Transesophageal cardioversion

Transesophageal

cardioversion

With the use of a novel quadripoiar esophageal electrode system, we have attempted 131 transesophageal cardioversions in 105 patients: 109 episodes were atriai fibrillation, 16 episodes were atrial flutter, 2 episodes were supraventricular tachycardia, and 4 episodes were ventricular tachycardia. The mean predicted transesophageai impedance (+ SEM) of 52.6 ? 1.1 Q was significantly lower than the mean predicted transthoracic impedance (+ SEM) of 63.1 I 1.6 Q (n = 104; p < 0.01). Of the 66 patients who presented with atriai fibrillation as the initial arrhythmia, successful transesophageal cardioversion (maximal delivered transesophageal energy of 100 J in 64 patients and 200 J in 4 patients) was recorded in 70 (79.5%); transesophageai cardioversion required a mean delivered energy of 63.1 f 4.2 J and a mean peak current of 20.3 + 0.6 A. Transthoracic countershock (maximal delivered energy of 360 J) was attempted in 17 of 16 patients when the transesophageal approach had been unsuccessful; countershock was successful in 10 patients, which yielded an overall success rate of 92.0% (mean successful delivered energy [transesophageai and transthoracic] of 65.3 + 7.6 J). All episodes of atrial flutter, supraventricular tachycardia, and ventricular tachycardia were successfully terminated with the use of the esophagus. This esophageal electrode system permits low-energy countershock of atrial and ventricular tachyarrhythmias. (AM HEART J 1993;125:396.)

Pascal P. McKeown, MD,a Simon Croal, DPhil,b J. Desmond Allen, MD,C John Anderson, DPhil,b and A. A. Jennifer Adgey, MDa Belfast, Northern Ireland

The management of cardiac tachyarrhythmias constitutes an important aspect of cardiologic practice. When pharmacologic therapy is ineffective or inappropriate the use of nonpharmacologic measures, such as electric countershock, is often necessary. The technique of direct-current cardioversion was first described by Lawn’ in 1962 and has been widely employed in the ensuing years.lM3Transthoracic cardioversion is, however, a procedure of low efficiency, since only a small fraction of the delivered energy actually passesthrough the heart; the rest is dissipated in extracardiac tissues.4-6The close anatomic relationship between the esophagus and the heart has been recognized for many years. The technique of esophageal ECG was first described in 1906 by Cremer,7 and in more recent years transesophageal pac-

From %he Regional Medical Belfast; bthe Bio-engineering stown, Co Antrim, Northern Queen’s University of Belfast. Supported ciation. Received

in part

Cardiology Centre, Royal Victoria Hospital, Department, University of Ulster, JordanIreland; and cthe Physiology Department,

by the Northern

for publication

Reprint requests: Professor ology Centre, Royal Victoria Northern Ireland.

Apr.

Ireland

24, 1992;

Chest

accepted

A. A. J. Adgey, MD, Hospital, Grosvenor

Heart

and

Stroke

Asso-

Aug. 28, 1992. Regional Medical CardiRd., Belfast BT12 6BA

ing and echocardiography have become well established.7-g In spite of reports that date back to the 195Os,which record the placement of an esophageal electrode in dogs for low-voltage alternating-current defibrillation of the heart,lO the widespread application of this technique has been limited by the lack of a suitable stable electrode system. As a result of experimental work in animals,‘l we have designed a novel esophageal electrode system and have investigated its efficacy in the countershock of atria1 and ventricular tachyarrhythmias. METHODS

The study wasapproved by the ResearchEthical Committee of Queen’s University of Belfast, and all patients gave informed written consent.Patients were entered into the study between April 1988 and April 1990,and during that period 111 patients who required countershock of atria1 or ventricular tachyarrhythmias were referred for transesophagealcardioversion. Three of these patients were excluded from the study at the time of referral becauseof a history of esophagealdisease.Three other patients wereeither unableto swallowthe electrode or unable to tolerate the electrode after successfulplacement and were thus excluded from further analysis. Overall, 131 transesophagealcardioversion procedures were attempted in 105 patients. Twenty patients, all of whom had undergoneinitial successfulcardioversion, subCopyright

396

1993

by Mosby-Year

000%8703/93/S1.00

+ .lO

Book, 4/l/42641

Inc.

