Journal of Electrocardiology 45 (2012) 357–360www.jecgonline.com
The role of pacing-induced dyssynchrony in left ventricular remodelingassociated with long-term right ventricular pacing
for atrioventricular blockRóbert Pap, MD, PhD,⁎ Rodrigo Gallardo, MD, Dóra Rónaszéki, MD, Gergely Ágoston, MD,Vassil B. Traykov, MD, László Sághy, MD, Albert Varga, MD, PhD, Tamás Forster, MD, DSc
2nd Department of Medicine and Cardiology Centre, University of Szeged, Szeged, Hungary
Received 30 November 2011
Abstract Aims: Patients with atrioventricular (AV) block can develop left ventricular (LV) dysfunction with
Keywords: Right ventricular pacing; Left ventricular dyssynchrony; Heart failure
Introduction
Right ventricular pacing (RVP) can lead to thedevelopment of heart failure (HF), even in some patientswith preserved left ventricular (LV) systolic function beforeimplantation.1,2 This adverse effect of RVP is thought to bedue to abnormal LV activation. It has been shown that LVdilatation, systolic dysfunction, and HF with long-termRVP are associated with LV mechanical dyssynchrony.3
However, it remains unclear whether dyssynchrony is acause or consequence of LV dysfunction and whether
uthor. 2nd Department of Medicine and CardiologySzeged, 6720 Szeged, Korányi fasor 6, [email protected]
dyssynchrony induced by RVP is the basis of adverse LVremodeling seen in the long term.
We aimed to clarify the relationship between RVP-induced LV dyssynchrony and changes in LV dimensionsand function during long-term RVP in patients with normalsystolic LV function and atrioventricular (AV) block.
Patients
Nineteen consecutive patients undergoing pacemakerimplantation for AV block were prospectively included.Exclusion criteria were as follows: LV ejection fraction (EF)less than 50%; dependence on a temporary pacemaker (nointrinsic rhythm ≥40 beats per minute); severe HF (NewYork Heart Association 3-4); significant pulmonary disease;acute coronary syndrome; severe, uncorrected valvular
AVB indicates AV block; LBBB, left bundle-branch block; RBBB, rightbundle-branch block.
358 R. Pap et al. / Journal of Electrocardiology 45 (2012) 357–360
disease; and low (b1 year) life expectancy. Patientcharacteristics are summarized in Table 1.
Methods
Patients underwent standard 12-lead electrocardiographicrecording and transthoracic echocardiography within24 hours before and after pacemaker implantation and at
Fig. 1. Determination of opposingwall delay by tissue Doppler imaging. Velocity curvphase velocities (white arrows) is measured. Definition of the ejection phase is basedillustration online.
the end of follow-up. All patients received dual-chamberpacemakers (2 patients single-lead VDD; others DDDdevices), programmed to factory settings. Patients wereseen in the device clinic at 6 weeks and at 6-month intervalsthereafter. Pacemaker function was checked, and settingswere changed accordingly. Patients were questioned aboutHF symptoms (undue fatigue or dyspnea on exertion of newonset) by personnel blinded to echocardiographic data.
Echocardiography
Standard M-mode, Doppler, 2-dimensional, and tissueDoppler echocardiography was carried out using ToshibaAplio CV ultrasound system equipped with Tissue DopplerImaging Quantification software (Toshiba Corp, Tokyo,Japan). All measurements were performed offline by anexaminer blinded to study data. Left ventricular volumes werecalculated using the biplane Simpson method. Color-codedtissue Doppler echocardiography acquisition and assessmentwere carried out according to the joint recommendations ofthe American Society of Echocardiography and the HeartRhythm Society.4 Three consecutive cardiac cycles wereanalyzed in apical 4-chamber, 2-chamber, and long-axisviews. The opposing wall delay: the difference between time-to-peak ejection phase velocities of opposing segments(ΔTpeak) was determined for each view, and the maximalΔTpeak was used as a measure of LV dyssynchrony (Fig. 1).
es of opposing LV segments are generated, and the delay between peak ejectionon pulsed-wave Doppler from the LV outflow tract (lower right panel). Color
Fig. 2. Left ventricular dyssynchrony expressed as ΔTpeak (see text) atbaseline, after implant, and at the end of follow-up (FU) for each patient.
