772

Intraoperative Radiofrequency Ablation of Chronic AtrialFibrillation: A Left Atrial Cnrative Approach by

Elimination of Anatomic "Anchor'' Reentrant Circuits

HANS KOTTKAMP. M.D:'± GERHARD HINDRICKS. M.D.,*tDIETER HAMMEL, M.D.,t RUDIGER AUTSCHBACH, M.D.,§

JORG MERGENTHALER, M.D..='= MARTIN BORGGREEE. M.D.,*GUNTER BREITHARDT, M.D..* FRIEDRICH-WILHELM MOHR. M.D.,§

atid HANS H. SCHELD, M.D.f

From the Departments of -^Cardiology and Anj;it)logy and IThoracic. Heart and Vessel Surger>\Hospital of the Westfalische Wilhelms-University. Mlinster, and the Departments of ?Cardiology and §Heart Surgery.

University of Leipzig, Heart Center, Leipzig, Germany

Intraoperative Ablation of Atrial Fibrillation, introduction: The percutaneous ap-proach to radiofrequency (RF) catheter ahlation lor curative treatment of atrial fihrillation (AF)is au investij>ational technique, and the optimal composition of le.siou lines is unknown. Wetested an intraoperative RF ahlation concept with elimination of left atrial anatomic "anchor"reentrant circuits.

Methods and Results: In 12 patients with an indication for valve surgery and chronic AF, aright atrial-transseptal approach was chosen for access to the left atrium. AF had been presentfor 4.3 ± 3.9 years; the left atria measured 56 ± 7 mm. Under direct vision, contigutiu.s lesionlines were placed endocardially with temperature-guided RF energy applications for treatmentof AF with a specially designed prohe. The lesion lines were placed hetween the mitral annulusand the left lower pulmonary vein, further to the left upper pulmonary vein, from there to theright upper pulmonary vein, and finally to the right lower pulmonary vein. The antiarrhythmicahlation procedure lasted 19 ± 4 minutes. One patient died postoperatively of low cardiacoutput. During follow-up of 11 ± 6 months, chronic AF was ahlated successfully in 9 of Upatients (82%). Six patients were in stable sinus rhythm or intermittent pacemaker rhythm, andthree patients were in sinus rhythm with intermittent atypical atrial flutter.

Conclusions: Intraoperative RF energy application for induction of contiguous lesion lines isfeasible. Elimination of anatomically defined "anchor" reentrant circuits within the left atriumprevented chronic AF in > 8(1% of the patients treated. Intraoperative validation of lesion lineconcepts for curative treatment of AF may he transferred to percutaneous ahlation techniques.(J Ciirdiovasc Etcctropliysiol. Vol. JO. pp. 772-7H0, June 1999)

atrial fibrillation, catheter ablation, antiarrhythmic surgery, electrophysiology, arrhythmias

Introduction

Within the last decade, percutaneous radiofre-quency (RF) catheter ablation with focal energy

Supported in pan by IMF Grant Hi-M-!I/96-36.

Address for correspondence: Hans Koukamp. M.D., UniversitiitLeipzig. Her/zentruin. Klinik fiir Inrn-rc Mcdi/in/Kardiologic. Rus-senstr. 19. D-04289 Leipi^ig. Germany. Fax: 49-341-865-1461:E-mail: Koilh^'>medizin.uni-leip/.ij!.de

Manuscript received 28 December 1998; Accepted for publication12 March 1999.

applications has become the treatment of firstchoice for most supraventricular imd AV ar-rhythmias.' *• A new concept of RF etiergy appli-cation evolved when atrial flutter or atrial reen-trant tachycardias involving atriotomy scars weretargeted for ablation.' '̂ In these cases, contigu-ous linear lesions are induced to form electri-cally insulated barriers within "isthmuses" ofmyocardium that are an integral part of the re-entrant circuit of the respective arrhythmia. Thefeasibility of percutaneous RF catheter ablationof atrial iibrillation (AF) in hutnans by induction

Kottkamp. et al. Intraoperative Ablation of Atrial Fibrillation 773

of long linear lesions was reported recently.'"^ '̂Overall, the percutaneous approach to RF cath-eter ablation of AF is still an investigationaltechnique with several problems. First, the opti-mal composition of right and/or left atrial lesionlines is unknown. Second, ciitiical failure of per-cutaneously induced lesion lines may be the re-suit of insufficient lesion extent and/or geometry,or ablation failure may be due to insufficientrealization of the proposed lesion lines, i.e., theinduction of /7W7contiguous lesions withoutachievement of linear conduction block. Third,percutaneous validation of contiguous lesionlines in patients with chronic AF seems verycomplex. Therefore, optitnized effective and safecatheter ablation of AF remains a challenge forcurative percutaneous treatment of this commonarrhythmia.