Volume Number

125

2,

Part

Transesophageal

1

397

conversion of SVTIVT

Fig. 1. Esophagealquadripolar electrode system.

sequently had repeat cardioversions attempted during 1988to 1990,becauseof either the development of a new arrhythmia or the recurrenceof the initial arrhythmia. The patient group included 68 men and 37 women,who ranged in age from 22 to 88 years (mean, 61.3 + 1.1 years). The most commonarrhythmia that required cardioversionwas atria1 fibrillation (109 episodes);there were fewer episodes of atria1flutter (16episodes),supraventricular tachycardia (2 episodes),and ventricular tachycardia (4 episodes). Those patients who were referred for transesophagealcardioversion of ventricular tachycardia were hemodynamitally stable (ventricular rate ranged from 115to 160beats/ min), but their arrhythmias were resistant to pharmacologic therapy. The underlying causesof the arrhythmias are detailed in Table I accordingto the first arrhythmia in each patient. Alcohol consumptionand medications(digoxin, P-blockers, calcium antagonists, and amiodarone) were recorded. The duration of the arrhythmia wasthe length of time since first documentation on an ECG. It is probable therefore that the actual duration of an arrhythmia was longer in somepatients. Echocardiography wasperformed in the majority of patients (88 of 105patients; 77 of 88 patients who presented with atria1 fibrillation as the initial arrhythmia), and left atria1 enlargement was determined by meansof M-mode echocardiograms,which were recorded with guidancefrom two-dimensional imageswith the transducer in the left parasternalposition. The anteroposterior dimensionof the left atrium was taken as the maximum distance from the leadingedgeof the posterior aortic wall to the leading edge of the posterior atria1wall asrecorded during diastole.Left atria1 enlargement was considered to be present if the left atria1 diameter exceeded 40 mm. Body mass index (weight/

height?) was calculated for all patients.12 The majority of patients were receiving anticoagulants orally before the procedure, except at the discretion of the referring

physician

or if the arrhythmia

was known to be of

short duration (i.e.,
Table I. Arrhythmias at initial presentation: Causes Cause

AF

AFL

Rheumatic Ischemic Post cardiac surgery Hypertension Cardiomyopathy Alcohol-induced Sinoatrial disease Lone Other* Total

23 20 8 4 7 2 4 7 13 ss

3 2 2 1

0 1

0 2 1 12

SVT

VT

1 0 0 0 0 0 0 0 0

i

AF, Atria1 fibrillation; AFL, atria1 flutter; SW’, supraventricular tachycardia; VT, ventricular tachycardia. *This group includes patients with mitral valve prolapse, congenital heart disease, myocarditis, and atria1 myxoma (cardioversion after surgical removal).

those patients in whom atria1 arrhythmias developedduring the hospital stay. Equipment Esophageal electrode. The esophageal systemconsistsof four separate electrode terminals, each of coiled goldplated copper wire, which are mounted on polyvinyl chloride tubing (Fig. 1). Several electrode sizesfor this esophagealsystem were available: the external diameter ranged from 5 to 9 mm. The majority of cardioversionswere performed with the 5 mm or 6 mm electrode size. Each electrode terminal is provided with a separate connection, which facilitates the recording of esophagealbipolar electrograms. For the purposesof countershock, all four electrode terminals are connected via a junction box to maximize the surface areafor energy transfer. The designof the electrode was influenced by experimental work on greyhound dogs, which indicated that the surface area of the esophagealsystemshould be maximized to reduce impedance measurementsand that the length of the electrode

398

McKeown

et al.

should not exceed that of the heart to minimize extracardisc distribution of energy.lr Cardiouerters. All cardioversions were performed with either a Liteguard 9 defibrillator (Temtech Ltd, Bangor, Northern Ireland), which hasthe facility for measurement and display of predicted impedanceas a standard feature, or a Liteguard 6 defibrillator (Temtech Ltd), which was modified to provide such predicted impedance measurements. The recording of these predicted impedancemeasurementsis done after a high-frequency [30 kHz] lowamplitude (100 PA) current passesbetween the respective electrodes, as initially described by Geddes.13This technique for impedancemeasurementhassincebeenshownto provide good correlation with actual impedance values, which were recorded during delivered shocks.14 These machines deliver critically damped sine-waveform shocks.The defibrillators were modified by the addition of a current meter, which registerspeak delivered current at the time of each shock. The current meter records the voltage measuredacrossa 1 Qresistor placed in series with the electrodes;according to Ohm’s law, the peak delivered current in amperesis identical to the peak voltage recorded in volts, sincethe resistancehasa constant value of 10. In 10patients the current meter failed to register the peak delivered current during successfulcardioversion. In thesepatients estimatedcurrents weretherefore calculated from current-resistance graphs, which were prepared as follows. The defibrillator wasdischargedrepeatedly (n = 6) into 11known resistances(from 20to 120Q)at energiesthat rangedfrom 30 to 360J, and the peak current at discharge was recorded; the mean peak current was then plotted againstthe known resistancefor the various energies.This methodology has previously been validated in clinical practice.‘” Procedure. The majority of attempted cardioversions (n = 124) were performed as elective procedures; the patients fasted overnight, and antiarrhythmic drugswerenot given on the morning of cardioversion. The remaining seven attempted cardioversionswere performed as emergency procedures, in three episodesof atria1 fibrillation because of clinical deterioration that necessitated attempted cardioversion and in four episodesof ventricular tachycardia becauseof lack of responseto pharmacologic therapy. These sevenpatients had not eaten for at least 4 hours before the procedure. Before insertion of the esophagealsystemdiazepamwas administered intravenously in aliquots (2.5 to 20 mg) to achieveadequaterelaxation of the patient. The patient was then placed in the left lateral position, and the esophageal systemwasadvancedthrough the mouth with the patient’s cooperation to a depth of between 35 and 40 cm from the tip of the system to the incisors. The distance that was chosenas the average distance from the teeth to the gastroesophagealjunction was 40 cm. No fluoroscopic equipment wasnecessaryfor positioning of the esophagealsystem. The closeproximity of the esophagealsystem to the heart is illustrated on a lateral chest roentgenogram,which was performed on one patient for illustration purposes (Fig. 2).

