Europace (2015) 17, 864–870 doi:10.1093/europace/euv032

CLINICAL RESEARCH Atrial fibrillation

The development of AF over time in patients with permanent pacemakers: objective assessment with pacemaker diagnostics demonstrates distinct patterns of AF Conn Sugihara1, Rick Veasey1, Nick Freemantle 2, Steve Podd 1, Steve Furniss 1, and Neil Sulke 1* 1 Cardiology Research Department, Eastbourne Hospital, East Sussex Healthcare NHS Trust, Kings Drive, East Sussex BN21 2UD, UK; and 2Department of Primary Care and Population Health, University College London, London, UK

Received 25 January 2015; accepted after revision 30 January 2015

Aims

Registration

----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords

Atrial fibrillation † Permanent pacemaker

Introduction Atrial fibrillation (AF) has conventionally been divided into paroxysmal AF, and several forms of persistent AF (persistent, long-standing persistent and permanent). However, defining AF by the duration of the longest detected episode or the clinical decision to perform DC cardioversion1 may have limitations. Further, this definition of AF subtypes allows progression to persistent AF, but never regression back to paroxysmal AF.

Published series describing the natural history of AF have depended on standard definitions.2 – 7 In the Euro Heart Survey, 15% of 1219 patients with initially paroxysmal AF were reclassified as having persistent AF after 1 year.8 Investigators observed clinical correlates for AF progression, and developed the HATCH score (2 points for heart failure or previous cerebrovascular event; one point for age .75, hypertension or chronic obstructive pulmonary disease) as a predictive tool for AF progression. It is in this context a schematic that in the 2010 ESC guideline on AF

* Corresponding author. Tel: +44 1323 435869; fax: +44 1323 435821; Email: [email protected] Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2015. For permissions please email: [email protected].

Downloaded from by guest on June 1, 2015

To describe the long-term patterns of atrial fibrillation (AF) in patients with permanent pacemakers. ..................................................................................................................................................................................... Methods A total of 2092 pacemaker Holter downloads were analysed in 323 patients with dual chamber permanent pacemakers, and results describing a cumulative 1031 patient-years of beat-to-beat monitoring. Four subtypes of AF were applied: (i) non-progressive low-burden PAF (NPLB-PAF, n ¼ 120): such patients never have .1% AF burden throughout follow-up; (ii) chronic progressive PAF (CP-PAF, n ¼ 55): AF burden increases but is never 100%; (iii) relapsing–remitting PAF (RR-PAF, n ¼ 78): AF burden has reduced at least once by more than 2% and is never 100%; (iv) persistent AF (PersAF, n ¼ 70): 100% AF burden for at least 28 days. Overall, mean AF burden rose 0.34% per year (P , 0.0001). After accounting for age, heart failure (HF) had a significant interaction with AF burden (P ¼ 0.0022), but HATCH score and CVA/TIA did not. There were no differences in the frequency or duration of monitoring between the four AF subtypes. Atrial fibrillation episode frequency discriminated between subtypes (P ¼ 0.0004). Eighteen of 70 (26%) patients with PersAF had pacemaker documented episodes of sinus rhythm (i.e. reversion to ‘paroxysmal AF’) after the onset of PersAF. ..................................................................................................................................................................................... Conclusion In this cohort, the development of AF over time appears more complex than current definitions suggest. Atrial fibrillation can remain low burden without progression, remit – relapse, or progress as is described in currently accepted definitions. More frequent episodes of AF indicated a favourable subtype. Persistent AF is not inevitable, and can revert to paroxysmal AF. ..................................................................................................................................................................................... Clinical Trial NCT02016950, http://clinicaltrials.gov/show/NCT02016950.

865

Distinct patterns of AF in patients with PPMs

What’s new? † Patients with AF and permanent pacemakers were studied over long-term follow-up † Persistent AF was not inevitable, and could revert to paroxysmal AF. † AF could remain low burden without progression, remit–relapse, or progress as is described in currently accepted definitions. † More frequent episodes of AF indicated a favourable subtype.

