OUP user menu

Beneficial effects of biventricular pacing in chronically right ventricular paced patients with mild cardiomyopathy

Irene E. van Geldorp, Kevin Vernooy, Tammo Delhaas, Martin H. Prins, Harry J. Crijns, Frits W. Prinzen, Barbara Dijkman
DOI: http://dx.doi.org/10.1093/europace/eup378 223-229 First published online: 4 December 2009


Aims To investigate whether cardiac resynchronization therapy (CRT) by means of biventricular (BiV) pacing can improve left ventricular (LV) function, remodelling and clinical status in chronically right ventricular (RV) paced patients with mild cardiomyopathy.

Methods and results Thirty-six chronically (10 ± 7 years) RV paced patients with left ventricular ejection fraction (LVEF) < 40% or LVEDD > 55 mm, without an established indication for CRT, were subjected to 6 months RV and BiV pacing in a patient-blinded, randomized crossover design. Treatment-effects of BiV pacing were evaluated for LV function, LV remodelling and clinical status. As compared with RV pacing, BiV pacing significantly improved LV function (LVEF 46 ± 12 vs. 39 ± 12% and LVFS 24 ± 7 vs. 21 ± 7%) and reduced LV end-diastolic and end-systolic diameters and volumes (LVEDD 56 ± 8 vs. 59 ± 8 mm, LVESD 43 ± 8 vs. 47 ± 9 mm, LVEDV 132 ± 65 vs.144 ± 62 mL and LVESV 77 ± 56 vs. 92 ± 55 mL, respectively). In 19 patients (53%) response to BiV pacing was clinically relevant, defined as LVESV reduction >15%. BiV pacing also significantly improved NYHA classification.

Conclusion BiV pacing following chronic RV pacing may improve LV function and reverse LV remodelling in patients with relatively mild LV dysfunction or remodelling. Hence, upgrade to BiV pacing might be considered in chronically RV paced patients with mild cardiomyopathy.

  • Pacing
  • Biventricular
  • LV function
  • LV remodelling
  • Heart failure


In right ventricular (RV) pacing, the sequence of electrical activation resembles the activation pattern as in left bundle branch block (LBBB).1 This asynchronous electrical pattern is accompanied by abnormal dyssynchronous mechanical interactions within the left ventricle (LV).2,3 Evidence is increasing that LBBB as well as RV pacing, are associated with impairment of LV function, structural remodelling of the LV, and an increased risk for heart failure.49 Cardiac resynchronization therapy (CRT) by means of biventricular (BiV) pacing, aims to reverse the deleterious effects that may originate from LV dyssynchrony. In the majority of patients with severe LV dysfunction and severe clinical heart failure associated with either LBBB,1015 or RV pacing induced dyssynchrony,16,17 BiV pacing improves clinical presentation, reduces mortality, reverses LV remodelling and improves LV function. Therefore, BiV pacing is nowadays strongly recommended in patients with ventricular dyssynchrony (QRS > 120 ms), severe LV dysfunction [LV ejection fraction (LVEF) <35%] and moderate to severe clinical heart failure (NYHA classification III–IV).18 The purpose of the present study was to investigate whether BiV pacing can improve LV function and can reverse LV remodelling in chronically RV paced patients with mild cardiomyopathy. The second aim was to investigate whether BiV pacing could improve, or at least maintain, clinical performance in those patients without severe heart failure.


Study population and inclusion criteria

From a database with ±1000 pacemaker patients in the Maastricht University Medical Centre, all patients with permanent pacemaker stimulation were screened for inclusion. Criteria for inclusion in the study were permanent RV pacing (>95% paced beats) accompanied with echocardiographic signs of LV remodelling [defined as LV end-diastolic diameter (LVEDD) >55 mm] or LV dysfunction (defined as LVEF <40%). At least one of these echocardiographic signs had to be present despite stable pharmacological treatment and pacemaker programming aimed at maximal ventricular filling and minimal ventricular pacing. Inclusion was regardless of clinical symptoms of heart failure. Nevertheless, patients with severe LV dysfunction in combination with moderate to severe heart failure symptoms (i.e. LVEF <35% and NYHA III–IV, respectively) were excluded from the current study, as they had an established indication for CRT.18 Other criteria for exclusion were myocardial infarction or cardiac surgery within the previous 6 months, or non-cardiac conditions that could limit exercise capacity and life expectancy (within the duration of the study).

