Paravolidakis et alProspective Randomized Tri al

Prospective Randomized Trial of External VersusInternal Transcatheter Cardioversion in Patients withChronic Atrial Fibrillation

K.E. Paravolidakis, T.M. Kolettis,G.N. Theodorakis, I.A. Paraskevaidis,T.S. Apostolou, and D.Th. KremastinosOnassis Cardiac Surgery Centre, 2nd Department of Cardiologyand General Hospital of Nikea, 2nd Department of Cardiology,Athens, Greece

Abstract. To evaluate the safety and long-term ef~cacy of

internal transcatheter cardioversion, forty patients with

chronic, lone atrial ~brillation were studied. The patients

were randomised to internal transcatheter cardioversion or

to conventional external cardioversion. In cases where the

procedure was unsuccessful, cross-over to the alternate

method was performed. Oral anticoagulation therapy was

started three weeks prior to the procedure and was main-

tained for another three weeks following successful

cardioversion.

Sinus rhythm was restored in 16/18 patients (88%) in the

internal cardioversion group, versus 9/22 patients (40%) in

the external cardioversion group (p , 0.01). In addition,

8/13 (61%) patients who were crossed-over to internal

cardioversion were successfully cardioverted to sinus

rhythm. In contrast, both patients who were crossed-over to

external cardioversion remained in atrial ~brillation.

During a mean follow-up period of 23 months, 13

(39.3%) patients maintained sinus rhythm. Using the inten-

tion to treat principle, the recurrence rate was not statisti-

cally different between the two methods.

It is concluded that internal cardioversion is more effec-

tive in acutely restoring sinus rhythm compared to external

cardioversion. However, both methods have similar long-

term recurrence rates.

Key Words. atrial ~brillation, cardioversion, external, in-

ternal

Chronic atrial ~brillation is a common arrhythmia af-fecting 2 to 4% of the population over 60 years ofage[1,2]. It can occur without any known underlyingdisease, the so-called “lone” atrial ~brillation [4]. Thisarrhythmia is associated with hemodynamic disadvan-tages as well as with an increased risk of thromboem-bolic events [3].

Although direct current cardioversion has been ex-tensively used for restoration of sinus rhythm, eitheras ~rst-line treatment or following failed pharma-cological conversion, failure rates up to 50% have beenreported [5–8]. In order to increase success rates, in-ternal catheter cardioversion has been described [9].

However, the energy used in this series was high, atthe range of 200 to 300 Joules. Therefore, the purposeof the present study was to evaluate the initial andlong-term results of internal catheter cardioversion,using lower de~brillation energy.

Patients and Methods

Patients with chronic, lone atrial ~brillation were stud-ied. The minimum arrhythmia duration required forstudy entry was 1 month, with a maximum of 24months. Patients without structural heart disease,documented by patient history, physical examination,12-lead electrocardiogram, maximal exercise stresstesting and two-dimensional echocardiogram wereconsidered eligible for the study. Detailed informedconsent was obtained from each patient.

Transesophageal echocardiography was performedin order to exclude the presence of thrombi in the leftatrium. Anticoagulant therapy with acenocoumarolwas started three weeks prior to the procedure aimingat an international normalised ratio (INR) of 2 to 4 andwas discontinued three weeks after successfulcardioversion. The patients were randomised to inter-nal transcatheter cardioversion or to conventional ex-ternal cardioversion. The randomisation was per-formed by toss of a coin. In cases where the procedurewas unsuccessful, cross-over to the alternate methodwas performed.

Procedure description

A 6 French USCI quadripolar electrode was intro-duced percutaneously through the femoral vein and

Address correspondence to: Konstantinos Paravolidakis, M.D.,Onassis Cardiac Surgery Center, 2nd Dept. of Cardiology, 356Sygrou Ave., 176 74 Athens, Greece. E-mail: elbee@ath.forth-net.gr

Received 13 November 1997; Accepted 25 February 1998

249

Journal of Interventional Cardiac Electrophysiology 1998;2:249–253

© Kluwer Academic Publishers. Boston. Printed in U.S.A.

positioned under _uoroscopic control in the tricuspidarea in order to record satisfactorily the His bundleactivity through the distal pole of the electrode. Thefourth pole was connected to the negative pole of ade~brillator while the positive pole was connected to a16-cm cutaneous patch (R2 Corp., Niles, Illinois, USA),placed on the patient’s back. Two surface ECG leadsand one endocardiac electrogram from the tricuspidarea were continuously recorded on a multichannel di-rect-writing recorder (Nihon-Koden Co, model RMC1100, Tokyo, Japan). A temporary pacing electrode wasplaced at the right ventricular apex for back-up emer-gency pacing. The procedure has been described pre-viously [9]. A light general anaesthesia with intrave-nous propofol was used.

