- Email: [email protected]
Identification of “substrate fibrillators” and “trigger fibrillators” by pacemaker diagnostics
Identification of “substrate fibrillators” and “trigger fibrillators” by pacemaker diagnostics Alexander Yang, MD,* Jaap Ruiter, MD,† Dietrich Pfeiffer, MD,‡ Bahman Esmailzadeh, MD,§ Armin Welz, MD,§ Marc Hochhäusler, MD,* Georg Nickenig, MD,* Berndt Lüderitz, MD,* Thorsten Lewalter, MD* *From the Department of Cardiology, University of Bonn, Bonn, Germany, †Medical Center Alkmaar, Alkmaar, The Netherlands, ‡Department of Cardiology, University of Leipzig, Leipzig, Germany, and §Department of Cardiac Surgery, University of Bonn, Bonn, Germany. BACKGROUND The population of patients likely to respond to selected pacing algorithms for maintenance of sinus rhythm is unknown.
day) than group B (3.8 ⫾ 1.2 per day, P ⫽ .006). Mean AF episode duration was shorter in group A (11.4 ⫾ 10.2 hours) than group C (41.4 ⫾ 27.5 hours, P ⫽ .03).
OBJECTIVES The purpose of this study was to identify patients with specific onset patterns of paroxysmal atrial fibrillation (AF).
CONCLUSION The coincidence of low PAC activity before AF onset, high AF burden, and extended arrhythmia episode duration appears to be the consequence of a high atrial substrate factor. In these “substrate fibrillators,” the efficacy of pacing algorithms for maintenance of sinus rhythm may be limited. In contrast, “trigger fibrillators” exhibiting low AF burden despite high PAC incidence may represent the target population for specific PAC-suppressing pacing algorithms.
METHODS Dual-chamber pacemakers with advanced diagnostic functions were implanted in 112 patients with conventional indications for antibradycardia pacing and a history of paroxysmal AF. Pacemaker diagnostic data were analyzed after 97.5 ⫾ 40.9 days. According to the frequency of premature atrial contractions (PACs) during the 5 minutes before AF onset, patients were assigned to one of three groups: group A (high PAC activity), group B (moderate PAC activity), or group C (low PAC activity). RESULTS AF burden was lower in group A (4.6% ⫾ 2.4%) than group B (15.8% ⫾ 3.0%, P ⫽ .003) and group C (15.5% ⫾ 3.1%, P ⫽ .003). Fewer AF episodes occurred in group A (2.1 ⫾ 1.3 per
Introduction Atrial fibrillation (AF) is a common arrhythmia in patients requiring a pacemaker1– 4 and is associated with increased morbidity and mortality, mainly due to thromboembolic complications.5–7 Electrophysiologic mechanisms involved in the initiation and perpetuation of AF are complex. They include the development of dispersion of atrial repolarization, myocardial conduction disturbances, and premature atrial contractions (PACs).8 In particular, PACs are believed to play a major role in the induction of AF. It has been reported that the number of PACs progressively increases during the last 2 minutes before the onset of AF, and that PACs are directly involved in arrhythmia initiation.9 Several prospective, randomized trials indicate that atrial-based pacing may prevent AF in patients with conventional pacemaker indications.1–3 Therefore, specific atrialpacing algorithms for secondary prevention of AF have been introduced into clinical practice. These algorithms, incorporated in the pacemaker software, target the modification of atrial substrate and the suppression of AF triggers Address reprint requests and correspondence: Dr. Alexander Yang, Department of Cardiology, University of Bonn, Sigmund-Freud-Straße 25, D-53105 Bonn, Germany. E-mail address: alexander.yang@web. de. (Received November 28, 2005; accepted January 25, 2006.)
KEYWORDS Atrial fibrillation; Pacing; Pacemaker; Trigger; Substrate (Heart Rhythm 2006;3:682– 688) © 2006 Heart Rhythm Society. All rights reserved.
(e.g., PACs, sudden bradycardia, short-long-short sequences).10,11 The investigators of the Atrial Dynamic Overdrive Pacing Trial (ADOPT) reported that preventive overdrive pacing decreases symptomatic AF burden in patients with sinus node dysfunction and a history of AF.12 However, other trials that studied the efficacy of numerous combined algorithms failed to prove benefit from preventive pacing13,14 The discrepant clinical data suggest that preventive pacing therapies are effective only in subgroups of patients. Furthermore, aggressive atrial pacing caused by the indiscriminate programming of combined algorithms may have proarrhythmic effects as well. Therefore, the efficacy of pacing strategies for prevention of AF may be increased by selective use of algorithms based on individual onset characteristics. Because AF-preventive pacing most notably aims at suppressing PACs, the objective of the present study was to identify patients with specific onset patterns of paroxysmal AF who are likely to respond to selected pacing algorithms.
Methods The study population was extracted from the VIP (Prevention of Atrial Fibrillation [Vorhofflimmern] by Individualized Pacemaker Programming) registry, a prospective multicenter trial that investigated the efficacy of selected pacing
1547-5271/$ -see front matter © 2006 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2006.01.030
Yang et al
Identification of “Fibrillators” by Pacemaker Diagnostics
683
Figure 1 a: Identification of a “trigger fibrillator” according to the PACs preceding onset histogram, exhibiting high activity of premature atrial contractions (PACs) during the last 5 minutes prior to onset of atrial fibrillation (AF). b: Intracardiac marker ECG and rate profile diagram derived from the pacemaker diagnostics of the same patient showing an episode of AF that is preceded by four PACs during the last 9 seconds before arrhythmia onset. Rate profile diagram: Œ atrial sense; PAC; 〫 tachycardic atrial sense; ● atrial pace; □ ventricular sense; ventricular pace.
algorithms for prevention of AF. The VIP registry included patients with a conventional indication for antibradycardia pacing and a history of paroxysmal AF documented by at least one ECG. Eligible patients received a dual-chamber pacemaker (Selection 900 AF2.0, Selection 9000, or PreventAF; Vitatron Medical BV, Arnhem, The Netherlands) with a bipolar atrial lead placed at the right atrial appendage. The VIP study consisted of three phases: maturation phase (6 weeks), diagnostic phase (3 months), and therapy phase (3 months).15 The study data were obtained from the diagnostic phase of the VIP study. During this phase, the patients were programmed to conventional dual-chamber pacing (DDD or DDDR, recommended lower rate limit 60 bpm) with all preventive pacing algorithms disabled. Parameters of AF onset recording were programmed by the local study participants, according to the VIP registry protocol. A persistent atrial rate ⬎200 bpm for at least six consecutive ventricular intervals was required to detect AF onset. The end of an AF episode was documented in case the atrial rate dropped to ⬍200 bpm for at least 30 ventricular beats. Patients were included in the present study if they completed the diagnostic phase and had at least two episodes of AF according to the pacemaker diagnostic counters. In addition, patients were eligible only if the onsets of AF were
properly sensed by the pacemaker. For this purpose, all detailed onset reports obtained by pacemaker interrogation were reviewed (up to 12 reports) by authors A.Y. and M.H. The detailed onset reports provide an insight into onset scenarios of single episodes of AF, including intracardiac marker ECGs, rate trends, and PAC trends. Data were rejected if at least one episode of atrial oversensing (far-field sensing, retrograde atrial conduction) or undersensing was detected. Patients also were excluded if they exhibited permanent AF or persistent AF throughout the whole follow-up (AF burden 100%). The following pacemaker-derived diagnostic parameters were considered: ● ● ● ●
Number of PACs Total AF burden (absolute value and percentage of storage period) Number of AF episodes Number of PACs per minute during the last 5 minutes before AF onset according to the PACs preceding onset histogram (Figures 1a and 2a).