359R. Pap et al. / Journal of Electrocardiology 45 (2012) 357–360
Data analysis
Left ventricular dyssynchrony induced by RVP wasexpressed as the difference of postimplant and preimplantmaximal ΔTpeak in milliseconds and in percentage of thepreimplant value. The change in LV end-systolic volume(ESV) during follow-up was monitored as a marker of LVremodeling.5,6 A 15% or greater increase in LV ESV wasprospectively defined as significant adverse LV remodeling.The change in LV ESV during follow-up was correlated withbaseline variables and pacing-induced change in dyssyn-chrony. Frequently paced patients were defined as thosehaving RVP 60% or greater of the time. Continuousvariables are expressed as mean ± SD and compared usingStudent t test. The relationship of continuous variables wasexplored by calculating Pearson correlation coefficient (r).Proportions were compared using χ2 test. P b .05 wasconsidered statistically significant.
Results
Baseline parameters and immediate effect of pacing
Baseline LV dyssynchrony (maximal ΔTpeak) correlatedwith age (r = 0.561, P = .015), but not with LV dimensions,EF, or QRS duration of baseline intrinsic rhythm. However,patients with left bundle-branch block morphology of the
Table 2Parameters at baseline and after long-term RVP
Baseline After implant P End of FU P (vs baseline)
FU indicates follow-up; EDV, end-diastolic volume; ΔTpeak, dyssynchronyparameter (see text).
intrinsic QRS tended to have more expressed LV dyssyn-chrony at baseline than those with right bundle-branch blockor normal QRS (102 ± 10 vs 62 ± 34 milliseconds,respectively; P = .131). Immediately after implantation,LV dyssynchrony increased in some of the patients, whereasit decreased in others with RVP compared to intrinsic rhythm(Fig. 2). The pacing-induced change in maximal ΔTpeak
ranged between 74 milliseconds and −56 milliseconds andwas inversely correlated with baseline dyssynchrony whenexpressed in milliseconds (r = −0.686, P = .010) orpercentage of the preimplant value (r = −0.655, P = .008).The overall mean dyssynchrony and LV volumes wereunchanged acutely by RVP (Table 2).
Adverse LV remodeling in the long term
One patient died, and 2 were lost to follow-up; theremaining 16 patients were followed up for 28 ± 3.6 months.Mean LV ESV increased, and EF decreased significantly bythe end of follow-up, compared with baseline (Table 2). Thepredefined end point of 15% or greater increase in LV ESVwas reached by 38% of the patients, 83% of whom reportedHF symptoms by the end of follow-up, compared with 20%of patients without significant LV remodeling (P = .013). Ofthe baseline and immediate postimplant parameters, nonecorrelated with long-term LV remodeling. However, patientswith 60% or greater cumulative RV pacing had a greaterchange in LV ESV from baseline (9.9 vs 0.08 mL, P = .027).Within this frequently paced group (75% of patients), thechange in dyssynchrony induced acutely by pacing(expressed in milliseconds or percentage) correlated withthe increase in LV ESV (Fig. 3; r = 0.727, P = .026 or r =0.641, P = .033). Patients who had a 15% or greater increasein LV ESV during follow-up had greater changein dyssynchrony induced immediately after implantationby RVP (28.4 vs −7.8 milliseconds, P = .037) and greaterdyssynchrony at the end of follow-up (92.4 vs 55.4milliseconds, P = .046). The mean LV dyssynchronymeasured immediately after initiating RVP did not changeduring the following 2 years (Fig. 2, Table 2).
Fig. 3. Correlation of the acute, RVP-induced change in dyssynchrony andLV ESV change during follow-up.
360 R. Pap et al. / Journal of Electrocardiology 45 (2012) 357–360
Discussion
Multiple studies have shown LV dyssynchrony inducedacutely by RVP in patients with normal ventricularactivation at baseline7,8 or LV systolic dysfunction.9,10
Furthermore, numerous reports show that dyssynchrony afterlong-term RVP is associated with LV dilation and reducedsystolic function.3,11 Biventricular pacing can mitigate thedetrimental long-term effects of RVP,6 further suggesting arole for LV dyssynchrony behind the deterioration of LVfunction. Although intuitive, a cause-and-effect relationshipbetween acutely induced LV dyssynchrony and long-termLV remodeling during RVP has not been demonstrated.
This study has shown a variable acute effect of RVP onLV synchrony in patients with preserved EF and AV block.Patients with very dyssynchronous LV activation duringintrinsic rhythm tended to have a decrease or no change indyssynchrony, whereas synchronously activated left ventri-cles were desynchronized by RVP. Our findings reproducedthose of Pastore et al,10 who also showed an inverse relationbetween baseline dyssynchrony and the change in dyssyn-chrony induced by RVP.
In our study, an increase in dyssynchrony induced acutelyby RVP predicted adverse LV remodeling after 2 years ofRVP. Moreover, LV dyssynchrony immediately afterimplantation was stable over time. These observationssuggest that RVP-induced dyssynchrony is the cause andnot the consequence of long-term LV remodeling.
