Usefulness of Overlapping of the E and A Waves of the TransmitralFlow as a Predictor of Responders to Cardiac

Resynchronization Therapy

Hitoshi Minamiguchi, MDa, Yasushi Sakata, MDa,*, Tomohito Ohtani, MDa, Isamu Mizote, MDa,Yasuharu Takeda, MDa, Hiroya Mizuno, MDa, Yuji Okuyama, MDa, Satoshi Nakatani, MDa,

Masashi Fujita, MDb, Tetsuya Watanabe, MDb, Masaaki Uematsu, MDb, and Issei Komuro, MDa

Atrioventricular (AV) dyssynchrony as well as ventricular-ventricular dyssynchrony

aDepartment oSchool of MediciRosai Hospital, A2012; revised man

See page 1617*CorrespondinE-mail addres

0002-9149/13/$ -http://dx.doi.org/1

plays an important role in the selection of candidates for cardiac resynchronizationtherapy (CRT), but no method for assessing the AV dyssynchrony has been established.The aim of this study was to investigate whether the degree of overlap of the E and Awaves can predict response to CRT. The study subjects were 48 consecutive patientsmaintaining sinus rhythm and intrinsic AV conduction who underwent de novo dual-chamber CRT device implantation. CRT responders were defined as those with reduc-tions in left ventricular end-systolic volume >15% at 6 months after CRT deviceimplantations. Twenty-three patients (48%) were CRT responders. In a multivariateanalysis, the overlap ratio of the E and A waves was the only independent predictor ofresponse to CRT (odds ratio 1.03, 95% confidence interval 1.01 to 1.06, p [ 0.01). Usinga cut-off value of 33%, patients with overlap ratios of the E and A waves ‡33% hada significantly higher rate of response to CRT than those with ratios <33% (73% vs 27%,p [ 0.002). In conclusion, the overlap ratio of the E and A waves before CRT deviceimplantation may predict CRT response. This simple method may be helpful in evalu-ating dyssynchrony in patients, particularly with severe reduced left ventricular wallmotion, because this method does not require any wall motion analysis. � 2013 ElsevierInc. All rights reserved. (Am J Cardiol 2013;111:1613e1618)

One of the reasons for a lack of a hemodynamic andfunctional improvement obtained with cardiac resynchroni-zation therapy (CRT) is that the assessment of atrioventric-ular (AV) dyssynchrony as well as ventricular-ventriculardyssynchrony before CRT device implantation has beencontroversial. Left ventricular (LV) diastolic filling timemeasured by transmitral flow for the cardiac cycle length(DFT/RR) has previously been proposed as a parameterof AV dyssynchrony before CRT device implantation.1

However, this parameter was not sufficient for predictingCRT response in large clinical trials.2,3 In this study, wefocused on the degree of overlap of the E and A waves oftransmitral flow before CRT device implantation becausetimely atrial contraction would be desirable, especially inpatients with heart failure and reduced LV ejection fractions(LVEFs). We hypothesized that patients with more overlapof the E and A waves of transmitral flow might be betterCRT candidates, and we retrospectively investigatedpredictors of response to CRT.

f Cardiovascular Medicine, Osaka University Graduatene, Suita, Japan; and bDivision of Cardiology, Kansaimagasaki, Japan. Manuscript received November 27,uscript received and accepted January 31, 2013.for disclosure information.g author: Tel: 81-6-6879-3640; fax: 81-6-6879-3639.s: yasushisk@cardiology.med.osaka-u.ac.jp (Y. Sakata).

see front matter � 2013 Elsevier Inc. All rights reserved.0.1016/j.amjcard.2013.01.333