For internal cardioversion, the protocol included astepwise energy increase of 50 Joules starting at 50Joules with a maximum energy delivery of 150 Joulesor until sinus rhythm appearance. If the 150 Joulesshock failed, the procedure was considered unsuccess-ful.

For external cardioversion, two 16-cm cutaneouspatches were placed in the anteroposterior con~gura-tion, as previously described [5]. The protocol includedthe delivery of one synchronised 200 Joules direct cur-rent shock and up to two 360 Joules shocks. If sinusrhythm was not restored by the randomly allocatedcardioversion mode, the patient was crossed over toalternate mode. The patients were monitored in theintensive coronary care unit for 24 hours after the pro-cedure. Successful cardioversion was de~ned as thepresence of sinus rhythm at the time of the dischargefrom the hospital.

Follow-up

All patients were followed up the ~rst month after theprocedure, and at three, six and twelve months there-after. The follow up included clinical examination anddetailed echocardiograpic study. The presence of sinusrhythm or the event of atrial ~brillation recurrencewas recorded by a 12-lead ECG.

Statistical Analysis

All data were expressed as mean 6SD. The Student’spaired and unpaired t-rest were used for statisticalcomparisons. The recurrence rate of atrial ~brillationafter cardioversion was estimated using the Kaplan-Meier method. A p value ,0.05 was considered statis-tically signi~cant. Discriminant analysis was used toidentify variables that might have affected long-termsinus rhythm maintenance.

Results

Forty-two patients were initially included in the study.One patient was excluded because of thrombi in the leftatrial appendage revealed by transesophageal echo-cardiogram and another patient converted to sinus

rhythm prior to the procedure. Thus, forty patients (23men, mean age 57 6 8 years) were enrolled in thestudy. Mean atrial ~brillation duration was 9 6 7months (range 1–24 months).

Twenty patients had a history of unsuccessful at-tempt of pharmacological conversion with quinidine amean of 3 6 2 months prior to study entry. Threeweeks prior to the procedure 19 patients were placedon oral amiodarone therapy 200 mg daily.

Eighteen patients were randomly assigned to inter-nal and 22 to external cardioversion. There were nosigni~cant differences between the two groups in age,atrial ~brillation duration, left atrial or left ventriculardimensions prior unsuccessful pharmacologic conver-sion attempts, or amiodarone treatment. (Table 1).

Immediate results

Sinus rhythm was restored in 16 of 18 patients (88%)who underwent internal cardioversion, compared with9 of 22 patients (40%) who underwent externalcardioversion (p , 0.01). Thirteen patients who failedexternal cardioversion, were crossed over to internalcardioversion; eight out of these 13 patients (61%),were successfully cardioverted with an internal shock.Similarly, two patients who failed internal cardiover-sion were crossed over to external cardioversion. Boththese patients failed to convert to sinus rhythm.

Energy requirements. In the internal cardioversiongroup, 2 patients converted to sinus rhythm with the~rst shock (50 Joules), 8 patients with the second shock(100 Joules), and 6 patients with the third shock (150Joules). In the external cardioversion group, 4 patientsconverted with the ~rst shock (200 Joules), 3 patientswith the second shock (360 Joules) and 2 patients withthe third shock (360 Joules). The mean energy used insuccessful shocks was 112 6 34 in the internalcardioversion group and 271 6 84 in the externalcardioversion group (p , 0.05). In the internal afterfailed external cardioversion group, 2 patients con-

Table 1. Comparison of clinical characteristics in patientstreated with internal or external de~brillation

Internal External(18 patients) (22 patients) p value

AGE (years) 55 6 8 59 6 8 NSAF DUR (months) 8 6 7 9 6 8 NSLA (mm) 43 6 6 44 6 5 NSESD (mm) 34 6 5 33 6 5 NSEDD (mm) 50 6 6 50 6 5 NSFS % 31 6 8 34 6 7 NSPrior pharmacologic 11 9 NS

attemptsAmiodarone therapy 9 10 NS

LA: Left Atrium; ESD: End Systolic Diameter; EDD: End DiastolicDiameter; FS: Fractional Shortening of Left Ventricle

250 Paravolidakis et al

verted to sinus rhythm with the second shock (100Joules) and 6 patients with the third shock (150 Joules).

Transient short-lasting episodes of any degree ofA-V block occurred in 8 patients of the internalcardioversion group. These episodes lasted from 5 sec-onds to 10 minutes and included ~rst degree A-V blockin 4 patients, second degree A-V in 1 patient and thirddegree A-V block in 3 patients. However, no patientrequired permanent pacemaker implantation. In con-trast, no patient in the external cardioversion grouprequired temporary ventricular pacing. No other pro-cedure-related complications were observed.

