Low-Energy Cardioversion of Spontaneous Atrial Fibrillation
Immediate and Long-term Results
Background Recent studies have suggested that induced atrial fibrillation (AF) could be successfully terminated by using a two-catheter electrode system and low energy (<400 V). This study evaluated the efficacy and safety of low-energy cardioversion in spontaneous chronic and paroxysmal AF.
Methods and Results Forty-two consecutive patients with spontaneous AF underwent low-energy electrical cardioversion. AF was chronic (≥1 month) with a mean duration of 9±19 months in 28 patients (group I) or paroxysmal with a history of recurrent attacks and a mean duration of the present episode of 7±16 days in 14 patients (group II). An underlying heart disease was present in 28 patients. A 3/3-ms biphasic shock was delivered between catheters positioned in the right atrium and the coronary sinus in 32 patients. In 10 patients, the left pulmonary artery branch was used. The catheters were connected to a custom external defibrillator. The shocks were synchronized to the R wave. Following a test shock of 60 V, the energy was increased in 40-V steps until a maximum of 400 V or restoration of sinus rhythm. Sinus rhythm was restored in 22 of the 28 patients (78%) of group I by using a mean leading-edge voltage of 297±57 V (mean energy, 3.3±1.3 J) and in 11 of 14 patients (78%) of group II by using a mean leading-edge voltage of 223±41 V (mean energy, 1.8±0.7 J). The energy required for terminating chronic AF was significantly (P<.001) higher than that required for terminating paroxysmal AF. Among the other variables studied, the duration of AF significantly affected the successful voltage. Ventricular proarrhythmia occurred in 1 patient with atrial flutter due to an unsynchronized shock. Of the 22 patients of group I in whom sinus rhythm was restored, 14 (63%) remained in sinus rhythm with a mean follow-up of 9±3 months. Pain level showed a good correlation with increasing voltage. However, a marked interindividual variation was noted.
Conclusions Atrial defibrillation using low energy between two intracardiac catheters with an electrical field between the right and left atria and the protocol used is feasible in patients with persistent spontaneous AF. The technique is safe provided synchronization to the R wave is achieved. A low recurrence rate of AF was seen in patients in whom sinus rhythm was restored.
Atrial fibrillation is a common arrhythmia that has gained a renewed interest in recent years. A number of patients complain of disabling symptoms, and the arrhythmia is associated with hemodynamic impairment and a decreased life expectancy.1 Antiarrhythmic therapy remains the treatment of choice and is efficient in maintaining SR and preventing recurrences in a portion of patients. However, there is a group of patients in whom the use of antiarrhythmic agents is unsuccessful or associated with intolerable side effects. Moreover, recent reports2 3 have raised concern regarding the safety of antiarrhythmic agents, as their use may result in an increased mortality rate compared with placebo-treated subjects or with similar groups of patients not on antiarrhythmic therapy. Therefore, nonpharmacological therapy for AF has become appealing. Such options include catheter ablation or modification of the atrioventricular node with or without a permanent pacemaker and an IAD.4 One of the prerequisites for the use of an atrial defibrillator is the ability to safely and efficiently terminate recurrences of AF. Recent technological advances have allowed the development of defibrillation leads with a large surface area and of a prototype of an IAD capable of delivering R wave–synchronized shocks following a programmed minimum RR interval, with a voltage ranging from 0 to 400 V and providing a direct readout of energy and catheter system impedance. Furthermore, the considerable advances made in the implantable cardioverter defibrillator for ventricular tachyarrhythmias have also made possible the development of an IAD.
Cooper et al5 report that AF may be cardioverted in an acute model of AF in sheep by using low-energy shocks and that the best electrode configuration is the use of two catheters, one in the RA and the other in the CS. Reports6 7 in a limited number of patients have shown that cardioversion of short-lived AF is possible in humans by using low-energy shocks. Alt et al8 report the termination of chronic AF in 10 of 13 patients, and Murgatroyd et al9 have shown the feasibility and safety of low-energy shock in terminating induced AF in 19 patients. The purpose of the present study was to evaluate the immediate and long-term results of low-energy atrial defibrillation in 42 patients with spontaneous chronic AF or long-lasting episodes of paroxysmal AF.