Methods

Patients were selected for CRT device implantationsaccording to current guidelines and criteria: (1) severe heartfailure (New York Heart Association functional class IIIor IV) despite optimal medical treatment, (2) an LVEF<35%, and (3) prolonged QRS duration (>120 ms). Of 80consecutive patients who underwent de novo dual-chamberCRT device implantations at 2 institutions from June 2007to June 2012, patients with external LV assist devices(n ¼ 4), acquired second- or third-degree AV block (n ¼ 7),persistent atrial tachycardia and/or atrial fibrillation(n ¼ 12), and upgraded pacemakers from right ventricularapical pacing (n ¼ 5) and those without follow-up data(n ¼ 4) were excluded from this study. Ultimately,48 patients maintained in sinus rhythm and with intrinsicAV conduction were retrospectively analyzed. The studyprotocol conformed to the Declaration of Helsinki and wasaccepted by each hospital’s ethics committee. Writteninformed consent was obtained from all patients. Allpatients underwent baseline evaluations and echocardio-graphic examinations, including evaluations of dyssyn-chrony before and 6 months after CRT.

Before hospital discharge, optimization of the AV andventricular-ventricular intervals was performed by echo-cardiography or an intracardiac electrocardiographicallyguided timing cycle optimization algorithm.4e7 Echocar-diographic images were all obtained before CRT deviceimplantation and 6 months after implantation. LV volumes

www.ajconline.org

Figure 1. Overlap ratio of the E and A waves. The overlap ratio of the E andA waves is measured as follows: ‘a’ is the interval between the point of theintersection of the E and A waves and the termination of the E wave, while‘b’ is the interval between the peak of the E wave and the termination of theE waves. The overlap ratio of the E and A waves is calculated as (a/b) �100%. When the E and A waves are completely separated, the overlap ratioof the E and A waves is 0%. When the E and A waves are totally fused, theoverlap ratio of the E and A waves is 100%.

Table 1Baseline characteristics before cardiac resynchronization therapy deviceimplantation

Variable Overall(n ¼ 48)

Responders(n ¼ 23)

Nonresponders(n ¼ 25)

pValue

Age (yrs) 59 � 14 64 � 13 55 � 14 0.035Men 34 (71%) 14 (61%) 20 (80%) 0.15CRT defibrillator 46 (96%) 21 (91%) 25 (100%) 0.13New York Heart

Associationclass III/IV

36/12 20/3 16/9 0.067

Nonischemiccardiomyopathy

37 (77%) 18 (78%) 19 (76%) 0.85

Heart rate (beats/min) 75 � 15 74 � 15 76 � 15 0.58PQ interval (ms) 177 � 40 173 � 35 180 � 45 0.56QRS duration (ms) 147 � 31 145 � 27 150 � 35 0.56Left bundle branch

block16 (33%) 11 (48%) 5 (20%) 0.048

B-type natriureticpeptide (pg/ml)

389 � 326 401 � 409 380 � 244 0.83

Medicationsb blockers 47 (98%) 22 (96%) 25 (100%) 0.29Angiotensin-converting

enzyme inhibitorsand/or angiotensinreceptor blockers

37 (77%) 19 (83%) 18 (72%) 0.38

Spironolactone 32 (67%) 16 (70%) 16 (64%) 0.68Diuretics 41 (85%) 17 (74%) 24 (96%) 0.030Amiodarone 17 (35%) 7 (30%) 10 (40%) 0.48Digoxin 8 (17%) 4 (17%) 4 (16%) 0.89Intravenous inotropic

agents10 (21%) 2 (9%) 8 (32%) 0.047

Data are expressed as mean � SD or as number (percentage).

1614 The American Journal of Cardiology (www.ajconline.org)

were determined using the method previously reported.8

The grade of mitral regurgitation was assessed according tothe guidelines of the American Society of Echocardiog-raphy.9 We calculated the overlap ratio of the E and Awaves. This overlap ratio was defined as in Figure 1. InFigure 1, ‘a’ is the interval between a point of intersection ofthe E and A waves and the termination of the E wave, while‘b’ is the interval between the peak of the E wave and thetermination of the E wave. The overlap ratio of the E and Awaves was calculated as (a/b) � 100%. When the E and Awaves were completely separated, the overlap ratio of the Eand A waves was 0%. When the E and A waves were totallyfused, the overlap ratio was 100%. We determined the levelof intraventricular and interventricular mechanical dyssyn-chrony with using M-mode echocardiography and tissueDoppler imaging. A septal-posterior wall motion delay ob-tained on M-mode imaging in the parasternal long-axis viewof �130 ms10 or the opposing wall delay between theanteroseptal-to-posterior wall or the septal-to-lateral wall of�65 ms was defined as intraventricular mechanical dys-synchrony.11 Interventricular electromechanical delay wasthe difference between the LV pre-ejection period and theright ventricular pre-ejection period. An interventricularelectromechanical delay of �40 ms was defined as inter-ventricular dyssynchrony.1 A reduction in the LV end-systolic volume of >15% at 6 months after CRT was usedas an objective measure of response to CRT in this study.2,3