February American

Heart

1993 Journal

Pregelled electrodes,either 12cm diameter R2 pads(R2 Corp., Niles, Ill.) or 11.5 X 8.5 cm Andover pads (Andover Coated Prod., Inc., Marblehead, Mass.) were usedas the external chest wall electrodes. Predicted transesophageal and transthoracic impedanceswererecorded in all patients with the technique describedabove. The transesophageal impedancerefers to that recorded betweenthe esophageal systemand the pregelledelectrode that wasplacedjust inside the apex; the transthoracic impedancerefers to that recordedbetweentwo pregelledelectrodesthat wereplaced in the conventional anterior-apex position. Transesophageal impedancewasrecorded in all patients; transthoracic impedancewasrecorded in 104of 105patients. In the other patient the implantation of a permanent pacemakerprecluded placement of the anterior electrode. Aliquots of methohexital sodiumwerethen administered to achieve light anesthesia (total dose 30 to 200 mg). Transesophagealcardioversion wast,hen attempted, starting with a delivered transesophagealenergy of 30 J. If successfulcardioversion wasnot effected with the first shock, further transesophagealshockswere administered in stepwiseincrements up to 100J (the protocol for the initial 15 attempted cardioversions was 30, 40, 50, 70, and 100 J; during the remainder of the study the protocol was 30,50, and 100 J). It was initially decided to limit the delivered transesophagealenergy to 100J to establishthe safety of the procedure.After 25 cardioversionshad beenattempted with the aforementionedprotocol (21 of 25 successful)and no complications had beenencountered, we decidedto explore the useof higher transesophagealenergies.The next delivered energy available with the defibrillators that were usedin this study was200J, and we therefore decidedthat one additional transesophagealshock of 200 J would be administeredto thosepatients with atria1 fibrillation when cardioversion had been unsuccessful with a delivered transesophagealshock of 100 J. Five patients received transesophagealshocksof 200J; one of thesefive received a 200 J shock becauseof a secondepisodeof atria1 fibrillation. In four of thesefive patients there were no adverse effects. However, the fifth patient had esophagealmucosal injury (details in Resultssection) and the limit of 100J delivered transesophagealenergy wasthen reinstituted. If a similar rhythm that required cardioversionrecurred in the samepatient, then the first energy selectedwasthat which was previously successful. In 19 of 20 episodesof atria1 fibrillation in which transesophagealcardioversion was unsuccessful,transchest cardioversion wasattempted with a maximal delivered energy of 360J. Transchestcardioversionwasnot attempted in the other patient, a 47-year-old man with atria1 fibrillation, becausehe had, somemonths earlier, failed to convert to sinusrhythm when this approach was used. At the time of dischargeall patients were given specific instructions to contact the cardiology unit if they experienced any esophagealsymptoms. Patients were subsequently reviewed 4 weekslater and specifically questioned about the development of esophagealsymptoms; at that time repeat ECG wasperformed to determine the maintenance of sinus rhythm.

Volume Number

125 2, Part

1

Transesophageal

conversion of SVTIVT

399

Fig. 2. Lateral chestroentgenogramdemonstratesthe closeanatomic relationship betweenthe esophageal electrode system (E) and the heart (H).

