OUP user menu

Impact of atrial antitachycardia pacing and atrial pace prevention therapies on atrial fibrillation burden over long-term follow-up

Anne M. Gillis, Margaret Morck, Derek V. Exner, Robert S. Sheldon, Henry J. Duff, Brent L. Mitchell, George D. Wyse
DOI: http://dx.doi.org/10.1093/europace/eup115 1041-1047 First published online: 21 May 2009

Abstract

Aims Selective atrial pacing algorithms have been developed for prevention of atrial tachycardia/atrial fibrillation (AT/AF). Although short-term studies have shown modest to minimal incremental benefit of these algorithms compared with conventional dual-chamber (DDD/R) pacing for prevention of AT/AF, the long-term effects of these algorithms are unknown. Accordingly, we compared atrial antitachycardia pacing (ATP) therapy and combined atrial ATP and atrial pace prevention (ATP + Prevention) algorithms to conventional DDD/R pacing for prevention of AT/AF over long-term follow-up.

Methods and results Seventy-one patients with AT/AF following pacemaker insertion were randomized to DDD/R pacing, DDD/R plus ATP pacing, or DDD/R plus ATP and prevention pacing and followed for 3 years. Atrial tachycardia/AF burden and an AF symptom scale were compared over time between groups. Atrial tachycardia/AF burden remained stable over 3 years in the DDD/R and ATP + Prevention groups. Atrial tachycardia/AF burden increased significantly over time in the ATP group. Patients not on class I or III antiarrhythmic drug therapy were more likely to experience an increase in AT/AF burden over time.

Conclusion Atrial ATP and atrial ATP in combination with atrial pace prevention algorithms do not suppress AT/AF over long-term follow-up compared with DDD/R pacing.

  • Atrial fibrillation
  • Atrial pacing algorithms
  • Dual-chamber pacing
  • Antitachycardia pacing

Introduction

Atrial fibrillation (AF) occurs frequently in the pacemaker population.14 A number of atrial pacing algorithms have been developed specifically for the prevention of AF. These include atrial overdrive pacing to suppress atrial premature beats, algorithms to prevent a pause following an atrial premature beat, transient overdrive pacing following termination of AF, and atrial antitachycardia pacing (ATP) for termination of atrial tachycardia or atrial flutter that might degenerate into AF.215 A number of clinical trials have evaluated the efficacy of one or more of these algorithms to prevent AF. Although some studies were initially promising,2,5,7,8,10 larger randomized clinical trials have not demonstrated substantial benefits of these therapies for the prevention of AF over relatively short-term follow-up.2,3,6,1215 However, the impact of such therapies over long-term follow-up has not been reported.

Clinical trials comparing atrial or dual-chamber pacing with ventricular pacing have observed a delay in the benefit of atrial pacing for prevention of AF ranging from 6 months to 2 years.16,17 The present study was designed to assess the efficacy rates of atrial ATP and atrial pace prevention therapies on AF burden over the long term. We tested the hypotheses that atrial ATP therapy or atrial ATP therapy in conjunction with three atrial pace prevention therapies would reduce the burden of atrial tachycardia/AF (AT/AF) over long-term follow-up.

Methods

Study population

The study population consisted of patients receiving a Medtronic AT 501/500 pacemaker in the setting of documented symptomatic bradycardia and previously documented AF. The vast majority of patients (97%) had sinus node dysfunction and paroxysmal AF as the indication for pacing. Chronotropic incompetence was not quantified at the time of pacemaker implantation. To be eligible, patients had to have experienced at least three episodes of AT/AF lasting longer than 5 min in duration within the 2-month period prior to randomization, they were on stable drug therapy for suppression or rate control of AF and had no medical illnesses that might impact survival. Exclusion criteria included permanent AF, a history of cardioversion for persistent AF, prior atrioventricular (AV) junction ablation, geographic isolation, or inability to give informed consent. Patients were enrolled between February 2001 and January 2004 and followed for 3 years. This study was approved by the Conjoint Health Research Ethics Board at the University of Calgary and was registered as a clinical trial at https://register.clinicaltrials.gov.

