Pacing and Clinical Electrophysiology

Many studies have shown that cardiac resynchronization therapy (CRT) reduces symptoms, improves exercise capacity, and decreases mortality in patients with heart failure and left ventricular (LV) dyssynchrony.^1–5 Although atrial fibrillation (AF) occurs in approximately 30% of heart failure patients,⁶ the arrhythmia was an exclusion criterion in most of the major CRT trials perhaps for the reason stated in the CARE-HF trial that such patients "cannot benefit from the atrial component of CRT."⁵ However, mounting evidence indicates that CRT in AF patients produces a benefit similar to patients in sinus rhythm^7–13 with aggressive control of the ventricular rate especially with atrioventricular (AV) node ablation.^7,14 Consequently, CRT is now increasingly applied to patients with persistent or permanent AF.

Apart from other more complex issues, AF raises one simple question: Do we need an atrial lead for CRT? There are obvious arguments against atrial lead implantation in persistent or permanent AF: (1) Increased hardware carrying a higher risk of large vein thrombosis or loss of lead integrity, (2) longer implantation and fluoroscopy time, (3) risk of atrial lead dislodgement or other complications associated with the atrial lead, and (4) increased costs. Costs may not really apply in most centers because CRT generators come only with connection ports for atrial and ventricular leads. Additionally, in many countries, the overall prices for CRT include all connected leads.

The above arguments are not an issue in patients with sinus rhythm. So, what are the arguments in favor of atrial lead implantation at the time of CRT implantation in patients with permanent or persistent AF?

First, information about the atrial rhythm may be helpful in patients with implantable cardioverter-defibrillators (ICDs) for the discrimination between ventricular and supraventricular tachycardia. Though so-called "enhanced detection criteria" such as sudden onset, cycle length stability, and analysis of the morphology of the sensed ventricular signal can discriminate between ventricular and supraventricular tachyarrhythmias, this may be particularly difficult in AF patients with "pseudo-regular" conduction, i.e., ventricular cycle length variations <40–50 ms. Additionally, AF may organize to atrial flutter, again with regular conduction to the ventricle. In these situations, it may be impossible to determine the appropriateness of tachycardia classification by the device based on ventricular electrograms alone. Despite the superior interpretation of stored dual-chamber electrograms by dual-chamber devices (e.g. shown by Kim et al. ¹⁵), atrial leads in AF patients are rarely implanted specifically for this reason.

More importantly, biventricular pacing without an atrial lead may be suboptimal, ineffective, or even detrimental in patients with "permanent" AF who revert to sinus rhythm during CRT (Fig. 1). Improved atrial mechanical function increases the risk of adverse hemodynamics leading to pacemaker syndrome especially in the setting of 1:1 retrograde ventriculoatrial (VA) conduction.^16,17 A recent study showed that CRT reduces atrial diameter, induces atrial reverse remodeling, and improves atrial pump function.¹⁸ These factors may increase the likelihood of spontaneous reversion to sinus rhythm, even after long-lasting AF—as reported in several cases.^19,20 The decision to implant an atrial lead, therefore, hinges on the probability of AF reverting to sinus rhythm in the future.

CARE-HF and COMPANION trials have not shown a reduction of AF in patients randomized to CRT versus those restricted to optimal medical therapy.^4,21 However, both trials excluded patients with AF at the time of enrollment and only a history of AF was allowed. Kies et al. implanted CRT devices in 74 patients with advanced heart failure and persistent (n = 20) or permanent (n = 54) AF.²² Eighteen patients were cardioverted at the time of CRT implantation. After 6 months, only five patients remained in sinus rhythm. No patient reverted spontaneously to sinus rhythm, leaving 93% of this patient group in AF. While these results seem to suggest that maintenance of sinus rhythm in AF-CRT patients is rare, there was no systematic attempt to maintain sinus rhythm. Furthermore, cardioversion was performed at a time sinus rhythm was least likely to be maintained. The ideal time for cardioversion in the setting of reverse LV remodelling would be after 3 or 6 months but not at the time of implantation. In contrast to these findings, Hügl et al. observed a significant gradual decrease in the average time that 84 patients with a CRT device spent in AF (AF burden).²³ During a period of 3 months, the AF burden decreased from a mean of 10 to 4 hours per day. This has been confirmed in 28 consecutive patients who received an upgrade procedure from dual-chamber pacing to CRT.²⁴ Using device-memory functions, these workers found only 14% of patients were free from AF with dual-chamber systems 3 months before upgrade compared to 90% 1 year after CRT implantation. At the same time, the number and duration of AF recurrences and the necessity of hospitalization for heart failure or arrhythmia were decreased. More recently, Delnoy et al. reported their experience in 96 CRT patients with permanent or persistent AF.²⁵ They implanted atrial leads in patients with AF lasting <2 years. The mean follow-up was 2 years and antiarrhythmic drug therapy after CRT was optimized with a significant increase in amiodarone administration. In this study, 25% of 96 AF patients were in sinus rhythm after 1 year. Obviously, eight patients received cardioversion while 16 patients reverted to sinus rhythm spontaneously. At 2 years, 21% of the AF group was in sinus rhythm. Similarly, preliminary data in abstract form, about 30 patients with AF for >3 years, showed that 57% of patients could be brought back to sinus rhythm after approximately one year of CRT.²⁶ However, the duration of ensuing sinus rhythm was not stated.

The incidence of successful cardioversion in CRT patients with persistent or seemingly "permanent" AF is unclear. Certainly, a single cardioversion at the time of implantation is insufficient. Most likely, chronic amiodarone treatment and cardioversion after signs of stabilized LV and LA reverse remodeling (e.g., smaller left atrial diameter) appear to be a promising approach. The potential of CRT to revert AF to sinus rhythm may be independent from the influence on new-onset AF. Of note, the label of "permanent" for AF is a matter of judgment and simply implies that no further attempt to achieve and maintain sinus rhythm is worthwhile. Such a label should not discourage attempts to restore sinus rhythm in AF-CRT patients with drug therapy (mostly amiodarone) and/or appropriately timed electrical cardioversion after the remodeling process is largely completed. The suggestion²⁵ to implant atrial leads in CRT patients with AF lasting <2 years may have merit. This recommendation could easily be extended to more or even almost all patients with AF because AV synchrony delivered by an atrial lead may dramatically improve heart failure. In the absence of an atrial lead, reversion to sinus rhythm may provoke pacemaker syndrome and hemodynamic deterioration. The additional effort and time required to implant an atrial lead on top of two ventricular leads seems, in fact, negligible. While awaiting return of sinus rhythm, CRT devices need to be programmed to VVI(R) or DDD(R) with reliable mode switching function (i.e., continuous detection of AF). Our suggestion to add an atrial lead in AF patients undergoing CRT can only be provisional since it requires confirmation by a prospective randomized study, which has not been performed so far.