The number of PACs per hour during sinus rhythm was calculated by dividing the total number of PACs by the total time (hours) of sinus rhythm (follow-up period minus AF burden). The mean AF episode duration (hours) was deter-
684
Heart Rhythm, Vol 3, No 6, June 2006
Figure 2 a: Identification of a “substrate fibrillator” according to the PACs preceding onset histogram. This patient had low activity of premature atrial contractions (PACs) during the last 5 minutes prior to onset of atrial fibrillation (AF); 326 (81%) of 402 AF episodes were preceded by ⬍2 PACs per minute during the last 5 minutes before arrhythmia onset. b: Intracardiac marker ECG and rate profile diagram obtained from the same patient showing an episode of AF that is preceded by a single PAC (sudden onset). Rate profile diagram: Œ atrial sense; PAC; 〫 tachycardic atrial sense; ● atrial pace; □ ventricular sense; ventricular pace.
mined by dividing total AF burden (hours) by the number of detected AF episodes. In addition, the rate of ventricular pacing (%) was retrieved because of the known influences of ventricular pacing on the incidence of AF.16 The patients were assigned into one of three groups, using the PACs preceding onset histogram. The PACs preceding onset histogram presents each individual’s number of AF episodes according to the PAC activity during the last 5 minutes before arrhythmia onset. Group A was composed of patients with high PAC activity before arrhythmia onset (⬍30% of AF episodes preceded by ⬍2 PACs per minute before AF onset). In group B patients, moderate PAC activity was observed (30%–70% of AF episodes preceded by ⬍2 PACs per minute before AF onset). Group C patients exhibited low PAC activity (⬎70% of the episodes were preceded by ⬍2 PACs per minute before onset) (Figures 1 and 2). The three groups were compared with respect to AF burden (%), mean AF episode duration, number of AF episodes per day, and number of PACs per hour during sinus rhythm. The study was approved by the local institutional review committees, and all patients provided written informed consent.
Statistical analysis Statistical analyses were performed using the SPSS 10.0 software package (SPSS Inc., Chicago, IL, USA). Continuous data are expressed as mean ⫾ SEM, except for baseline characteristics and standard pacing data, which are given as mean ⫾ SD. For analysis of baseline characteristics and standard pacing data, the significance of differences in means among the three subgroups was tested with one-way analysis of variance. Kruskal-Wallis H-test was used to analyze AF burden, mean arrhythmia episode duration, number of AF episodes, and PAC frequency because these variables were not normally distributed. As appropriate, the Mann-Whitney U-test, adjusted with Bonferroni correction for multiple comparisons, was performed for post hoc analyses. Adjusted P values were calculated by multiplying each P value by the factor 3. Chi-square test was used to compare categorical data. P ⬍.05 was considered significant.
Results Baseline characteristics A total of 112 patients (54 men and 58 women; mean age 71.1 ⫾ 9.4 years) were enrolled in the study and followed-up for 97.5 ⫾ 40.9 days. Comparing the three groups, there was no significant difference with regard to demographic data, distri-
Yang et al
Identification of “Fibrillators” by Pacemaker Diagnostics
bution of indications for antibradycardia pacing, antiarrhythmic drug therapy, or underlying cardiac disease. Baseline characteristics are given in Table 1.
Standard pacing parameters The pacemakers were programmed to DDD in 45 patients (40.2%) and to DDDR in the remaining 67 subjects (59.8%). There was no difference with regard to the number of patients programmed to rate-adaptive pacing in the three groups (group A: n ⫽ 10 [55.6%]; group B: n ⫽ 25 [59.5%]; group C: n ⫽ 32 [61.5%]). The lower rate limit was set at a mean of 61.5 ⫾ 4.6 bpm and was similar in the three groups (group A: 60.6 ⫾ 2.9 bpm; group B: 61.2 ⫾ 3.8 bpm; group C: 61.7 ⫾ 5.1 bpm; P ⫽ .43). Atrial sensitivity and atrioventricular (AV) delay were programmed to a mean of 0.54 ⫾ 0.28 mV and 194.9 ⫾ 31.0 ms, respectively. Both atrial sensitivity (group A: 0.57 ⫾ 0.35 mV; group B: 0.52 ⫾ 0.26 mV; group C: 0.55 ⫾ 0.28 mV; P ⫽ .78) and AV delay (group A: 201.7 ⫾ 39.4 ms; group B: 193.6 ⫾ 28.6 ms; group C: 193.5 ⫾ 29.9 ms; P ⫽ .6) were similar among the three groups.
Analysis of pacemaker memory functions
A mean of 41.1 ⫾ 9.9 PACs per hour was prevalent in the entire study group. There was a significant relationship between the total number of PACs and the PAC activity prior to AF, because the number of PACs per hour progressively decreased from group A to group C (group A: 115.9 ⫾ 52.0; group B: 43.5 ⫾ 9.8; group C: 13.4 ⫾ 5.9; P ⬍.0001; Figure 3a). Analysis of AF characteristics was based on a total of 48,001 AF episodes. A mean AF burden of 13.8% ⫾ 1.9% was documented in the entire study group. Group A patients had a significantly lower AF burden (4.6% ⫾ 2.4%) than subjects in the other two groups (group B: 15.8% ⫾ 3.0%, adjusted P ⫽ .003; group C: 15.5% ⫾ 3.1%, adjusted P ⫽ .003; Figure 3b). AF episodes lasted for a mean of 23.7 ⫾ 12.9 hours and were less sustained in patients with high PAC activity before arrhythmia onset compared with AF episodes in subjects with low PAC activity (11.4 ⫾ 10.2 hours vs 41.4 ⫾ 27.5 hours, adjusted P ⫽ .03; Figure 3c). In addition, fewer AF episodes occurred in group A (2.1 ⫾ 1.3 AF episodes per day) compared with group B (3.8 ⫾ 1.2 AF episodes per day, adjusted P ⫽ .006). There was no difference in the AF recurrence rate between groups A and B (2.5 ⫾ 0.7 per day; Figure 3d). The percentage of ventricular pacing was similar in all three groups, with a mean of 79.6% ⫾ 7.0% in group A, 77.3% ⫾ 4.6% in group B, and 76.4% ⫾ 4.0% group C (P ⫽ .53).