Study protocol

Consenting patients were randomized to one of three treatment limbs: conventional dual-chamber (DDD/R) pacing group, DDD/R plus atrial ATP ON (ATP group), or DDD/R plus atrial ATP and three atrial pace prevention therapies ON (ATP + Prevention group). Randomization was performed using a random number generator in blocks of 12. The majority of patients (60%) were randomized 2 months following pacemaker implantation, 15 patients (21%) were randomized at the 6-month follow-up visit, and 19% were randomized within 30 months of pacemaker implantation. The 2-month baseline period reflects AF burden measured over the 2-month period just prior to randomization.

A group with only atrial pace prevention therapies ON was not included since data at the time of the trial design in late 2000 suggested that these therapies in the absence of atrial ATP had minimal effect on AF outcomes.14 Other studies suggesting a potential benefit of atrial pace prevention therapies were not published until the study was underway. Patients were assessed every 3 months for the first year following randomization and every 6 months thereafter up to 3 years. Atrial tachycardia/AF burden was determined at each follow-up visit. Patients completed the University of Toronto Atrial Fibrillation Severity Scale annually.18,19 Patients were censored from further follow-up prior to completion of the 3 year follow-up period in the case of death or severely symptomatic AT/AF requiring major changes in therapy (i.e. initiation of Class I or III antiarrhythmic drug therapy, need for electrical cardioversion, programming on atrial ATP therapy, or need for AV junction ablation). A total sample size of 120 patients (40/group) were calculated assuming a 33% reduction in AT/AF burden (estimated to be 4 h/day at the end of follow-up in the DDD/R group) with ATP and prevention therapies programmed on (power 0.80, α = 0.05). The study was terminated early due to phasing out of the AT500/501 pacemaker.

Pacemaker programming

Patients received a Medtronic AT 500/501 dual-chamber pacemaker. The features of this device have been described in detail previously.12,14,20,21 All atrial leads were positioned in the right atrial appendage. Pacemakers were programmed to DDD/R with a lower rate of 70 bpm and an upper rate dependent on the patient's age and underlying heart disease. The paced AV interval was programmed to 240 ms and the sensed AV interval to 210 ms. Mode switching was enabled. The AT 500/501 pacemaker has diagnostic features that permit detection and storage of information about the date, time of onset, duration, and rhythm classification of all episodes of AF or AT. Information on the number of daily AT/AF events and total duration of AT/AF episodes per day are stored by the device for up to 14 months.11,2022 The pacemaker was programmed to collect marker channel annotations and stored intracardiac atrial electrograms on the first 30 and the last five episodes of AT/AF.11,2022 During follow-up, each episode of AT/AF with a stored atrial electrogram was reviewed to confirm appropriate detection and classification of AT/AF.

The device has three pacing algorithms designed to prevent AT/AF recurrence. These include an atrial overdrive algorithm designed to maintain the atrial pacing interval just above the intrinsic rate (atrial preference pacing), a mode designed to prevent short and long intervals after a premature atrial contraction (atrial rate stabilization), and a post-mode-switch overdrive pacing mode designed to prevent early recurrence of AT/AF after episode termination.11,12,14 The programming parameters are summarized in Table 1. For episode termination, two rate-adaptive atrial ATP therapies are available (Burst+ and Ramp) to treat AT episodes. If the initial atrial ATP therapies programmed were ineffective (efficacy rate <60%), more aggressive ATP therapies could be programmed during follow-up.

View this table:
Table 1

Programming parameters at randomization

ParameterDDD/RATPATP + Prevention
Detection
 AF interval200200200
 AT interval200–300200–300200–300
 A sensitivity0.450.450.45
Bradycardia pacing
 Lower rate707070
Prevention
 Atrial rate stabilization
  Increment (%)OffOff25
  PVARP (ms)220
  Minimum pacing interval (ms)500
 PMOP
  Overdrive rate (bpm)OffOff80
  Overdrive period5 min
 Atrial pacing preference
  DecrementOffOff50 ms
  Search beats10
 Atrial ATP
  AT/AF duration to initiate (min)11
  ATP therapy 1OffBurst +Burst +
  ATP therapy 2RampRamp
  ATP therapy 3RampRamp
  • AF, atrial fibrillation; AT, atrial tachycardia; ATP, atrial antitachycardia pacing; DDD/R, dual-chamber pacing; PVARP, post-ventricular atrial refractory period; PMOP, post-mode switch overdrive pacing.