Statistical analysis. Analysis was performed with the Statistical Package for the Social Sciences (SPSS Inc., Chicago, Ill.) on a Digital ProfessionalComputer, Digital Equipment Corp., Maynard, Mass.). For discretevariables, statistical analysiswasperformed with chi squaretests; for continuous variables a Student’s t test was employed. Resultsare presentedasmean + SEM. Differenceswere considered statistically significant when p was ~0.05. RESULTS Impedance measurements. The mean transesoph-

ageal impedance of 52.6 f 1.1 D (range, 27 to 82 a) was considerably lower than the mean transthoracic impedance of 63.1 f 1.6 D (range, 28 to 105 Q) (n = 104; p < 0.01). This significant difference between the transesophageal and transthoracic impedances was observed with all esophageal electrode sizes. All arrhythmias. Overall, during this series of 131 attempted cardioversions, a total of 313 shocks (successful and unsuccessful) were delivered, including 280 shocks that were delivered by the esophageal approach and 33 shocks that were delivered by the transthoracic approach. One hundred eleven of the

131 episodes (84.7 % ) were successfully terminated with the esophageal system; this included all of the episodes of atria1 flutter, supraventricular tachycardia, and ventricular tachycardia and 89 of 109 episodes of atria1 fibrillation (Table II). The number of transesophageal shocks required to effect this success ranged from 1 to 5 (mean, 1.9 f O.l), the energy ranged from 30 to 200 J (mean, 59.3 f 3.3 J), and the peak delivered current ranged from 11.4 to 33.0 A (mean, 19.5 f 0.5 A). Transthoracic cardioversion of atria1 fibrillation was attempted in 19 of 20 episodes in which the esophageal approach had failed and was successful in 12, which yielded an overall combined success rate for all arrhythmias of 94.6% (Table II). The number of shocks (both transesophageal and transthoracic shocks) required to achieve this success ranged from 1 to 7 (mean, 2.3 +- O.l), the energy ranged from 30 to 300 J (mean, 77.1 + 5.9 J), and the peak delivered current ranged from 11.4 to 50.2 A (mean, 21.2 f 0.7 A). Atrial flutter. All 16 episodes of atria1 flutter were successfully terminated with transesophageal shocks

400

McKeown et al.

American

February 1993 Heart Journal

Fig. 3. Broad complex tachycardia. Bipolar esophageal recording demonstrates intermittent ret,rograde 2:l ventriculoatrial conduction, thus confirming the ventricular origin of the tachycardia. A, Atria1 depolarization; V, ventricular depolarization; Oes, esophageal recording.

Table II. Conversion rates for individual arrhythmias Transesophageal

Overall

No. successful

Percent

AF AFL

89/109 16/16

81.7 100

SVT VT Total

212

Arrhythmia

Abbreviations *One patient, version, did text).

- 414 111/131 as in Table I. who had an unsuccessful not undergo an attempt

100 -100

84.7

No. successful

Percent

101/108 16/16 212 - 414 123/130

attempt at transesophageal at transthoracic cardioversion

93.5’ 100

100 -100

94.6

cardio(see

(Table II); on 14 occasions the resulting rhythm was sinus rhythm, which required 30 J in each of 10 episodes and 50 J in 4 episodes (30,50 J shocks in two episodes and a single 50 J shock in two episodes), and on the other two occasions atria1 fibrillation resulted (on both occasions after shocks of 30 J delivered energy). Peak delivered currents ranged from 11.4 to 18.0 A (mean, 15.3 + 0.6 A) with a mean number of 1.1 * 0.1 shocks (range, 1 to 2). On the two occasions when atria1 fibrillation was induced, sinus rhythm was subsequently restored after further 30, 50, and 100 J transesophageal shocks. Supraventricular tachycardia. The two episodes of supraventricular tachycardia were corrected to sinus rhythm by transesophageal cardioversion; the first with 30 J (peak delivered current, 18.1 A) and the other with 30, 50 J shocks (peak delivered current, 16.0 A). Ventricular tachycardia. The four episodes of ventricular tachycardia were all terminated with 30 J shocks that were delivered via the esophageal system.

Peak currents ranged from 13.1 to 18.2 A (mean, 15.0 f 2.0 A). During these episodes, all of which presented as broad complex tachycardias, the esophageal recording clearly showed the nature of the underlying atria1 rhythm and thus confirmed the ventricular origin of the tachycardia (Fig. 3). Atrial fibrillation. Of the 88 patients who presented with atria1 fibrillation as their first arrhythmia, 55 were men and 33 were women, with a mean age of 61.1 f 1.2 years (range, 22 to 88 years). Transesophageal success, During the first attempted transesophageal cardioversion, 70 of 88 (79.5 “; ) patients whose initial arrhythmia was atria1 fibrillation were successfully converted to sinus rhythm. The success rates, which were assessed according to the underlying cause, are listed in Table III. The maximum transesophageal energy delivered was 100 J in 84 patients and 200 J in four patients. The mean energy during successful transesophageal cardioversion was 63.1 + 4.2 J (range, 30 to 200 J), the mean peak current was 20.3 + 0.6 A (range, 11.9 to 33.0 A), and the mean number of shocks per patient was 2.2 + 0.1 (range, 1 to 5). The mean transesophageal impedance among the 70 patients who had successful cardioversion was 52.1 -+ 1.4 Q compared with a mean transesophageal impedance of 56.5 + 2.6 12for the 18 patients who had unsuccessful cardioversion by the esophageal approach (p > 0.05). Patients who had successful cardioversion by this route had a lower mean body weight (71.0 + 2.0 kg vs 79.5 +- 3.4 kg; p = 0.05) and lower mean body mass index (24.3 +- 0.5 kg/m2 vs 27.0 it 1.1 kg/m?; p < 0.05) than those in whom the technique was unsuccessful. There was no association between the success of transesophageal cardioversion and either duration of arrhythmia or left atria1 size.