Continuous variables are expressed as mean � SD.Factors were determined using chi-square tests, and between-group comparisons were made using Mann-Whitney U testsfor continuous variables and Fisher’s exact tests for dichot-omous variables. Only variables with p values <0.05 onunivariate analysis were entered into a multivariate logisticregression analysis to identify independent factors of CRTresponse. A p value <0.05 was considered statisticallysignificant. Receiver-operating characteristic curves weregenerated, and the area under the curve was determined asa measure of the ability to predict a positive response at any

cut-off value. JMP version 9.0 (SAS Institute Inc., Cary,North Carolina) was used for all statistical tests.

Results

Patient characteristics are listed in Table 1. b blockerswere administered in 47 patients: carvedilol (average dose8.5 � 6 mg) in 42 patients, bisoprolol (2.8 � 0.3 mg) in 2patients, metoprolol (40 � 20 mg) in 2 patients, and atenolol(12.5 mg) in 1 patient. The various analyses of the echo-cardiographic measurements are listed in Table 2. At6-month follow-up, 23 patients (48%) were CRT responders.Three patients in the nonresponder group experiencedprogression to persistent atrial fibrillation, whereas allpatients in the responder group maintained sinus rhythm.The 2 groups were generally comparable in terms of baselinedemographic, clinical, and echocardiographic characteristics(Tables 1 and 2). The overlap ratio of the E and A waves washigher in the responder group than in the nonrespondergroup (44 � 34% vs 19 � 23%, p ¼ 0.006). New YorkHeart Association functional class and electrocardiographicand echocardiographic parameters 6 months after CRTdevice implantation are listed in Table 3. Significantreductions in heart rate were more frequently observed in theresponder group than in the nonresponder group (67 � 10 vs76 � 10 beats/min, p ¼ 0.004), although the incidence ofpatients increasing the doses of b blockers was similarbetween the responder group and the nonresponder group

Table 3Exercise tolerance and electrocardiographic and echocardiographic parameters at 6 months after cardiac resynchronization therapy device implantation

Variable Overall (n ¼ 48) Responders (n ¼ 23) Nonresponders (n ¼ 25) p Value

New York Heart Association class I/II/III/IV 6/21/14/7 6/14/3/0 0/7/11/7 0.002Heart rate (beats/min) 72 � 11 67 � 10 76 � 10 0.004PQ interval (ms) 158 � 47 139 � 30 173 � 53 0.01QRS duration (ms) 141 � 22 141 � 23 142 � 22 0.95B-type natriuretic peptide (pg/ml) 324 � 361 135 � 129 505 � 418 0.0003Echocardiographic parametersLVEF (%) 31 � 12 40 � 9 22 � 6 0.0001Deceleration time of the E wave (ms) 185 � 66 202 � 52 170 � 75 0.11DFT/RR 0.48 � 0.08 0.50 � 0.07 0.46 � 0.08 0.08Overlap ratio of the E and A waves (%) 13 � 23 10 � 16 16 � 28 0.36

Data are expressed as mean � SD or as numbers.

Figure 2. Representative examples of a responder and a nonresponder. (A) Representative patient from the responder group. The overlap ratio of the E and Awaves before CRT device implantation was 80%. After CRT device implantation, this ratio improved to 20%. (B) Representative patient from the nonrespondergroup. The overlap ratio of the E and A waves before CRT device implantation was 0%. After CRT device implantation, this ratio was also 0%.