Study outcomes and data analysis

The effect of pacing therapies on AT/AF burden over time was the primary efficacy endpoint.11,20,21 Burden was defined as the quantity of AT/AF (h/day) retrieved from the device data logs. Baseline AT/AF burden was measured 2 months prior to randomization. The proportion of patients in each group progressing to permanent AF and changes in the University of Toronto AF Severity Scale were compared over time.18,19 Permanent AT/AF was defined as an AT/AF burden of 24 h/day for at least 6 months.22 Factors influencing AT/AF burden including antiarrhythmic drug therapy were compared. Differences in AT/AF burden were compared by factorial analysis of variance. Differences in proportions were compared using a χ2 analysis. Data are presented as mean ± SD or median values where appropriate. A value of P < 0.05 for two-sided comparisons was considered significant.

Results

Study population

The characteristics of the study population are shown in Table 2. Seventy-one patients were randomized to DDD/R (n = 21), DDD/R plus atrial ATP therapy (n = 24), or DDD/R plus atrial ATP and AF prevention therapies (n = 26). Despite randomization, the ATP group tended to have more patients with valvular heart disease compared with the other two groups (P = 0.06). Significantly fewer patients in the ATP group were on class I/III antiarrhythmic drug therapy compared with the DDD/R group (P < 0.05).

View this table:
Table 2

Characteristics of study population

DDD/RATPATP + Prevention
Number212426
Age (years)68 ± 771 ± 874 ± 7
Male12 (57%)7 (29%)10 (39%)
Indications for pacing
 Sinus node disease21 (100%)22 (92%)26 (100%)
 AV block0 (0%)2 (8%)0 (0%)
Cardiovascular disease13 (62%)18 (75%)19 (73%)
 Hypertension8 (38%)14 (58%)14 (54%)
 Congestive heart failure3 (14%)7 (29%)5 (19%)
 Coronary artery disease9 (43%)9 (38%)8 (31%)
 Cardiomyopathy0 (0%)2 (8%)0 (0%)
 Valvular1 (5%)4 (17%)0 (0%)
LV function
 Normal18 (86%)16 (67%)20 (77%)
Cardiovascular drugs
 Class I/III antiarrhythmic18 (86%)12 (58%*)18 (69%)
 Statin10 (48%)6 (25%)12 (46%)
 ACE/ARB11 (52%)13 (54%)14 (54%)
 β-Blocker10 (48%)7 (29%)12 (46%)
 Calcium channel blocker7 (33%)13 (54%)14 (54%)
  • ACE, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; ATP, atrial antitachycardia pacing; AV, atrioventricular; DDD/R, dual-chamber pacing; LV, left ventricle.

  • *P < 0.05.

Patients were followed for 3.0 ± 1.0 years. One patient developed a lead fracture and data was censored during this time interval until the lead was replaced. On the basis of the review of AT/AF episodes with stored intra-atrial electrograms, atrial undersensing, or farfield R wave oversensing resulting in inaccurate detection of AT/AF or inaccurate classification of an episode termination was not observed. Four patients died during follow-up, one in the DDD/R, two in the ATP, and one in the ATP + Prevention groups. The cause of death was cardiovascular in one due to an intracerebral bleed. Five patients were censored at 1.4 ± 0.8 years prior to study completion due to a major change in therapy, three in the DDD/R, one in the ATP, and one in the ATP + Prevention groups. Amiodarone was initiated in one patient, electrical cardioversion was performed in two patients, atrial ATP was programmed on in one patient for the treatment of atrial flutter, and one patient underwent AV junction ablation. Twelve patients (17%) developed permanent AF during follow-up.