Volume125 Number 2, Part

Transesophageal conversion of SVT/VT

1

Forty-three (48.9 % ) of these patients underwent successful transesophageal cardioversion with the use of low-energy shocks, which were defined as 150 J. Interestingly, this group did have a significantly lower transesophageal impedance than the group in which patients either required higher energies to effect cardioversion or had a failed cardioversion (mean impedance, 48.7 k 1.8 Q vs 57.1 t 1.6 Q; p < 0.01). Overall success.Eighty of 87 (92.0 % ) patients, who underwent attempted transesophageal/transthoracic cardioversion, were successfully converted to sinus rhythm (Table III). Seventy of these patients underwent successful transesophageal cardioversion and are discussed above. Ten of the 17 patients who underwent additional attempted transthoracic cardioversion were also restored to sinus rhythm; this required the use of 200 J in six cases and 300 J in four cases (cumulative energy range, 380 to 680 J). The success rate with regard to cause is recorded in Table III. The mean shock energy during successful cardioversion with transesophageal and transchest shocks was 85.3 + 7.8 J (range, 30 to 300 J), the mean successful peak current was 22.3 + 0.9 A (range, 11.9 to 50.2 A), and the mean number of shocks was 2.5 f 0.1 (range, 1 to 7). Seven patients, including six men and one woman, did not have successful cardioversion by either route, despite the use of energies of up to 360 J. Maintenance of sinus rhythm on review at 4 weeks. Of the 80 patients who underwent successful

cardioversion, either by the transesophageal or the transthoracic approach, 79 survived for review at 4 weeks; the other patient died of staphylococcal septicemia, which was unrelated to the cardioversion procedure. Of these 79 patients, 46 (58.2%) were in sinus rhythm at the time of review. Complications. One woman experienced pain on swallowing 8 hours after successful restoration of sinus rhythm by the esophageal approach; she had received transesophageal shocks of 30,50,100, and 200 J. Esophagoscopy revealed moderate superficial mucosal damage on the lower anterior wall of the esophagus. She fasted, apart from the administration of oral antacids. Intravenous fluids were prescribed until her symptoms resolved completely within the next 72 hours. Repeat esophagoscopy was performed 4 weeks later and showed complete resolution of the mucosal injury. The woman has since been seen at review over the following 18 months and remains free of symptoms. In a separate study elective esophagoscopy was performed in three patients within 72 hours of transesophageal countershock for ventricular tach-

401

Table III. Atria1 fibrillation

(initial presenting arrhythmia): Success rates in relation to cause Transesophageal success Cause

No.

Rheumatic Ischemic Post cardiac surgery Hypertension Cardiomyopathy Alcohol-induced Sinoatrial disease Lone Other Total

17123 14/20 s/s

73.9 70 100

21123 15120

314 717

75 100

414 717

212

100

414 617 - 9/13 70/88

Percent

Overall success

100 85.7 - 69.2 79.5

No.

B/8

212 414 717 12/12* ao/a7*

Percent 91.3 75 100 100 100 100 100 100 -100 92.0

*One patient, who had an unsuccessful attempt at transesophageal cardioversion, did not undergo an attempt at transthoracic cardioversion (see text).

yarrhythmias that were induced during programmed ventricular stimulation. These patients were free of symptoms. Two had received transesophageal shocks of 50 J, and the other had received transesophageal shocks of 50 and 100 J. There was no endoscopic evidence of esophageal injury.16 Three patients died within 1 month after the cardioversion procedure of causes that were unrelated to the procedure; two patients died of end-stage heart failure, and the other died of staphylococcal septicemia, conditions which predated cardioversion. The patient who had staphylococcal septicemia died 1 week after successful transesophageal cardioversion; he had received shocks of 30,50, and 100 J; at autopsy no esophageal damage was noted. In one patient, a 65-year-old man with atria1 fibrillation, a right hemiparesis developed 4 days after countershock, from which he recovered with mild to moderate functional impairment. This patient had not started to receive prophylactic anticoagulation therapy before cardioversion because the referring consultant had believed that the arrhythmia was of short duration. No deaths were encountered as a result of the procedure. DISCUSSION