Study protocol

The Institutional Review Board approved the protocol and all patients provided written informed consent at enrolment in the study (for study design see Figure 1). Patients underwent an upgrade to a BiV pacing device and had routine post-implant pacemaker evaluations. Post-implant all patients were BiV paced to detect and solve possible problems with pacing leads or pacemaker systems within the run-in period. Two to four weeks after implantation patients were randomly assigned to either RV or BiV pacing. At the same time-point individual optimization of the pacemaker settings was performed using echocardiography. The atrioventricular-delay in both RV and BiV pacing was optimized to the shortest delay that still provided maximal ventricular filling, assessed using pulsed Doppler analysis of the transmitral flow. Patients with permanent atrial fibrillation were exempt from the atrioventricular-delay procedure. Optimal interventricular-interval in BiV pacing was determined by maximizing the aortic velocity-time-integral. After 6 months of either BiV pacing or RV pacing, patients crossed over to the other pacing configuration for which the same optimization procedure was performed as at randomization. Thereafter, they were paced in the other pacing configuration for another 6 months. Patients were blinded for pacemaker configurations. Changes in pacing mode were performed by an independent physician, using an external programmer. Clinical and echocardiographic evaluations were performed before enrolment in the study (prior to upgrade procedure; measurement 1), and after the first and second phases (6 and 12 months after randomization; measurement 2 and 3, respectively).

Figure 1

Study design. Subsequent to inclusion and baseline evaluation after chronic right ventricular (RV) pacing, patients underwent an upgrade procedure to a biventricular (BiV) pacing device and a run-in period before optimization of pacing settings. Thereafter patients were randomized to either BiV pacing or RV pacing, followed by a crossover to the other mode. At the end of each 6 months pacing period left ventricular (LV) function, LV remodelling and clinical status were evaluated.

Clinical evaluation

Symptoms of heart failure were classified according to the NYHA classification for heart failure. For further assessment of clinical status, a 12-lead electrocardiogram and an exercise tolerance test (i.e. treadmill-test, 6 min hall-walk distance, or bicycle-test) were performed. Approximate oxygen cost during exercise testing was derived from the peak work rate achieved by the patient and was defined in metabolic equivalents (METs).19 The self-administered ‘Minnesota Living with Heart Failure Questionnaire’ was used for scoring the quality-of-life on a scale from 0 (best) to 105 (worst).20

Echocardiographic evaluation

All measurements were performed according to the guidelines of the American Society of Echocardiography. An experienced physician performed the echocardiographic evaluation using Philips Sonos 5500 or Philips iE 33 (Philips Medical Systems, The Netherlands) with 1.8–3.5-MHz transducers. Standard 2D and Doppler data from three consecutive beats were digitally stored for offline analysis with a customized software package (Xcelera, Philips Medical Systems, The Netherlands). During post-processing a single reviewer (K.V.) systematically measured LV end-diastolic (LVEDD) and LV end-systolic diameters (LVESD) in parasternal long-axis views. End-diastolic and end-systolic LV volumes (LVEDV and LVESV, respectively) and LVEF were calculated from the apical two- and four-chamber images with Simpson's biplane method.21 Additionally, LV fractional shortening (LVFS) was used as a parameter for LV pump function [LVFS = (LVEDD−LVESD)/LVEDD]. Using colour-flow Doppler in the apical four-chamber view, mitral regurgitation was assessed and its severity was semi-quantitatively graded on a scale from 0 (no regurgitation) to 4 (severe regurgitation).22 The time delay between the onset of aortic and pulmonic flow was used to define interventricular asynchrony. The time difference in onset of systolic motion between septum and LV lateral wall, determined using pulse wave tissue Doppler, was used as a measure of intraventricular dyssynchrony.