Atrial tachycardia/atrial fibrillation burden over time

The AT/AF burden for each randomized group at baseline (2 months prior to randomization) and at each annual follow-up visit is shown in Figure 1. Atrial tachycardia/AF burden ranged from 0.03 to 20.3 h/day (median 1.8 h/day), 0.03 to 9.0 h/day (median 0.7 h/day), and 0.04 to 23.0 h/day (median 0.8 h/day) at baseline in the DDD, ATP, and ATP + Prevention groups, respectively. Atrial tachycardia/AF burden decreased slightly at 1 year compared with baseline and then tended to return to baseline values in the DDD and ATP + Prevention groups, but these differences were not statistically significant. Atrial tachycardia/AF burden increased over time in the ATP group, and at 3 years, this increase was statistically significant compared with the baseline (P < 0.0005) and 1 year (P < 0.002) follow-up assessment (Figure 1). After 1 year of follow-up, the AT/AF burden decreased ≥30% in 52% of the patients in the DDD/R group, in 37% of the patients in the ATP group, and in 40% of the patients in the ATP + Prevention group. The proportion of patients with a reduction in burden by ≥30% did not vary significantly over time. Permanent AF developed at 3 years of follow-up in 14% of the DDD/R group, 25% of the ATP group, and 8% of the ATP + Prevention group (P = 0.22).

Figure 1

Median atrial tachycardia/atrial fibrillation (AT/AF) burden at baseline and 1, 2 and 3 years of follow-up in patients randomized to dual-chamber pacing (DDD/R), DDD/R and atrial antitachycardia pacing (ATP) or DDD/R and atrial ATP plus atrial pace prevention (ATP + Prevention) therapies. The AT/AF burden increased significantly at 3 years compared with baseline and 1 year follow-up in the ATP group. Data are boxplots with 25th and 75th percentiles at the top and bottom borders; the bars indicate the range of the data.

Drug use and atrial tachycardia/atrial fibrillation burden

Sixty-eight per cent of patients were on class I/III antiarrhythmic drugs. Atrial tachycardia/AF burden was similar at baseline between the group on antiarrhythmic drug therapy and the group not on antiarrhythmic therapy (Figure 2). During long-term follow-up, the AT/AF burden significantly increased in the group not on class I/III antiarrhythmic drug therapy (P < 0.001). An increase in AT/AF burden was observed at 2 and 3 years of follow-up in each of the treatment subgroups not on class I/III antiarrhythmic drug therapy (Figure 2). Factorial ANOVA indicated that the use of class I/III antiarrhythmia drugs and not the three different pacing therapies determined the differences in AT/AF burden noted at 2 and 3 years (P < 0.05).

Figure 2

Impact of the use of Class I/III antiarrhythmic drug therapy on atrial tachycardia/atrial fibrillation (AT/AF) burden over time. Atrial tachycardia/atrial fibrillation burden remained stable over time in the group on antiarrhythmic drug therapy, whereas AT/AF burden increased significantly over time in the group not on antiarrhythmic drug therapy. Data are boxplots with median data (horizontal bar) and 25th and 75th percentiles at the top and bottom borders; the bars indicate the range of the data.

Per cent ventricular pacing

The per cent of atrial and ventricular pacing during follow-up is shown in Table 3. The AV interval was increased at follow-up if required to promote intrinsic AV conduction. Thus, the per cent of ventricular pacing decreased over time. Annual ambulatory ECG monitoring demonstrated frequent ventricular pseudofusion during pacing (35% at year 1). The relationship between AT/AF burden and the proportion of ventricular pacing is shown in Figure 3. A significant linear relationship between % ventricular pacing and AT/AF burden was not observed. However, the proportion of patients with AT/AF burden >1 h/day was significantly greater in patients paced in the ventricle ≥40% of the time (Figure 3). The burden of AT/AF increased over time in the group paced in the ventricle ≥40% of the time.

Figure 3

Relationship between atrial tachycardia/atrial fibrillation (AT/AF) burden at 1 (A) and 2 years (B) follow-up and per cent ventricular pacing. A linear relationship between per cent ventricular pacing and AT/AF burden was not observed. The proportion of patients with AT/AF burden ≥1 h/day was greater in the groups paced in the ventricle ≥40% of the time at years 1 and 2 compared with those paced <40% of the time (P < 0.05). The proportion of patients with AT/AF burden ≥1 h/day increased over time in the group paced ≥40% of the time.