The use of external electrical countershock for arrhythmias other than ventricular fibrillation was first described in 1956 by Zoll et a1.17who successfully reverted ventricular tachycardia with the use of alternating-current shocks. The use of synchronized direct-current cardioversion was described 6 years later by Lawn,’ and since that time the value of transchest cardioversion has been well validated in many

402

McKeown

et al.

studies.‘-3 Overall success rates with the use of this technique have been reported to vary from 68’ c’ to 100%) with high successrates for atria1 flutter and lower rates for atria1 fibrillation.Zs X Nevertheless, in transthoracic cardioversion only a small fraction of the delivered energy actually passes through the heart. Guyton and Satterfield5 in 1951 recorded that three times the voltage was needed for external defibrillation as compared with the use of electrodes applied directly to the heart. In 1932, Kouwenhoven et a1.4reported that when electrodes were applied directly to the head and tail of dogs, as little as 9% of the total current actually passed through the heart. In further experiments Kouwenhover-i6 applied electrodes at the suprasternal notch and the apex and calculated that with this combination, 20 % to 25 9;)of the countershock actually passed through the heart. The close anatomic relationship between the esophagus and the heart permits ready accessto the heart from the esophagus for both investigation and management of cardiac disorders. Since the early part of this century the esophagus has been used to record the electrical activity of the heart. The initial placement of an electrode in the esophagus is attributed to Cremer’ who in 1906 succeeded in recording an esophageal ECG with a silver electrode, which was 10 cm long and 1.5 cm wide. Not surprisingly, the patient was a professional sword swallower. The first major clinical contribution to the technique of esophageal ECG was published in 1936 by Brown,18*1gwho examined tracings from a total of 142 human subjects and described in great detail the characteristic tracings that were recorded from the auricle in normal subjects and in patients during auricular arrhythmias. Various esophageal electrodes have subsequently been designed for the purposes of ECG recording; of particular note is the “pill electrode,” which was proposed by Arzbaecher.20 Successful transesophageal cardiac pacing has also been achieved.8* 21-24 It has been well known for many years that low energies are required for ventricular defibrillation when the electrodes are applied directly to the heart.” Whipple and PentonlO suggested the use of an esophageal approach to defibrillation and reported successful defibrillation in mongrel dogs with the use of transesophageal shocks of alternating current. This approach to cardioversion of atria1 arrhythmias in human beings was adopted by McNally et a1.25who achieved successful cardioversion in 13 patients with the use of low energies that ranged from 15 to 60 J. Five of these patients subsequently reverted to atria1 fibrillation and required energies that ranged from

American

February 1993 Hean Journal

150 to 215 J to effect successful transthoracic cardioversion.‘” There have since been occasional reports of t,he use of this technique for treatment of atria1 arrhythmias.“-“” Volkmann et al.‘j’ have also reported successful t,ermination of ventricular tachycardia with this approach. Lukoseviciute and Peculiene”” have examined the use of transesophageal cardioversion in patients with atria1 arrhythmias that are resistant to conventional transthoracic cardioversion and have achieved notable successin this area when t.he delivery of high voltage transesophageal shocks (up to 6 kV, which equates to 320 J delivered energy) was permitted. In spite of these sporadic reports, the technique has not become generally accepted. This is mainly due to the lack of a suitable stable esophageal electrode system. The electrode system that was used in this study was designed on the basis of experimental work in greyhound dogs.I1 In the animal model it was established t,hat the surface area of the esophageal electrode should be maximized to reduce the impedance recorded between this system and an external electrode. It was also concluded that the length of the esophageal system should not exceed that of the heart to minimize extracardiac distribution of current. The quadripolar system was thus designed to meet these criteria and to permit the facility for recording esophageal ECGs. In this study all episodesof atria1 flutter, supraventricular tachycardia, and ventricular tachycardia were successfully terminated with the esophageal approach. The facility of esophageal recording proved useful in the management of the broad complex tachycardias, because the esophageal recording permitted clear identification of the underlying atria1 activity before cardioversion. Analysis of the data has shown a 79.5 “0 successrate (70 of 88 patients) for transesophageal cardioversion of atria1 fibrillation, which was increased to 92.0% when additional transthoracic countershock was undertaken in those patients resistant to the esophageal approach. This success rate is virtually identical to that reported by Dalzell et a1.33who, in a similar population, achieved successful transthoracic cardioversion in 73 of 80 patients (91.2%) who experienced atria1 fibrillation; they recorded the mean energy of the successful shock (or maximal if unsuccessful) to be 222 + 12 .J. In this study a mean energy of 85.3 +- 7.8 J was required with the transesophageall transthoracic approach. Low-energy shocks (50 J or less) were successful in 43 of 88 patients who underwent attempted transesophageal cardioversion of atria1 fibrillation. This subgroup of patients had a significantly lower mean