Pacemaker implantation

All patients were implanted with Guidant (St. Paul, MN, USA) BiV pacemaker systems (B.D.). When indicated an implantable cardioverter-defibrillator (ICD) with an additional LV pacing lead was implanted. LV pacing leads were inserted by a transvenous approach via the coronary sinus into either the lateral or postero-lateral vein whenever possible. Well functioning right atrial and apical RV leads of the prior pacemaker system were used for the upgraded system. In case of ICD implantation an RV coil lead was placed in the RV apex. After the upgrade procedure, lead positions of RV and LV leads were determined from chest X-ray films (posterior-anterior and lateral view).

Statistical analysis

Statistical data-analyses were performed (M.P., I.v.G.) using SPSS version 16.0 (SPSS Inc., Chicago, IL, USA). A two-tailed P-value of <0.05 was considered significant. Reduction in LVESV was considered as major efficacy variable for the study. The smallest difference with clinical significance was considered to be a change in LVESV of >10% with BiV pacing as compared with RV pacing. To reach over 80% power with an estimated 10% loss of patients during follow-up, the study had to include 37 patients (type 1 error: 0.05, 2-tailed; and type 2 error: 0.10). Baseline characteristics at study enrolment were assessed. Comparisons of baseline characteristics between both randomization groups were performed using either unpaired Student t-test for continuous or χ2 test for discrete variables. The efficiency (described in terms of treatment-effects) of BiV pacing treatment as compared with RV pacing, was assessed from crossover data using one-way analysis of variance (ANOVA) for repeated measures. The stimulation mode (RV vs. BiV pacing) was defined as within-subjects factor. To allow for testing on treatment-period interaction (carry-over effect), the randomization sequence (RV → BiV vs. BiV → RV) was entered as a between-groups factor. Responders were defined as patients with an LVESV-difference (RV paced LVESV vs. BiV paced LVESV) >15% in favour of BiV pacing, since this degree of LVESV reduction has been used as a cut-off value for response in other trials.2326 Comparison between responders and non-responders was performed with the use of unpaired Student t-test or χ2 test. Pearson's product-moment correlation coefficient was used to measure the strength of relationship between treatment-effects and baseline LVEF, LVESV, interventricular asynchrony and intraventricular dyssynchrony during RV pacing, the duration of RV pacing prior to, and the age at upgrade. Additionally, to detect possible predictors for the response to BiV pacing analysis of covariance (ANCOVA) was performed on the treatment-effect of BiV pacing on LVESV. In this ANCOVA analysis NYHA classification at inclusion, QRS duration in RV pacing, LV lead position, and baseline echocardiographic characteristics were defined as covariates (independent variables).


Study population

From the entire pacemaker population 706 patients were paced in VVI or DDD mode. Within this group 93 patients were eligible. However, 11 of them had an indication for CRT following the guidelines and were therefore excluded from this study. Two patients recently suffered myocardial infarction, and in 31 patients exercise capacity and life expectancy were limited due to non-cardiac causes. Among the remaining 49 patients matching all study-criteria, 40 patients provided written informed consent. Thereafter, one patient was withdrawn from the study, as he preferred to have no replacement and upgrade of his pacemaker system prior to end-of-life of the system. Two patients were excluded from the study because transvenous LV lead implantation failed and surgical intervention was unfavourable. In the period between October 2004 and October 2006, 37 patients were enrolled after successful implantation of a BiV pacing device (ICD; n = 12). After the upgrade procedure and the run-in period, 18 patients were randomized to receive RV pacing in the first 6 months (group A) and to crossover to BiV pacing for the second period of 6 months. The other 19 patients (group B) were randomized to both pacing treatments in reversed order (BiV pacing first and RV pacing thereafter). In one patient from group A, LV function and remodelling worsened considerably during RV pacing in the first period. Within 3 months after programming to BiV pacing he died from end-stage heart failure. Since paired data (BiV vs. RV) from this patient were not available, this case was excluded from analysis. From the patients included, 19 patients met both echocardiographic inclusion criteria (LVEF < 40% and LVEDD > 55 mm). The other patients had echocardiographic signs of either remodelling [LVEDD > 55 mm (n = 13)], or impaired LV function [LVEF < 40% (n = 4)]. All subject characteristics at study enrolment are summarized in Table 1. On average, patients had relatively mild LV dysfunction (LVEF 36 ± 10%) and they were without severe symptoms of heart failure after chronic RV pacing (10 ± 7 years). Patient characteristics were not significantly different between both randomization groups.

