Reproducibility of internal atrial defibrillation threshold in paroxysmal and persistent atrial fibrillation

Abstract

Background

Several pharmacological or technical factors may affect atrial defibrillation threshold (ADFT) for internal cardioversion (ICV) in the treatment of atrial fibrillation (AF).

Methods

We evaluated the reproducibility of ADFT in lone paroxysmal (electrically induced AF, 10 pts, 51 ± 4 years) or persistent AF (15 pts, 64 ± 7 years). The AF pattern (F–F interval) was characterised before each ICV attempt. A first step-up synchronised ICV test (ICV1, biphasic shock waveform 6 ms/6 ms) with increasing energy levels from 0.2 to 20 J was performed by a dual-lead defibrillation system (right atrium–coronary sinus configuration) connected to an external cardioverter defibrillator. After 30  min of stable sinus rhythm, a new sustained AF was induced (>20 min duration) and ICV protocol was repeated (ICV2). The AF cycle length was recorded for 30  s from the lateral wall of right atrium in basal condition and before each cardioversion attempt.

Results

The mean values of AF cycle length before a successful shock were similar in both AF populations (paroxysmal AF: pre-ICV1 175 ± 21 ms vs pre-ICV2 181 ± 20 ms (p=NS); persistent AF pre-ICV1 194 ± 25 ms vs pre-ICV2 202 ± 15 ms (p=NS)). No significant differences were observed between the two successful ICV tests concerning intensity, energy and impedance levels. The value of ADFT energy was reproducible in paroxysmal AF population (SD differences 1.2, coefficient of variability 9.6%). In persistent AF group only the impedance was reproducible (SD differences 2.6 Ω, coefficient of variability 4.5%), but not the energy requirements (SD differences 9.6, coefficient of variability 44.3%).

Conclusions

ADFT is reproducible in paroxysmal AF patients, while a high coefficient of variability is present in persistent AF, possibly related to different patterns of re-entrant circuits in the reinduced AF. This observation is important in order to evaluate factors influencing ICV-ADFT correctly in AF patients.

internal cardioversion, atrial fibrillation, electrophysiological study

Introduction

Atrial fibrillation (AF) is one of the most frequently occurring cardiac arrhythmias, often resulting in hospital admission. Clinical symptoms such as embolic stroke and possible high risks of advanced heart failure underline the need to restore sinus rhythm as quickly as possible [1].

Several studies have shown that the transvenous internal cardioversion of AF can be performed with a considerably lower energy level compared with the external approach [2–,5]. Pain perception has a major impact on quality of life in patients with an implantable atrial defibrillator and with present energy levels many patients may suffer considerable discomfort. Consequently several attempts have been proposed to decrease the atrial defibrillation threshold (ADFT), to reduce the energy requirement for a successful cardioversion and to improve the tolerability of this therapeutic procedure, with encouraging results [6–,11]. With lead configurations embracing both atria (right atrium; coronary sinus) energy requirements could be minimised to a level below 3 J. This internal low energy cardioversion allows delivery of the shock energy in conscious patients with mild sedation. However, the presence of spontaneous variability of ADFT in the same patient with AF is still unknown. Thus evaluation of reproducibility of ADFT is a necessary preliminary step before assessment of different technical or therapeutic approaches to AF.

The aim of our study was to evaluate the short-term reproducibility of the ADFT in patients with paroxysmal and persistent AF treated with low-energy internal atrial cardioversion.

Methods

Study population

Patients with a history of frequent paroxysmal AF episodes scheduled for electrophysiological study or with persistent AF (<1 month) were considered for enrolment in the study.

Exclusion criteria were age >80 years, New York Association functional class III or IV heart failure, mean ventricular rate during AF <45 beats/min, recent (<6 months) myocardial infarction or unstable angina pectoris, serum potassium blood level <3.5 mEq/L), severe renal or hepatic failure and severe hypoxia (pO₂ < 55 mmHg).

The investigation conforms to the principles set forth in the Declaration of Helsinki and the Local Ethics Committee approved the study protocol. All the patients gave written informed consent to undergo internal cardioversion and accepted the study protocol.