Methods This study complies with the Declaration of Helsinki, and was approved by the appropriate national Ethics Committee and institutional review board. It is not possible to consistently apply currently accepted definitions of AF to pacemaker-acquired data, so a new classification of AF subtypes based on pacemaker-derived AF data was prospectively defined. Boundaries between four mutually exclusive groups were selected based upon the limits of pacemaker AF diagnostic technology. A large pacemaker data set was examined, and patients were divided into groups based on these subtypes. The clinical and electrophysiological characteristics of each group were then described.

Atrial fibrillation subtypes (1) Non-progressive low-burden paroxysmal AF (NPLB-PAF): AF burden was never .1%. This subtype was defined to select patients with the lowest AF burden. (2) Chronic progressive paroxysmal AF (CP-PAF): AF burden was .1%, never decreased and never reached 100%. (3) Relapsing– remitting paroxysmal AF (RR-PAF): AF burden was never 100%, and there was at least one absolute reduction in AF burden of ≥2% during the follow-up period. The cutoff of 2% absolute reduction in AF burden was selected because the lower limit of detection of AF burden for most PPMs is 1% (i.e. an apparent 1% absolute change in AF burden may not represent a true change). This subtype was defined to describe patients who had detectable reductions in AF burden during follow-up.

Data collection and analysis All patients with a known diagnosis of AF and a dual chamber DDDRP PPM equipped with full AF diagnostic capabilities were identified. All had complete retrospective analysis of all available pacemaker download data in January 2014. All available stored intracardiac electrocardiograms were extracted and checked for appropriate arrhythmia detection. If there was any evidence of inappropriate AF detection, or a pacemaker programming change occurred that had an impact on AF detection, that patient was excluded from the analysis. Otherwise, all available consecutive PPM downloads are reported. The minimum duration of monitoring for any time period for comparison was 28 days. Atrial fibrillation burden was defined as the proportion of monitored time a device detected AF. Atrial fibrillation burden data were available for all downloads and all patients had more than one AF burden recorded. AF episode frequency (number of device-detected AF episodes per year) was also analysed. AF episode frequency was defined as the number of episodes of AF detected by the device, divided by the duration of the monitoring period. Atrial fibrillation burden and AF episode frequency are different measures; it is possible to have many short AF episodes or few long episodes equating to the same AF burden. Complete review of the medical records of all patients was undertaken by clinicians blinded to pacemaker data. Age was defined as the age at the most recent available pacemaker download. The presence of any of the following clinical comorbidities were recorded: heart failure (HF), hypertension (HTN), stroke or transient ischaemic attack (CVA/TIA), chronic obstructive pulmonary disease (COPD), thyroid disease (any history of treated hypothyroidism or hyperthyroidism), valve disease (severe aortic or mitral disease or any valve surgery), vascular disease (coronary or peripheral vascular disease), renal disease (end-stage renal failure), cancer (any treated cancer except completely excised non-melanoma skin cancers), obstructive sleep apnoea (OSA), and previous cardiac surgery. The CHADS2, CHA2DS2Vasc and HATCH scores were calculated for each patient. Clinical characteristics were collected at the time of the last available pacemaker download. Based on all available pacemaker AF burden data across long-term follow up, each patient was classified as NPLB-PAF, CP-PAF, RR-PAF, or PersAF. Patients with PersAF were further analysed for AF burden data prior to and after the first detected episode of 100% AF burden.

Statistical analysis Data analysis was performed using SPSS statistical software (version 21, IBM Corp, New York, USA), and SAS 9.3 (SAS Institute, Cary, NC, USA). The median and interquartile ranges are displayed for descriptive data. Categorical variables are reported as observed number of patients (percentage). A P-value of ,0.05 was considered significant. The relationship between AF burden and clinical characteristics was examined in a multivariable generalized mixed linear model accounting for multiple measures within a subject. A candidate set of predictive variables (HATCH score, CVA/TIA, HF, and age) were modelled with AF burden as the response variable. Non-significant candidate explanatory variables were removed using backward selection. Data were then analysed according to four groups: NPLB-PAF, CP-PAF, RR-PAF, or PersAF. The number, mean duration of follow-up, mean number of downloads, mean and median AF burdens were compared (Table 2). Clinical comorbidities and demographics were described and compared per group (Table 3).