Long-term results

Figure 1 shows the Kaplan-Meier analysis of long-termsinus rhythm maintenance. The atrial ~brillation re-currence rates were not statistically different betweenthe two groups. In contrast, patients who cardiovertedwith internal after unsuccessful external cardiover-sion, had signi~cantly higher recurrence rates com-pared to either group. During the follow-up period, nothromboembolic event was recorded.

Variables associated with outcome

Mean left atrial size before the procedure in the groupof patients remaining in sinus rhythm, irrespectively ofthe cardioversion mode, was 41 6 3 mm, while for therelapsed patients was 45 6 5 mm (p , 0.01).

Mean fractional shortening of patients remaining insinus rhythm was 3565%, which was signi~cantly

higher compared to a mean of 31 6 8% in patients inwhom atrial ~brillation recurred (p , 0.05). Discrimi-nant analysis of clinical parameters which might haveaffected sinus rhythm maintenance are shown in Table2. The duration of atrial ~brillation, the total energydelivered during the procedure, and the mode ofcardioversion did not affect sinus rhythm maintenance.In contrast, smaller left atrial size, higher fractionalshortening and amiodarone therapy were independentpredictors of chronic sinus rhythm maintenance.

Discussion

Since its ~rst description by Lown in 1962 [1] direct-current cardioversion has been extensively used forsinus rhythm restoration in patients with chronic atrial~brillation [5,6]. However, this method is associatedwith a signi~cant failure rate, varying from 20 to 50%in previous reports [5–7,10]

Fig. 1. Kaplan Meier life table analysis showing the recurrence rates in the 3 patient-groups.

Table 2. Discriminant analysis of clinical parameters whichmay affect the sinus rhythm maintenance

Left atrial size 0.007Left ventricular fractional shortening before 0.01

de~brillationAtrial fibrillation duration NSMethod of de~brillation (internal vs external) NSTotal energy NS

Prospective Randomized Trial 251

Our results indicate that internal catheter cardio-version is signi~cantly more effective in acutely restor-ing sinus rhythm in patients with chronic atrial ~brilla-tion compared the conventional external cardioversion.One could argue that an unusually high (60%) failurerate was observed in our study, somewhat higher toprevious reports [5–7,10]. One possible reason for fail-ure of external cardioversion is the high impedancebetween the electrodes, due to larger body surfacearea or due to patch-electrode placement [8]. Nonethe-less, although body surface area was not recorded inour study population, the patch con~guration used inour study was used in most previous reports [6,7].

Our results are in accordance with those of Levy etal. [9,11]. They prospectively compared internal cathe-ter to conventional external cardioversion in 112 pa-tients with chronic atrial ~brillation and reportedsigni~cantly higher success rates with the former tech-nique. However, high energy shocks, at the range of200 to 300 Joules were used in these studies [9,11]. Thepresent study evaluated the feasibility of internalcatheter cardioversion using lesser amounts of energy,at the range of 50 to 150 Joules. The initial successrates for internal cardioversion in our study (88%)were comparable to those of Levy et al. [11] (91%).Thus, our data suggest that lower energy should beused, since the higher energy delivery does not resultin a substantially better outcome.

In accordance to previous reports [7,14–18] weshowed that left atrial size and left ventricular systolicfunction are important factors in_uencing long-termsinus rhythm maintenance. The mode of cardioversiondid not appear to affect long-term sinus rhythm main-tenance. This ~nding con~rms that the randomisationperformed in our study resulted into two comparablegroups.

We found that fractional shortening before cardio-version was a predictor of long-term effectiveness, al-though no structural heart disease was identi~ed inany patient. This may re_ect the duration of atrial~brillation, since there is evidence suggesting thatchronic, long-lasting atrial ~brillation may lead to someform of slowly progressive intrinsic cardiomyopathy[19,20].

We believe that three weeks of anticoagulation post-cardioversion is suf~cient, since there is evidence thatby this period the mechanical function of the atria isrestored [6,19]. We report a high incidence of high de-gree atrioventricular block after internal cardiover-sion. However, it is reassuring that no patient requireda permanent pacemaker. Mansourati et al reportedtransient atrioventricular block in 11% of cases, but inaccordance with our ~ndings, no case of permanentpacing [21]. It appears that this is a transient “stun-ning” effect on the A-V node, which nevertheless, man-dates temporary ventricular pacing.

Compared to lone, atrial ~brillation associated withstructural heart disease (such as valvular heart diseaseor congestive heart failure) is associated with signi-

~cantly lower success rates with external cardiover-sion. Internal cardioversion is a promising techniqueand should be evaluated in this group of patients.