Volume

125

Number

2, Part 1

transesophageal impedance than patients who required higher energies for or had a failed attempt at cardioversion. What are the mechanisms for reduction in impedance measurements and energy requirements for transesophageal cardioversion of arrhythmias? The esophagus lies within a few millimeters of the left atria1 wall and thus has a close anatomic relationship with the heart. The impedance reduction is related at least in part to the reduction in interelectrode distance that is provided by this approach; it also minimizes the passage of current through tissues of high resistivity, which include lung tissue.34 The reduction in energy requirements is believed to be based on many factors. The direction of current flow may be relevant. The concept of a preferred axis for cardiac defibrillation was suggested by Moore et a1.35 who demonstrated higher success rates for ventricular defibrillation in dogs if the electrodes were positioned longitudinally (at the base of the heart and the apex) rather than transversely (opposing electrodes on the right and left ventricles). Elam et a1.36,37 have also suggested that the plane-parallel electrode geometry of the esophagus to chest wall electrode combination is a relevant factor. It seems evident from our study that the significant reduction in impedance with the esophageal approach must be a factor, as demonstrated by the lower mean transesophageal impedance in those patients who underwent successful low-energy cardioversion of atria1 fibrillation. In their early work McNally et a1.25 had administered twenty 200 J shocks in rapid succession to six anesthetized dogs; the animals were kept alive for 3 days and then killed. No adverse effects were reported. In human subjects who underwent attempted transesophageal cardioversion Lukoseviciute and Peculiene32 reported no adverse effects despite the administration of shocks up to 6 kV, equating to delivered energies in excess of 300 J. In our study we encountered no complications in those patients who had received a maximal transesophageal shock of 100 J; esophageal mucosal injury was induced in one of five patients who had received a transesophageal shock of 200 J. In conclusion, we have shown that this esophageal electrode system is easily positioned in human beings. The significant reduction in impedance that is encountered with the application of this technique permits low-energy countershock of atria1 and ventricular tachyarrhythmias. This approach appears to be safe if the transesophageal energy delivery is limited to 100 J. It has been shown to provide goodquality esophageal electrograms, which prove to be useful in the diagnosis of mechanisms of arrhyth-

Transesophageal

conuersion

of

SVTWT

403

mias. We believe that this versatile system will prove to be a valuable addition to the field of arrhythmia management. We thank the staff of Brunswick Manufacturing Company, Massachusetts, for their assistance in the technical design and manufacture of the esophageal electrode system used in this study.

REFERENCES

1. Lown B, Amarasingham R, Neuman J. New method for terminating cardiac arrhythmias. JAMA 1962;182:548-55. 2. Resnekov L, McDonald L. Appraisal of electroconversion in treatment of cardiac dysrhythmias. Br Heart J 1968;30:786811. 3. Collins RE, Giuliani E. Atria1 defibrillation in the US: experience at the Mayo Clinic. In: Tacker WA, Geddes LA, eds. Proceedings of the Cardiac Defibrillation Conference. West Lafayette,-Indiana: Purdue University, 1975:21-5. 4. Kouwenhoven WB. Hooker DR. Lanaworthv OR. The current flowing through the heart under conditions of electric shock. Am J Physiol 1932;100:344-50. 5. Guyton AC, Satterfield J. Factors concerned in electrical defibrillation of the heart, particularly through the unopened chest. Am J Phvsiol 1951:167:81-7. 6. Kouwenhoven WB. The development of the defibrillator. Ann Intern Med 1969;71:449-58. I. Cremer M. Ueber die direkte Ableitung der Aktionsstrome des menschlichen Herzens vom Oesophagus und uber das Elektrokardiogramm des Fotus. Muenchener Med Wochenschr 1906;53:811-3. 8. Benson DW, Dunnigan A, Benditt DG, Schneider SP. Transesophageal cardiac pacing: history, application, technique. Clin Prog Pacing Electrophysiol 1984;2:360-72. 9. Roelandt JRTC, Sutherland GR. Oesophageal echocardiography. Br Heart J 198&6&l-3. 10. Whipple GH, Penton GB. Transesophageal ventricular defibrillation [Abstract]. Clin Res Proc 1956;4:105. 11. Croal S, Anderson JMcC, Allen JD. Electrical defibrillation of the canine ventricle using oesophageal and praecordial electrodes. Ir J Med Sci 1990;159:28. 12. Khosla T, Lowe CR. Indices of obesity derived from body weight and height [Abstract]. Br J Prev Sot Med 1967;21: 12218. 13.

14.

15.

16.

17.

18.

19.