Downloaded from by guest on June 1, 2015

depicted inevitable AF progression despite antiarrhythmic drugs and AF ablation.9 Assuming that AF will deteriorate regardless of intervention has important clinical implications. Antiarrhythmic therapy might only be considered as a temporizing rather than a curative measure. Patients with ‘advanced’ AF might be offered different treatment to patients perceived to have paroxysmal AF. Even patients who have not yet developed persistent AF, but are considered at risk of doing so, may be offered different treatments. Indeed the investigators who developed the HATCH score suggested that ‘cardioversion and ablation may be avoided in patients with a high HATCH score’.8 Pacemaker AF diagnostics are considered the gold standard for AF detection,10 and are both objective and more descriptive of AF than currently employed techniques.11 We conducted a retrospective analysis of patients with AF and permanent pacemakers (PPMs) to observe patterns of AF as defined by their implanted device.

(4) Persistent AF (PersAF): 100% AF burden detected by their pacemaker for at least 28 days at least once during follow-up. This category was chosen to select only patients who unequivocally had persistent AF.

866

C. Sugihara et al.

Table 1 Monitoring intensity and AF burden for population and subtypes NPLB-PAF

CP-PAF

RR-PAF

PersAF

Total

n

120

55

78

70

323

Median (IQR) duration of beat-to-beat pacemaker follow-up (days)

1213 (689– 1668)

735 (374– 1307)

1190 (615–1903)

1054 (417– 1546)

1072 (523–1630)

Median (IQR) number of pacemaker downloads Median AF burden (IQR) (%)

5 (3– 8) 0.1 (0.0–0.3)

5 (4–7) 1.0 (0.1– 8.1)

7 (4–9) 4.9 (1.0– 28.0)

6 (4– 10) 98.0 (5.1– 100.0)

6 (4–8) 1.0 (0.0– 30.4)

...............................................................................................................................................................................

AF, atrial fibrillation; NPLB-PAF, non-progressive low-burden paroxysmal AF; CP-PAF, chronic progressive paroxysmal AF; RR-PAF, relapsing –remitting paroxysmal AF; PersAF, persistent AF.

Table 2 Pacing indication for population and subtype Pacing indication

NPLB-PAF

CP-PAF

RR-PAF

PersAF

Total

Atrioventricular block (%) Sick sinus syndrome (%)

50 (41.7) 58 (48.3)

14 (25.5) 30 (54.5)

18 (23.1) 47 (60.3)

20 (28.6) 36 (51.4)

102 (31.6) 171 (52.9)

Atrial fibrillation (%)

10 (8.3)

8 (14.5)

12 (15.4)

14 (20.0)

44 (13.6)

2 (1.7)

3 (5.5)

1 (1.3)

0 (0.0)

6 (1.9)

...............................................................................................................................................................................

Other (%)

AF, atrial fibrillation; NPLB-PAF, non-progressive low-burden paroxysmal AF; CP-PAF, chronic progressive paroxysmal AF; RR-PAF, relapsing –remitting paroxysmal AF; PersAF, persistent AF.

NPLB-PAF

CP-PAF

RR-PAF

PersAF

Total

34.4 (9.5–74.0) 26.1 (0.5–99.0)

29.0 (4.0–75.3) 46.1 (2.0–99.3)

55.6 (21.3–88.0) 70.0 (6.0– 98.5)

6.8 (0.4– 52.3) 90.0 (27.0–99.5)

35.0 (6.0– 77.0) 35.0 (6.0– 77.0)

............................................................................................................................................................................... Median atrial pacing (IQR) (%) Median ventricular pacing (IQR) (%)

AF, atrial fibrillation; NPLB-PAF, non-progressive low burden paroxysmal AF; CP-PAF, chronic progressive paroxysmal AF; RR-PAF, relapsing–remitting paroxysmal AF; PersAF, persistent AF.

The relationship between AF episode frequency and AF subtype was compared accounting for multiple measures within a subject, with the denominator degrees of freedom derived from the patient stratum.