Table 2Echocardiographic parameters before cardiac resynchronization therapy device implantation

Variable Overall (n ¼ 48) Responders (n ¼ 23) Nonresponders (n ¼ 25) p Value

LVEF (%) 26 � 7 28 � 6 23 � 8 0.04LV end-diastolic volume (ml) 247 � 88 239 � 78 254 � 97 0.56LV end-systolic volume (ml) 184 � 82 173 � 66 195 � 94 0.37Left atrial dimension (mm) 46 � 8 45 � 8 47 � 7 0.45Mitral regurgitation grade �II 28 (58%) 13 (57%) 15 (60%) 0.81Deceleration time of the E wave (ms) 182 � 64 190 � 49 173 � 75 0.35Tei index 0.56 � 0.15 0.56 � 0.15 0.56 � 0.16 0.97DFT/RR 0.47 � 0.07 0.47 � 0.07 0.46 � 0.07 0.87DFT/RR <0.4 9 (19%) 4 (17%) 5 (20%) 0.82Overlap ratio of the E and A waves (%) 30 � 31 44 � 34 19 � 23 0.006Septal posterior wall motion delay >130 ms 20/33 (61%) 13/19 (68%) 7/14 (50%) 0.28Interventricular mechanical delay >40 ms 14/36 (39%) 9/20 (45%) 5/16 (31%) 0.40Ts (anteroseptal-posterior, septal-lateral) >65 ms 14/27 (52%) 7/15 (47%) 7/12 (58%) 0.55

Data are expressed as mean � SD or as number (percentage).Ts ¼ the time to peak myocardial systolic velocity during the ejection phase.

Heart Failure/Transmitral Flow and Cardiac Resynchronization Therapy 1615

(39% vs 36%, p ¼ NS). New York Heart Associationfunctional class improved in 91% of the responders and only40% of the nonresponders (p ¼ 0.0002). Representativecases in the 2 groups are shown in Figure 2.

On multivariate logistic regression analysis, the overlapratio of the E and A waves (odds ratio 1.03, 95% confidenceinterval 1.01 to 1.06, p¼ 0.01) was an independent predictorof response to CRT. On the basis of the receiver-operating

Figure 3. Although the DFT/RR ratio is the same, the DFT/RR ratio alone cannot distinguish between the following patients. (A) A patient in whom the Eand A waves of transmitral flow are separated. (B) A patient in whom the E and A waves of transmitral flow are overlapped. After optimization of CRT,extending the diastolic filling time would be desirable in the latter patient (B), but there is little possibility of extending the diastolic filling time in the formerpatient (A).

1616 The American Journal of Cardiology (www.ajconline.org)

characteristic curve analysis, the overlap ratio of the E and Awaves of transmitral flow achieved an area under the curveof 0.70 (p ¼ 0.01) for the ability to predict CRT response,although DFT/RR was not statistically significant. Anoverlap ratio of the E and A waves of �33% had optimalsensitivity (68%) and specificity (76%) for predictingresponders to CRT. Using a cut-off value of 33%, patientswith overlap ratios of the E and A waves of �33% hada significantly higher response rate than those with overlapratios of the E and A waves of <33% (73% vs 27%, p ¼0.002). This cut-off value applied to 10 patients (21%) ofCRT super-responders, defined as having significant reduc-tions in LV end-systolic volume of >30%. Patients withoverlap ratios of the E and A waves of �33% had a signif-icantly higher rate of super-response than those with overlapratios of the E and A waves of <33% (36% vs 8%, p ¼0.01). Twenty-two patients had overlap ratios of the E and Awaves �33%, and 26 patients did not. DFT/RR wassignificantly smaller (0.44 � 0.06 vs 0.49 � 0.08, p ¼0.008), and left bundle branch block morphology wassignificantly more frequently observed in patients withoverlap ratios of the E and A waves of �33% than in thosewith ratios <33% (64% vs 8%, p <0.0001), while the otherbaseline and echocardiographic parameters were similarbetween the groups.