This prospective study included 42 patients with spontaneous, non–self-terminating AF for whom cardioversion was indicated. In no patient was AF electrically or mechanically induced. Informed consent was obtained from all patients. AF was defined as chronic if it was present for ≥1 month; paroxysmal AF was defined as recurrent attacks of arrhythmia.9 Only patients with episodes of paroxysmal AF with a duration ≥24 hours were included. Whenever possible, antiarrhythmic therapy was discontinued for five half-lives. However, the presence of antiarrhythmic therapy was not an exclusion criterion. All patients with paroxysmal or chronic AF underwent a transesophageal echocardiogram to rule out the presence of intracardiac thrombus and received subcutaneous heparin until the coagulation time was twice or more that of control subjects. All patients underwent a workup that included a history and physical examination, 12-lead ECG, chest x-ray, 24-hour ambulatory monitoring, M-mode and bidimensional echocardiogram, laboratory tests (including creatinine, serum potassium, and red blood count), and thyroid function evaluation. This study protocol received approval from the ethics committee of the University of Marseilles.
The 42 patients included 26 men and 16 women aged 42 to 85 (mean, 67±10) years; weights ranged from 54 to 107 (mean, 74±14) kg. Mean left atrial diameter was 44±7 (range, 26 to 62) mm. Underlying heart disease was present in 28 patients, including valvular heart disease (9), hypertension (13), dilated cardiomyopathy (2), hypertrophic cardiomyopathy (1), atrial septal defect (2, one of whom had had surgery), and hemochromatosis (1). The patients were subdivided into two groups: chronic AF (group I) and paroxysmal AF (group II). The patients’ clinical characteristics are shown in Table 1⇓. There were no significant differences between the two groups except for the duration of the current episode of AF.
Defibrillator shocks were delivered through two 6F catheters (Elecath), each with a surface area of 2.83 cm2. One of these was introduced through the left subclavian vein and positioned in the CS. The other catheter was introduced through the right femoral vein and positioned in the RA such that the electrodes had contact with the anterolateral wall. The RA catheter served as the cathode and the left atrial catheter as the anode. When the CS could not be successfully catheterized, the anodal catheter was positioned in the left branch of the pulmonary artery. An additional catheter was positioned in the right ventricular apex in order to obtain satisfactory R wave synchronization and to provide postshock ventricular pacing. Catheter locations are shown in Fig 1⇓. The electrodes used for defibrillation were connected to a custom external atrial defibrillator (XAD, In Control Inc) capable of delivering a 3/3 biphasic shock waveform with a leading-edge voltage that could be programmed between 10 and 400 V. The other programmable parameter was the RR interval preceding the shock, which was set to be ≥500 ms for safety reasons.10 After each shock, the defibrillator displayed the measured voltage, energy, and impedance. Cardioversion was performed under local anesthesia (lidocaine, 1%) in the fasting state without sedation and after informed consent. The patient was informed that sedation would be provided at any time should shock delivery become intolerable. At each voltage level, the patient was asked to rate the pain level according to the following scale: (1) could not feel the shock, (2) shock felt but described as not painful, (3) mildly painful shock, (4) moderately painful shock, or (5) severely painful shock. When needed, sedation was provided with midazolam at an initial dose of 2.5 mg and a total dose of 0.10 mg/kg body wt. The defibrillation protocol included a test shock of 20 V and a first shock of 60 V; the energy was then increased in 40-V steps to a maximum of 400 V. The end point was restoration of SR or completion of the defibrillation protocol. After restoration of SR, oral anticoagulation was instituted and continued for a minimum of 1 month. At the time of the study, 24 patients were admitted on antiarrhythmic therapy aimed at preventing recurrences, including amioda-rone in 21 patients and a class I agent (disopyramide or hydroquinidine) in 2 patients. One patient was on a β-blocking agent. Eighteen patients were on no antiarrhythmic treatment. Following cardioversion, all patients were put on class I or III prophylactic antiarrhythmic therapy. Patients were followed up, and a physical examination and 12-lead ECG were obtained at 1, 3, 6, 9, and 12 months or earlier if symptomatic AF recurred.
Statistical analysis used Student’s unpaired t test, and multivariate analysis used the Pearson correlation.
At the time of cardioversion, 41 patients had AF and 1 had atrial flutter. This patient was reported in this series for reasons described in detail below. Of the 42 patients, atrial arrhythmia was terminated with low-energy shocks in 33 (78%). SR was obtained immediately following the shock in 24 patients (Fig 2⇓) and following a short (<30 s) run of persisting AF in 9 patients. In the remaining 9 patients, the complete defibrillation protocol failed to terminate AF (8 patients) or atrial flutter (1 patient). A successful termination of AF following a low-energy shock is shown in Fig 2⇓. The impedance, RR interval preceding shocks, leading-edge shock voltage, and energy of the total number of shocks are presented in Table 2⇓. The data regarding leading-edge voltage and energy of the successful shocks in the two groups are shown in Table 3⇓. Although the total success rate was similar in the two groups, the leading-edge voltage and the energy required were significantly higher in the chronic AF group (P<.001). Because the CS could not be catheterized in 10 patients, the catheter was positioned in the left branch of the pulmonary artery. The leading-edge voltage and the energy required to successfully terminate AF were significantly (P<.02) higher when the anodal electrode catheter was in the pulmonary artery rather than in the CS (Table 4⇓). Twenty of the 32 patients (62.5%) in the CS group had chronic AF as opposed to 8 of the 10 patients (80%) in the pulmonary artery group.