Results All patients (2546) with PPMs in a single centre were screened. In total, 4095 individual device downloads were examined. A total of 323 patients with AF and the appropriate pacemaker data were identified. In the 323 patients with evaluable data, 2092 downloads are reported, of which 1838 (88%) had AF episode data available (as not all PPMs capture AF episode data). A cumulative 1031 patientyears of beat-to-beat arrhythmia data (mean 3.2 years per patient) were analysed. In a multivariable model with HATCH, age, CVA/TIA, and HF as predictive variables, AF burden was found to increase on average by 0.34% for every 1 year increase in age (P , 0.0001). After accounting for age at pacemaker download, only HF was found to have a significant interaction with AF burden (P ¼ 0.002).

Atrial fibrillation subtypes There were 120 patients with NPLB-PAF, 55 with CP-PAF, 78 with RR-PAF, and 70 with PersAF (Figure 1). There were no significant differences between the duration or frequency of arrhythmia monitoring between the NPLB-PAF, CP-PAF, RR-PAF, or PersAF patient groups (Table 1). The majority (52.9%) of patients had their PPMs initially implanted for sick sinus syndrome. Approximately, one-third (31.6%) had atrioventricular block (AVB) on pacemaker implant, and 13.6% had AF as a pacing indication. Six patients (1.9%) had other indications such as hypertrophic cardiomyopathy, ventricular tachyarrhythmias, or heart failure. Patients with NPLB-PAF were more likely to have had AVB as a pacing indication (Table 2). Median atrial pacing and ventricular pacing were both 35.0%. However, there was a wide range in the proportion of atrial and ventricular pacing seen both in the population as a whole and within each subtype (Table 3). Fourteen clinical comorbidities and three composite scores were compared (Table 4). There were no statistically significant differences

Downloaded from by guest on June 1, 2015

Table 3 Atrial and ventricular pacing for population and subtype

867

Distinct patterns of AF in patients with PPMs

between AF subtypes except for in HF, HTN, and the HATCH score. Heart failure was significantly less prevalent in NPLB-PAF than in any other subtype (P ¼ 0.001). The AF burden and episode frequency in the 88% of downloads where AF episode data were available are presented in Table 5. Median AF burden was highest (99.8%) in PersAF, which reflects the fact that most patients who had 100% AF burden remained at

Persistent AF (PersAF) 22%

Relapsing remitting PAF (RR-PAF) 24%

Non-progressive low burden PAF (NPLB-PAF) 37%

that level for many years. For AF data prior to the onset of 100% AF, the median AF burden was comparable to other groups (3.5%). Atrial fibrillation episode frequency was highest in patients with RR-PAF. Overall, AF episode frequency discriminated between AF subtypes (P ¼ 0.0004). Twenty-two percent of the population was classified as PersAF. Although the majority remained in AF once persistent AF occurred, 18 of 70 (26%) had sustained reduction in AF burden (to ,50%) after the onset of 100% AF (Figure 2). PPM downloads for these patients were re-examined, and all patients had proven sinus rhythm after sustained AF on intracardiac electrograms. Twelve of these 18 patients had a left atrial ablation which coincided with their decline in AF burden. However, in one patient, AF burden declined spontaneously 4 years after an AF ablation which had been assumed to have failed. In three patients, AF burden remained persistently low or non-existent on multiple PPM downloads after DC cardioversion. Finally, in 3 of 18 (17%) patients, AF burden was observed to decline spontaneously, without any antiarrhythmic therapy or intervention.

Discussion

Figure 1 Proportion of population by atrial fibrillation subtype.

Understanding of the natural history of AF requires repeated consistent and objective assessment for AF in the same individual over the long term. Previous studies have accepted very low levels of AF monitoring, with as few as a single 12-lead ECG per year being described. The Euro Heart Survey analysis for AF progression8 reported 1 year follow-up for 1219 patients, which may appear comparable to the 1031 patient-years of follow-up in our study. However, we report automated arrhythmia data based on intracardiac electrograms,

Table 4 Clinical comorbidities, demographics, and composite scores for population and subtype Total

NPLB-PAF

CP-PAF

RR-PAF

PersAF

P-value

Mean age (years)

77.4

76.3

76.2

78.7

78.7

0.244

Gender (female), n (%) Heart failure, n (%)