View this table:
Table 3

Device reported data during follow-up

% Ventricular pacingDDD/RATPATP + Prevention
 Baseline74.4 ± 29.272.2 ± 28.480.l ± 24.7
 1 year72.2 ± 32.865.7 ± 31.279.3 ± 24.2
 2 year61.0 ± 36.758.8 ± 33.173.4 ± 32.5
 3 year59.9 ± 36.546.7 ± 29.164.6 ± 39.2
Atrial pacing
 Baseline83.9 ± 22.474.3 ± 29.582.7 ± 21.9
 1 year91.5 ± 15.777.7 ± 21.597.2 ± 4.1
 2 year90.9 ± 15.175.0 ± 28.992.8 ± 11.6
 3 year95.6 ± 4.371.9 ± 27.489.4 ± 22.7
AT/AF Episodes/Day
 Baseline5.2 ± 11.05.4 ± 15.34.5 ± 7.9
 1 year2.3 ± 4.54.9 ± 13.17.0 ± 23.9
 2 year7.1 ± 11.66.4 ± 18.45.5 ± 13.5
 3 year6.4 ± 15.65.8 ± 15.26.8 ± 15.5
  • Data are mean ± 1 SD. AT/AF, atrial tachycardia/atrial fibrillation; ATP, atrial antitachycardia pacing; DDD/R, dual-chamber pacing.

Atrial fibrillation severity scale

The AF severity score was similar between the three groups at baseline: 28 ± 11, 28 ± 10, and 24 ± 10 in the DDD/R, ATP, and ATP + Prevention groups, respectively (P = NS). The AF severity score did not change significantly over time in any group. A linear correlation was observed between the AF severity scale and the per cent of time the heart rate was >100 bpm during the longest episode of AF detected during follow-up r = 0.38, P < 0.002.

Discussion

Previous studies evaluating the impact of atrial pacing algorithms or atrial ATP in combination with atrial pacing algorithms for prevention of AF failed to show a substantial incremental benefit compared with dual-chamber pacing alone. Most trials were relatively short ranging from 3 months of therapy in ATTEST (Atrial Therapy Efficacy and Safety Trial)12 to 6 months of therapy in ASPECT (Atrial Septal Pacing Efficacy Clinical Trial)14 and ADOPT (Atrial Dynamic Overdrive Pacing Trial).13 Since clinical trials of atrial or dual-chamber pacing vs. ventricular pacing have observed a delay in the benefit of atrial pacing for prevention of AF ranging from 6 months to 2 years,16,17 we hypothesized that long-term therapy might be required to demonstrate a benefit of selective atrial pacing algorithms for AF suppression.

Atrial fibrillation prevention therapies

In the present study, atrial ATP or atrial ATP in combination with atrial pace prevention algorithms did not reduce the burden of AT/AF over long-term follow-up compared with DDD/R pacing. These results and the previous short-term studies indicate that atrial ATP therapy and AF pace prevention therapies cannot be recommended for the general pacemaker population with AF. Some data suggest that individual patients may benefit from these therapies.5,7,9,11 For example, patients with recurrent atrial flutter or atrial tachycardia demonstrated to be reliably pace-terminated may experience a significant reduction in AT/AF burden.11,23 However, not all patients with atrial tachyarrhythmias that appear to be organized respond to atrial ATP therapy.23,24 Thus, the success of this therapy and its continued long-term use should be individualized. Some patients with frequent atrial premature beats just prior to onset of AT/AF may respond to overdrive pacing.7 Transient overdrive pacing after termination of AT/AF may delay AT/AF recurrence in some patients.9 However, since the majority of patients do not experience a substantial reduction in AT/AF burden with these algorithms, more research is needed to identify patient characteristics that predict responsiveness.

Atrial ATP therapy alone was associated with a significant increase in AT/AF burden at 3 years compared with baseline. This response raises concerns about proarrhythmia secondary to atrial ATP therapy. Nevertheless, this response was not observed in the group treated with atrial ATP therapy and atrial pace prevention therapies, and despite randomization, there were some potentially important baseline differences among the three groups. There were slightly more patients in the ATP group with valvular heart disease compared with the other two groups. Patients in the ATP group were less likely to be on class I/III antiarrhythmic drug therapy compared with the DDD/R group. Together these slight imbalances in clinical characteristics may have resulted in the ATP group having a higher susceptibility to progression of AF. Our conservative interpretation is that there is no beneficial effect of ATP on AT/AF burden.