-

-

-

Geddes LA. Characteristics of defibrillating electrodes and living tissue. In: Proceedings of the Cardiac Defibrillation Conference. West LafayetG, Indiana: Purdue University, 1975:45-53. Dalzell GWN, Magee H, Anderson J, Adgey AAJ. Predicted and actual impedance in cardioversion [Abstract]. Circulation 1986;74:11-388. Kerber RE, Carter J, Klein S, Grayzel J, Kennedy J. Open chest defibrillation during cardiac surgery: energy and current requirements. Am J Cardiol 1980;46:393-6. MC Keown PP, Khan MM, Croal S, Allen JD, Anderson J, Adgey AAJ. Electrocardiographic recording, atrial pacing, and defibrillation in the course of electrophysiologic studies using the esophagus. Coronary Artery Dis 1992;3:383-91. Zoll PM, Linenthal AJ, Gibson W, Paul MH, Norman LR. Termination of ventricular fibrillation in man by externally applied electric countershock. N Engl J Med 1956;254:727-32. Brown WH. A study of the esophageal lead in clinical electrocardiography. Part I. The application of the esophageal lead to the human subject with observations on the Ta-wave, extrasystoles and bundle-branch block. AM HEART J 1936;12:1-45. Brown WH. A study of the esophageal lead in clinical electrocardiography. Part II. An electrocardiographic study of auricular disorders in the human subject by means of the esophageal lead. AM HEART J 1936;12:307-38.

404

McKeown

et al.

20. Arzbaecher R. A oil1 electrode for the studv of cardiac ” arrhythmias. Med I&strum 1978;12:277-81. 21 Shafiroff BGP, Linder J. Effects of external electrical pacemaker stimuli on the human heart. J Thorac Surg 1957;33:54450. 22. Gallagher JJ, Smith WM, Kerr CR, Kasell J, Cook L, Reiter M, Sterba R, Harte M. Esophageal pacing: a diagnostic and therapeutic tool. Circulation 1982;65:336-41. 23. Burack B, Furman S. Transesophageal cardiac pacing. Am J Cardiol 1969;23:469-72. 24. Andersen HR, Pless P. Trans-esophageal pacing. PACE 1983;6:674-9. 25. McNally EM, Meyer EC, Langendorf R. Elective countershock in unanesthetized patients with use of an esophageal electrode. Circulation 1966;33:124-7. 26. Fields J, Shachtman J, Craig SD, Berkovits BV, Roger A. Special electrode device for trans-esophageal cardioversion IAbstractl. Digest 7th Int Conf Med Biol Ene Stockholm. i967:77. 1 27. Lorkiewicz Z, Szelagowicz B, Gracz-Kaysiewicz H, Ponizynski A, Ratajczak C. Low energy transesophageal cardioversion without general anaesthesia. Kardiol Pol 1968;11:125-30. 28. Richter G, Muller S. Erste Erfahrungen bei der elektrischen Kardioversion mit Hilfe einer Osophaguselektrode. Z Gesamte Inn Med 1968;23:724-6. 29. Armand P, Paecht A, Arion H. La cardioversion par voie oesophagienne [Abstract]. Arch Ma1 Coeur 1970;63:595.

American

30.

31.

32.

33.

34.

35.

36.

37.

February 1993 Heart Journal

Volkmann H, Muller S, Reuter A. Erfahrungen bei der elektrischen Kardioversion mittels Osophaguselektrode. Deutsche Gesundh-Wesen 1977;32:965-70. Volkmann H, Paliege R, Muller S, Wiese C, Kuhnert H. Der transosophageale Zugangsweg zum Herzen zur Elektrotherapie tachykarder Rhythmusstorungen. Z Gesamte Inn Med 1981;36:951-7. Lukoseviciute AJ, Peculiene IR. Transesophageal cardioversion of resistant atria1 arrhythmias. Resuscitation 1980;8:15965. Dalzell GWN, Anderson J, Adgey AAJ. Factors determining success and energy requirements for cardioversion of atria1 fibrillation. Q J Med 1991;78:85-95. Geddes LA, Baker LE. The specific resistance of biological material: a compendium of data for the biomedical engineer and physiologist. Med Biol Eng 1967;5:271-93. Moore TW, Ferris CD, Khazei AH, Cowley RA. Preferred cardiac axis for electrical stimulation. Bull Maryland Univ Sch Med 1967;52:3-5. Elam JO, Reque EV, Rattenborg CC. Esophageal electrocardiography and low-energy ventricular defibrillation. In: Safar P, ed. Advances in cardiopulmonary resuscitation. New York: Springer Verlag, 1977:167-74. Elam JO. Esophageal site for ECG and ventricular defibrillation [Abstract]. Med Instrum 1978;12:54.