View this table:
Table 1

Patient characteristics at baseline

Total (n = 36)Group A (n = 17)Group B (n = 19)P-value
Gender (M/F)28/813/415/40.858
Age (years)65 ± 1067 ± 1064 ± 110.374
History of RV pacing (years)10 ± 79 ± 711 ± 70.562
Aetiology of pacing indication0.790
 Spontaneous AV-block, n (%)17 (47%)7 (41%)10 (53%)
 Surgically induced AV-block, n (%)3 (8%)2 (12%)1 (5%)
 His-ablation (in permanent atrial fibrillation), n (%)11 (31%)6 (35%)5 (26%)
 Brady/Tachy syndrome (in atrial fibrillation), n (%)5 (14%)2 (12%)3 (16%)
 Coronary artery disease, n (%)13 (36%)8 (47%)5 (26%)0.196
 Atrial fibrillation (permanent), n (%)19 (15) (53%)7 (7) (41%)12 (8) (63%)0.187
 Hypertension, n (%)6 (17%)4 (24%)2 (11%)0.296
 Diabetes mellitus, n (%)3 (8%)1 (6%)2 (11%)0.615
NYHA classification0.617
 Class I, n (%)9 (25%)4 (24%)5 (26%)
 Class II, n (%)19 (53%)10 (59%)9 (47%)
 Class III, n (%)8 (22%)3 (18%)5 (26%)
RV paced QRS duration (ms)195 ± 26196 ± 29193 ± 230.078
LVEDD (mm)60 ± 759 ± 760 ± 70.698
LVESD (mm)48 ± 848 ± 848 ± 90.984
LVEDV (mL)155 ± 72147 ± 63163 ± 800.520
LVESV (mL)104 ± 6498 ± 61109 ± 690.633
LVEF (%)36 ± 1036 ± 1136 ± 90.938
LVFS (%)20 ± 620 ± 621 ± 60.552
  • AV-block, atrioventricular block; LVEDD and LVESD, left ventricular end-diastolic and end-systolic diameter, respectively; LVEDV and LVESV, left ventricular end-diastolic and end-systolic volume; LVEF, left ventricular ejection fraction; LVFS, left ventricular fractional shortening; RV, right ventricular.


In one patient the LV lead was placed epicardially via a minimal invasive thoracotomy after transvenous implantation of the lead failed. In another patient the initial procedure was discontinued because of a coronary sinus dissection, and re-intervention was performed 1 month later. In three other patients re-intervention was needed because of, stimulation of the diaphragm by the LV lead, RV lead dislodgement (n = 1) or impending LV lead dislodgement (n = 1). Minor complications were pocket haematoma (n = 3) and temporary stimulation of the diaphragm (n = 1). All complications mentioned were solved before pacemaker optimization and randomization.