Limitations
This is a small series of highly selected patients from asingle institution, the results therefore need to be confirmedby studies with larger number of patients from multiplecenters. The lack of a significant correlation betweenintrinsic QRS morphology, QRS width, and baselinedyssynchrony might have been influenced by the lownumber of patients, although the weakness of such anassociation is supported by previous studies in HF.12 Inaddition, intrinsic QRS width was remarkably narrow in thispopulation (Table 1); this might have contributed to the lackof correlation. Factors not addressed in this study, such assubtle differences in pacing lead position, might havecontributed to the variable effect of RVP on LVdyssynchrony.13 Furthermore, it is still not clear from theseresults why some patients experience a deterioration of LVfunction, whereas others do not with long-term RVP.
Clinical implications
Right ventricular pacing has variable effects on LVdyssynchrony in patients with AV block and preserved LV
systolic function. An increase in LV dyssynchrony at theinitiation of RVP is associated with LV remodeling in thelong term in frequently paced patients. Efforts to minimizeinduced dyssynchrony when ventricular pacing is unavoid-able may reduce LV dysfunction and HF after RVP.
References
1. Sweeney MO, Hellkamp AS, Ellenbogen KA, et al. Adverse effect ofventricular pacing on heart failure and atrial fibrillation among patientswith normal baseline QRS duration in a clinical trial of pacemakertherapy for sinus node dysfunction. Circulation 2003;107:2932.
2. Zhang XH, Chen H, Siu CW, et al. New-onset heart failure afterpermanent right ventricular apical pacing in patients with acquired high-grade atrioventricular block and normal left ventricular function.J Cardiovasc Electrophysiol 2008;19:136.
3. Tops LF, Schalij MJ, Holman ER, van Erven L, van der Wall EE, BaxJJ. Right ventricular pacing can induce ventricular dyssynchrony inpatients with atrial fibrillation after atrioventricular node ablation. J AmColl Cardiol 2006;48:1642.
4. Gorcsan III J, Abraham T, Agler DA, et al. Echocardiography forcardiac resynchronization therapy: recommendations for performanceand reporting—a report from the American Society of Echocardiogra-phy Dyssynchrony Writing Group endorsed by the Heart RhythmSociety. J Am Soc Echocardiogr 2008;21:191.
5. McManus DD, Shah SJ, Fabi MR, Rosen A,WhooleyMA, Schiller NB.Prognostic value of left ventricular end-systolic volume index as apredictor of heart failure hospitalization in stable coronary artery disease:data from theHeart andSoulStudy. JAmSocEchocardiogr 2009;22:190.
6. Yu CM, Chan JY, Zhang Q, et al. Biventricular pacing in patientswith bradycardia and normal ejection fraction. N Engl J Med2009;361:2123.
7. Liu WH, Chen MC, Chen YL, et al. Right ventricular apical pacingacutely impairs left ventricular function and induces mechanicaldyssynchrony in patients with sick sinus syndrome: a real-time three-dimensional echocardiographic study. J Am Soc Echocardiogr2008;21:224.
8. Fornwalt BK, Cummings RM, Arita T, et al. Acute pacing-induceddyssynchronous activation of the left ventricle creates systolicdyssynchrony with preserved diastolic synchrony. J CardiovascElectrophysiol 2008;19:483.
9. Schmidt M, Brömsen J, Herholz C, et al. Evidence of left ventriculardyssynchrony resulting from right ventricular pacing in patients withseverely depressed left ventricular ejection fraction. Europace2007;9:34.
10. Pastore G, Noventa F, Piovesana P, et al. Left ventricular dyssynchronyresulting from right ventricular apical pacing: relevance of baselineassessment. Pacing Clin Electrophysiol 2008;31:1456.
11. Hong WJ, Yung TC, Lun KS, Wong SJ, Cheung YF. Impact of rightventricular pacing on three-dimensional global left ventricular dyssyn-chrony in children and young adults with congenital and acquired heartblock associated with congenital heart disease. Am J Cardiol2009;104:700.
12. Bleeker GB, Schalij MJ, Molhoek SG, et al. Relationship between QRSduration and left ventricular dyssynchrony in patients with end-stageheart failure. J Cardiovasc Electrophysiol 2004;15:544.
13. Takemoto Y, Hasebe H, Osaka T, et al. Right ventricular septal pacingpreserves long-term left ventricular function via minimizing pacing-induced left ventricular dyssynchrony in patients with normal baselineQRS duration. Circ J 2009;73:1829.