Discussion The present study demonstrated that patient groups with different onset patterns of paroxysmal AF can be distinguished within the pacemaker population. These patients also differ with regard to their clinical presentation of AF. There was a lower AF burden in patients with high PAC activity before AF onset than in subjects with either moderate or low PAC activity. The decreased AF burden in group A resulted from fewer AF recurrences compared
685
with group B and a lower mean episode duration compared with group C. In addition, PAC activity prior to arrhythmia onset correlated with the total number of PACs. These differences could not be explained by age, gender, indication for pacing, antiarrhythmic drug therapy, or underlying cardiac disease. The coincidence of low PAC activity, high AF burden, and long arrhythmia episode duration in group C can be interpreted as the presence of a high atrial substrate factor. In these “substrate fibrillators,” AF is suggested to be mainly mediated by severe structural and electrophysiologic changes in the atrial myocardium, with few or even no PACs involved in the initiation of AF. In contrast, group A consisted of patients with low AF burden, decreased mean AF episode duration, and low AF recurrence rate, despite predominantly high PAC activity before AF onset. Alterations in atrial substrate may be less pronounced in these patients, and high trigger activity seems mandatory for development of AF. Onset mechanisms may serve as a marker for the temporal progression of paroxysmal AF. Whereas early stages of the disease can be described by low AF burden in the presence of a high trigger activity, progressive atrial remodeling may result in increases in AF burden and arrhythmia episode duration despite declining PAC activity. However, long-term follow-up studies are necessary to verify this hypothesis. In our study group, substrate AF was more common than trigger AF. This finding is not surprising, given that the pacemaker population comprises mainly older patients, and symptomatic bradycardia usually is the consequence of progressive electrophysiologic changes in atrial substrate. Bradycardia, tachycardia, autonomic dysregulation, and PACs are considered triggers of AF, whereas substrate factors can account for sustaining the arrhythmia. Alterations in atrial electrophysiology, such as dispersions of repolarization and conduction disturbances, are based on complex mechanical, structural, and signaling processes.8 On a cellular level, interstitial fibrosis, myocytic hypertrophy, increases in connective tissue elements,17 and altered gap junctions are often present.18 These changes may be responsible for the decrease in action potential duration, depression of action potential plateau,19 increase in stretch-activated channels,20 and alterations in ionic currents, particularly the L-type calcium current.21 Furthermore, atrial remodeling is aggravated by AF itself and finally results in the development of sites where unidirectional blocks occur. We presume that substantial structural and electrophysiologic alterations exist in “substrate fibrillators,” characterized by increased AF burden and long-lasting AF episodes that are induced by just a few PACs or even in the absence of PACs. Clinical trials of pacing for maintenance of sinus rhythm by activation of AF-preventive algorithms have shown inconsistent results. The ADOPT investigators randomized 288 patients with sinus node disease and recurrent AF to
686 Table 1
Heart Rhythm, Vol 3, No 6, June 2006 Comparison of baseline characteristics Parameter
Group A (n ⫽ 18)
Group B (n ⫽ 42)
Group C (n ⫽ 52)
Age (yr) Gender Male Female Indication for pacing Sinus node dysfunction Atrioventricular block Other Antiarrhythmic drug therapy Beta-blocker* Class I drug* Class III drug* History of cardiac disease Idiopathic atrial fibrillation Structural heart disease Hypertensive heart disease* Coronary artery disease* Other*
74.4 ⫾ 7.5
71.3 ⫾ 10.6
69.7 ⫾ 8.7
7 (38.9%) 11 (61.1%)
26 (61.9%) 16 (38.1%)
21 (40.4%) 31 (59.6%)
13 (72.2%) 5 (27.8%) 0
31 (73.8%) 8 (19.1%) 3 (7.1%)
39 (75.0%) 10 (19.2%) 3 (5.8%)
8 (44.4%) 1 (5.5%) 4 (22.2%)
29 (69.0%) 3 (7.1%) 16 (38.1%)
25 (48.1%) 4 (7.7%) 12 (23.1%)
23 19 9 5 7
26 26 18 10 3
12 6 3 3 1
(66.7%) (33.3%) (16.7%) (16.7%) (5.6%)
(54.8%) (45.2%) (21.4%) (11.9%) (16.7%)
(50.0%) (50.0%) (34.6%) (19.2%) (5.8%)
All differences were statistically nonsignificant.
*Multiple specifications possible.
either DDDR pacing or DDDR plus continuous overdrive pacing. Overdrive pacing reportedly was accompanied by a higher amount of atrial stimulation (92.9% vs 67.9%, P ⬍.0001) and a 25% reduction in symptomatic AF burden (1.87% vs 2.50%). Symptomatic AF burden decreased progressively in both groups but was lower in the treatment group at each follow-up visit (1, 3, and 6 months after pacemaker implantation; P ⫽ .005).12 Other trials have not shown benefit of preventive algorithms for prevention of AF.13,14 However, these studies differed considerably with regard to the number and types of programmed algorithms. This aspect is of importance because whether aggressive atrial pacing, as generated by activation of numerous combined algorithms, also provides proarrhythmic effects is unknown. The VIP registry investigated the selective use of preventive pacing therapies on the basis of different AF induction scenarios. Patients with a predominantly high PAC activity before arrhythmia onset were programmed to PACassociated preventive pacing algorithms, whereas subjects with few PACs underwent continuous overdrive pacing. A significant improvement in AF burden was observed only in the patients with high PAC activity.15 The VIP registry underlines that different mechanisms of AF onset affect the susceptibility to preventive pacing algorithms. The data of the current study are derived from advanced pacemaker diagnostic features. Unlike ambulatory Holter ECGs or external event recorders, pacemakers provide continuous arrhythmia monitoring over virtually unlimited periods. The Automatic Interpretation for Diagnostic Assistance (AIDA) trial was one of the first studies to demonstrate that pacemaker memory functions are capable of accurately recording atrial tachyarrhythmias. The AIDA trial reported a 51% incidence of atrial tachyarrhythmias in patients with dual-chamber
pacing. Interestingly, 58% of the patients remained clinically asymptomatic and, of these, 73% had no history of AF before pacemaker implantation.4 In patients with a conventional indication for antibradycardia pacing and a history of paroxysmal AF, an 88% recurrence rate of AF was reported during long-term follow-up.22 Other studies have investigated onset characteristics of AF derived from pacemaker data. It has been shown that approximately two thirds of AF episodes were preceded by ⬍2 PACs per minute.23 Guyomar et al24 re-
Figure 3 Number of premature atrial contractions (PACs) (a), burden of atrial fibrillation (AF) (b), mean AF episode duration (c), and AF recurrence rate (d) in patients with high (group A, n ⫽ 18), moderate (group B, n ⫽ 31), and low (group C, n ⫽ 49) PAC activity during the last 5 minutes before AF onset. Group C patients had lower numbers of PACs overall compared with group A and group B patients. AF burden was markedly decreased in group A compared with the two other groups. AF episodes were less sustained in group A than in group C. Fewer AF episodes occurred in group A than in group B. *Adjusted P ⬍.0001 vs either group A or group B. †Adjusted P ⫽ .003 vs either group B and group C. ‡Adjusted P ⫽ .03 vs group C. §Adjusted P ⫽ .006 vs group B.
Yang et al
Identification of “Fibrillators” by Pacemaker Diagnostics
ported that the most common mechanisms of AF initiation were sudden onset (24%), increasing frequencies of PACs plus short runs (17%), and increased PAC activity plus sudden onset (12%). In contrast to our results, there was no correlation between the onset mechanisms of AF and the duration of the AF episodes. This discrepancy may be explained by the number of AF episodes included in both studies. In the previous study, analysis of AF onset mechanisms was limited to 12 AF episodes per patient, or a total of 318 episodes. In contrast, the present study was based on all AF episodes recorded during follow-up, with a total of 48,001 AF recurrences. In summary, these studies underline that pacemaker diagnostic features raise the diagnostic yield in individuals with cardiac pacemakers and paroxysmal AF.
Study limitations The findings of the present study refer to patients with antibradycardia cardiac pacing and paroxysmal AF. Therefore, these observations may not be generally applied to subjects without symptomatic bradycardia or those with persistent AF. The present study focused on the identification of electrophysiologic changes in subsets of pacemaker patients with paroxysmal AF. Echocardiographic data were not available in all patients and, therefore, were not considered during analysis. Differences in atrial diameter and/or volume between the groups of patients are likely but may not exist, because altered electrophysiologic properties are not necessarily accompanied by significant macroscopic changes as visualized by echocardiography. Hence, it should be emphasized that the terms “trigger fibrillators” and “substrate fibrillators” are solely based on electrophysiologic characteristics. Further investigations, including studies of long-term follow-up of “trigger fibrillators” and “substrate fibrillators,” are necessary to confirm our results. It can be argued that the follow-up was too short. However, AF indices are likely to change with increasing duration of diagnostic follow-up because of the natural course of electrophysiologic remodeling or beneficial effects of pacing, respectively. Thus, data derived from long follow-up periods may be less representative for “substrate AF” and “trigger AF”. Moreover, from a clinical point of view, prevention of paroxysmal AF should be initiated as soon as possible. Therefore, we consider a 3-month diagnostic follow-up adequate. Ventricular desynchronization imposed by ventricular pacing has gained increasing attention because of its impact on the incidence of AF.15 It is possible that the high rate of ventricular pacing in our study group influenced the results. However, the three subgroups exhibited no difference in the rate of ventricular pacing.