Biventricular pacing improves left ventricular function and reverses left ventricular remodelling

There was no significant effect of the treatment sequence (RV → BiV vs. BiV → RV), as no interaction between randomization sequence and within-subjects factor could be detected for any of the parameters evaluated (no carry-over effect). Thus, treatment-effects of BiV pacing could be calculated from the BiV paced and RV paced measurements previous to and 6 months after cross-over (baseline measurements are not included in this analysis). Figure 2 shows the results on echocardiographic measurements of 6 months BiV pacing vs. 6 months RV pacing. The positive treatment-effects of BiV pacing in LVEF (+7%; P < 0.001) and LVFS (+3%; P = 0.001) are indicative for a significantly better LV function with BiV pacing as compared with RV pacing. LV dilatation was significantly less in BiV pacing as compared with RV pacing, as expressed by smaller diameters and volumes in BiV pacing as compared with RV pacing (LVEDD: −4%, P < 0.001; LVESD: −8%, P < 0.001; LVEDV: −9%, P = 0.001; LVESV: −19%, P < 0.001, respectively). The severity-grade of mitral regurgitation was similar in both pacing configurations (1.3 ± 0.9 and 1.1 ± 0.7 during RV and BiV pacing, respectively; P = 0.107). In 19 patients (53%) the LVESV-difference (RV paced LVESV vs. BiV paced LVESV) was >15% in favour of BiV pacing. These patients are referred to as ‘responders’. Only one patient (3%) was defined as an ‘adverse-responder’ as in this patient the LVESV difference between both pacing-modes was >15% in favour of RV pacing. In 20 patients (56%) BiV pacing was beneficial for LV function indicated by a positive treatment-effect in LVEF >5%. In none of the patients LVEF was adversely influenced >5% by BiV pacing. In the responders the mean LVESV reduction (LVESV treatment-effect: −32 ± 10%) was paralleled by a significantly larger treatment-effect in LVEF as compared with non-responders (+9 ± 7 vs. +4 ± 5%, respectively; P = 0.030) and the number of patients with LVEF-treatment-effect >5% was 14 (74%) in the responder-group.

Figure 2

Left ventricular function and remodelling. Comparison of left ventricular (LV) function (A) and LV remodelling (B) after 6 months biventricular (BiV) pacing vs. 6 months right ventricular (RV) pacing. LVEF, LV ejection fraction; LVFS, LV fractional shortening; LVEDD, LV end-diastolic diameter; LVESD, LV end-systolic diameter; LVEDV, LV end-diastolic volume; LVESV, LV end-systolic volume.

Clinical outcome

NYHA classification was significantly better (P = 0.007) after BiV pacing as compared with RV pacing (Table 2). In six (17%) patients NYHA classification was one class less in BiV pacing [NYHA II vs. NYHA III (n = 2) and NYHA I vs. NYHA II (n = 4) for BiV vs. RV pacing, respectively]. In one patient (3%) the difference was even two NYHA classification levels in favour of BiV pacing (NYHA I in BiV pacing vs. NYHA III in RV pacing). The mean LVESV reduction in all these seven patients was 19 ± 22%; four patients were responders (LVESV reduction >15%) and one patient was the ‘adverse-responder’ (LVESV increase >15%). Quality-of-life-scores and quantity of METs as derivative of performance on exercise testing were not significantly different during BiV pacing and RV pacing (Table 2).

View this table:
Table 2

Clinical outcome

RV pacingBiV pacingP-value
NYHA classification0.007
 Class I, n (%)11 (31%)16 (44%)
 Class II, n (%)18 (50%)16 (44%)
 Class III, n (%)7 (19%)4 (11%)
Quality-of-life score (total)30 ± 2427 ± 230.427
Exercise test (METs)9 ± 410 ± 30.205
  • BiV, biventricular; METs, metabolic equivalents; RV, right ventricular.