Antiarrhythmic drug use

Patients not on class I/III antiarrhythmic drug therapy were more likely to experience a significant increase in AT/AF burden over time. This observation highlights the importance of continuing antiarrhythmic drug therapy following pacemaker implantation if a rhythm control approach for AF management is planned.

Ventricular pacing and atrial tachycardia/atrial fibrillation burden

Ventricular pacing even in the DDD/R mode increases the probability of AF development particularly if the cumulative ventricular pacing exceeds 40%.25 Algorithms to minimize ventricular pacing have been shown to prevent the development of persistent AF.26 Such algorithms were not available in the AT501 pacemaker. In the present study, efforts were made to minimize ventricular pacing by programming AV delays to promote intrinsic AV conduction during pacemaker follow-up. Sweeney et al.25 reported that the incidence of AF following pacemaker implantation increased in patients paced in the ventricle more than 40% of the time. Blanc et al.8 reported that selective atrial pace prevention algorithms were more likely to suppress AT/AF if ventricular pacing was minimized. In the present study, ambulatory ECG monitoring was performed annually to assess appropriate detection of AT/AF and to assess ventricular pacing. At the 1 year assessment, 35% of patients had frequent ventricular pseudofusion observed during DDD pacing and frequent ventricular pseudofusion continued to be observed in patients during long-term follow-up. Since ventricular pseudofusion occurred frequently during pacing, the pacing counters may not represent an accurate measure of ventricular pacing. This may explain the lack of a clear relationship between per cent ventricular pacing and AT/AF burden.

Ventricular rate during atrial fibrillation and symptoms

We observed a relationship between ventricular rate during AF and symptoms. This observation suggests that heart rate control is an important requirement for reduction in symptoms in patients with AF. Nevertheless, the relative importance of ventricular rate, regularity of the ventricular rate, and loss of AV synchrony in producing symptoms is poorly understood.

Limitations

The number of patients recruited was less than the estimated sample size due to replacement of the AT 500/501 with newer pacing technology. Thus, the study may have been underpowered to detect an effect of the AF pace prevention algorithms. Although the sample size was small, every patient randomized had documented AT/AF following pacemaker implantation. Previous studies had been limited by enrolling as many as 30–40% of patients without AT/AF recurrence following pacemaker implantation.12,14 The majority of patients (81%) were enrolled within 6 months of pacemaker implantation. However, some were enrolled up to 30 months following pacemaker implantation which might have impacted AT/AF recurrence rates and the progression of AT/AF burden over time. It is possible that a septal atrial lead position might have been associated with improved outcomes.27 However, the ASPECT study did not demonstrate a benefit of AF pace prevention algorithms in association with atrial septal pacing.14 The pacemaker technology employed in the present study did not have features designed to minimizing ventricular pacing. Thus, a large proportion of ventricular pacing may have promoted AF in some patients. Whether the AF prevention algorithms assessed in this study in combination with algorithms to minimize ventricular pacing would increase the magnitude of AT/AF prevention requires further study. However, the results of the present study do apply to patients with high-grade AV block requiring frequent conventional DDD/R pacing. The presence of significant chronotropic incompetence was not assessed in the present study. Whether programming rate response influenced the study outcomes is unknown.

Conclusions

Atrial ATP therapy and atrial ATP therapy in combination with atrial pace prevention algorithms do not provide incremental benefit for reducing the burden of AT/AF over long-term follow-up compared with conventional DDD/R pacing alone. However, atrial ATP therapy may be beneficial for treating atrial tachycardia or atrial flutter in individual patients. Class I/III antiarrhythmic drug therapy should be continued following pacemaker implantation if a rhythm control strategy for AF management is planned.

Conflict of interest: A.M.G. and D.G.W. are consultants to Medtonic Inc., A.M.G., D.V.E., H.J.D., and B.M. have received speakers honoraria from Medtronic Inc., A.M.G. receives research funding from Medtronic Inc., and D.V.E. receives research funding from Medtronic Inc. and St Jude. The University of Calgary receives funding for fellowship training in clinical electrophysiology from Medtronic of Canada and St Jude, Canada.

Funding

This work was supported by the Canadian Institutes for Health Research and Heart and Stroke Foundation of Alberta. A.M.G. and H.J.D. are Medical Scientists and D.V.E. is a Scholar of the Alberta Heritage Foundation for Medical Research.

References

View Abstract