Cox and co-workers"*'^ developed an exten-sive surgical concept of induction of multipleatdal incisions in both atria for curative treat-ment of AF. Their concept proved that properlyplaced multiple biatrial atriototnies can alter thearrhythmia substrate sufficiently to prevent fur-ther AF episodes in most patients. However,widespread application of the surgical maze pro-cedure may bave been precluded by tbe time-consuming nature of this operation, with itsassociated long cardiopulmonary and aorticcross-clamp times.

In the present study, tbe operation theater onone side and RF energy application on the otherside were combined to test a new concept ofintraoperative RF ablation of chronic AF. In pa-tients witb an arrhytbmia-independent indicationfor valve surgery and chronic AF, a right atrial-transseptal approach was chosen for access to tbeleft atrium. A left atrial approach to intraopera-tive RF energy application was applied that wasrestricted to elitninate anatomically determinedmacroreentrant circuits involving tbe pulmonaryveins and mitral annulus. Experimental investi-gations in freshly excised pig left atria with tbisregimen demonstrated the smooth induction ofcontiguous transmural lesion lines without endo-cardial disruption or peribration.-" Using thisconcept, the feasibility of intraoperative induc-tion of long contiguous lesions under direct vi-sion with a hand-held probe using RF energy wastested. In addition, a specific pathophysiologichypotbesis was tested that might be applied topercutaneous catheter ablation strategies in tbefuture.

Methods

Patient Characteristics

All patients included in this .study bad anindication for valve surgery irrespective of AF.Inclusion criterion for the additional arrbythtniaprocedure was long-lasting AF for more than 2years tbat was either chronic persistent or parox-ysmal. In cases of chronic paroxysmal AF, theepisodes were long-lasting and interrupted onlyrarely by short periods of sinus rbythm. Addi-tional criteria for inclusion were symptoms ofpalpitations and tachycardia that were related toirregular or tacbyarrhythniic ventricular responseduring AF. previous thromboembolic events,and/or hemodynamic aspects of atrial contractilefunction. Effective anticoagulation for at least 2months before surgery was a prerequisite for thearrhythmia procedure. Exclusion criteria for ad-ditional AF ablation were age > 75 years, pre-vious cardiac surgery, myocardial infarctionwithin the last 3 months, and other severe con-comitant noncardiac diseases affecting perioper-ative risk. Presurgical evaluation included leftand right beart catheterization, ecbocardiogra-pby, and 48-bour Holter monitoring.

Twelve patients (mean age 67 ± 9 years) metthe inclusion criteria and were included in thisstudy (Table 1). Seven patients had cotiibinedmitral valve disease, 3 patients had mitral valveregurgitation, 1 patient had mitral stenosis, and 1patient bad cotnbined aortic valve disease. Ad-ditional coronary artery disease was diagnosed in3 patients. Ten patients had chronic persistent AFwith an arrhythmia history of 2 to 13 years. Twopatients had the paroxysmal form of chronic AFwitb frequent long-lasting episodes for 4 and 3years, respectively. The left atrium was ecbocar-diographically enlarged in all patients and mea-sured 43 to 69 mm (mean 56 ± 7). All patientswere preoperatively in New York Heart Associ-ation Functional Class II or III. All patients gavewritten informed consent for the investigationalnature of this study. The study was approved bythe local ethics committees.

Surgieal Procedure

Tbe operation was petformed through a medianstemotomy. Hypothemiic (32°C) caidiopulnionarybypass was instituted with cannulation of the supe-rior and inferior caval veins and cross-clamping ofthe aorta. Cardiac arrest was accomplished using

774 Journal of Cardiovascular Electrophysiology Vot. JO, No. 6, June 1999

PI.No.

1

2

34

5

6

7

8

9

to

1!