Determinants of response to biventricular pacing

No evident determinant for the response to BiV pacing was found. The degree of difference between RV paced and BiV paced LVESV was neither affected by baseline LVESV (Pearson's R = −0.5, ANCOVA P = 0.163), nor by baseline LVEF (R = 0.3; P = 0.506) (Figure 3). Also, the duration of chronic RV pacing (10 ± 7 years) prior to enrolment in the current study (R = 0.1; P = 0.777), and patients’ age at upgrade (R < 0.1; P = 0.308) did not affect the treatment-effect on LVESV. LV lead-tips were located at the lateral free wall (n = 17), posterior free wall (n = 10), anterior free wall (n = 2), lateral-base (n = 2), posterior-base (n = 3), and anterior-base (n = 2). The position of the LV lead-tip was not a significant covariate for the response to BiV pacing (P = 0.596). The duration of the QRS complex was significantly shorter (−16%, P < 0.001) during BiV pacing than during RV pacing (163 ± 19 ms, range 130–200 vs. 195 ± 26 ms, range 140–240 ms). However, baseline QRS duration and measurements of interventricular and LV intraventricular dyssynchrony could not predict the response to BiV pacing (R = 0.2 and P = 0.802; R = −0.2 and P = 0.633; and R = −0.1 and P = 0.523, respectively). The response to BiV pacing was not influenced by the NYHA classification at baseline (P = 0.271), nor by the presence of atrial fibrillation as a comorbidity (P = 0.501).

Figure 3

Response to biventricular pacing. Percentage reduction in left ventricular end-systolic volume (LVESV) after biventricular pacing plotted against the baseline LVESV (A) and baseline left ventricular ejection fraction (LVEF) (B).

Patients in NYHA I and II, subgroup analysis

In patients with mild symptoms of heart failure (NYHA I and II, n = 28) the treatment-effects of BiV pacing were similar to compared with the entire study cohort. In this subgroup, mean LVESV reduction was 19% (P < 0.001) and LVESV reduction was >15% in 15 patients (54%). All parameters of remodelling were smaller in BiV pacing as compared with RV pacing: LVEDD 56 ± 7 vs. 58 ± 7 mm (4%, P = 0.002), LVESD 43 ± 8 vs. 46 ± 9 mm (8%, P < 0.001), LVEDV 124 ± 53 vs. 134 ± 49 mL (8%, P = 0.025), and LVESV 68 ± 47 vs. 83 ± 47 mL (19%, P < 0.001). Also regarding LV function, BiV pacing was better in comparison with RV pacing, indicated by a higher mean LVEF (48 ± 12 vs. 41 ± 12%, P < 0.001) and higher LVFS (24 ± 7 vs. 21 ± 8%, P = 0.004). In 16 patients (57%) from this subgroup the difference between RV and BiV paced LVEF was >5% in favour of BiV pacing.


The present study shows that BiV pacing following chronic RV pacing improves LV function, reverses LV remodelling and improves NYHA classification in patients with cardiomyopathy associated with RV pacing, although not having an established indication for CRT.

Biventricular pacing in asymptomatic and mildly symptomatic patients with relatively mild left ventricular dysfunction

On average, patients from this study had less severe LV dysfunction and LV remodelling, as well as fewer clinical symptoms of heart failure, as compared with patients in other studies on CRT. Even though patients in the present study had these better conditions, response to BiV pacing was evident with clinically relevant reverse remodelling (LVESV reduction >15%) in 53% of the patients, and improvement in LV function in the majority of them. The finding that BiV pacing is beneficial in mildly symptomatic and even in asymptomatic patients (NYHA I–II subgroup) is supported by earlier investigations.2729 The present study shows that improvement of LV function and reversion of LV remodelling after BiV pacing may even occur in patients with less severe LV dysfunction (LVEF 36 ± 10%) as compared with other studies (i.e. REVERSE study LVEF 27 ± 7%).2729 Furthermore, the response to BiV pacing does not seem to be related to the severity of either LV dysfunction or LV remodelling.

Biventricular pacing and clinical status

In patients with severe heart failure BiV pacing has been shown to be beneficial regarding clinical status by many studies.13,16,17 The present study shows no evident benefit of BiV pacing in exercise performance or quality of life, despite significant beneficial treatment-effects of BiV pacing on LV function and remodelling. This finding might be explained by the fact that the parameters used were probably too rough to define modest benefits in those patients with a good clinical condition, leaving not much room for clinical improvement. Nevertheless, NYHA classification was improved by BiV pacing. Though, the clinical benefit of BiV pacing was not always paralleled by echocardiographic benefit.

