 14.3 � 0.7. On average, 108 � 11 ml of contrast wassed. CRT device and transvenous LV lead implantation wereuccessful in all but 1 patient (absence of suitable veins onenography, and this patient was referred for surgical LV leadmplantation), without major complications. All remaining 20atients received a combined CRT-implantable cardioverter-efibrillator device (Contak Renewal 4 CRT-D, Guidant, andnsync Marquis or Sentry CRT-D, Medtronic Inc.). Two typesf LV leads were used, Easytrack 4517-90 (Guidant) or AttainTW bipolar 4194 (Medtronic Inc.). Five patients underwentpgrading of a previously implanted implantable cardioverter-efibrillator to a CRT-debrillator device. During implantation,nvasive balloon occlusion venography was performed in all1 patients.

n the area of latest mechanical activation (match). (B) Absence of theselatest mechanical activation. The upper graphs depict dyssynchronyy (SL delay) and standard deviation of time to peak myocardial systolicvolume (LVESV) and LV end-diastolic volume (LVEDV) before and

yopathy, NYHA class III heart failure and wide QRS complex (157 ms,branches (A, B, C). These findings are confirmed by invasive venographymechanical activation (yellow) is located in the lateral wall. The LV leadof the LV, represented (green) on TSI, obtained immediately after CRT

d withirea ofral delaystolic

cardioment sidethe latetivation

observed, with an increase in LV ejection fraction from 13% to 28%.

1027Heart Failure/Venous Anatomy and Dyssynchrony in CRT Candidates

CBA

D FE

G H

I J

PVLV

PIV

CS

LMV PIV

GCV

LMVAIV

LMV

AIV

LMV LMV

LMV

AIV

AIV

tpt3ipioivhtanbcv

pgcesmeC

iiiscpmL(cpd

aatfI(tLu4agTacddivi

D

UemMtrtcoisrespppaoiMsvla

eIidvDascbtodrblcufmLspblsmpw3ps

1028 The American Journal of Cardiology (www.AJConline.org)

Identification of the coronary sinus and its major tribu-aries on MSCT and invasive venography in the entire studyopulation is reported in Table 1. The posterior interven-ricular vein was not observed with invasive venography inpatients (14%), whereas on MSCT this vein was observed

n all patients. The left marginal vein was observed in 12atients (57%) on MSCT and in 13 patients (62%) withnvasive venography. The number of visible side branchesf the major tributaries are reported in Table 1. Figure 1llustrates the excellent agreement between MSCT and in-asive venography in a patient with heart failure with aistory of coronary artery bypass grafting. The diameters ofhe coronary sinus and tributaries, measured with MSCTnd invasive venography are depicted in Figure 2. No sig-ificant differences in measured diameters were observedetween MSCT and invasive venography, and excellentorrelations were observed between noninvasively and in-asively measured diameters (Figure 2).

At baseline, all patients had severe LV dilatation with de-ressed LV function, accompanied by functional mitral regur-itation (Table 2); also, a significant extent of LV dyssyn-hrony was observed.11–14 Immediately after implantation, LVnd-diastolic volume remained unchanged, whereas LV end-ystolic volume decreased significantly with an improve-ent in LV ejection fraction and mitral regurgitation. The

xtent of LV dyssynchrony was significantly reduced afterRT implantation.

Using a 12-segment LV model with 3-dimensional TSImaging, the area of latest mechanical activation was lateraln 8 patients (38%), anterolateral in 9 (43%), posterolateraln 2 (9.5%), and inferolateral in 2 (9.5%). From the fluoro-copic images obtained during CRT implantation, the pre-ise LV lead position was determined. If the LV lead wasositioned in a cardiac vein draining the area of latestechanical activation, this was considered a match betweenV lead position and area of latest activation. In 12 patients

57%), a match was noted between the area of latest me-hanical activation on tri-plane TSI and the final LV leadosition. In these patients, a significant decrease in LVyssynchrony was observed (Figure 3).