154 (47.7) 52 (16.1)

59 (49.2) 9 (7.5)

29 (52.7) 11 (20.0)

41 (52.6) 13 (16.7)

25 (35.7) 19 (27.1)

0.143 0.004a

Hypertension, n (%)

147 (45.5)

42 (35)

26 (47.3)

43 (55.1)

36 (51.4)

0.025a

Diabetes, n (%) Stroke or TIA, n (%)

26 (8) 40 (12.4)

9 (7.5) 11 (9.2)

5 (9.1) 6 (10.9)

10 (12.8) 14 (17.9)

2 (2.9) 9 (12.9)

0.166 0.321

COPD, n (%)

17 (5.3)

5 (4.2)

3 (5.5)

4 (5.1)

5 (7.1)

0.851

Vascular disease, n (%) Thyroid disease, n (%)

90 (27.9) 28 (8.7)

29 (24.2) 8 (6.7)

17 (30.9) 7 (12.7)

26 (33.3) 10 (12.8)

18 (25.7) 3 (4.3)

0.495 0.161

Cancer, n (%)

7 (2.2)

1 (0.8)

1 (1.8)

3 (3.8)

2 (2.9)

0.526

Renal failure, n (%) Valve disease, n (%)

1 (0.3) 29 (9.0)

1 (0.8) 11 (9.2)

0 (0.0) 4 (7.3)

0 (0.0) 9 (11.5)

0 (0.0) 5 (7.1)

0.638 0.773

...............................................................................................................................................................................

Obstructive sleep apnoea, n (%)

1 (0.3)

1 (0.8)

0 (0.0)

0 (0.0)

0 (0.0)

0.638

Cardiac surgery, n (%) CHADS2 score (mean)

35 (10.8) 1.6

11 (9.2) 1.3

6 (10.9) 1.6

9 (11.5) 1.9

9 (12.9) 1.7

0.877 0.067

CHA2DS2Vasc score (mean)

3.2

2.8

3.3

3.7

3.3

0.096

HATCH score (mean)

1.7

1.3

1.7

1.9

1.7

0.025a

Values presented as n (% of AF subtype) unless otherwise stated. P-values are for the hypothesis that there are no differences between subtypes for the relevant parameter. AF, atrial fibrillation; NPLB-PAF, non-progressive low-burden paroxysmal AF; CP-PAF, chronic progressive paroxysmal AF; RR-PAF, relapsing –remitting paroxysmal AF; PersAF, persistent AF. a P-value , 0.05.

Downloaded from by guest on June 1, 2015

Chronic progressive PAF (CP-PAF) 17%

868

C. Sugihara et al.

Table 5 Median AF burden and episode frequency per AF subtype and prior to the onset of PersAF NPLB-PAF

CP-PAF

RR-PAF

PersAF

Prior to onset of PersAF

............................................................................................................................................................................... Number where AF episode frequency data available

116

53

75

70

34

Median AF burden (IQR) (%) Median AF episode frequency per year (IQR)

0.1 (0.0–0.1) 0 (0– 5)

1.0 (0.1– 4.7) 30 (1– 300)

4.9 (0.1–28.2) 110 (4– 1108)

99.8 (4.2– 100) 6 (2 –56)

3.5 (0.1– 53.4) 25 (2– 160)

AF, atrial fibrillation; NPLB-PAF, non-progressive low burden paroxysmal AF; CP-PAF, chronic progressive paroxysmal AF; RR-PAF, relapsing–remitting paroxysmal AF; PersAF, persistent AF.

100% 90% 80%

AF burden

70% 60% 50%

30% 20% 10% 0% 0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Years after development of 100% AF burden

Figure 2 Atrial fibrillation burden data for 18 of 70 (26%) of patients with PersAF who had sustained reduction in AF burden after the onset of 100% AF burden. Twelve of 18 patients had AF ablation temporally associated with decline in AF ablation (solid lines). Six of 18 patients had decline in AF burden not associated with AF ablation (interrupted lines). AF, atrial fibrillation; PersAF, persistent atrial fibrillation.

and this represents significantly improved objective data for analysis. All data in this study is based on continuous beat-to-beat rather than intermittent monitoring.