Conclusion and clinical implications Advanced pacemaker diagnostic functions provide detailed insight into the electrophysiologic characteristics of paroxysmal AF. Apart from selecting patients who are likely to
687
benefit from pharmacologic therapy and ablative procedures, pacemaker diagnostics may help to identify responders to selective AF-preventive pacing algorithms. In the current study, “trigger fibrillators” were differentiated from “substrate fibrillators” according to the onset scenarios of AF. We conclude that “trigger fibrillators” may represent the target group for preventive pacing therapies. These patients in particular could benefit from specific PAC-suppressing algorithms. Vice versa, the efficacy of preventive algorithms is presumed to be limited in “substrate fibrillators,” characterized by low PAC activity and high AF burden.
References 1. Andersen HR, Nielsen JC, Thomsen PE, Thuesen L, Mortensen PT, Vesterlund T, Pedersen AK. Long-term follow-up of patients from a randomised trial of atrial versus ventricular pacing for sick-sinus syndrome. Lancet 1997;350:1210 –1216. 2. Kerr CR, Connolly SJ, Abdollah H, Roberts RS, Gent M, Yusuf S, Gillis AM, Tang AS, Talajic M, Klein GJ, Newman DM. Canadian Trial of Physiological Pacing: effects of physiological pacing during long-term follow-up. Circulation 2004;109:357–362. 3. Lamas GA, Lee KL, Sweeney MO, Silverman R, Leon A, Yee R, Marinchak RA, Flaker G, Schron E, Orav EJ, Hellkamp AS, Greer S, McAnulty J, Ellenbogen K, Ehlert F, Freedman RA, Estes NA III, Greenspon A, Goldman L. Ventricular pacing or dual-chamber pacing for sinus-node dysfunction. Mode Selection Trial in sinus-node dysfunction. N Engl J Med 2002;346:1854 –1862. 4. Defaye P, Dournaux F, Mouton E. Prevalence of supraventricular arrhythmias from the automated analysis of data stored in the DDD pacemakers of 617 patients: the AIDA study. The AIDA Multicenter Study Group. Automatic Interpretation for Diagnosis Assistance. Pacing Clin Electrophysiol 1998;21:250 –255. 5. Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation 1998;98:946 –952. 6. Kannel WB, Abbott RD, Savage DD, McNamara PM. Epidemiologic features of chronic atrial fibrillation: the Framingham Study. N Engl J Med 1982;306: 1018 –1022. 7. Wolf PA, Abbott RD, Kannell WB. Atrial fibrillation as independent risk factor for stroke: the Framingham Study. Stroke 1991;22:983–988. 8. Allessie MA, Boyden PA, Camm AJ, Kleber AG, Lab MJ, Legato MJ, Rosen MR, Schwartz PJ, Spooner PM, Van Wagoner DR, Waldo AL. Pathophysiology and prevention of atrial fibrillation. Circulation 2001;103:769 –777. 9. Hnatkova K, Waktare JE, Murgatroyd FD, Guo X, Baiyan X, Camm AJ, Malik M. Analysis of the cardiac rhythm preceding episodes of paroxysmal atrial fibrillation. Am Heart J 1998;135:1010 –1019. 10. Mitchell AR, Sulke N. How do atrial pacing algorithms prevent atrial arrhythmias? Europace 2004;6:351–362. 11. Yang A, Lüderitz B, Lewalter T. Atrial fibrillation: profit from cardiac pacing? Z Kardiol 2005;94:141–151. 12. Carlson MD, Ip J, Messenger J, Beau S, Kalbfleisch S, Gervais P, Cameron DA, Duran A, Val-Mejias J, Mackall J, Gold M. A new pacemaker algorithm for the treatment of atrial fibrillation: results of the Atrial Dynamic Overdrive Pacing Trial (ADOPT). J Am Coll Cardiol 2003;42:627– 633. 13. Padeletti L, Pürerfellner H, Adler SW, Waller TJ, Harvey M, Horvitz L, Holbrook R, Kempen K, Mugglin A, Hettrick DA. Combined efficacy of atrial septal lead placement and atrial pacing algorithms for prevention of paroxysmal atrial tachyarrhythmia. J Cardiovasc Electrophysiol 2003;14:1189 –1195. 14. Lee MA, Weachter R, Pollak S, Kremers MS, Naik AM, Silverman R, Tuzi J, Wang W, Johnson LJ, Euler DE. The effect of atrial pacing therapies on atrial tachyarrhythmia burden and frequency: results of a randomised trial in patients with bradycardia and atrial tachyarrhythmias. J Am Coll Cardiol 2003;41:1926 – 1932. 15. Lewalter T, Yang A, Pfeiffer D, Ruiter J, Schnitzler G, Markert T, Asklund M, Przibille O, Welz A, Esmailzadeh B, Linhart M, Lüderitz B. Individualised selection of pacing algorithms for the prevention of recurrent atrial fibrillation: results from the VIP registry. Pacing Clin Electrophysiol 2006;29:124 –134. 16. Sweeney MO, Hellkamp AS, Ellenbogen KA, Greenspon AJ, Freedman RA, Lee KL, Lamas GA. Adverse effect of ventricular pacing on heart failure and atrial fibrillation among patients with normal baseline QRS duration in a clinical trial of pacemaker therapy for sinus node dysfunction. Circulation 2003;107: 2932–2937.
688 17. Legato MJ, Bull MB, Ferrer MI. Atrial ultrastructure in patients with fixed intra-atrial block. Chest 1974;65:252–261. 18. Mary-Rabine L, Albert A, Pham TD, Hordof A, Fenoglio JJ Jr, Malm JR, Rosen MR. The relationship of human atrial cellular electrophysiology to clinical function and ultrastructure. Circ Res 1983;52:188 –199. 19. Van Wagoner DR, Pond AL, Lamorgese M, Rossie SS, McCarthy PM, Nerbonne JM. Atrial L-type Ca2⫹ currents and human atrial fibrillation. Circ Res 1999;85:428 – 436. 20. Hu H, Sachs F. Stretch-activated ion channels in the heart. J Mol Cell Cardiol 1997;29:1511–1523. 21. Mewes T, Ravens U. L-type calcium currents of human myocytes from ventricle of non-failing and failing hearts and from atrium. J Mol Cell Cardiol 1994;26: 1307–1320.
