Biventricular Pacing in Pacemaker Dependency: One Size Does not Fit all

This editorial refers to ‘Preventing ventricular dysfunction in pacemaker patients without advanced heart failure: results from a multicentre international randomized trial (PREVENT-HF)’ by M. Stockburger et al., published in this issue on pages 633–641. Cardiac resynchronization therapy (CRT) has become standard therapy for heart failure patients with decreased left ventricular (LV) systolic function, with the indication extending recently to less symptomatic patients.1 Implantation rates in Europe have increased steadily with a clear difference between countries.2 Recent evidence suggests that CRT might also be useful in patients without heart failure who are pacemaker-dependent.3 It is, however, unclear if this can be translated into the whole population of pacemaker-dependent patients. In this issue of the European Journal of Heart Failure, Stockburger et al. describe the results of PREVENT-HF, a small randomized trial describing the outcome of biventricular stimulation in patients with atrioventricular (AV) block and no overt heart failure, in which no benefit of biventricular stimulation was demonstrated.4 This is in contrast to the larger randomized PACE study by Yu et al., in which a preservation of LV function was demonstrated in patients receiving a biventricular device whereas right ventricular (RV) stimulation led to a reduction in LV ejection fraction (LVEF) and an increase in LV end-systolic volumes.3 In HOBIPACE, a study that included patients with mild heart failure, biventricular stimulation appeared to be beneficial over RV stimulation.5 A retrospective analysis of the PROSPECT study demonstrated that a substantial percentage of patients had an LVEF >35% but showed a similar improvement to patients with more severely depressed LV function.6 The results of PREVENT-HF will stimulate more discussion and research into the use of biventricular pacing in patients with preserved LVEF, rather than providing a clear conclusion. Of note, the study is severely limited by small sample size, the significant number of patient drop-outs, crossovers, and lack of paired echocardiographic data leading to ~30% potential data loss (33 out of 108 patients), heterogeneous population (with LV end-diastolic volumes ranging from <50 to >250 mL), and underpowered design with the estimation of sample size based on a very old study of a systolic heart failure population. A possible extension of CRT to patients with heart failure with preserved ejection fraction (HFPEF) has recently been suggested7 even though this is not supported by a subgroup analysis of the PACE study.3 Patients with HFPEF, have a similarly poor prognosis to those with a low LVEF, but treatment options for these patients are limited.8 Recently, acute RV apical (RVA) pacing in bradycardic patients was shown to impair LV diastolic function in patients with normal LVEF, and was particularly noted in those with pre-existing LV diastolic dysfunction.9 Optimal programming of AV intervals and heart rate might therefore be even more critical in this patient group, as the improvement in diastolic filling is possibly the main mechanism of response.7 What is the rationale behind biventricular pacing in ventricular pacing-dependent patients with normal LV function? Right ventricular pacing induces interventricular and LV-intraventricular dyssynchrony similar to intrinsic left bundle branch block (LBBB). This can lead to an increase in LV volumes, and a decrease in LVEF and mitral regurgitation. Some patients even develop severely depressed LV function within a short time frame, a condition that has been coined ‘dyssynchronopathy’ in a canine model of AV block that develops rapid heart failure.10 Development of a dyssynchronopathy might also explain the well-known but ill-defined group of super-responders to CRT.11 It should be noted however, that this only occurs in a subgroup of patients. It is unclear what factors lead to dyssynchronopathy, but they could range from genetic and risk factor vulnerability to environmental effects. Figure 1 summarizes these ideas. If this condition develops, it is also questionable whether full recovery is possible. With the lack of predictive markers, it might be feasible to implant a biventricular device with a low threshold to prevent heart failure. What is the evidence against routine placement of a biventricular device in patients that are ventricular pacing-dependent? Complications of LV lead placement are not uncommon. Although there were few complications in the PREVENT-HF and PACE studies,3,4 the success of LV lead placement is ~90% in most studies and the rates of subacute complications such as lead dislodgements or lead failure and phrenic nerve stimulation are ~10%. With the rapid development of better tools for LV lead delivery (e.g. coronary vein subselector), improved design of LV leads (e.g. downsizing and development of a quadripolar lead), and the advancement of implantation skills, most of these complications can now be avoided or modified by post-implant programming. Battery longevity is another worrisome disadvantage of biventricular devices, especially if high LV lead thresholds have to be used. Battery life in combination with higher costs of the device could put a significant burden on healthcare costs, if biventricular stimulation becomes the standard of care to prevent heart failure in ventricular-pacing dependent patients. Why not wait for deterioration in LV function due to RV pacing in ventricular pacing-dependent patients before upgrading? Upgrading to a biventricular device is associated with significant complications, reported to approach 20% in recent registry data.12 Furthermore, structural remodelling in heart failure is associated with fibrosis and cell death in both the atria and ventricles that might not be fully reversible once it has developed.13 Where do we go from here? It might be too early to schedule all patients who are expected to be continuously paced in the ventricles for a biventricular device. Until we find good predictors of poor prognosis, a practical approach needs to be used taking into account the age of the patient, systolic and possibly diastolic LV function, and co-morbidities. In cases of mild systolic dysfunction or severe diastolic dysfunction, biventricular stimulation might be a good choice. If, however, a standard DDD system is chosen, frequent echocardiograms are needed to detect worsening of LV function. In the case of otherwise unexplained worsening of LV function, early upgrade to a biventricular device should be considered. Upgrading is, however, associated with significant complications and possibly incomplete recovery. Of note, the relationship between pacing-induced LBBB, development of dyssynchrony and LV remodelling, and subsequent evolution of heart failure is complex. A recent cross-sectional study observed that the prevalence of LV intraventricular dyssynchrony in patients with preserved LVEF who had RVA pacing-induced LBBB was 54% compared with 73% in systolic heart failure patients with coexisting LBBB.14 From the analysis of the PACE study, about half of the patients who received RVA pacing developed early systolic dyssynchrony, and this phenomenon was associated with a seven-fold risk of developing adverse LV remodelling at 1 year (data presented in Clinical Trial Update session of European Society of Cardiology meeting 2010). Therefore, the quest for predictive markers for RV pacing-induced or intrinsic LBBB-induced heart failure is an important research goal in this field. Conflict of interest: none declared.