In our study, 17 patients (35%) had LVEFs �30%, and31 (65%) had LVEFs <30%. Among patients with LVEFs�30%, there was a tendency toward a difference but nosignificant difference in the response rate between patientswith overlap ratios of the E and A waves of �33% andthose without (83% vs 37%, p ¼ 0.06). However, amongpatients with LVEFs <30%, those with overlap ratios ofthe E and A waves of �33% had a significantly greater

number of CRT responders than those with ratios <33%(69% vs 20%, p ¼ 0.006).

Discussion

Our results demonstrate that determining the overlapratio of the E and A waves of transmitral flow before CRTdevice implantation can be a simple tool for predictingresponse to CRT in patients with chronic heart failure. Theoverlap ratio of the E and A waves was higher in theresponder group than in the nonresponder group, whileDFT/RR and even the other conventional parameters for theprediction of response did not differ between the groups. Amultiple regression analysis showed that this overlap ratiowas an independent predictor of CRT response after anadjustment of the parameters showing significant differencesbetween the responder and nonresponder groups. Thisoverlap ratio was also a factor for predicting CRT super-response. Moreover, even among patients with low LVEFs,those with higher overlap ratios included more respondersthan those with lower ratios.

The evidence that the overlap of the E and A waves oftransmitral flow is superior to DFT/RR for predicting CRTresponse can be supported by the limitation of DFT/RR forexamining AV dyssynchrony. DFT/RR has been previouslyproposed as a parameter of AV dyssynchrony,1 but itsusefulness remains controversial.2,3 One of the reasons isthat DFT/RR cannot distinguish between a separation andoverlap (Figure 3) of the E and A waves. The former casewould not have any beneficial effects, because there is littlepossibility of extending the diastolic filling time by opti-mization of the CRT device, but the latter case might havebeneficial effects from optimization. In fact, DFT/RR

Heart Failure/Transmitral Flow and Cardiac Resynchronization Therapy 1617

significantly improved after CRT device implantation in theresponder group (0.47 � 0.07 before vs 0.50 � 0.07 after,p ¼ 0.04), but DFT/RR did not in the nonresponder group(0.46 � 0.07 before vs 0.46 � 0.08 after, p ¼ NS).Therefore, assessment of the overlap ratio of the E and Awaves of transmitral flow may be more effective for eval-uating AV dyssynchrony than DFT/RR.

Unexpectedly, the overlap ratio of the E and A waveswas the only independent predictor of CRT response in ourstudy, while the other conventional parameters were not.This may be explained by 3 mechanisms. First, overlap ofthe E and A waves of transmitral flow reflected electricalventricular dyssynchrony.1 An intraventricular conductiondelay increases the pre-ejection interval of the left ventriclebecause of the delayed systole period. This causes abnor-malities of diastolic function with shortening of theventricular filling period by fusion of the E and A wavesbecause the next E wave is delayed by an increase in theduration of the ventricular systole. In these patients, LVpacing may change the diastolic filling timing, using atimely atrial contraction more effectively. In the baselinecharacteristics of our study, left bundle branch block wasmore frequent in the higher overlap ratio group than in thelower overlap ratio group (64% vs 8%, p <0.0001). Thisevidence may support our speculation.

Second, the overlap ratio of the E and A waves is simpleto measure. Mechanical dyssynchrony has been measuredwith conventional echocardiographic techniques, includingM-mode, Doppler, and tissue Doppler imaging. However,no single echocardiographic measurement of dyssynchronyappears to have had a clinically relevant impact in largeclinical trials.2,3 Two-dimensional speckle-tracking imagingand real-time 3-dimensional echocardiography are prom-ising methods to decrease nonresponse.12e17 However, thereproducibility of these echocardiographic parametersremains controversial. Therefore, a simple and reproduciblemethod is still desirable in clinical practice, and transmitralflow velocity is suitable for that purpose.