Each patient arrived in the electrophysiology laboratory in the fasting, non-sedated state, with all antiarrhythmic drugs (including digoxin) discontinued for at least five half-lives (6 months for amiodarone). In patients with chronic AF the usual anticoagulation guidelines of American College Chest Physicians [12] were followed.

Under fluoroscopic guidance in the EP lab , a 6-Fr standard quadripolar catheter (DAIG Corporation, Minnetonka, MN, USA) was inserted through right femoral vein puncture and positioned into the right ventricle for synchronising the shocks and, if needed, for backup ventricular pacing. Furthermore two “single coil” catheters for internal cardioversion (Rhythm Technologies, Jacksonville, FL, USA) were inserted through left subclavian vein and right femoral vein punctures and positioned into the coronary sinus and antero-lateral right atrial wall. The two defibrillation catheters were connected with an external defibrillator (HVS- 02, Ventritex, Inc., Sunnyvale, CA, USA) able to deliver 6 ms/6 ms biphasic shocks, using both capacitors of the defibrillator to create two sequential 6 ms shocks of opposite polarity, a 0.2  ms intershock delay and trailing-edge voltage of first phase equal to the leading-edge voltage of the second phase.

In patients with paroxysmal AF atrial pacing 1–3 extrastimuli delivered at decreasing coupling intervals or atrial burst were used to induce AF; 20 min of continuous AF was required prior to ADFT determination, defined as the energy at which successful conversion to sinus rhythm was achieved.

A step up protocol was used to determine ADFT starting from an energy level of 0.5 J and increasing by 0.5 J at 2 min intervals; joules, volts and impedance were computed for each attempt, following a previously reported protocol [13]. A shock was considered successful if stable sinus rhythm was restored within 3 s; in this case the shock sequence was terminated.

After at least 20 min of stable sinus rhythm, AF was reinduced following the abovementioned protocol and a new series of shocks was delivered after 20 min of stable AF, starting at leading-edge voltage corresponding to two protocol steps below the previous effective shock voltage. If the shock was unsuccessful in sinus rhythm restoration, we restarted the step-up protocol from the beginning.

The short duration of sinus rhythm after the electrical cardioversion is not considered significantly to influence the electrophysiological parameters of the atria with respect to electrical remodeling in patients with persistent AF [14].

In order to improve the tolerability of the procedure, the patients were sedated by a midazolam infusion (2.5 mg over 30 s), repeated if necessary in 1 mg increments (maximal dose: 12 mg) until patients developed slurred speech and were not easily arousable by verbal and physical stimuli (Ramsay sedation score 5) [15].

Cardiac rhythm was continuously monitored during the procedure; we did not routinely use sedation if not requested by the patient.

Electrograms were acquired and recorded with a 32 channel polygraph (LabSystem DUO Bard, C.R.Bard Inc., Worcester, MA, USA) at baseline, before and after each delivered shock.

In each patient at least 30 s of AF was used to determine the average A–A interval during AF (AF cycle length), using electrograms acquired from antero-lateral right atrial wall.

Statistical analysis

The results are presented as the mean value ± standard deviation of the mean. The reproducibility of the ADF was assessed following the method proposed by Bland and Altman [16]. The variability was expressed as standard deviation of differences (SDD) between the repeat measurements and by calculation of a coefficient of variability (CV) defined as the SDD between pairs of measurements. This coefficient represents SDD in percentage relative to the mean of the sample.

Results

Study population

Ten patients with a history of frequent episodes of paroxysmal AF scheduled for electrophysiological study and 15 patients with persistent AF were enroled in the study. The demographic characteristics of study population were reported in Table 1.

Table 1

Clinical characteristics of the study populations

	Paroxysmal AF	Persistent AF
N. of patients	10	15
Age (years)	58 ± 10	59 ± 15
Sex (M/F)	4/6	9/6
Left atrium diameter (mm)	46 ± 5	45 ± 7
LV ejection fraction (%)	50 ± 5	49 ± 8
AF duration (weeks)	–	10.3 ± 6
No heart disease	6	4
Arterial hypertension (n)	4	7
Ischaemic heart disease (n)	5	1
Valvular heart disease (n)	0	2

Mean ± SD; LV, left ventricle; AF, atrial fibrillation; M, male; F, female, N., number.