12

Mean

Age(years)

67

72

6766

75

66

65

44

72

58

71

75

67 ± 9

Vitium

Comb.MVD

MI.CAD

MIComb.

MVDComb.

MVD.AI

Comb.AVD

Comb.MVD

Comb.MVD

MI.CAD

MST.CAD

Comb.MVD

Comb.MVD

AF Type

Persistent

Persistent

ParoxysmalPersistent

Persistent

Persistent

Persistent

Persistent

Persistent

Persistent

Persistent

Piiroxysmal

APDuration(years)

2

3

32

3

2

13

2

2

12

3

4

4.3 ± 3.9

ni Chara

LA(mm)

63

65

5661

50

45

69

53

50

56

53

48

56 ± 7

TABI.K 1cteristics and

Operation

MV repl.

MV repl..RCXgraft

MV repairMV repl.

MV repl.

AV repl.

MV repl.

MV repl.

MV repair.LADgraff

MV repl..LAD/RCXgrafts

MV repair

MV repl.

Operalion Dyf;:

BypassTime(min)

115

107

10995

67

107

85

99

73

138

97

153

104 ± 2 4

ACCTime(min)

84

76

6874

45

74

63

73

55

94

67

108

73 ± 17

AblationDuration

(min)

27

22

1714

19

17

19

19

18

28

16

16

19 ± 4

Follow-Up(moulh.s)

20

19

1715

*

9

8

8

7

6

3

3

11 ± 6

KhytlmiOutcome

SR/PM.paroxysmalAFlu

AF

SR. AFiu abl.SR/PM

*

SR.paroxysmalAFlu

SR

SR

SR

AF

SR

SR

* Patient 5 died during the pcriuperative period (see lext for details).ACC lime = mnic cross-clamp time; AF = atrial tibrillation: AFlu = atrial flutter; AFlu abl. ^ catheter ablation of atrial IKitter; AI = aonicinsufficiency; AV repl. = aortic valve replacement; comb. AVD ^ combined aortic valve disease; bypass time = cardiopulmonary bypasslime: CAD = coronury artery disease; LA = left atrium: LAD = left anterior descending coronary artery: MI = mitral insufficiency: MST =mitral stenosis; MV = mitral valve: MV repl. = mitral valve replacement; comb. MVD = combined mitral valve disease: PM = pacemakerrhythm; RCX = ramtis circumflexus: SR = sinus rhythm.

retrograde cold blood cardioplegia. The rightatrium was opened witb a longitudinal incision atthe right atrial free wall in a craniocaudal direction.Care was taken to avoid damage to the sinus noderegion and the sinus node blood supply. The inter-atrial septum was dissected through the fossa ovalisfor access to the left atrium, with the lower borderat a sufficient distance to the tricuspid annulus andthe triangle of Koch to avoid damage to the AVnode. No additional left atrial free-wall incisionswere made, and left atrial appendectomy was notperformed. Mitral valve replacement or mitralvalve repair was perfonned before the RF abladonprocedure.

RF Catheter Ahlation

The right atrial-transseptal approach to the leftatrium allowed access to the mitral annulus as

well as exposure of tbe orifices of tbe pulmonaryveins. RF alternating current was administeredusing a continuous sinusoidal unmodulatedwaveform of 500 kHz (modified HAT 200S.Sulzer-Osypka GmbH, Grenzach-Wyhlen. Ger-many). Energy was delivered in a unipolar modebetween the 4-nim (first two patients) or 10 mmtip electrode of a specially designed ablationprobe {Sulzer-O.sypka GmbH) (Fig. 1) and a10X16 cm external backplate electrode tbat wasplaced underneath the patient's back. The abla-tion probes had a thermistor embedded centrallyin tbe distal part of the tip electrode for contin-uous monitoring of catheter tip temperature. Theablation procedure was done in a bloodless field.Temperature guided energy applications wereperformed with a preselected catheter tip temper-ature of 60" to 65"C (lO-niin tip electrode) or 70°to 75°C (4-mm tip electrode). Based on catbeter

Kottkamp, et al. Intraoperative Ablation of Atrial Fibrillution 775

Figure 1. Specially de.signed ablation probe with a T-shaped lO-mm tip electrode (Sulzer-Osypka). Magnificationof the electrode tip is shown on the right. The ablation probehad a thermistor embedded centrally in the distal part of thetip electrode.