A total of 334 shocks were delivered to the 42 patients. No complications occurred following R wave–synchronized shocks. However, in one 49-year-old woman with mitral valve disease who had a history of paroxysmal AF and flutter, ventricular tachycardia that rapidly degenerated into ventricular fibrillation was induced following a nonsynchronized shock. At the time of electrical cardioversion she was on amiodarone (200 mg/d) and presented with paroxysmal atrial flutter in the hemodynamic laboratory. Atrial pacing failed to terminate atrial flutter or convert it to AF. This patient was excluded from the protocol but was reported in order to provide complete information regarding the technique. Low-energy cardioversion was attempted to obtain an adequate hemodynamic evaluation of the valvular disease in SR. Synchronization on surface ECG was faulty, and shock occurred on the T wave (Fig 3⇓). As the RR was short (240 ms), the programmed RR was reduced in order to trigger the external device. The shock induced ventricular fibrillation that was immediately and successfully terminated by a 300-J external shock, and the patient did well.
Cardiac enzymes including creatine phosphokinase and its myocardium-specific isoenzymes CK-MB measured before, immediately after, and 2 and 6 hours following the defibrillation protocol did not show any significant change (45±36, 46±31, 46±34, and 47±32 IU, respectively).
Data regarding shock-related pain are available for 30 patients who agreed to undergo the protocol without sedation. The test shock (20 V) was well tolerated and rated a mean of 1.6±3.5 on the subjective scale. Only 1 patient described the test shock as moderately painful. At 60, 100, 140, 180, and 220 V, shock-related pain rated a mean of 2.1, 2.5, 3.1, 3.4, and 3.7, respectively (Table 5⇓). However, 2 patients at 180 V and 4 at 220 V described the shock as severely painful and were sedated. At 260 V and above, the shocks reached a score ranging from a mean of 3.9 to a mean of 4.5 at 340 V and 400 V, respectively. t test analysis showed a good correlation between the level of discomfort and increasing voltage. The difference was significant between 140- and 220-V shocks (P<.02) and 220- and 300-V shocks (P<.01). A marked interindividual variation was noted, eg, 5 patients found the 60-V shock mildly painful while 5 patients felt no discomfort with a 180-V shock. Fig 4⇓ shows the distribution of patients according to leading-edge voltages and shock-related pain levels.
Outcome of Patients and Follow-up
Of the 6 patients with chronic AF who failed low-energy cardioversion, SR was restored during the same session in 4 patients after external cardioversion in 1 and high-energy internal cardioversion in 3. The remaining 2 patients failed high-energy internal cardioversion. Of the 22 patients in group I (chronic AF) who were successfully cardioverted with low-energy shocks, AF recurred within the first 24 hours in 1 patient and in 2 patients on days 2 and 3, respectively. At hospital discharge, 19 of the 22 patients were in SR. With a mean follow-up of 9±3 (range, 7 to 13) months, 14 patients (63%) remained in SR. Recurrence of AF was observed in 5 patients, within 1 week in 3 and after 1 month and 13 months after discharge in the 2 remaining patients. All 19 patients discharged in SR were on class I or III antiarrhythmic drugs. Among group II (paroxysmal AF) patients who failed low-energy internal cardioversion, SR was restored in 2 by using external cardioversion (260 J) and in 1 by using high-energy internal cardioversion (300 J). Thus, using the three modalities for cardioversion, SR was restored in 40 of 42 (95%) patients.
Of the 24 patients who were on antiarrhythmic therapy at the time of the procedure, low-energy cardioversion resulted in 22 successful results (91%) compared with 11 in the 18 patients (61%) who were not on antiarrhythmic therapy. This was also the case for each of the two groups (chronic and paroxysmal) studied separately.