Conclusion

Internal cardioversion is a safe and effective methodfor restoring sinus rhythm in patients with chronicatrial ~brillation and is associated with higher successrates compared to external cardioversion. TransientA-V block occurs frequently, however, permanentpacemaker implantation was not encountered in ourseries. Energies up to 150 Joules appear suf~cient forsatisfactory results. However, recurrence rates aresimilar in both methods. Failed external cardioversion,increased left atrial size and relatively compromisedleft ventricular function may be associated with higheratrial ~brillation recurrence rates.

We are in indebted to Miss Eleni Binou for her meticulous secre-tarial assistance.

References

1. Lown BN. Electrical reversion of cardiac arrhythmias. BrHeart J 1967;29;149–489.

2. Alpert J, Petersen P, Godtfredsen J. Atrial ~brillation natu-ral history complications and management. Ann Rev Med1988;39(4):41–52.

3. Cairns J, Connolly S. Nonrheumatic atrial ~brillation: Riskof stroke and role of antithrombotic therapy. Circulation1991;84:469–481.

4. Kopencky S, Gersh B, McCoon M, et al. The natural historyof lone atrial ~brillation. N Engl J Med 1987;317:669–674.

5. Morris J, Peter R, McIntosh H. Electrical conversion ofatrial ~brillation. Ann Int Med 1966;65:2216–231.

6. Falk R, Podrid PH. Electrical cardioversion of atrial ~brilla-tion. In: Falk R, Podrid PH, eds. Atrial ~brillation. Mecha-nisms and Management. New York: Raven Press Ltd.,1992.

7. Lundstrom T, Ryden L. Chronic atrial ~brillation. Long-term results of direct current conversion. Acta Med Scand1988;223;53–59.

8. Ewy G. Optimal technique of electrical cardioversion ofatrial ~brillation. Circulation 1992;86(5):1645–1647.

9. Levy S, Lacombe P, Cointe R. High energy transcathetercardioversion of chronic atrial ~brillation. J Am CollCardiol 1988;12:514–518.

10. Szekely P, Sideris D, Batson G. Maintenance of sinus rhythmafter atrial de~brillation. Br Heart J 1970;32:741–746.

11. Levy S, Lauribe PH, Dolla E, et al. A randomized compari-son of external and internal cardioversion of chronic atrial~brillation. Circulation 1992;86:1415–1420.

12. Van Gelder T, Grijns H, Van Gilst W, et al. Prediction ofuneventful cardioversion and maintenance of sinus rhythmfrom direct current electrical cardioversion of chronic atrial~brillation and _utter. Am J Cardiol 1991;68:41–46.

13. Sutton M, Kingma J, Koomen E, et al. Recurrence of parox-ysmal atrial ~brillation or _utter after successful cardiover-sion in patients with normal LV function. Am J Cardiol1993;71:710–713.

252 Paravolidakis et al

14. Waris E, Kreus K, Salokannel J. Factors in_uencing persist-ence of sinus rhythm after DC shock treatment of atrial~brillation. Acta Med Sand 1971;189:161–166.

15. Henry W, Morganroth J, Pealman A, et al. Relation betweenechocardiographically determined left atrial size and atrial~brillation. Circulation 1976;53(2):273–279.

16. Brodsky M, Allen B, Capparelli E, et al. Factors determin-ing maintenance of sinus rhythm atrial ~brillation chronicatrial ~brillation with left atrial dilatation. Am J Cardiol1989;63:1065–1068.

17. Mitchell J, Shapiro W. Atrial function and the hemodynamicconsequences of atrial ~brillation in man. Am J Cardiol1969;23:55–567.

18. Hoglund C, Rosluhamer G. Echocardiographic left atrialdimension as a predictor of maintaining sinus rhythm after

conversion of atrial ~brillation. Acta Med Scand 1985;217:411–415.

19. Van Gelder J, Crijns H, Blanksma P, et al. Time course ofhemodynamic changes and improvement of exercise toler-ance after cardioversion of chronic atrial ~brillation unasso-ciated with cardiac valve disease. Am J Cardiol 1993;72:560–566.

20. Gosseling A, Crijns H, Hamer H, et al. Changes in left andright atrial size after cardioversion of atrial ~brillation: Roleof mitral valve disease. J Am Coll Cardiol 1993;22:1666–1672.

21. Mansourati J, Larlet JM, Salaun G, Maheu B, Blanc JJ.Safety of high energy internal cardioversion in atrial ~bril-lation. PACE 1997;20(I):1919–1923.

Prospective Randomized Trial 253