AF electrophysiological characteristics and atrial defibrillation threshold

The procedure was successful in all patients in restoring sinus rhythm both at baseline and after AF reinduction. No patient required extra sedation or an anaesthetic.

No significant differences in AF cycle length, before successful shocks, were detected in the two study populations.

In paroxysmal AF patients the mean value of basal AF cycle length was 194 ± 25 ms compared with 202 ± 15 ms of reinduced AF (p=NS); in persistent AF the mean value of basal AF cycle length was 175 ± 21 ms compared with 181 ± 20 ms of reinduced AF (p=NS).

No significant difference was observed between the two successful cardioversion attempts in terms of mean value of intensity, energy and impedance levels in both paroxysmal and persistent (Table 2) AF populations.

Table 2

Voltage, shock impedance and delivered energy of successful shocks for baseline (Bas) and reinduced (Reind) atrial fibrillation (AF)

	Voltage (V)		Energy (J)		Impedance (Ω)
	Bas	Reind	Bas	Reind	Bas	Reind
Paroxysmal AF patients
1	163	117	2.72	1.85	61	64
2	104	118	1.59	1.75	67	69
3	322	254	5.95	4.73	55	54
4	202	200	3.49	3.66	59	55
5	262	262	3.95	4.05	67	65
6	202	117	3.41	1.97	60	61
7	309	255	7.17	4.49	43	57
8	167	203	2.6	3.32	65	62
9	261	262	3.95	4	67	66
10	67	167	0.88	2.6	78	65
Mean	205.9	195.5	3.571	3.242	62.2	61.8
SD	83.9	62.3	1.8	1.1	9.2	5.0
Persistent AF patients
1	300	240	5.5	3.8	55	56
2	360	360	9.2	9.4	55	54
3	300	240	6	3.5	60	64
4	480	540	16.6	18.1	71	68
5	300	300	5.4	5.8	67	64
6	360	420	7.8	10.9	68	68
7	360	360	9.1	8.9	55	57
8	480	360	18.6	10.3	47	47
9	240	300	3.3	4.3	68	68
10	540	420	20	11.4	59	63
11	540	360	20.4	8.5	58	61
12	480	420	18.1	14	48	48
13	420	360	11.4	9.2	61	55
14	360	420	8.5	11.8	61	61
15	300	220	18.6	10.3	47	47
Mean	388.0	354.6	11.9	9.3	58.6	58.7
SD	95.8	85.3	6.1	3.9	7.7	7.4

No significant differences were found comparing shock voltage (paroxysmal AF p = 0.56, persistent AF 0.109), delivered energy (paroxysmal AF p = 0.054, persistent AF 0.855) and impedance (paroxysmal AF p = 0.92, persistent AF 0.425) in baseline and reinduced AF.

As expected the ADFT of patients with paroxysmal AF was significantly lower compared with patients having persistent AF (p<0.05).

The evaluation of ADFT reproducibility showed that the value of ADFT was reproducible in paroxysmal AF population; in persistent AF group only the impedance was reproducible, while a significant variability in the energy requirements was detected (Table 3, Fig. 1).

Figure 1

Reproducibility of atrial defibrillation threshold in patients with persistent and paroxysmal AF.

Table 3

coefficient of variability of atrial defibrillation threshold in paroxysmal and persistent AF

	SD differences	Coefficient of variability (%)
Paroxysmal AF (J)	1.2	5
Paroxysmal AF (V)	55	9.6
Paroxysmal AF (impedance (Ω))	6.7	0.6
Persistent AF (J)	4.7	44.3
Persistent AF (V)	75.4	20.3
Persistent AF (impedance (Ω))	2.6	4.5

Discussion

Low-energy intra-atrial defibrillation is a new therapeutic option for restoring sinus rhythm in patients with AF. The success rate is quite high when right atrium–coronary sinus or right atrium–pulmonary artery electrode configurations are used, although the former is associated with a slightly lower ADFT. One of the most important clinical problems of this therapeutic method concerns its tolerability, especially with respect to a possible implant of an automatic device. The subjective discomfort is mainly related to the level of energy required for successful cardioversion [3]; several authors have reported the possibility of reducing the ADFT by modifying the lead configuration, the shock waveform, the degree of synchronisation of the arrhythmia or using antiarrhythmic drugs [6–,11].