Variables and Successful Voltage
An attempt was made to correlate the voltage resulting in successful cardioversion with variables such as type of AF (paroxysmal versus chronic), age, gender, weight, duration of current episode of AF, left atrial size, tested defibrillation impedance, creatine phosphokinase and potassium levels, and presence or absence and type of underlying heart disease. Among these, only the type and duration of AF predicted the successful voltage level. The latter was lower when AF was paroxysmal (P<.001) as opposed to chronic and when the duration (in days) was shorter (P<.004). There was no significant difference between patients with chronic and paroxysmal AF regarding defibrillation shock impedance.
Restoring SR is a desirable end point in patients with chronic or paroxysmal non–self-terminating AF. It eliminates symptoms when AF is poorly tolerated, prevents left ventricular dysfunction, improves exercise capacity, and may reduce the risk of embolic complications. Electrical external cardioversion is the technique of choice for the restoration of SR. However, the energy required to terminate AF is high, and the technique is not always successful. We have reported11 a catheter technique that can terminate chronic AF in patients who failed external cardioversion. The technique uses the proximal electrode of a quadripolar catheter as the cathode and a backplate as the anode. The energy is 200- or 300-J monophasic shocks. In another study, we showed12 that internal cardioversion is superior to external cardioversion and that the recurrence rate of AF is not higher following internal cardioversion. Thus, the use of internal high-energy cardioversion after failure of external cardioversion results in a higher number of patients in SR at 1 year. Other reports have confirmed the high success rates of high-energy internal cardioversion.13 14 Unfortunately, both external and internal cardioversion techniques use high-energy shocks and require general anesthesia. In an elegant work, Cooper et al5 studied different electrode configurations and waveforms to convert AF in sheep. They found that the electrode configuration that resulted in the lowest defibrillation threshold comprised catheters in the RA appendage and the CS. They also found that biphasic waveform shocks resulted in lower atrial defibrillation thresholds than monophasic ones.
Preliminary Comparison With Previous Studies
Reports in humans6 9 have confirmed that low-energy cardioversion using the electrode configuration found to be most effective in sheep, ie, RA and CS,5 is effective. Most of the data appear in abstracts and concern particular subsets of patients. Keane et al6 successfully restored SR in 8 of 9 patients with chronic AF by using spring-coil electrodes with a mean energy of 6.7±2.2 J. Successful cardioversion of electrically induced AF was obtained by Johnson et al7 in 6 patients with a mean of 2.5±1.4 J by using biphasic shocks. Alt et al8 report successful atrial defibrillation in 10 of 13 patients with chronic AF with a mean energy of 3.7±1.7 J. Murgatroyd et al9 have recently reported on 19 patients with stable AF lasting >5 minutes, in all of whom SR was restored with a leading-edge voltage ranging from 140 to 340 V (0.7 to 4.4 J). AF was spontaneous in only 4 of these patients; in the remaining 15, AF was induced by catheter manipulation or electrical stimulation.9 Therefore, the data available on catheter-based atrial defibrillation using low energy in humans are limited. The present series represents the largest group of catheter-based, low-energy cardioversion of AF in humans. It included only patients with spontaneous non–self-terminating AF and confirms that low-energy atrial defibrillation may be achieved in 78% of patients. Although the success rate was similar in patients with chronic or paroxysmal AF, the leading-edge voltage required was significantly (P<.001) lower with paroxysmal (223±41 V) than chronic (297±57 V) AF. This represents an important new finding in light of consideration of the automatic atrial defibrillator as a therapeutic alternative.
No ventricular proarrhythmia was observed with synchronized shock. Ventricular fibrillation was induced with an unsynchronized shock in a flutter patient. Although this patient represents a protocol violation in several respects, ie, flutter as opposed to fibrillation and preceding RR interval programmed <500 ms, we included her in our series. Unlike the 41 other patients of our series in whom the technique was performed in the electrophysiological laboratory, this patient was in the hemodynamic laboratory, and a display of the synchronization signal to the QRS complex could not be obtained. Synchronization occurred on the T wave instead of the R wave, which explains the proarrhythmia.
The importance of R wave synchronization is emphasized and requires a lead in the right ventricle. Ayers et al10 have also studied in sheep the conditions under which ventricular fibrillation might be induced during synchronized shocks delivered during AF. They found that short (<300 ms) RR intervals preceding the shock were associated with a low but definite risk of ventricular fibrillation. Shocks delivered following RR intervals >300 ms were not associated with proarrhythmia. Based on these findings, the shocks were delivered after an RR interval of ≥500 ms. However, this represents a limitation in patients with AF and rapid ventricular response.