LUPRUPV

RLPV

Figure 2. Sctienuitic drawing of the left atrium (posteriorview) and the geometry of the radiofrequency energy-in-duced left atrial linear lesions for treatment of at rial fibril-lation. The first lesion line (I) connected the posterior mitralannulus with the left lower pulmonary vein (LLPV). The.second line ill) connected the LLPV with the left upperpulmtmary vein (LUPV). From there, a third line (III)coupled the LUFV with the right upper pultnonary vein(RUPV). Finally, the RUPV and right lower pulmonary vein(RLPV) orifices were connected with a contiguous lesionline (fV). FO - fossa avails: MV — mitral valve: RFprobe — .specially designed prohe for temperature-guidedradiofrequency energy application.

Figure 3. Contiguous lesion lines induced endocardiallywithin the left atrium for ablation of atrial fibrillation. (Top)Lesion line beginning at the ostium of the lower left pulmo-nary vein extending toward the upper left pulmonary vein.(Bottom) Lesion line beginning at the upper left pulmonaryvein and extending toward the upper right pulmonary vein.

Stability, induction of a well-visualized lesionline, and achievement of the preselected cathetertip temperature, the ablation probe was with-drawn after 20 to 30 seconds. Power, impedance,and catheter tip temperature were monitored con-tinuously during the energy applications.

The first part of the RF application protocolconsisted of placing a contiguous lesion line ex-tending from the posterior aspect of tbe mitralannulus to the left lower pulmonary vein (Fig. 2).The second line connected the left lower andupper pulmonary vein orifices. From there, athird line coupled the left and right upper pul-monary veins. Finally, the right upper and lowerpulmonary vein orifices were connected with a

776 Journal of Cardiovascular Electrophysiology Vol. 10, No. 6, June 1999

contiguous lesion line. Care was taken to ad-vance the tip of tbe ablation probe a few milli-meters inside tbe proximal parts of the pulmo-nary veins for ablation of muscle strands thatextended into the walls of tbe pulmonary veins.

Postoperative Management and Follow-Up

After the operation, all patients were moni-tored with continuous electrocardiographic(ECG) recordings for the first 7 postoperativedays followed by a 24-bour ECG. In cases ofpostoperative AF that occurred within the first 10days after the operation, sotalol 240 to 320 mgwas given and combined with electrical cardio-version, if necessary. In patients in whom sotalolwas applied and stable sinus rbythm could beachieved, this therapy was routinely witbdrawnafter 3 months. The patients were seen in tbeoutpatient clinic 3, 6, 12, and 18 months after tbeoperation, at which time 12-lead and 24-hourECGs were performed. All patients and referringphysicians were instructed to document on ECGif recurrences of an arrhythmia were suspected.Oral anticoagulant therapy was prescribed for atleast 6 montbs.

Results

Operative Course

Mitral valve replacement was performed in 8patients, mitral valve repair in 3, and aortic valvereplacement in 1 (Table I). Additional coronaryartery bypass grafting was done in three patients.Cardiopulmonary bypass time for tbe completeoperation including valve replacement/repair, theantiarrbythmic treatment, and bypass grafting, ifnecessary, measured 104 ± 24 minutes, aorticcross-clamp time 73 ± 17 tiiinutes, and time forantiarrhythmic treatment by placing the linearlesion lines 19 ± 4 minutes. The linear lesionscould be placed in all patients without difficulty(Fig. 3). Lesion width measured 4 to 5 mm.There were no intraoperative complications, es-pecially no left atrial perforation during RF en-ergy application or visible clot formation.

Early Postoperative Period

ln all 12 patients, sinus rhythm was docu-mented directly after the operation (Fig. 4). Incases of sinus bradycardia, atrial pacing (AAI or

{months]

Figure 4. (Top) Thirty-day follow-up period. (Bottom)Long-term follow-tip afler intraoperative radiofrequencyablation of atrial fihrillation (AF). AFlu — atrial flutter;ECV = electrical cardioversion; PM = pacemaker; RF =radiofrequency catheter ablation: Sot. = sotalol; Sot.ex = sotalol withdrawal; SR = sinus rhythm; SR/DDD-PM = sinus rhythm changing with DDDpacemaker rhythm.See text for discussion.