However, only few and contradictory data are available about the reproducibility of the ADFT in the single patient in different AF settings, making evaluation of the different technical or therapeutic reported experiences difficult.

Lau and colleagues [17] evaluating the effect of iv d-sotalol pre-treatment on defibrillatory threshold tested its reproducibility in 17 patients (11 chronic and 6 acute AF) by reinducing AF by rapid atrial pacing after a successful shock. The authors reported that the coefficient of correlation between the first and the second shock was 0.87 (p < 0.001); however this is a poor index for reproducibility assessment.

Boriani et al. [18] performed a transvenous low-energy atrial cardioversion in a series of fully conscious patients (30 patients with persistent AF and 5 patients with paroxysmal AF): comparing ADFTs using Student's t test they did not observe any statistical difference between consecutive successful shocks. More recently they reported [19] that in patients with chronic persistent AF who undergo internal atrial cardioversion, reinduction of sustained AF after effective cardioversion results in a type of AF with comparable atrial defibrillation threshold and atrial cycle length.

On the contrary Santini and coworkers [20] evaluating the change in the ADFT at rest and during exercise in chronic and reinduced AF observed a significantly lower ADFT in the second shock than the first (6.32 ± 2.09 J vs 7.40 ± 0.87 J, p<0.05).

Recently, Ammer and colleagues [21] assessed the ADFT in patients with recurrent AF using repeated internal cardioversion, showing a significant reduction in defibrillation energy requirements for internal cardioversion in the second AF episode, 6.67 ± 3.09 J (first) vs 3.83 ± 2.62 J (second); p = 0.003, related to a shorter arrhythmia duration.

Probably, the reason for these conflicting results is related to the heterogeneity in characteristics of patient populations (i.e. paroxysmal vs persistent AF), study protocols and statistical analysis.

In our study, we specifically evaluated the reproducibility of ADFT in two different AF populations using the Bland and Altman test, currently considered the gold-standard method in the evaluation of the variability of physiological parameters [22].

In patients with clinical paroxysmal AF, episodes during electrophysiological study were induced and showed a relatively more organized arrhythmia: in this setting we observed good reproducibility of energy requirements for successful cardioversion.

On the contrary in patients with permanent persistent AF, with less organized atrial activity, we demonstrated a non-significant reduction in ADFT with the second internal cardioversion (11.9 vs 9.3 J), associated with a low reproducibility (CV 44.3%) of effective energy shock values.

The reason for this difference could reflect a different degree of electrical remodelling in the two AF populations: in persistent AF a short period of sinus rhythm is probably not enough to reverse the electrophysiological changes particularly associated with a long lasting arrhythmia [23,,24]. The more evident heterogeneity of atrial activation in patients with persistent AF may be expected to show higher ADFTs compared with induced, short-lasting AF, and low ADFT reproducibility.

Limitation of the study

Recently Tse et al. [25] reported that successful defibrillation of AF is similar to that in the ventricle and may be a matter of chance. In their study the use of single shock conversion data to assess ADFT can predict 60% and 45% chance of success for cardioversion in patients with acute and persistent AF, respectively. This observation may account for the lower reproducibility of ADFT in patients with persistent AF observed in our study.

Conclusion

In conclusion our results suggest that it may be difficult to compare the ADFT values between repeated internal cardioversions in a patient with persistent AF, when the efficacy of different interventions employed to reduce ADFT are evaluated. Before comparing different technical or pharmacological approaches to ADFT reduction, it is necessary to consider the clinical characteristic of the arrhythmia (persistent vs paroxysmal) and that there is at least a 20% spontaneous variability between consecutive ADFT in persistent AF patients.