Moreover, positioning the LV lead in a vein draining therea of latest mechanical activation (match) resulted in ancute decrease in LV end-systolic volume (average reduc-ion of 26 ml or 13.3%) with an increase in LV ejectionraction (average increase of 12%) after CRT implantation.n contrast, LV dyssynchrony persisted in the 8 patients38%) with a mismatch between the LV lead position andhe area of latest mechanical activation, and improvement inV ejection fraction and reduction in LV end-systolic vol-me were not observed in these patients (Figure 3). Figureshows preimplantation venograms and MSCT images of

n 85-year-old patient with heart failure. A large left mar-inal vein is identified and used for LV lead implantation.his LV lead position matched the area of latest mechanicalctivation (identified with 3-dimensional TSI), and syn-hrony was restored acutely after CRT (septal-to-lateralelay 90 ms at baseline, 10 ms after implantation, standardeviation of 12 LV segments 45 ms at baseline, 5 ms aftermplantation). In addition, a reduction of LV end-systolicolume from 191 to 153 ml was observed with an increase

n LV ejection fraction from 13% to 28%. a

iscussion

ntil recently, invasive venography was the only techniquenabling clinicians to evaluate the coronary sinus and itsajor tributaries. In several preliminary studies, 16-sliceSCT was proposed as a valid non-invasive alterna-

ive.7,16,17 These findings were recently confirmed in a ret-ospective analysis of 100 patients using 64-slice MSCTechnology.8 All studies report a significant variability in theardiac venous system, particularly in patients with a historyf myocardial infarction.7,8,16,17 These results may havemportant implications for the decision of transvenous orurgical LV lead positioning.8,18 In contrast to previouseports on MSCT venography, the present study includedxclusively patients with heart failure who were actuallycheduled for CRT implantation, allowing for a direct com-arison between MSCT venography and invasive venogra-hy during CRT implantation. In agreement with previouslyublished series, the left marginal vein was not observed inlarge percentage of patients.7,8,9 Moreover, all tributaries

f the coronary sinus that were eventually used for LV leadnsertion during CRT implantation were visualized on

SCT. Furthermore, no significant differences were ob-erved between invasive venography and 64-slice MSCTenography regarding diameters of the different vessels, inine with a preliminary comparison between 16-slice MSCTnd invasive venography.17

In selected patients, CRT reduced symptoms, improvedxercise capacity, and reduced morbidity and mortality.19,20,21

n these large randomized trials, CRT is successful in approx-mately 70% of patients. Positioning the LV lead in a veinraining the area of latest mechanical activation may be aaluable strategy to further improve the success rate.3 Tissueoppler techniques have proven their efficacy to identify the

rea of latest mechanical activation.4,5,12,13,14,22 In the presenttudy, tri-plane TSI was used to locate the area of latest me-hanical activation. This approach may be particularly usefulecause it provides a true 3-dimensional representation ofhe left ventricle with a 3-dimensional dyssynchrony mapbtained during a single heartbeat. Moreover, automatedyssynchrony calculations are possible that may potentiallyeduce observer variability. In 57% of patients, a matchetween area of latest mechanical activation and final LVead position was noted, and CRT resulted in acute resyn-hronization with acute reduction in LV end-systolic vol-me and improvement in LV ejection fraction. Theseavourable effects were not observed in patients with aismatch between area of latest mechanical activation andV lead position. These observations agree with recenttudies highlighting the importance of matching the LV leadosition with the site of latest activation.23,24,25 This relationetween the response to CRT and the match between area ofatest mechanical activation and LV lead position under-cores the important role of non-invasive imaging to deter-ine LV lead implantation strategy. With tri-plane TSI, the

recise area of latest mechanical activation can be derived,hereas MSCT depicts the cardiac venous system on a-dimensional LV volume set. Transvenous LV lead im-lantation is preferred if a suitable tributary of the coronaryinus matches the area of latest mechanical activation. In the