Heart Rhythm, Vol 3, No 6, June 2006 22. Israel CW, Grönefeld G, Ehrlich JR, Li YG, Hohnloser SH. Long-term risk of recurrent atrial fibrillation as documented by an implantable monitoring device: implications for optimal patient care. J Am Coll Cardiol 2004;43: 47–52. 23. Yang A, Hochhäusler M, Schrickel J, Bielik H, Shlevkov N, Schimpf R, Schwab JO, Esmailzadeh B, Schneider C, Mellert F, Welz A, Saborowski F, Lüderitz B, Lewalter T. Advanced pacemaker diagnostic features in the characterization of atrial fibrillation: impact on preventive pacing algorithms. Pacing Clin Electrophysiol 2003;26:310 –313. 24. Guyomar Y, Thomas O, Marquie C, Jarwe M, Klug D, Kacet S, Carlioz R, Ferrier A, Fossati F, Guerin S, Heuls S, Graux P. Mechanisms of onset of atrial fibrillation: a multicenter, prospective, pacemaker-based study. Pacing Clin Electrophysiol 2003;26:1336 –1341.
day) than group B (3.8 ⫾ 1.2 per day, P ⫽ .006). Mean AF episode duration was shorter in group A (11.4 ⫾ 10.2 hours) than group C (41.4 ⫾ 27.5 hours, P ⫽ .03).
OBJECTIVES The purpose of this study was to identify patients with specific onset patterns of paroxysmal atrial fibrillation (AF).
CONCLUSION The coincidence of low PAC activity before AF onset, high AF burden, and extended arrhythmia episode duration appears to be the consequence of a high atrial substrate factor. In these “substrate fibrillators,” the efficacy of pacing algorithms for maintenance of sinus rhythm may be limited. In contrast, “trigger fibrillators” exhibiting low AF burden despite high PAC incidence may represent the target population for specific PAC-suppressing pacing algorithms.
METHODS Dual-chamber pacemakers with advanced diagnostic functions were implanted in 112 patients with conventional indications for antibradycardia pacing and a history of paroxysmal AF. Pacemaker diagnostic data were analyzed after 97.5 ⫾ 40.9 days. According to the frequency of premature atrial contractions (PACs) during the 5 minutes before AF onset, patients were assigned to one of three groups: group A (high PAC activity), group B (moderate PAC activity), or group C (low PAC activity). RESULTS AF burden was lower in group A (4.6% ⫾ 2.4%) than group B (15.8% ⫾ 3.0%, P ⫽ .003) and group C (15.5% ⫾ 3.1%, P ⫽ .003). Fewer AF episodes occurred in group A (2.1 ⫾ 1.3 per
Introduction Atrial fibrillation (AF) is a common arrhythmia in patients requiring a pacemaker1– 4 and is associated with increased morbidity and mortality, mainly due to thromboembolic complications.5–7 Electrophysiologic mechanisms involved in the initiation and perpetuation of AF are complex. They include the development of dispersion of atrial repolarization, myocardial conduction disturbances, and premature atrial contractions (PACs).8 In particular, PACs are believed to play a major role in the induction of AF. It has been reported that the number of PACs progressively increases during the last 2 minutes before the onset of AF, and that PACs are directly involved in arrhythmia initiation.9 Several prospective, randomized trials indicate that atrial-based pacing may prevent AF in patients with conventional pacemaker indications.1–3 Therefore, specific atrialpacing algorithms for secondary prevention of AF have been introduced into clinical practice. These algorithms, incorporated in the pacemaker software, target the modification of atrial substrate and the suppression of AF triggers Address reprint requests and correspondence: Dr. Alexander Yang, Department of Cardiology, University of Bonn, Sigmund-Freud-Straße 25, D-53105 Bonn, Germany. E-mail address: alexander.yang@web. de. (Received November 28, 2005; accepted January 25, 2006.)
KEYWORDS Atrial fibrillation; Pacing; Pacemaker; Trigger; Substrate (Heart Rhythm 2006;3:682– 688) © 2006 Heart Rhythm Society. All rights reserved.
(e.g., PACs, sudden bradycardia, short-long-short sequences).10,11 The investigators of the Atrial Dynamic Overdrive Pacing Trial (ADOPT) reported that preventive overdrive pacing decreases symptomatic AF burden in patients with sinus node dysfunction and a history of AF.12 However, other trials that studied the efficacy of numerous combined algorithms failed to prove benefit from preventive pacing13,14 The discrepant clinical data suggest that preventive pacing therapies are effective only in subgroups of patients. Furthermore, aggressive atrial pacing caused by the indiscriminate programming of combined algorithms may have proarrhythmic effects as well. Therefore, the efficacy of pacing strategies for prevention of AF may be increased by selective use of algorithms based on individual onset characteristics. Because AF-preventive pacing most notably aims at suppressing PACs, the objective of the present study was to identify patients with specific onset patterns of paroxysmal AF who are likely to respond to selected pacing algorithms.
Methods The study population was extracted from the VIP (Prevention of Atrial Fibrillation [Vorhofflimmern] by Individualized Pacemaker Programming) registry, a prospective multicenter trial that investigated the efficacy of selected pacing
1547-5271/$ -see front matter © 2006 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2006.01.030
Yang et al
Identification of “Fibrillators” by Pacemaker Diagnostics
683
Figure 1 a: Identification of a “trigger fibrillator” according to the PACs preceding onset histogram, exhibiting high activity of premature atrial contractions (PACs) during the last 5 minutes prior to onset of atrial fibrillation (AF). b: Intracardiac marker ECG and rate profile diagram derived from the pacemaker diagnostics of the same patient showing an episode of AF that is preceded by four PACs during the last 9 seconds before arrhythmia onset. Rate profile diagram: Œ atrial sense; PAC; 〫 tachycardic atrial sense; ● atrial pace; □ ventricular sense; ventricular pace.
algorithms for prevention of AF. The VIP registry included patients with a conventional indication for antibradycardia pacing and a history of paroxysmal AF documented by at least one ECG. Eligible patients received a dual-chamber pacemaker (Selection 900 AF2.0, Selection 9000, or PreventAF; Vitatron Medical BV, Arnhem, The Netherlands) with a bipolar atrial lead placed at the right atrial appendage. The VIP study consisted of three phases: maturation phase (6 weeks), diagnostic phase (3 months), and therapy phase (3 months).15 The study data were obtained from the diagnostic phase of the VIP study. During this phase, the patients were programmed to conventional dual-chamber pacing (DDD or DDDR, recommended lower rate limit 60 bpm) with all preventive pacing algorithms disabled. Parameters of AF onset recording were programmed by the local study participants, according to the VIP registry protocol. A persistent atrial rate ⬎200 bpm for at least six consecutive ventricular intervals was required to detect AF onset. The end of an AF episode was documented in case the atrial rate dropped to ⬍200 bpm for at least 30 ventricular beats. Patients were included in the present study if they completed the diagnostic phase and had at least two episodes of AF according to the pacemaker diagnostic counters. In addition, patients were eligible only if the onsets of AF were
properly sensed by the pacemaker. For this purpose, all detailed onset reports obtained by pacemaker interrogation were reviewed (up to 12 reports) by authors A.Y. and M.H. The detailed onset reports provide an insight into onset scenarios of single episodes of AF, including intracardiac marker ECGs, rate trends, and PAC trends. Data were rejected if at least one episode of atrial oversensing (far-field sensing, retrograde atrial conduction) or undersensing was detected. Patients also were excluded if they exhibited permanent AF or persistent AF throughout the whole follow-up (AF burden 100%). The following pacemaker-derived diagnostic parameters were considered: ● ● ● ●
Number of PACs Total AF burden (absolute value and percentage of storage period) Number of AF episodes Number of PACs per minute during the last 5 minutes before AF onset according to the PACs preceding onset histogram (Figures 1a and 2a).
