Because ventricular contraction is triggered by electrical activation, it seems logical to imagine that ventricles that are paced electrically might demonstrate differences in patterns of contraction from those that are activated by the intrinsic conduction system. However, attempts to define the relationship between the electrocardiographic event and the mechanical event have proven to be more challenging than one might expect. Echocardiography, magnetic resonance imaging, contrast ventriculography, and electrocardiographic imaging (ECGI) have all been used to characterize the complexities of left ventricular (LV) contraction and relaxation. Among these modalities, echocardiography has the advantage of being noninvasive as well as widely available. Its role in identifying optimal pacing sites on the basis of the characterization of dyssynchrony continues to evolve. Echocardiography offers a remarkable variety of techniques and measurements, with no clear gold standard for the characterization of LV dyssynchrony or for the prediction of clinical response.

The term dyssynchrony itself is difficult to define. Electrical dyssynchrony is not synonymous with mechanical dyssynchrony . Prolonged QRS duration (electrical dyssynchrony), as measured on standard 12-lead electrocardiography, remains today the most widely used method for identifying dyssynchrony and candidates for cardiac resynchronization therapy (CRT). ECGI provides some interesting insights into the complexities of dyssynchrony. ECGI is a functional, noninvasive mapping modality in which activation sequences are reconstructed from 250 body surface electrocardiographic traces and thoracic computed tomographic images. It also provides insight into lines of block that can result in the delayed activation of anatomic adjacent areas. Jia et al reported findings from a small number of patients with severe heart failure and left bundle branch block. ECGI of these patients revealed them to be a heterogeneous group, with variable LV activation patterns, regions of delayed conduction, and lines of block. The electrical heterogeneity is paralleled by mechanical heterogeneity.

In the past decade, there has been a paradigm shift in the field of clinical pacing. “Physiologic pacing” in patients with sinus node dysfunction has evolved to no longer simply mean atrioventricular (AV) synchronous pacing. Optimal ventricular pacing mandates minimal ventricular pacing in patients with sinus node dysfunction or predominantly intact AV conduction. Initial studies of hemodynamic parameters and quality-of-life measures established some benefits of AV synchronous pacing over single-chamber ventricular pacing. As clinical endpoints (the incidence of atrial fibrillation, heart failure, stroke, and mortality) were studied, it became apparent that even AV synchronous ventricular pacing was associated with increased morbidity and mortality. In the Mode Selection Trial (MOST), patients with sinus node dysfunction were randomized to ventricular only pacing or dual chamber pacing (DDD) (AV synchronous pacing). A secondary analysis of this trial demonstrated an increased incidence of atrial fibrillation and an increased risk for hospitalization for heart failure with increasing percentage of right ventricular (RV) pacing. Patients in this trial had normal or near normal ejection fractions. Next, the Multicenter Automatic Defibrillator Implantation Trial (MADIT) II showed a trend toward increased rates of hospitalization for heart failure in patients who received dual-chamber defibrillators compared with those patients who received single-chamber defibrillator. Because of concerns about the apparent deleterious effects of RV pacing, the Dual Chamber and VVI Implantable Defibrillator (DAVID) trial examined the composite endpoints of time to death and first hospitalization for heart failure. Patients were randomized to dual-chamber pacing at a lower rate of 70 beats/min or ventricular pacing at a lower rate of 40 beats/min. Because ventricular pacing was activated only when the intrinsic heart rate was slowed markedly in the latter group, these patients were paced less frequently than those in the former group. Mortality and heart failure hospitalizations were higher in the DDD group. These 3 large clinical trials, along with several small studies, have highlighted the long-term deleterious effects of RV apical pacing. These adverse effects include LV systolic and diastolic dysfunction, congestive heart failure, activation of the sympathetic nervous system, functional mitral regurgitation, increased risk for atrial fibrillation, and abnormalities in myocardial histopathology.

All current pacemakers have features that promote intrinsic ventricular conduction in patients in whom ventricular pacing is not required. This is accomplished by prolonging AV delays or with specific algorithms that allow pacemakers to automatically switch between atrial based pacing (AAI) and DDD with the occurrence of any second-degree or third-degree AV block (Managed Ventricular Pacing; Medtronic, Inc, Minneapolis, MN) In the Search AV Extension and Managed Ventricular Pacing for Promoting Atrio-Ventricular Conduction (SAVEPACe) trial, the use of this algorithm resulted in a 40% relative risk reduction in the development of persistent atrial fibrillation. Pacemakers implanted for the indication of sinus node dysfunction should be programmed to minimize ventricular pacing, regardless of the ventricular pacing site. Even patients who have intermittent AV block may benefit from algorithms that promote intrinsic ventricular depolarization.