bsence of a suitable coronary sinus tributary in the area of

lpwcbaTbdhvprfr

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

1029Heart Failure/Venous Anatomy and Dyssynchrony in CRT Candidates

atest mechanical activation, a surgical approach may bereferred. In the present study, the LV lead was positionedithout taking the LV dyssynchrony information into ac-

ount. Accordingly, only 57% of patients exhibited a matchetween LV lead position and area of latest mechanicalctivation. These patients had substantial benefit from CRT.he remaining 43% of patients, however, had a mismatchetween LV lead position and area of latest activation andid not benefit from CRT. Potentially, superior results couldave been obtained if the preimplantation information pro-ided by tri-plane TSI and MSCT had influenced the im-lantation strategy in these patients. At present, CT venog-aphy is still associated with substantial radiation, buturther technological development will most likely result ineduction of radiation burden.

1. Abraham WT. Cardiac resynchronization therapy. Prog CardiovascDis 2006;48:232–238.

2. Mehra MR, Greenberg BH, Cardiac resynchronization therapy. Caveatmedicus! J Am Coll Cardiol 2004;43:1145–1148.

3. Bax JJ, Abraham T, Barold SS, Breithardt OA, Fung JW, Garrigue S,Gorcsan J III, Hayes DL, Kass DA, Knuuti J, Leclercq C, Linde C,Mark DB, Monaghan MJ, Nihoyannopoulos P, Schalij MJ, StellbrinkC, Yu CM. Cardiac resynchronization therapy: part 1 issues beforedevice implantation. J Am Coll Cardiol 2005;46:2152–2167.

4. Yu CM, Bax JJ, Monaghan M, Nihoyannopoulos P. Echocardio-graphic evaluation of cardiac dyssynchrony for predicting a favourableresponse to cardiac resynchronization therapy. Heart 2004;90(suppl6):vi17–vi22.

5. Bank AJ, Kelly AS. Tissue Doppler imaging and left ventriculardyssynchrony in heart failure. J Cardiac Fail 2006;12:154–162.

6. Goitein O, Lacomis JM, Gorcsan J III, Schwartman D. Left ventricularpacing lead implantation: potential utility of multimodal image inte-gration. Heart Rhythm 2006;3:91–94.

7. Jongbloed MR, Lamb HJ, Bax JJ, Schuijf JD, de Roos A, van der WallEE, Schalij MJ. Noninvasive visualization of the cardiac venous sys-tem using multislice computed tomography. J Am Coll Cardiol 2005;45:749–753.

8. Van de Veire N, Schuijf J, De Sutter J, Devos D, Bleeker G, de RoosA, van der Wall E, Schalij M, Bax J. Non-invasive visualization of thecardiac venous system in coronary artery disease patients using 64-slice computed tomography. J Am Coll Cardiol 2006;48:1832–1838.

9. von Lüdinghausen M. The venous drainage of the human myocardium.Adv Anat Embryol Cell Biol 2003;168(suppl I):1–105.

0. Rodevand O, Bjornerheim R, Aakhus S, Kjekshus J. Left ventricularvolumes assessed by different new three-dimensional echocardio-graphic methods and ordinary biplane technique. Int J CardiovascImaging 1998;14:55–63.

1. Bax JJ, Bleeker GB, Marwick TH, Molhoek SG, Boersma E, SteendijkP, van der Wall EE, Schalij MJ. Left ventricular dyssynchrony predictsresponse and prognosis after cardiac resynchronization therapy. J AmColl Cardiol 2004;44:1834–1840.

2. Van de Veire N, Bleeker G, De Sutter J, Ypenburg C, Holman E, van

der Wall E, Schalij M, Bax JJ. Tissue synchronization imaging accu-

rately measures left ventricular dyssynchrony and predicts response tocardiac resynchronization therapy. Heart 2007;93:1034–1039.