The natural history of atrial fibrillation The 2010 ESC guideline9 suggests that AF burden should inexorably increase over the long term, regardless of medical intervention. Our data confirm that this does occur, but only in a minority of patients. More than one-third of patients never suffered more than 1% AF burden during follow-up. Indeed many patients with conventionally defined ‘persistent’ AF had spontaneous improvement in AF burden. Even patients with very high AF burden had large reductions in AF during long-term follow-up.

Atrial fibrillation subtypes To enhance the currently accepted definitions of AF using beat-to-beat arrhythmia assessment, we categorized AF subtypes

according to long-term patterns of AF measured continuously over several years. The single largest patient group had low-burden non-progressive paroxysmal AF (LBNP-PAF). Their mean CHA2DS2-Vasc score was 2.8, illustrating the high embolic risk for this group. All patients with LBNP-PAF in our study were appropriately anticoagulated. It is probable, however, that the majority of patients with LBNP-PAF outside this study are not. With such a low AF burden, these individuals are unlikely to have their AF detected despite being at high risk of preventable cerebrovascular disease.12 The RR-PAF, seen in 1 in 6 of our cohort, appears inconsistent with the current perception of AF progression. These patients had similar comorbidities, demographics, and follow-up as those with more progressive forms of AF. The median AF burden in patients with RR-PAF was comparable to patients destined to develop persistent AF. The only distinguishing feature was that they had discrete, frequent episodes of AF. This suggests that RR-PAF has a distinctly different pattern of AF which is not equivalent to CP-PAF observed at a

Downloaded from by guest on June 1, 2015

40%

869

Distinct patterns of AF in patients with PPMs

different point in time. It is likely that individuals with RR-PAF have different triggers and substrate for the initiation and maintenance of AF. Chronic progressive paroxysmal AF (CP-PAF) and persistent AF (PersAF) are the subtypes that approximate best to the standard model for AF progression. However, patients with CP-PAF did not appear to have aggressively worsening AF, and many with PersAF were observed to have reductions in AF burden during follow-up. Importantly, the two groups combined represented less than half of this elderly population with a high prevalence of comorbid disease. Continuous AF for more than a month as proved by continuous intracardiac monitoring is ‘persistent’, however, it is comparatively uncommon in this cohort. Remarkably, more than a quarter of this group had sustained, clinically relevant reductions in AF burden (Figure 2). The reductions in AF burden for these patients were large, sustained, and unrelated to any methodological issue with the definition of AF burden.

Atrial fibrillation subtypes and clinical comorbidities

Atrial fibrillation subtypes in clinical practice The AF patterns reported in this analysis describe patterns of AF over all available follow-up. Although these descriptions are useful to illustrate the true variation in the natural history of AF, they cannot necessarily be applied prospectively to individual patients. However, these findings have clear implications in clinical practice. There appear to be important patient subsets with either a very low burden of AF, or frequent but non-sustained episodes of AF. The electrophysiological basis for these patterns must be driven by the balance between triggers to initiate AF and the substrate to maintain it. Long-term AF monitoring may indicate the AF arrhythmia mechanism, even when clinical comorbidities appear to be unhelpful predictors.13

Study limitations Any long-term description of the time-course for AF can be criticized. It should be noted that all patients in this analysis had a primary indication for permanent pacing. Both atrial and ventricular pacing has been recognized to affect AF burden and patterns.14 – 19 Hence we do not suggest that all our findings are generalizable to

Future directions The expansion in remote home monitoring for paced patients has created a large and growing database of pacemaker arrhythmia data. These large data sets are attractive targets for a similar analysis to our own, and may provide more insight relevant to a larger population of patients. Similarly, there is great potential in analysing arrhythmia data from implantable loop recorders,20,21 which offer the best available method for long-term arrhythmia surveillance in patients who do not have an indication for PPMs.

Clinical implications Atrial fibrillation does not always progress. Conventional descriptions conceal the true variety and complexity of AF, and fail to detect patient groups who have a favourable long-term phenotype. Clinical comorbidities are a poor predictor of AF type, and should not be used to inform patients of the likely development of their arrhythmia. High AF episode frequency appears to signify a nonprogressive phenotype. Persistent AF appears to be neither inevitable nor permanent.