Biventricular pacing in patients with a history of chronic right ventricular pacing

Although in most studies benefits of CRT are shown in patients with dyssynchrony due to an ‘intrinsic’ LBBB (ischaemic or idiopathic), in the current study beneficial treatment-effects of BiV pacing are shown in patients with RV pacing-induced dyssynchrony (paced LBBB). Improvement in LV function and reversion of LV remodelling are described for BiV upgrade in chronically RV paced patients by other studies too,2,16,17,30 but the history of chronic RV pacing was longer in the current study (10 ± 7 years). Delnoy et al.31 have shown similar clinical and echocardiographic improvement when CRT was applied in patients with preceding chronic RV pacing compared with patients with native LBBB. In the current study the response to BiV pacing does not seem to be related to the number of years of chronic RV pacing prior to the upgrade. To appreciate these findings, it is important to keep in mind that also in LBBB patients dyssynchrony may have existed for several years before the presence of symptoms becomes the final indication for ‘de novo CRT’. As CRT restores LV electrical and mechanical synchrony, and therefore improves LV function, it seems reasonable that response may be independent of aetiology and time-span of the underlying dyssynchronous activation pattern.

Implications of results

To date, the presence of severe LV dysfunction (LVEF < 35%) and moderate to severe clinical heart failure (NYHA classification III–IV) are widely used as criteria for CRT. Although the new American guidelines are less tenacious to the presence of clinical heart failure symptoms, severe LV dysfunction is still a strict criterion: it is advised to consider CRT in slightly symptomatic patients (NYHA II) only if LVEF <35%.32 Improvement of clinical condition and improvement in echocardiographic parameters in patients with mild heart failure suggest that CRT prevents or slows the progression to severe heart failure. In the prevention of LV dysfunction, LV remodelling and heart failure, it might be valuable to start BiV pacing in patients with an indication for permanent ventricular stimulation. However, the complication rate and unnecessary costs are important issues in a mildly symptomatic patient cohort and make (upgrade to) BiV pacing in every patient with a conventional pacemaker indication or an LBBB unfavourable. Nevertheless, our suggestion is to monitor LV function routinely and to carefully consider (upgrade to) BiV pacing from the moment that echocardiographic signs of remodelling or deterioration of LV function occur, whereas disregarding NYHA classification in this concern.

Limitations of the study

As this was a single-centre study, the study was limited to a relatively small number of patients. Nonetheless, in a crossover-design every patient acts as his/her own control, so paired analysis can be performed and fewer patients are needed to reach a powered level of significance. Using a crossover-design, placebo effects of study-inclusion and pacemaker upgrade to a BiV device can be excluded because, data from the crossover-phase were used for the assessment of treatment-effects. Furthermore, patients were blinded for pacemaker configurations. Conversely, in a single blind trial unintended bias by investigators cannot be totally excluded. To minimize this bias the investigators were unaware of the results of previous measurements in the individual patient.


We conclude that BiV pacing following chronic RV pacing may improve LV function and reverse LV remodelling even in patients with relatively mild LV dysfunction or remodelling, and no severe clinical symptoms of heart failure. In addition, in these patients clinical status can be improved by BiV pacing. Hence, upgrade to BiV pacing might be considered in patients with mild cardiomyopathy associated with chronic RV pacing, even though they have no established indication for CRT following current guidelines.


This work was supported by Boston Scientific Corporation (Natick, MA, USA).


We gratefully acknowledge the patients who participated in this study. We would also like to thank our echocardiographist Jos Habets BSc, for performing all echocardiographic investigations in the study.

Conflict of interest: F.P. is consultant to Medtronic Inc. (Minneapolis, MN, USA) and receives research grants from Medtronic Inc. (Minneapolis, MN, USA), Boston Scientific Corp. (Natick, MA, USA), and EBR Systems (Sunnyvale, CA, USA).


View Abstract