DDD) was performed with the temporary wiresplaced at the time of the operation until completesinus node recovery or until implantation of apermanent pacemaker (see following).

Patient 5 developed hemodynamic deteriora-tion and ventricular fibrillation 24 hours after theoperation. Mitral valve replacement and the in-traoperative antiarrhythmic treatment were un-eventful. After successful deflbrillation, the he-modynamic situation could not be stabilizeddespite treatment with intravenous catecbol-amine. The patient underwent immediate reop-eration. At that time, the previously mild aorticinsufficiency appeared aggravated, and the leftventricie was adynamic. There were no signs ofcardiac tamponade. the mitral valve replacementwas in place, and the left atrium was normalwitbout signs of perforation. The aortic valvewas replaced during the reoperation, but the pa-

Kottkamp, et al. Intraoperative Ablalion of Atrial Fibrillation 777

tient died of unresponsive low cardiac output.Patient 2 developed bleeding from the distalanastomosis of the venous bypass graft to thecircumflex coronary artery with resulting cardiactamponade. Tbe patietrt underwent reoperationon the .second postoperative day, and the postop-erative period was uneventful.

In 6 of 11 patients, AF recurred in the periop-erative period (Fig. 2). In one of these (patient 8),AF was self-terminating. In five patients (pa-tients 1, 2, 4, 10, and II), sotalol (240 to 320mg/day) was given to convert AF and was suc-cessful in three (patients 1. 4, and II). In twopatients (patients 2 and 10), sotalol was ineffec-tive, so electrical cardioversion also was applied.However, both patients had early recurrence ofAF tbat persisted during the follow-up period. Inthree patients (patients I, 4, and 10), DDD pace-makers were implanted between days 8 and 14after the operation because of sinus bradycardia.

Long-Term Follow-Up

Eleven patients were discharged from tbe hos-pital. Afler a follow-up of 3 to 20 months (mean1 I ± 6), six patients were in stable sinus rbythm(patients 7, 8, 9, II, and 12) or in sinus rhythmchanging with DDD pacemaker rhythm (patient4). Tbree patients had sinus rbythm with parox-ysmal atypical atrial flutter. In one oi" these pa-tients (patient 3), atrial flutter became persistentand led to tachycardia due to 2:1 conduction. Inthe otber two patients (patients I and 6), infre-quent episodes of self-terminating atrial flutterwitb normal ventricular response were observed.The patient with persistent atrial flutter under-went an invasive electrophysiologic study, andso-called incisionai reentry around the scar at theright atrial free wall was diagnosed. In the samesession, the lower end of the incision was con-nected to the inferior caval vein with four RFpulses. During the second pulse, atrial flutterterminated. Complete conduction block aroundtbe inferior end of fbe incision was verified witbstimulation and mapping techniques. After abla-tion, stable sinus rbythm was documented in thispatient without recurrences of AF or atrial flutter.In two patients (patients I and 4), stitalol treat-ment was discontinued after 3 months withoutrecunences of AF during follow-up periods of 20and 15 months, respectively. Overall, chronic AFwas ablated successfully in 9 of I 1 patients(82%); the other two patients were still inchronic persistent AF (patients 2 and 10).

Discussion

Main Findings

First, intraoperative RF energy application us-ing a specially designed hand-held probe forplacement of contiguous linear lesion lineswithin tbe left atrium was feasible under directvision. No complications occurred that were re-lated to RF energy application. Second, furtherepisodes of chronic AF could be prevented withour approach of elimination of anatomically de-fined left atrial "anchor" reentrant circuits in >80% of the patients treated. The time needed forperforming this concept with placement of thefour lesion lines between the mitral annulus andthe pulmonary veins measured only 14 to 28minutes. Third, the efficacy of the proposed le-sion line concept could be validated, because theinduction of contiguous linear lesion lines couldbe visualized during the operation. Therefore,intraoperatively proven concepts of this lesiongeometry or of new lesion concepts can be ap-plied later to percutaneous catheter ablation tech-niques.