Conclusions Objective arrhythmia monitoring offers unique and clinically relevant insights into the natural history of AF. The currently accepted model of inevitable progression from paroxysmal to persistent AF was not reflected in our data. Our data suggest that the history of AF is more complex than the currently accepted model of progression from paroxysmal to persistent AF. Furthermore, clinical comorbidities did not correlate with AF patterns. The elderly, and patients with comorbid disease, appear no more likely to have a progressive phenotype of AF than other patients. Expanding access to long-term arrhythmia monitoring for patients with AF may improve management and outcomes. Future efforts to characterize AF patterns require more detailed data on AF (and sinus rhythm) episode duration and frequency distribution.

Downloaded from by guest on June 1, 2015

When patients were objectively split on the basis of AF phenotypes, previously recognized predictors for AF progression could not identify patients with persistent AF. In this cohort, the absence of heart failure appeared to increase the chance of having non-progressive low-burden paroxysmal AF. However, the most notable finding is that no comorbidity had any clinically meaningful discriminatory power to detect persistent AF. The disjunction between this analysis and prior work may reflect the necessarily subjective data other investigators have been forced to utilize. Unfortunately, it seems that the HATCH score may be biased by the fact that if a clinician expects to find persistent AF, they are often able to diagnose it. This has implications for the use of antiarrhythmic therapies in patients perceived to have paroxysmal AF.

patients who are not paced. However, by using more complete, objective, and accurate data than previously described, our findings may contribute to a better understanding of the natural history of AF. Permanent pacemakers have differing methods for detection and classification of atrial arrhythmias. In order to describe long-term follow-up in the maximum number of patients, any device capable of automatically recording atrial arrhythmia burden was acceptable in this study. Programming features, pacing algorithms, lead site, and sensing capabilities may all affect between-subject comparisons of arrhythmia data. However, all AF burden analyses were undertaking on a within-subject basis, which mitigated the potential effect of this heterogeneity. Furthermore, existing research in this field has previously relied on surface ECGs. The technical limitations inherent in intermittent surface ECG analysis are an order of magnitude greater than in beat-to-beat PPM arrhythmia detection. ESC guidance states that antiarrhythmic therapy has little effect on the natural history of AF. Therefore, exposure to drug and ablative treatment was not considered in this analysis, but is the subject of ongoing work.

870

Acknowledgements Data collection and analysis: Sarah Bridgland, Jet Van Zalen. Editing assistance: Dr Nik Patel, Dr Guy Lloyd. All are affiliated to the Cardiology Research Department, Eastbourne Hospital, East Sussex Healthcare NHS Trust, UK.

Funding This work was supported by the Eastbourne Cardiology Research Charity Fund. Conflict of interest: none declared.