Patients with advanced or complete heart block, and those receiving biventricular pacing systems for CRT, are paced virtually 100% of the time. It is these patients in whom the question of pacing site is important. Patients with normal or near normal LV function or those in whom CRT is not indicated currently receive single RV pacing leads. These are most often implanted in the RV apex (RVA) because of stability, accessibility, and operator preference. RVA pacing results in a dyssynchronous LV contraction. RV outflow tract (RVOT) and the RV mid septum have been proposed as alternate sites because they are associated with more favorable hemodynamics. However, the research supporting this conclusion is limited, with most studies focusing on acute hemodynamic and echocardiographic endpoints. A pooled analysis of 9 studies, comparing the acute and chronic hemodynamic effects of RVOT pacing and RVA pacing, demonstrated only a modest advantage to RVOT pacing in patients without systolic heart failure. When taken individually, the studies included in this meta-analysis yield conflicting results in terms of cardiac output, ejection fraction, and LV end-diastolic pressure. For example, in a study by Giudici et al, pacing from the RVOT improved cardiac output compared with RVA pacing in 89 patients. In a smaller study (16 patients) by Victor et al, there was no advantage to RVOT pacing on the basis of chronic follow-up of echocardiographic and radionuclide ventriculographic data. The Right Ventricular Outflow Tract Versus Apical Pacing (ROVA) study randomized 103 patients with chronic atrial fibrillation and heart failure to RVA or RVOT pacing. Quality-of-life scores at 3 months were not different between the two groups. Studies examining the effect of RV mid septum pacing on hemodynamic parameters of cardiac output (acutely) and LV ejection fraction (chronically) are favorable to RV mid septum pacing compared with RVA pacing but include a small number of patients.

Even if the RV pacing site associated with minimal dyssynchrony could be reliably identified, there are as yet no large trials demonstrating improved clinical outcomes. In fact, several small randomized studies showed no improvements in LV function or mortality when RVOT pacing was compared with RVA pacing. Gong et al were able to show that although RVOT pacing was associated with more synchronous LV contraction than RVA pacing, there was no difference in LV function. In that study, tissue Doppler imaging and 2-dimensional echocardiography were used to assess LV systolic and diastolic synchrony, LV volumes, and function. In another study, 122 patients with standard bradycardia indications for pacing were randomly assigned to RVA pacing or RVOT pacing. The follow-up period was a remarkable 10 years. RVOT pacing offered no mortality benefit over RVA pacing.

The study by Liu et al reported in this issue of JASE involved the use of real-time 3-dimensional echocardiography to study the acute impact of RVOT pacing on LV diastolic and systolic function. Ejection fraction and systolic mechanical synchrony were shown to be adversely affected with RVOT pacing compared with intrinsic depolarization. Liu et al’s findings provide additional data supporting the concept that when it comes to pacing, less is more. However, the study population included only patients with sick sinus syndrome, without left bundle branch block, and with intact AV conduction. Currently, the pacemakers in these patients would be programmed to minimize ventricular pacing.

An area of disappointment has been the failure to identify reproducible echocardiographic predictors of clinical response to CRT. The Predictors of Response to CRT (PROSPECT) study suggests that echocardiographic measures of ventricular dyssynchrony in patients receiving cardiac resynchronization devices are unable to identify a single measure of dyssynchrony that could predict response to CRT. Interobserver and intraobserver variability between the two highly experienced core labs were surprisingly and disappointingly high. Interobserver variability ranged from 6.5% to 72.1%, depending on the parameter studied. Three-dimensional echocardiography was not used in this study. Clearly, better methods for predicting which patients with heart failure are likely to respond to CRT are needed.

There remain many unanswered questions in pacing. Is there a single optimal pacing site in patients who do not have indications for CRT? Is there a maximal AV delay beyond which the benefits of minimal ventricular pacing are outweighed by the potentially deleterious effects on hemodynamics? Should patients with normal or near normal ventricular function and intraventricular conduction delays receive CRT pacemakers? What, if any, clinical benefits does selective site pacing offer over RVA pacing? Finally, what is the most practical and reproducible means for identifying the optimal site or sites in a particular patient? The reproducibility of this technique and the relationship of these measures to clinical endpoints will need to be studied. The ultimate clinical hypothesis, that pacing from a site associated with minimal dyssynchrony improves clinical endpoints, remains to be proved.