3. Yu CM, Chau E, Sanderson JE, Fan K, Tang MO, Fung WH, Lin H,Kong SL, Lam YM, Hill MR, Lau CP. Tissue Doppler echocardio-graphic evidence of reverse remodeling and improved synchronicityby simultaneously delaying regional contraction after biventricularpacing therapy in heart failure. Circulation 2002;105:438–445.

4. Yu CM, Zhang Q, Wing-Hong Fung J, Chi-Kin Chan H, Chan Y-S,Wai-Kwok Yip G, Kong S-L, Lin H, Zhang Y, Sanderson JE. A noveltool to assess systolic asynchrony and identify responders of cardiacresynchonization therapy by tissue synchronization imaging. J AmColl Cardiol 2005;45:677–684.

5. Thomas JD. How leaky is that mitral valve? Simplified Dopplermethods to measure regurgitant orifice area. Circulation 1997;95:548–550.

6. Abbara S, Cury RC, Nieman K, Reddy V, Moselewski F, Schmidt S,Ferencik M, Hoffmann U, Brady TJ, Achenbach S. Noninvasive eval-uation of cardiac veins with 16-MDCT angiography. Am J Roentgenol2005;185:1001–1006.

7. Mühlenbruch G, Koos R, Wildberger JE, Günther R, Mahnken AH.Imaging of the cardiac venous system: comparison of MDCT andconventional angiography. Am J Roentgenol 2005;185:1252–1257.

8. Singh JP, Houser S, Heist EK, Ruskin JN. The coronary venousanatomy. A segmental approach to aid cardiac resynchronization ther-apy. J Am Coll Cardiol 2005;46:68–74.

9. Freemantle N, Tharmanathan P, Calvert MJ, Abraham WT, Ghosh J,Cleland JG. Cardiac resynchronization for patients with heart failuredue to left ventricular systolic dysfunction–a systematic review andmeta-analysis. Eur Heart J 2006;8:433–440.

0. Cleland JGF, Daubert J-C, Erdmann E, Freemantle N, Gras D, Kap-penberger L, Tavazzi L, on behalf of the CARE-HF Study Investiga-tors. Longer-term effects of cardiac resynchronization therapy on mor-tality in heart failure (the CARE-HF trial extension phase). EurHeart J 2006;27:1928–1932.

1. Rivero-Ayerza M, Theuns DAMJ, Garcia-Garcia HM, Boersma E,Simoons M, Jordaens L. Effects of cardiac resynchronization therapyon overall mortality and mode of death: a meta-analysis of randomizedcontrolled trials. Eur Heart J 2006;27:2682–2688.

2. Van de Veire N, De Sutter J, Van Camp G, Vandervoort P, LancellottiP, Cosyns B, Unger P, Gillebert TC, on behalf of the Belgian Multi-center Registry on Dyssynchrony. Global and regional parameters ofdyssynchrony in ischemic and non-ischemic cardiomyopathy. Am JCardiol 2005;95:1020–1023.

3. Murphy RT, Sigurdsson G, Mullamalla S, Agler D, Popovic ZB,Starling RC, Wilhoff BL, Thomas JD, Grimm RA. Tissue synchroni-zation imaging and optimal left ventricular pacing site in cardiacresynchronization therapy. Am J Cardiol 2006;97:1615–1621.

4. Becker M, Franke A, Breithardt O, Ocklenburg K, Kaminski T, Kra-mann R, Knackstedt C, Stellbrink C, Hanrath P, Schauerte P, Hoff-mann R. Impact of left ventricular lead position on the efficacy ofcardiac resynchronization therapy. A two-dimensional strain echocar-diography study. Heart 2007;93:1197–1203.

5. Becker M, Kramann R, Franke A, Breithardt O-A, Heussen N, Knack-stedt C, Stellbrink C, Schauerte P, Kelm M, Hoffmann R. Impact ofleft ventricular lead position in cardiac resynchronization therapy onleft ventricular remodelling. A circumferential strain analysis based on

2D echocardiography. Eur Heart J 2007;28:1211–1220.