Pain related to shocks is an important issue both regarding patients’ tolerance of low-energy cardioversion with an external device or an automatic implantable defibrillator. The protocol used in this study showed a correlation between energy delivered and pain-related shocks. However, an important interindividual variation was noted. It is likely that the discomfort is increased when the defibrillation protocol includes shocks at increasing energy levels, which may generate anxiety in the patients. The termination of AF with energy levels <1 J was associated with little if any discomfort by Murgatroyd et al.9 Such a correlation could not be established in our study. Further studies with a specially adapted protocol will be needed to evaluate discomfort related to low-energy shocks.
Low-energy cardioversion of AF may have a number of clinical implications. Using temporary electrode catheters and an external device, this technique may be useful in terminating mechanically or electrically induced AF during electrophysiological studies, thereby allowing completion of the study. Internal cardioversion may also be useful in patients who fail external cardioversion. It may be of particular value in patients with severe obstructive lung disease, as the technique does not require general anesthesia. Low-energy cardioversion provides information that is necessary to study the feasibility of an IAD, which may be an interesting nonpharmacological therapy for AF. The present study showed that with the RA-CS configuration, patients with non–self-terminating paroxysmal AF require a mean energy of <2 J, and for selected patients of this group an atrial defibrillator may be an option.15
The energy required to terminate AF in a patient represents the minimum energy required to restore SR. Reproducibility was not tested because it would have required induction of AF after the first successful shock, and the latter could not be performed because the ethics committee of our university would not allow the intentional induction of AF in this study. An important issue for the use of an IAD is the pain possibly associated with the shock. The step-up defibrillation protocol generated anxiety in the nonsedated patients that impaired the appropriate evaluation of the shock-related discomfort. A specially designed protocol to address this issue may be required.
This study, using biphasic shocks and two catheter electrodes in the RA and the CS or the left pulmonary artery, showed that low-energy cardioversion is feasible in patients with chronic and persistent non–self-terminating AF of both long and short duration. The energy required to terminate paroxysmal AF was lower than that needed to terminate chronic AF. The technique is safe provided consistent synchronization to the R wave is achieved. This new technique of internal cardioversion may be useful in patients who fail external cardioversion and in a subset of patients with persistent AF in whom general anesthesia may be hazardous. It paves the way for an IAD. Further studies are required to test the safety and tolerability of the implanted device.
Selected Abbreviations and Acronyms
|IAD||=||implantable atrial defibrillator|
This work was supported in part by a grant from the Ministry of Health, Paris, France, and by a grant from In Control Inc, Redmond, Wash. We would like to thank Dr Jerry Griffin for his constructive criticisms, Steve Adler for his help with the statistical analysis, and Patricia Herpe for preparing this manuscript.
- Received December 28, 1995.
- Revision received January 13, 1997.
- Accepted January 15, 1997.
- Copyright © 1997 by American Heart Association
Coplen SE, Antman EM, Berlin JA, Hewitt P, Chalmers TC. Efficacy and safety of quinidine therapy for maintenance of sinus rhythm after cardioversion: a meta-analysis of randomized control trials. Circulation. 1990;82:1106-1116.
Lévy S, Camm AJ. An implantable atrial defibrillator: an impossible dream? Circulation. 1993;87:1769-1772.
Cooper RAS, Alferness CA, Smith WM, Ideker RA. Internal cardioversion of atrial fibrillation in sheep. Circulation. 1993;87:1673-1686.
Keane D, Sulke N, Cooke R, Jackson G, Sowton E. Endocardial cardioversion of atrial flutter and fibrillation. PACE Pacing Clin Electrophysiol. 1993;16:928. Abstract.
Johnson EE, Yarger MD, Wharton JM. Monophasic and biphasic waveforms for low-energy internal cardioversion of atrial fibrillation in humans. Circulation. 1993;88(suppl I):I-592. Abstract.
Ayers GM, Alferness CA, Ilina M, Wagner DO, Sirokman WA, Adams JM, Griffin JC. Ventricular proarrhythmic effects of ventricular cycle length and shock strength in a sheep model of transvenous atrial defibrillation. Circulation. 1994;89:413-422.
Lévy S, Lauribe P, Dolla E, Kou W, Kadish A, Calkins H, Paganelli F, Moyal C, Bremondy M, Schork A, Shyr Y, Das S, Shea M, Gupta N, Morady F. A randomized comparison of external and internal cardioversion of chronic atrial fibrillation. Circulation. 1992;86:1415-1420.
Chauvin M, Koenig A, Theolade R, Brechenmacher C. La place du choc électrique interne dans le traitement de la fibrillation auriculaire permanente: résultats préliminaires à propos de 28 cas. Arch Mal Coeur. 1991;84:377-382.