The number of PACs per hour during sinus rhythm was calculated by dividing the total number of PACs by the total time (hours) of sinus rhythm (follow-up period minus AF burden). The mean AF episode duration (hours) was deter-
684
Heart Rhythm, Vol 3, No 6, June 2006
Figure 2 a: Identification of a “substrate fibrillator” according to the PACs preceding onset histogram. This patient had low activity of premature atrial contractions (PACs) during the last 5 minutes prior to onset of atrial fibrillation (AF); 326 (81%) of 402 AF episodes were preceded by ⬍2 PACs per minute during the last 5 minutes before arrhythmia onset. b: Intracardiac marker ECG and rate profile diagram obtained from the same patient showing an episode of AF that is preceded by a single PAC (sudden onset). Rate profile diagram: Œ atrial sense; PAC; 〫 tachycardic atrial sense; ● atrial pace; □ ventricular sense; ventricular pace.
mined by dividing total AF burden (hours) by the number of detected AF episodes. In addition, the rate of ventricular pacing (%) was retrieved because of the known influences of ventricular pacing on the incidence of AF.16 The patients were assigned into one of three groups, using the PACs preceding onset histogram. The PACs preceding onset histogram presents each individual’s number of AF episodes according to the PAC activity during the last 5 minutes before arrhythmia onset. Group A was composed of patients with high PAC activity before arrhythmia onset (⬍30% of AF episodes preceded by ⬍2 PACs per minute before AF onset). In group B patients, moderate PAC activity was observed (30%–70% of AF episodes preceded by ⬍2 PACs per minute before AF onset). Group C patients exhibited low PAC activity (⬎70% of the episodes were preceded by ⬍2 PACs per minute before onset) (Figures 1 and 2). The three groups were compared with respect to AF burden (%), mean AF episode duration, number of AF episodes per day, and number of PACs per hour during sinus rhythm. The study was approved by the local institutional review committees, and all patients provided written informed consent.
Statistical analysis Statistical analyses were performed using the SPSS 10.0 software package (SPSS Inc., Chicago, IL, USA). Continuous data are expressed as mean ⫾ SEM, except for baseline characteristics and standard pacing data, which are given as mean ⫾ SD. For analysis of baseline characteristics and standard pacing data, the significance of differences in means among the three subgroups was tested with one-way analysis of variance. Kruskal-Wallis H-test was used to analyze AF burden, mean arrhythmia episode duration, number of AF episodes, and PAC frequency because these variables were not normally distributed. As appropriate, the Mann-Whitney U-test, adjusted with Bonferroni correction for multiple comparisons, was performed for post hoc analyses. Adjusted P values were calculated by multiplying each P value by the factor 3. Chi-square test was used to compare categorical data. P ⬍.05 was considered significant.
Results Baseline characteristics A total of 112 patients (54 men and 58 women; mean age 71.1 ⫾ 9.4 years) were enrolled in the study and followed-up for 97.5 ⫾ 40.9 days. Comparing the three groups, there was no significant difference with regard to demographic data, distri-
Yang et al
Identification of “Fibrillators” by Pacemaker Diagnostics
bution of indications for antibradycardia pacing, antiarrhythmic drug therapy, or underlying cardiac disease. Baseline characteristics are given in Table 1.
Standard pacing parameters The pacemakers were programmed to DDD in 45 patients (40.2%) and to DDDR in the remaining 67 subjects (59.8%). There was no difference with regard to the number of patients programmed to rate-adaptive pacing in the three groups (group A: n ⫽ 10 [55.6%]; group B: n ⫽ 25 [59.5%]; group C: n ⫽ 32 [61.5%]). The lower rate limit was set at a mean of 61.5 ⫾ 4.6 bpm and was similar in the three groups (group A: 60.6 ⫾ 2.9 bpm; group B: 61.2 ⫾ 3.8 bpm; group C: 61.7 ⫾ 5.1 bpm; P ⫽ .43). Atrial sensitivity and atrioventricular (AV) delay were programmed to a mean of 0.54 ⫾ 0.28 mV and 194.9 ⫾ 31.0 ms, respectively. Both atrial sensitivity (group A: 0.57 ⫾ 0.35 mV; group B: 0.52 ⫾ 0.26 mV; group C: 0.55 ⫾ 0.28 mV; P ⫽ .78) and AV delay (group A: 201.7 ⫾ 39.4 ms; group B: 193.6 ⫾ 28.6 ms; group C: 193.5 ⫾ 29.9 ms; P ⫽ .6) were similar among the three groups.
Analysis of pacemaker memory functions
A mean of 41.1 ⫾ 9.9 PACs per hour was prevalent in the entire study group. There was a significant relationship between the total number of PACs and the PAC activity prior to AF, because the number of PACs per hour progressively decreased from group A to group C (group A: 115.9 ⫾ 52.0; group B: 43.5 ⫾ 9.8; group C: 13.4 ⫾ 5.9; P ⬍.0001; Figure 3a). Analysis of AF characteristics was based on a total of 48,001 AF episodes. A mean AF burden of 13.8% ⫾ 1.9% was documented in the entire study group. Group A patients had a significantly lower AF burden (4.6% ⫾ 2.4%) than subjects in the other two groups (group B: 15.8% ⫾ 3.0%, adjusted P ⫽ .003; group C: 15.5% ⫾ 3.1%, adjusted P ⫽ .003; Figure 3b). AF episodes lasted for a mean of 23.7 ⫾ 12.9 hours and were less sustained in patients with high PAC activity before arrhythmia onset compared with AF episodes in subjects with low PAC activity (11.4 ⫾ 10.2 hours vs 41.4 ⫾ 27.5 hours, adjusted P ⫽ .03; Figure 3c). In addition, fewer AF episodes occurred in group A (2.1 ⫾ 1.3 AF episodes per day) compared with group B (3.8 ⫾ 1.2 AF episodes per day, adjusted P ⫽ .006). There was no difference in the AF recurrence rate between groups A and B (2.5 ⫾ 0.7 per day; Figure 3d). The percentage of ventricular pacing was similar in all three groups, with a mean of 79.6% ⫾ 7.0% in group A, 77.3% ⫾ 4.6% in group B, and 76.4% ⫾ 4.0% group C (P ⫽ .53).
