OUP user menu

Base over apex: does site matter for pacing the right ventricle?

Michael D. Gammage
DOI: http://dx.doi.org/10.1093/europace/eun087 572-573 First published online: 10 April 2008

Pacing for the relief of bradycardia might be considered a ‘mature’ therapy, having now been available for 50 years, so it is refreshing to see that there are still many unanswered questions relating to the practice of pacing the ventricle. Although it seems clear from population-based studies that ventricular pacing increases the risk of left ventricular dysfunction and that the cumulative percentage of pacing is directly related to that risk,1 it is unclear whether it is simply pacing the ventricle or the site of ventricular pacing that is the real problem. Perhaps, the most compelling evidence that right ventricular (RV) apical pacing may be deleterious came from the DAVID study2 which compared back-up VVI pacing at 40 ppm and DDDR pacing at 70 ppm. The VVI group did not require much ventricular pacing compared with the DDDR group. Although many patients had impaired LV function, only 12% had NYHA class III or IV symptoms and patients who were paced more frequently were more likely to have an adverse outcome in terms of heart failure. The debate about RV pacing is further confused by clinical anecdote (many patients seem to remain perfectly well despite RV apical pacing) and variable results from previous studies.35 These discrepancies have usually been put down to duration of pacing, differences in study design, variation in pacing site studied, and inadequate duration of follow-up.

The paper from Kypta et al.6 in this issue of Europace seems to go against the recent trend in reports in that the authors have shown no significant difference in cardiac function in patients paced from either the RV apex or the RV septum. This study has the benefit of a well-defined target site in the RV, an appropriately long follow-up period to allow for ventricular remodeling, and a reasonably large number of patients. So why do the results differ from those of Tse et al.3? Some answers may lie in the population studied and the methodology used by Kypta et al. Although broadly similar in demographics, more patients in the apical pacing arm presented with complete heart block and a narrow QRS complex and less were on ACE inhibitor or AIIRB therapy, suggesting the possibility of less pre-existing remodelling in this group. Although not supported by the pre-implant echo data, it is possible that underlying or co-existent cardiac disease may influence the remodelling process in response to RV pacing. With regard to methodology, N-terminal pro-brain natriuretic peptide (NT-pBNP) was measured and showed no differences at 3 or 18 months. Although a good surrogate for overall LV function, the range of measurements was large and it might have been better to study individual change in NT-pBNP (ΔNT-pBNP) as a more sensitive indicator of the remodelling process, as the implication of changes in NT-pBNP in the absence of overt heart failure remains unclear. Left ventricular ejection fraction was measured by echocardiography (unlike the study of Tse et al., 3 where radionuclide ventriculography was used); although widely utilized few studies in RV paced patients have managed to show differences with this technique. It may be that the tools routinely used for these studies are too blunt to show differences in a process that may take years to occur, although one might also ask whether such subtle changes are of functional significance.

A recent paper by Varma7 may offer some insight into this conundrum. Varma studied patients with normal and abnormal LV function, measuring LV activation delays in response to RV apical pacing. The results showed that RV apical pacing in patents with normal LV function increased mean QRS duration to 151 ms and delayed inferolateral LV activation by a mean of 76 ms, effects that were almost indistinguishable to those seen with intrinsic conduction in spontaneous LBBB. In patients with underlying LV dysfunction, however, the effects of RV apical pacing were much greater. What was, perhaps, most interesting was that the effects of RV apical pacing varied considerably in patients with normal LV function. This study was, of course, acute, but might help to explain why some patients seem to do very well with RV apical pacing. Whether those with the greatest extension of QRS and/or the greatest LV activation delay are at most risk of developing LV dysfunction remains to be shown. Studies in young patients with congenital heart block have shown that medium to long-term pacing is associated with deterioration of LV function8 and that development of LV dysfunction and heart failure hospitalizations is correlated with wider paced QRS.9

Another interesting point to note is that no study of non-RV apical sites has ever shown a detrimental effect of these sites when compared with RV apical pacing; in other words, there is no evidence to suggest that the patient will be less well served by non-RV apical pacing. This point is addressed by Kypta et al.6 in their discussion. Given the large range of ventricular function seen in patients presenting with a need for ventricular pacing, it seems unsurprising that a ‘one size fits all’ approach to pacing the RV would be appropriate; an individually selected, targeted approach would seem more likely to be successful. Several large, multi-centre studies are currently examining the medium-term effects of pacing from defined, selected sites within the RV (Protect-Pace, Optimize-RV, RASP) but will be unlikely to report before 2010. How to undertake a simple and effective pre-implant or on-table assessment of the best pacing site that will optimize LV function in an individual patient remains a matter for further study.

So how to advise the implanting physician faced with a patient with high-degree heart block and a normal ventricle? There have been advocates for the use of left ventricular or biventricular pacing in such patients, but no evidence currently exists to support this strategy or indeed to confirm that this will be as safe as RV pacing. Many algorithms now exist to reduce RV pacing and the use of these is also advocated wherever possible,10 but this approach is not suitable for the patient with permanent high degree heart block in whom pacing the ventricle for the majority of the time is mandatory. Perhaps, the simplest message would be to retain an open mind about the importance of RV pacing site, to minimize ventricular pacing wherever possible, and to attempt to select a pacing site in any individual patient that may result in the minimum reduction of LV function. As there seems to be no detriment to pacing the RV at sites remote from the apex, it would seem logical that the time has come to abandon the RV apex as the default site and train the next generation of implanters to routinely place leads elsewhere.

Conflict of interest: none declared.


  • The opinions expressed in this article are not necessarily those of the Editors of Europace or of the European Society of Cardiology.

  • doi:10.1093/europace/eun085