Third, our parameter does not require wall motion anal-ysis. Echocardiography, in particular advanced quantitativetechniques for regional wall motion that include tissueDoppler imaging, 2-dimensional speckle-tracking imaging,and real-time 3-dimensional echocardiography, may helpunderstand the complexity of LV mechanics,12e17 but inpatients with severe heart failure, poor LV wall motion oftenhampers an accurate assessment of LV dyssynchrony.17 Ina subanalysis of patients with LVEFs <30% in our study,those with overlap ratios of the E and A waves of �33% stillincluded significantly more CRT responders than those withratios <33%. This result demonstrates the advantage of ourparameter in patients with poor LV wall motion. In ourstudy, no single echocardiographic parameter of mechanicaldyssynchrony used in the large clinical trials2,3 was apredictor of response to CRT. However, a combination ofthe overlap ratio of the E and A waves and the indexes ofmechanical dyssynchrony, especially in borderline cases,may help in making a clinical decision.

Our study did have several limitations. First, the fusion ofthe E and A waves is dependent on heart rate. In this study,there was no difference in heart rate between the responderand nonresponder groups (74 � 15 vs 76 � 15 beats/min,

p ¼ NS), and the overlap ratio of the E and A waves was notcorrelated with heart rate before CRT device implantation(r ¼ 0.34). Therefore, at least in our study, the overlap ratioof the E and A waves was not affected by heart rate.

Second, the optimization of the AV and ventricular-ventricular intervals after CRT was not determined. Somecases were optimized by echocardiography and some bydevice algorithms. However, we used the same optimizationprogram for all patients, using echocardiography or thedevice algorithm after CRT device implantation.

Third, this simple method was not available in thepatients with atrial fibrillation, because the A waves of thetransmitral flow were not evaluated in these patients.

Fourth, the total number of responders was smaller thanexpected (48%), compared with previous CRT clinicaltrials. This is because our retrospective study populationincluded many patients with severe progressive LVdysfunction. LV end-systolic volume and left atrial volumebefore CRT device implantation were 184 � 82 and 100 �44 ml, respectively, and end-stage inotrope-dependent heartfailure was observed in 21% of the patients in our study.

Finally, this study was a retrospective study, and thesample size was relatively small. A prospective analysis ina large-scale study is desirable.

Acknowledgment: We are grateful for the excellent tech-nical assistance provided by Keiko Katsuki, Kumiko Mor-imoto, Naoko Kobayashi, Yayoi Ohno, Sanae Ohata,Sonoko Yamasaki, Naomi Fujita, and Tongsuk Son. Wealso thank Drs. Masaharu Masuda, Machiko Kanzaki,Yasuhiro Ichibori, and Tatsunori Taniguchi for referringtheir patients and collecting their data and John Martin forediting the manuscript.

Disclosures

The authors have no conflicts of interest to disclose.

1. Cazeau S, Bordachar P, Jauvert G, Lazarus A, Alonoso C, VandrellMC, Mugica J, Ritter P. Echocardiographic modeling of cardiac dys-synchrony before and during multisite stimulation: a prospective study.Pacing Clin Electrophysiol 2003;26:137e143.

2. Chung ES, Leon AR, Tavazzi L, Sun JP, Nihoyannopoulos P, MerlinoJ, Abraham WT, Ghio S, Leclercq C, Bax JJ, Yu CM, Gorcsan J, St.John SM, De Sutter J, Murillo J. Results of the Predictors of Responseto CRT (PROSPECT) trial. Circulation 2008;117:600e605.

3. Seo Y, Ito H, Nakatani S, Takami M, Naito S, Shiga T, Ando K,Wakayama Y, Aonuma K; J-CRT Investigators. The role of echocar-diography in predicting responders to cardiac resynchronizationtherapy. Cir J 2011;75:1156e1163.

4. Ritter P, Padeletti L, Gillio-Meina L, Gaggini G. Determination of theoptimal atrioventricular delay in DDD pacing. Comparison betweenecho and peak endocardial acceleration measurements. Europace1999;12:126e130.

5. Ishikawa T, Sumita S, Kimura K, Kikuchi M, Kosuge M, Kuji N, EndoT, Sugano T, Sigemasa T, Kobayashi I, Tochikubo O, Usui T.Prediction of optimal atrioventricular delay in patients with implantedDDD pacemakers. Pacing Clin Electrophysiol 1999;22:1365e1371.