Lesion Geometry for Curative Treatment of AF

In the first major biatrial clinical mappingstudy of AF by Cox et al.,'' 13 patients under-going surgical correction of Wolff-Parkinson-White syndrome were investigated. Simulta-neous mapping oi' the right and left atria with 156bipolar epicardial electrodes resulted in a rela-tively large recording area with a limited spatialresolution. Overall, Cox et al. demonstrated rightatrial reentrant circuits in 6 of 13 patients,whereas left atrial reentrant circuits could not bedetected. Konings et al.-- used a different tiiap-ping approach. A small epicardial electrode(3.6-cm diameter) containing 244 unipolar elec-trodes placed at the lateral right atrium resultedin high spatial resolution at the expense of over-all spatial sampling. Based on the spatiotemporalcomplexity of atrial activation during AF, threetypes of AF were defined. Patients were classi-fied according to the predominant activation pat-tern; therefore, the classification itrto dii'ierenttypes was part of a continuous spectrum.-- Takentogether, the multiple wavelet hypothesis of Moeet al.-^ could be confirmed in experimental--* andclinical-'-^ mapping studies, but the role of pref-erential pathways, periodicities, and other factorsfor the perpetuation of AF is still unclear.

778 Journal of Cardiovascular Electropbysiology Vol. 10, No. 6, June 1999

Cox et al.'** concluded that the reentrant cir-cuits of AF were so fleeting and unstable that themost accurate way to prevent AF was to interruptall of tbe potential pathways for right and leftatrial macroreentrant circuits. Therefore, theseinvestigators developed an extensive concept ofplacement of multiple atriai incisions in bothatria for curative treatment of AF. In their ap-proach, both atrial appendages were excised andthe pulmonary veins were isolated. In addition,multiple biatrial incisions were strategicallyplaced, first to interrupt potential reentrant cir-cuits and second to direct the impulse from tbesinus node area to the AV node.'"''' Tbis so-called maze procedure cured AF without tbeneed for antiarrhythmic medication in 58 of 65patients (89%) and with additional antiarrhyth-mic drugs in another 6 patients (9%) of theirstudy group.''' In patients undergoing only themaze procedure without previous cardiac surgery(n = 49), the cardiopulmonary bypass time mea-sured 184 minutes (range 130 to 256), and theaortic cross-clamping time was 69 minutes(range 50 to 102). More widespread applicationof the maze procedure may have been precludedby the time consuming nature of this operation.

Swartz and co-workers'"^ described the percu-taneous catheter application in a biatrial maze-like procedure with placement of multiple bia-trial linear lesions in a patient with chronic AF.Significant problems were encountered witb thelong procedure times and substantial risk of sys-temic thromboembolism. From a pathophysio-logic point of view, Swartz et al.'-'̂ attempted toreplicate the maze procedure by percutaneousmeans. In the same year, Haissaguerre et al."*reported successful catheter ablation with linearlesions restricted to the righr atrium in a patientwitb paroxysmal AF. Subsequently, Hais-saguerre et al.'^ induced linear right atrial lesionsof increasing complexity in 45 patients with par-oxysmal AF and achieved a success rate of 13%without antiarrhythmic drugs. Tbe success ratewas increased to 47% when additional linearlesions were applied within the left atrium.'^Recently, Gaita et al.-'' reported catheter ablationof idiopathic AF restricted to ablation lineswithin the right atrium.

The lesion geometries that have been appliedpercutaneously so far seem to be empirical, andno lesion geometry concept for percutaneous ap-plication has been validated. Therefore, interpre-tation of tbe moderately successful resultsachieved by percutaneous means is difficult for

several reasons. With current mapping and cath-eter technology, validation of the effects of RFenergy applications in terms of induction of con-tiguous transmurai lesion lines is highly prob-lematic. No stimulation techniques can be ap-plied in patients with chronic AF, i.e., validationof induction of complete conducfion block withstimulation techniques is not feasible. Even inpatients with paroxysmal AF who are treatedduring normal sinus rhythm, validation of leftatrial lesion lines is difficult and time consuming.Consequently, clinical failure of percutaneouslyinducted lesion lines may be the result of insuf-ficient lesion extent and/or geometry, or ablationfailure may be due to insufficient realization ofthe proposed lesion lines, i.e., the induction of/iW7contiguous lesions without achievement oflinear conduction block. Therefore, at this fime,percutaneous catheter abiation of AF is still aninvesfigational procedure.

