The usual reason for pacing is to prevent bradycardia, which is most commonly due to sick sinus syndrome or atrioventricular (AV) nodal disease. The need for pacing for either of these indications increases progressively with age. Atrial fibrillation (AF) is the most common serious arrhythmia. The incidence of AF also increases progressively with age, and patients with pacemakers may be even more prone to AF than comparable patients without pacemakers.1 Therefore, there has been continuing interest in discovering whether certain modes of pacing might be effective at preventing AF in patients with pacemakers.
The simplest question to ask is what chamber should be paced to decrease the chance of AF. Most, though not all, retrospective studies indicate that right atrial pacing decreases the incidence of AF compared with pacing in the ventricle.2 Andersen et al.3 compared single chamber right atrial pacing with single chamber right ventricular pacing in patients with sick sinus syndrome. They demonstrated significant benefit for atrial pacing with improved survival and a decreased incidence of thromboembolic events, AF, and congestive heart failure, especially after longer follow-up. The Mode Selection Trial in Sinus Node Dysfunction (MOST),4 in which all patients had dual chamber rate-responsive (DDDR) pacemakers implanted but were randomized to ventricular or dual chamber pacing, showed significant but less striking benefits in the incidence of AF and congestive heart failure. A possible explanation for the difference in the results between the two trials is that unintentional and unnecessary pacing of the right ventricle in the dual chamber group negated some of the benefits of sequential AV pacing compared with ventricular pacing. Two groups have examined the potential benefit of pacing at more than one site in the right atrium with the goal of activating the right and left atrium more synchronously. Saksena et al. found a small but not significant improvement in time to first AF and frequency of AF with dual site atrial pacing compared with conventional high right atrial pacing.5 Friedman et al. documented significant shortening of the P wave duration but observed only a tendency toward less AF with dual site compared with single site atrial pacing in a small study.6 Notable in both these studies is the significant improvement in AF with atrial pacing regardless of whether one or two sites is paced; the incremental benefit of dual site pacing is much smaller. There are theoretical reasons to think that pacing at a single site in the inter-atrial septum might accomplish similar atrial synchronization. Padeletti et al.7 showed that the incidence of AF was less with interatrial septal pacing compared with high right atrial pacing in a randomized prospective trial.
The next question is how fast to pace the atrium. Is a minimal rate required to prevent AF or is a minimal per cent atrial pacing needed? Will prevention algorithms that increase pacing rate in response to premature atrial contractions, prior episodes of atrial tachycardia, or following exercise be of any added benefit? The Atrial Dynamic Overdrive Pacing Trial (ADOPT) used an algorithm that increased the pacing rate to maintain atrial pacing regardless of the atrial rhythm. Compared with DDDR pacing with a lower rate of 60, patients had fewer symptomatic episodes reported by trans-telephonic monitoring, but device monitoring of atrial arrhythmia burden assessed by mode-switching duration was not affected.8 Blanc et al. compared prevention algorithms with DDD pacing at 70/min and showed no significant benefit in arrhythmias recorded by the pacemakers.9 With different outcome measures, it is not certain whether this study actually has a different result from the ADOPT trial. If indeed there is no efficacy, there are two possible explanations: pacing at 70 versus 60/min in the DDD mode may substantially decrease the incidence of AF, or the theoretically beneficial algorithms might actually be detrimental.
Another issue that may be important in these studies is the effect of unintended right ventricular pacing. Concerns about deleterious effects of right ventricular pacing come, in part, from studies using devices capable of pacing the right ventricle in patients with no indication for the pacing. The Dual Chamber and VVI (ventricular inhibited) Implantable Defibrillator Trial (DAVID)10 compared DDDR pacing with VVI pacing with dual chamber implantable cardioverter-defibrillators (ICDs) in patients with an indication for an ICD, ejection fraction ≤40%, but no indication for pacing. The unexpected result was an increased incidence of the composite end-point of hospitalization for CHF and death in patients programmed to DDDR mode compared with those programmed to VVI mode. In the VVI group there was almost no pacing at all in most patients (1% overall), but in the DDDR group at nominal AV delay settings, many patients had right ventricular pacing a high percentage of the time (60% of all ventricular beats) despite having no AV block indication for pacing. Interestingly, in their study of AF prevention algorithms, Blanc et al.9 found that a subgroup with less ventricular pacing showed some benefit of atrial arrhythmia prevention algorithms.
There has also been limited investigation of pacing to prevent AF in patients with no bradycardia indication for pacing at all. Lau et al.11 found that there was a significant improvement in the total duration of atrial arrhythmia episodes with dual site atrial overdrive pacing in patients with drug-refractory paroxysmal AF. However, there was not any symptomatic improvement associated with the lesser atrial arrhythmia burden.
For patients with sinus node disease, the easiest way to pace only the atrium is to implant an atrial pacemaker. This is a safe approach in properly selected patients,3 but many practitioners remain reluctant to implant an atrial pacemaker because of the small potential for developing heart block in the future and because of the perception that atrial leads may be less reliable. With a dual chamber pacemaker, programming a long fixed AV delay or using AV delay hysteresis, which lengthens the AV delay when ventricular sensing is occurring, can avoid right ventricular pacing. Recently, more sophisticated pacing algorithms have been developed that can maintain more AV conduction in patients with intermittent AV block.12 These approaches may not work for every patient because there are a small minority of patients who cannot tolerate atrial pacing with a long AV delay because of symptoms caused by poorly timed atrial contraction that are similar to pacemaker syndrome caused by ventricular pacing.
In patients with sick sinus syndrome, pacing the atrium is definitely of benefit. Pacing in the interatrial septum or at two sites may be a small additional advantage in delaying recurrence of AF. Pacing the atrium more, whether by increasing the lower rate limit or by means of an algorithm to provide consistent atrial pacing, is probably helpful, but results are not uniform among studies. The added benefit of more sophisticated algorithms to increase pacing in response to premature atrial beats or following paroxysmal atrial arrhythmias remains to be proven. On-going trials are evaluating the effect of AF prevention algorithms on prevention of AF in patients with sick sinus syndrome and paroxysmal AF, with an end-point of device logging of the AF burden. In some of the trials, prevention algorithms are combined with newer algorithms to minimize ventricular pacing. Even therapy effectiveness defined with the precision made possible by device logging of atrial arrhythmias does not guarantee that there will be clinical benefit in terms of morbidity or mortality, which would require larger trials than those presently in progress.