Discussion The present study demonstrated that patient groups with different onset patterns of paroxysmal AF can be distinguished within the pacemaker population. These patients also differ with regard to their clinical presentation of AF. There was a lower AF burden in patients with high PAC activity before AF onset than in subjects with either moderate or low PAC activity. The decreased AF burden in group A resulted from fewer AF recurrences compared
685
with group B and a lower mean episode duration compared with group C. In addition, PAC activity prior to arrhythmia onset correlated with the total number of PACs. These differences could not be explained by age, gender, indication for pacing, antiarrhythmic drug therapy, or underlying cardiac disease. The coincidence of low PAC activity, high AF burden, and long arrhythmia episode duration in group C can be interpreted as the presence of a high atrial substrate factor. In these “substrate fibrillators,” AF is suggested to be mainly mediated by severe structural and electrophysiologic changes in the atrial myocardium, with few or even no PACs involved in the initiation of AF. In contrast, group A consisted of patients with low AF burden, decreased mean AF episode duration, and low AF recurrence rate, despite predominantly high PAC activity before AF onset. Alterations in atrial substrate may be less pronounced in these patients, and high trigger activity seems mandatory for development of AF. Onset mechanisms may serve as a marker for the temporal progression of paroxysmal AF. Whereas early stages of the disease can be described by low AF burden in the presence of a high trigger activity, progressive atrial remodeling may result in increases in AF burden and arrhythmia episode duration despite declining PAC activity. However, long-term follow-up studies are necessary to verify this hypothesis. In our study group, substrate AF was more common than trigger AF. This finding is not surprising, given that the pacemaker population comprises mainly older patients, and symptomatic bradycardia usually is the consequence of progressive electrophysiologic changes in atrial substrate. Bradycardia, tachycardia, autonomic dysregulation, and PACs are considered triggers of AF, whereas substrate factors can account for sustaining the arrhythmia. Alterations in atrial electrophysiology, such as dispersions of repolarization and conduction disturbances, are based on complex mechanical, structural, and signaling processes.8 On a cellular level, interstitial fibrosis, myocytic hypertrophy, increases in connective tissue elements,17 and altered gap junctions are often present.18 These changes may be responsible for the decrease in action potential duration, depression of action potential plateau,19 increase in stretch-activated channels,20 and alterations in ionic currents, particularly the L-type calcium current.21 Furthermore, atrial remodeling is aggravated by AF itself and finally results in the development of sites where unidirectional blocks occur. We presume that substantial structural and electrophysiologic alterations exist in “substrate fibrillators,” characterized by increased AF burden and long-lasting AF episodes that are induced by just a few PACs or even in the absence of PACs. Clinical trials of pacing for maintenance of sinus rhythm by activation of AF-preventive algorithms have shown inconsistent results. The ADOPT investigators randomized 288 patients with sinus node disease and recurrent AF to
686 Table 1
Heart Rhythm, Vol 3, No 6, June 2006 Comparison of baseline characteristics Parameter
Group A (n ⫽ 18)
Group B (n ⫽ 42)
Group C (n ⫽ 52)
Age (yr) Gender Male Female Indication for pacing Sinus node dysfunction Atrioventricular block Other Antiarrhythmic drug therapy Beta-blocker* Class I drug* Class III drug* History of cardiac disease Idiopathic atrial fibrillation Structural heart disease Hypertensive heart disease* Coronary artery disease* Other*
74.4 ⫾ 7.5
71.3 ⫾ 10.6
69.7 ⫾ 8.7
7 (38.9%) 11 (61.1%)
26 (61.9%) 16 (38.1%)
21 (40.4%) 31 (59.6%)
13 (72.2%) 5 (27.8%) 0
31 (73.8%) 8 (19.1%) 3 (7.1%)
39 (75.0%) 10 (19.2%) 3 (5.8%)
8 (44.4%) 1 (5.5%) 4 (22.2%)
29 (69.0%) 3 (7.1%) 16 (38.1%)
25 (48.1%) 4 (7.7%) 12 (23.1%)
23 19 9 5 7
26 26 18 10 3
12 6 3 3 1
(66.7%) (33.3%) (16.7%) (16.7%) (5.6%)
(54.8%) (45.2%) (21.4%) (11.9%) (16.7%)
(50.0%) (50.0%) (34.6%) (19.2%) (5.8%)
All differences were statistically nonsignificant.
*Multiple specifications possible.
either DDDR pacing or DDDR plus continuous overdrive pacing. Overdrive pacing reportedly was accompanied by a higher amount of atrial stimulation (92.9% vs 67.9%, P ⬍.0001) and a 25% reduction in symptomatic AF burden (1.87% vs 2.50%). Symptomatic AF burden decreased progressively in both groups but was lower in the treatment group at each follow-up visit (1, 3, and 6 months after pacemaker implantation; P ⫽ .005).12 Other trials have not shown benefit of preventive algorithms for prevention of AF.13,14 However, these studies differed considerably with regard to the number and types of programmed algorithms. This aspect is of importance because whether aggressive atrial pacing, as generated by activation of numerous combined algorithms, also provides proarrhythmic effects is unknown. The VIP registry investigated the selective use of preventive pacing therapies on the basis of different AF induction scenarios. Patients with a predominantly high PAC activity before arrhythmia onset were programmed to PACassociated preventive pacing algorithms, whereas subjects with few PACs underwent continuous overdrive pacing. A significant improvement in AF burden was observed only in the patients with high PAC activity.15 The VIP registry underlines that different mechanisms of AF onset affect the susceptibility to preventive pacing algorithms. The data of the current study are derived from advanced pacemaker diagnostic features. Unlike ambulatory Holter ECGs or external event recorders, pacemakers provide continuous arrhythmia monitoring over virtually unlimited periods. The Automatic Interpretation for Diagnostic Assistance (AIDA) trial was one of the first studies to demonstrate that pacemaker memory functions are capable of accurately recording atrial tachyarrhythmias. The AIDA trial reported a 51% incidence of atrial tachyarrhythmias in patients with dual-chamber
pacing. Interestingly, 58% of the patients remained clinically asymptomatic and, of these, 73% had no history of AF before pacemaker implantation.4 In patients with a conventional indication for antibradycardia pacing and a history of paroxysmal AF, an 88% recurrence rate of AF was reported during long-term follow-up.22 Other studies have investigated onset characteristics of AF derived from pacemaker data. It has been shown that approximately two thirds of AF episodes were preceded by ⬍2 PACs per minute.23 Guyomar et al24 re-
Figure 3 Number of premature atrial contractions (PACs) (a), burden of atrial fibrillation (AF) (b), mean AF episode duration (c), and AF recurrence rate (d) in patients with high (group A, n ⫽ 18), moderate (group B, n ⫽ 31), and low (group C, n ⫽ 49) PAC activity during the last 5 minutes before AF onset. Group C patients had lower numbers of PACs overall compared with group A and group B patients. AF burden was markedly decreased in group A compared with the two other groups. AF episodes were less sustained in group A than in group C. Fewer AF episodes occurred in group A than in group B. *Adjusted P ⬍.0001 vs either group A or group B. †Adjusted P ⫽ .003 vs either group B and group C. ‡Adjusted P ⫽ .03 vs group C. §Adjusted P ⫽ .006 vs group B.
Yang et al
Identification of “Fibrillators” by Pacemaker Diagnostics
ported that the most common mechanisms of AF initiation were sudden onset (24%), increasing frequencies of PACs plus short runs (17%), and increased PAC activity plus sudden onset (12%). In contrast to our results, there was no correlation between the onset mechanisms of AF and the duration of the AF episodes. This discrepancy may be explained by the number of AF episodes included in both studies. In the previous study, analysis of AF onset mechanisms was limited to 12 AF episodes per patient, or a total of 318 episodes. In contrast, the present study was based on all AF episodes recorded during follow-up, with a total of 48,001 AF recurrences. In summary, these studies underline that pacemaker diagnostic features raise the diagnostic yield in individuals with cardiac pacemakers and paroxysmal AF.
Study limitations The findings of the present study refer to patients with antibradycardia cardiac pacing and paroxysmal AF. Therefore, these observations may not be generally applied to subjects without symptomatic bradycardia or those with persistent AF. The present study focused on the identification of electrophysiologic changes in subsets of pacemaker patients with paroxysmal AF. Echocardiographic data were not available in all patients and, therefore, were not considered during analysis. Differences in atrial diameter and/or volume between the groups of patients are likely but may not exist, because altered electrophysiologic properties are not necessarily accompanied by significant macroscopic changes as visualized by echocardiography. Hence, it should be emphasized that the terms “trigger fibrillators” and “substrate fibrillators” are solely based on electrophysiologic characteristics. Further investigations, including studies of long-term follow-up of “trigger fibrillators” and “substrate fibrillators,” are necessary to confirm our results. It can be argued that the follow-up was too short. However, AF indices are likely to change with increasing duration of diagnostic follow-up because of the natural course of electrophysiologic remodeling or beneficial effects of pacing, respectively. Thus, data derived from long follow-up periods may be less representative for “substrate AF” and “trigger AF”. Moreover, from a clinical point of view, prevention of paroxysmal AF should be initiated as soon as possible. Therefore, we consider a 3-month diagnostic follow-up adequate. Ventricular desynchronization imposed by ventricular pacing has gained increasing attention because of its impact on the incidence of AF.15 It is possible that the high rate of ventricular pacing in our study group influenced the results. However, the three subgroups exhibited no difference in the rate of ventricular pacing.