Elimination of Anatomic Left Atrial "Anchor"Reentrant Circuits

Tbe data presented by Haissaguerre et al."and Giata et al.--'' indicate that less extensivelesion line concepts may be effective in somepatients with AF when compared witb the mazeoperation. Wijffels et al.-'' investigated the per-petuation and termination of AF in chronicallyinstrumented goats. In their experiments, termi-nation of AF was associated significantly with anincrease in cycle length and conduction velocityof the fibrillation wavelets resulting in a progres-sive widening of the excitable gap.-'' This mighthave resulted from replacement of small, func-tional, so-called random reentrant circuits bylarger anatomically determined circuits.-'* Fur-ther evidence for a critical role of anatomicallydefined reentrant circuits in AF was reportedrecently by Allessie's group^'' in a rabbit biatrialdilatafion model of AF. Tbe succe.s.sful elimina-tion of all potential anatomically defined circuitsinvolving the caval veins, the tricu.spid and mitralannuli, and the orifices of the coronary sinus andthe pulmonary veins resulted in an increase ofthe intra-atrial pressure required to sustain AFfrom 9.2 to > 20 cm H2O in most animals.-' Inaddition, Elvan et al.^" reported that epicardialRF catheter ablation of the atria in dogs reducedthe inducibility and duration of AF.

The crucial role of the left atrium for theperpetuation of AF already has been indicated bythe so-called left atrial isolation procedure.-'' Ad-

Kottkamp, et al. Intraoperative Ablation of Atrial Fibrillation 779

ditional evidence that the isolated right atrial freewall cannot sustain AF in some patients despiteaggressive stimulation techniques came fromKuck and co-workers (personal communication).In their study, percutaneous isolation of the rightatrial free wall with three RF-induced linear le-sion lines resulted in sinus rhythm of the rightatrium but perpetuation of AF of the left atrialfree wall and the interatrial septum. Recently,Sueda and co-workers^" reported their results ofAF operation in patients with mitral valve dis-ease using a combination of left atrial incisions,cryoablation, and excision of the left atrial ap-pendage. Using this approach, chronic AF wasreduced significantly or eliminated at discbargein 86% of their patients compared with 27% inthe control group.

In the present study, the operafion theater on oneside and RF energy application on the other sidewere combined to test a new concept of intraoper-ative RF ablation of chronic AF with a right atrial-trans.septal access to the left atrium. A left atrialapproach to intraoperative RF energy applicationwas applied that was restricted to eliminate leftatrial anatomically detennined macroreentrant "an-chor" circuits involving the pulmonary veins andthe mitral annulus. During a follow-up of 3 to 20months (mean II ± 6 ) , chronic AF was success-fully ablated in > 80% of the patients. Noteworthy,the patients of the pre.sent study were in chronic AFfor 4.3 ± 3.9 years and had significandy enlargedleft atria. In 6 of 11 patients, AF recurred during theearly postoperative period but was self-terminatingor could be cirrdioverted with transient sotalol ther-apy in 4 of these patients. Even after the mazeoperation, transient AF is not unusual within theearly postoperative peritxl and may be attributed loshoitcned effective refractory peritxls and the so-called sterile postoperative jiericarditis that tran-siently allows shorter functional reentrant circuitsto sustain over a limited time.'*^'"' In our study,transient recurrences of AF did not occur after theperioperative 4-week period. Experimental inve.sti-gations using our approach in freshly excised pigleft atria demonstrated the smooth induction ofcontiguous transmural lesion lines witbout endo-cardial disruption or perforation.-" Clinically, thesites of energy applicafion were well visualized aspale lesion lines; therefore, the contiguous nature oftbe applied lesion lines could be validated underdirect vision. As a result, the lesion geometiy thatwas tested in our study might now be transferred topercutaneous catheter abiation techniques as a val-idated concept.