References

8. de Vos CB, Pisters R, Nieuwlaat R, Prins MH, Tieleman RG, Coelen R-JS et al. Progression from paroxysmal to persistent atrial fibrillation: clinical correlates and prognosis. J Am Coll Cardiol 2010;55:725 –31. 9. Camm AJ, Kirchhof P, Lip GY, Schotten U, Savelieva I, Ernst S et al. Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Europace 2010;12:1360 –420. 10. Brignole M, Vardas P, Hoffman E, Huikuri H, Moya A, Ricci R et al. Indications for the use of diagnostic implantable and external ECG loop recorders. Europace 2009;11: 671 –87. 11. Ziegler PD, Koehler JL, Mehra R. Comparison of continuous versus intermittent monitoring of atrial arrhythmias. Heart Rhythm 2006;3:1445 –52. 12. Hohnloser SH, Capucci A, Fain E, Gold MR, van Gelder IC, Healey J et al. A Symptomatic atrial fibrillation and Stroke Evaluation in pacemaker patients and the atrial fibrillation Reduction atrial pacing Trial (ASSERT). Am Heart J 2006;152:442 –7. 13. Ricci RP. Disease management: atrial fibrillation and home monitoring. Europace 2013;15(Suppl 1):i35 –9. 14. Lamas GA, Lee KL, Sweeney MO, Silverman R, Leon A, Yee R et al. Ventricular pacing or dual-chamber pacing for sinus-node dysfunction. N Engl J Med 2002;346: 1854 –62. 15. Stambler BS, Ellenbogen KA, Orav EJ, Sgarbossa EB, Estes NAM, Rizo-Patron C et al. Pacemaker Selection in the Elderly Trial Investigators. Predictors and clinical impact of atrial fibrillation after pacemaker implantation in elderly patients treated with dual chamber versus ventricular pacing. Pacing Clin Electrophysiol 2003;26:2000 –7. 16. Connolly SJ, Kerr CR, Gent M, Roberts RS, Yusuf S, Gillis AM et al. Effects of physiologic pacing versus ventricular pacing on the risk of stroke and death due to cardiovascular causes. New Engl J Med 2000;342:1385 –91. 17. Nielsen JC, Kristensen L, Andersen HR, Mortensen PT, Pedersen OL, Pedersen AK. A randomized comparison ofatrial and dual-chamber pacing in 177 consecutive patients with sick sinus syndrome. J Am Coll Cardiol 2003;42:614 –23. 18. Sweeney MO, Bank AJ, Nsah E, Koullick M, Zeng QC, Hettrick D et al. Minimizing ventricular pacing to reduce atrial fibrillation in sinus-node disease. N Engl J Med 2007;357:1000 –8. 19. Kristensen L. Incidence of atrial fibrillation and thromboembolism in a randomised trial of atrial versus dual chamber pacing in 177 patients with sick sinus syndrome. Heart 2004;90:661 –6. 20. Hindricks G, Pokushalov E, Urban L, Taborsky M, Kuck K-H, Lebedev D et al. Performance of a new leadless implantable cardiac monitor in detecting and quantifying atrial fibrillation results of the XPECT trial. Circulation 2010;3:141 –7. 21. Eitel C, Husser D, Hindricks G, Fru¨hauf M, Hilbert S, Arya A et al. Performance of an implantable automatic atrial fibrillation detection device: impact of software adjustments and relevance of manual episode analysis. Europace 2011;13:480–5.

Downloaded from by guest on June 1, 2015

1. Fuster V, Ryde´n LE, Cannom DS, Crijns HJ, Curtis AB, Ellenbogen KA et al. ACC/ AHA/ESC 2006 guidelines for the management of patients with atrial fibrillation: full text. Europace 2006;8:651 –745. 2. Kerr CR, Humphries KH, Talajic M, Klein GJ, Connolly SJ, Green M et al. Progression to chronic atrial fibrillation after the initial diagnosis of paroxysmal atrial fibrillation: results from the Canadian Registry of Atrial Fibrillation. Am Heart J 2005;149: 489 –96. 3. Kato T, Yamashita T, Sagara K, Iinuma H, Fu L-T. Progressive nature of paroxysmal atrial fibrillation: observations from a 14-year follow-up study. Circulation J 2004; 68:568–72. 4. Nieuwlaat R. Atrial fibrillation management: a prospective survey in ESC member countries: The Euro Heart Survey on Atrial Fibrillation. Eur Heart J 2005;26: 2422 –34. 5. Krahn AD, Manfreda J, Tate RB, Mathewson FAL, Cuddy TE. The natural history of atrial fibrillation: incidence, risk factors, and prognosis in the Manitoba follow-up study. Am J Med 1995;98:476 –84. 6. Pappone C, Radinovic A, Manguso F, Vicedomini G, Ciconte G, Sacchi S et al. Atrial fibrillation progression and management: a 5-year prospective follow-up study. Heart Rhythm 2008;5:1501 –7. 7. Levy S, Maarek M, Coumel P, Guize L, Lekieffre J, Medvedowsky J-L et al. Characterization of different subsets of atrial fibrillation in general practice in France: the ALFA study. Circulation 1999;99:3028 –35.

C. Sugihara et al.

The development of AF over time in patients with permanent pacemakers: objective assessment with pacemaker diagnostics demonstrates distinct patterns of AF.

To describe the long-term patterns of atrial fibrillation (AF) in patients with permanent pacemakers...
156KB Sizes 0 Downloads 6 Views