Conclusion and clinical implications Advanced pacemaker diagnostic functions provide detailed insight into the electrophysiologic characteristics of paroxysmal AF. Apart from selecting patients who are likely to
687
benefit from pharmacologic therapy and ablative procedures, pacemaker diagnostics may help to identify responders to selective AF-preventive pacing algorithms. In the current study, “trigger fibrillators” were differentiated from “substrate fibrillators” according to the onset scenarios of AF. We conclude that “trigger fibrillators” may represent the target group for preventive pacing therapies. These patients in particular could benefit from specific PAC-suppressing algorithms. Vice versa, the efficacy of preventive algorithms is presumed to be limited in “substrate fibrillators,” characterized by low PAC activity and high AF burden.
References 1. Andersen HR, Nielsen JC, Thomsen PE, Thuesen L, Mortensen PT, Vesterlund T, Pedersen AK. Long-term follow-up of patients from a randomised trial of atrial versus ventricular pacing for sick-sinus syndrome. Lancet 1997;350:1210 –1216. 2. Kerr CR, Connolly SJ, Abdollah H, Roberts RS, Gent M, Yusuf S, Gillis AM, Tang AS, Talajic M, Klein GJ, Newman DM. Canadian Trial of Physiological Pacing: effects of physiological pacing during long-term follow-up. Circulation 2004;109:357–362. 3. Lamas GA, Lee KL, Sweeney MO, Silverman R, Leon A, Yee R, Marinchak RA, Flaker G, Schron E, Orav EJ, Hellkamp AS, Greer S, McAnulty J, Ellenbogen K, Ehlert F, Freedman RA, Estes NA III, Greenspon A, Goldman L. Ventricular pacing or dual-chamber pacing for sinus-node dysfunction. Mode Selection Trial in sinus-node dysfunction. N Engl J Med 2002;346:1854 –1862. 4. Defaye P, Dournaux F, Mouton E. Prevalence of supraventricular arrhythmias from the automated analysis of data stored in the DDD pacemakers of 617 patients: the AIDA study. The AIDA Multicenter Study Group. Automatic Interpretation for Diagnosis Assistance. Pacing Clin Electrophysiol 1998;21:250 –255. 5. Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation 1998;98:946 –952. 6. Kannel WB, Abbott RD, Savage DD, McNamara PM. Epidemiologic features of chronic atrial fibrillation: the Framingham Study. N Engl J Med 1982;306: 1018 –1022. 7. Wolf PA, Abbott RD, Kannell WB. Atrial fibrillation as independent risk factor for stroke: the Framingham Study. Stroke 1991;22:983–988. 8. Allessie MA, Boyden PA, Camm AJ, Kleber AG, Lab MJ, Legato MJ, Rosen MR, Schwartz PJ, Spooner PM, Van Wagoner DR, Waldo AL. Pathophysiology and prevention of atrial fibrillation. Circulation 2001;103:769 –777. 9. Hnatkova K, Waktare JE, Murgatroyd FD, Guo X, Baiyan X, Camm AJ, Malik M. Analysis of the cardiac rhythm preceding episodes of paroxysmal atrial fibrillation. Am Heart J 1998;135:1010 –1019. 10. Mitchell AR, Sulke N. How do atrial pacing algorithms prevent atrial arrhythmias? Europace 2004;6:351–362. 11. Yang A, Lüderitz B, Lewalter T. Atrial fibrillation: profit from cardiac pacing? Z Kardiol 2005;94:141–151. 12. Carlson MD, Ip J, Messenger J, Beau S, Kalbfleisch S, Gervais P, Cameron DA, Duran A, Val-Mejias J, Mackall J, Gold M. A new pacemaker algorithm for the treatment of atrial fibrillation: results of the Atrial Dynamic Overdrive Pacing Trial (ADOPT). J Am Coll Cardiol 2003;42:627– 633. 13. Padeletti L, Pürerfellner H, Adler SW, Waller TJ, Harvey M, Horvitz L, Holbrook R, Kempen K, Mugglin A, Hettrick DA. Combined efficacy of atrial septal lead placement and atrial pacing algorithms for prevention of paroxysmal atrial tachyarrhythmia. J Cardiovasc Electrophysiol 2003;14:1189 –1195. 14. Lee MA, Weachter R, Pollak S, Kremers MS, Naik AM, Silverman R, Tuzi J, Wang W, Johnson LJ, Euler DE. The effect of atrial pacing therapies on atrial tachyarrhythmia burden and frequency: results of a randomised trial in patients with bradycardia and atrial tachyarrhythmias. J Am Coll Cardiol 2003;41:1926 – 1932. 15. Lewalter T, Yang A, Pfeiffer D, Ruiter J, Schnitzler G, Markert T, Asklund M, Przibille O, Welz A, Esmailzadeh B, Linhart M, Lüderitz B. Individualised selection of pacing algorithms for the prevention of recurrent atrial fibrillation: results from the VIP registry. Pacing Clin Electrophysiol 2006;29:124 –134. 16. Sweeney MO, Hellkamp AS, Ellenbogen KA, Greenspon AJ, Freedman RA, Lee KL, Lamas GA. Adverse effect of ventricular pacing on heart failure and atrial fibrillation among patients with normal baseline QRS duration in a clinical trial of pacemaker therapy for sinus node dysfunction. Circulation 2003;107: 2932–2937.
688 17. Legato MJ, Bull MB, Ferrer MI. Atrial ultrastructure in patients with fixed intra-atrial block. Chest 1974;65:252–261. 18. Mary-Rabine L, Albert A, Pham TD, Hordof A, Fenoglio JJ Jr, Malm JR, Rosen MR. The relationship of human atrial cellular electrophysiology to clinical function and ultrastructure. Circ Res 1983;52:188 –199. 19. Van Wagoner DR, Pond AL, Lamorgese M, Rossie SS, McCarthy PM, Nerbonne JM. Atrial L-type Ca2⫹ currents and human atrial fibrillation. Circ Res 1999;85:428 – 436. 20. Hu H, Sachs F. Stretch-activated ion channels in the heart. J Mol Cell Cardiol 1997;29:1511–1523. 21. Mewes T, Ravens U. L-type calcium currents of human myocytes from ventricle of non-failing and failing hearts and from atrium. J Mol Cell Cardiol 1994;26: 1307–1320.
Heart Rhythm, Vol 3, No 6, June 2006 22. Israel CW, Grönefeld G, Ehrlich JR, Li YG, Hohnloser SH. Long-term risk of recurrent atrial fibrillation as documented by an implantable monitoring device: implications for optimal patient care. J Am Coll Cardiol 2004;43: 47–52. 23. Yang A, Hochhäusler M, Schrickel J, Bielik H, Shlevkov N, Schimpf R, Schwab JO, Esmailzadeh B, Schneider C, Mellert F, Welz A, Saborowski F, Lüderitz B, Lewalter T. Advanced pacemaker diagnostic features in the characterization of atrial fibrillation: impact on preventive pacing algorithms. Pacing Clin Electrophysiol 2003;26:310 –313. 24. Guyomar Y, Thomas O, Marquie C, Jarwe M, Klug D, Kacet S, Carlioz R, Ferrier A, Fossati F, Guerin S, Heuls S, Graux P. Mechanisms of onset of atrial fibrillation: a multicenter, prospective, pacemaker-based study. Pacing Clin Electrophysiol 2003;26:1336 –1341.