6. Bordachar P, Lafitte S, Reuter S, Sanders P, Jais P, Haissaquerre M,Roudaut R, Garrigue S, Clementy J. Echocardiographic parameters ofventricular dyssynchrony validation in patients with heart failureusing sequential ventricular pacing. J Am Coll Cardiol 2004;44:2157e2165.

7. Baker JH, Mckezie J III, Beau S, Greer GS, Porterfield J, Fedor M,Greenberg S, Daoud EG, Corbisiero R, Bailey JR, Porterfield L.

1618 The American Journal of Cardiology (www.ajconline.org)

Acute evaluation of programmer-guided AV/PV and VV delay opti-mization comparing an IEGM method and echocardiogram forcardiac resynchronization therapy in heart failure patients and dual-chamber ICD implants. J Cardiovasc Electrophysiol 2007;18:185e191.

8. Takeda Y, Sakata Y, Mano T, Ohtani T, Kamimura D, Tamaki S,Omori Y, Tsukamoto Y, Aizawa Y, Komuro I, Yamamoto K.Competing risks of heart failure with preserved ejection fraction indiabetic patients. Eur J Heart Fail 2011;13:664e669.

9. Zoghbi WA, Enriquez-Sarano M, Foster E, Grayburn PA, Kraft CD,Levine RA, Nihoyannopoulos P, Otto CM, Quinones MA, RakowskiH, Stewart WJ, Waggoner A, Weissman NJ. Recommendations forevaluation of the severity of native valvular regurgitation with two-dimensional and Doppler echocardiography. J Am Soc Echocardiogr2003;16:777e802.

10. Pitzalis MV, Iacoviello M, Romito R, Massari F, Rizzon B, Luzzi G,Guida P, Andriani A, Mastropasqua F, Rizzon P. Cardiac resynchro-nization therapy tailored by echocardiographic evaluation of ventricularasynchrony. J Am Coll Cardiol 2002;40:1615e1622.

11. Bax JJ, Bleeler 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;44:1834e1840.

12. Kapetanakis C, Kearney MT, Siva A, Gall N, Cooklin M, MonaghanMJ. Real-time three-dimensional echocardiology: a novel technique to

quantify global left ventricular mechanical dyssynchrony. Circulation2005;112:992e1000.

13. Takeuchi M, Jacobs A, Suqenq L, Nishikage T, Nakai H, Weinert L,Salgo IS, Lang RM. Assessment of left ventricular dyssynchrony withreal-time 3-dimensional echocardiography: comparison with Dopplertissue imaging. J Am Soc Echocardiogr 2007;20:1321e1329.

14. Lim P, Buakhamsri A, Popovic ZB, Greenberg NL, Patel D, ThomasJD, Grimm RA. Longitudinal strain delay index by speckle trackingimaging: a new marker of response to cardiac resynchronizationtherapy. Circulation 2008;118:1130e1137.

15. Suffoletto MS, Dohi K, Cannesson M, Saba S, Gorcsan J III. Novelspeckle-tracking radial strain from routine black-and-white echocar-diographic images to quantify dyssynchrony and predict response tocardiac resynchronization therapy. Circulation 2006;113:960e968.

16. Marsan NA, Bleeker GB, Ypenburg C, Van Bommel RJ, Ghio S, Vande Veire NR, Delgado V, Holman ER, van der Wall EE, Schalij MJ,Bax JJ. Real-time three-dimensional echocardiography as a novelapproach to assess left ventricular and left atrium reverse remodelingand to predict to response to cardiac resynchronization therapy. HeartRhythm 2008;5:1257e1264.

17. Sonne C, Sugeng L, Takeuchi M, Weinert L, Childers R, Watanabe N,Yoshida K, Mor-Avi V, Lang RM. Real-time 3-dimensional echocar-diographic assessment of left ventricular dyssynchrony: pitfalls inpatients with dilated cardiomyopathy. JACC Cardiovasc Imaging2009;2:802e812.