Study Limitations and Outlook

In the present study, 12 consecutive patientswho fulfilled tbe inclusion criteria fomied the studypopulation. The percentage of patients who mighthave converted to .stable sinus rhythm after valvesurgery alone is unknown because no control groupexisted in our study. However, the patients were inchronic AF for a mean of 4.3 ± 3.9 years, the leftatria were significantly enlarged (mean 56 ± 7mm), and tbe patients were relatively old (mean67 ± 9 years); therefore, the possibility of achiev-ing stable sinus rhythm with valve surgery alonewas small. In addition, a similar patient populationwith valvular heart di.sease recently was investi-gated using a modified maze prtx^edure byKawaguchi et al.^' In their .study, sustained AF wassignificantly less frequent in the maze operationgroup (12% at 1 year) compared with the controlgroup (86%0, indicating that valvular surgei-y alonedid not lead to stable sinus rhythm in this compa-rable patient population.

Currently, no widespread application of a cur-ative treatment of AF exists. The extensive mazeoperation and its modifications have good re-sults, but they require very long operation times.At this time, percutaneous RF catheter ablationof AF is still an investigational approach, withhigh procedure and fluoroscopy times and a sig-nificant risk of systemic embolism. The pivotalquestion for curative treatment of AF is bow toacbieve the results of the maze procedure withthe lea.st damaging concept, i.e., the smallestnumber and extent of lesion lines. In the presentstudy, a pathophysiologic concept of eliminationof left atrial anatomic "anchor" reentrant circuitsachieved successful ablafion of AF in > 80% ofthe patients with a duration of antiarrhythmictreatment of only 19 ± 4 minutes. The "ideal"curative approach for treatment of AF might beto transfer the current concept to minimally in-vasive surgery without the need for stemotomyor to percutaneous catbeter ablation as soon ascatheter technology has improved sufficiently.

References

1. Jackman WM. Wang X, Friday KJ, et al: Catheterablalion of accessory atrioventricular pathways (Wolff-Parkinson-White syndrome) by radiofrequency current.N Engl J Med 1991:324:1605-1611.

2. Kuck KH. Schliiter M. Geiger M, et al: Radiofrequencycurrent catheter ablalion of accessory atrioventricularpathv^ays. Lancet 1991:337:1557-1561.

3. Jackman WM, Beckman KJ. McClelland JH, et al:

780 Journal of Cardiovascular Electrophysiology Vot. 10, No. 6, June 1999

Treatment of supraventricular tachycardia due to atrio-ventricular nodal reentry by radiofrequency catheterablation of slow-pathway conduction. N Engl J Med1992;327:3i3-318.

4. Haissaguerre M. Gaita F. Fischer B. et a!: Fliminationof atrioventricular nodal reenlranl tachycardia usingdiscrete slow potentials to guide application of radio-frequency energy. Circulation 1992,85:2162-2173,

5. Walsh EP. Saul PJ, Hulse EJ, et al: Transcatheter ab-lation of ectopic atrial tachycardia in young patientsusing radiofrequency current. Circulation 1992:86:1138-1146,

6. Lesh MD. Van Hare GF. Epstein LM. et al: Radiofre-quency catheter ablation ol' airial arrhythmias-—Resultsand mechanisms. Circulation I994;89:1074-1089.

7. Cosio FG, Lopez-Gil M, Goicolea A, et al: Radiofre-quency ablation of the inferior vena cava-tricuspidvalve isthmus in common atrial flutter. Am J Cardiol1993;71:7O3-7O9.

8. KJrkorijin G, Moncada E, Chevalier P, et al: Radiofre-quency ablation of atdal flutter: Efficacy of an anatomi-cally guided approach. Circulation 1994:9O:28(W-2814,

9. Olgin JE, Kalman JM. Fit/patrick AP, et al: Role ofright atrial endocardial structures as barriers to conduc-tion during human type I atiial flutter: Activation andentrainment mapping guided by intracardiac echocardi-ography. Circulation 1995:92:1839-1848.

10. Poty H, Saoudi N. Abdei Aziz A, et al: Radiofrequencycatheter ablation of type I airial flutter: Prediclion oflate success by electrophysiok)gical criteria. Circulation1995;92:I389-1392.

11. Cauchemez B, Haissaguerre M. Fischer B. el al: Elec-trophysiological effects of catheter ablation of inferiorvena cava-tricuspid annulus isthmus in common atrialflutter. Circulation l996;93:284-294.

12. Nakagawa H. Lazzara R, Khastgir T, ct al: Role of thetricuspid annulus and the eustachian valve/ridge onatrial flutter: Relevance to catheter ablation of the septalislhmus and a new technique for rapid identification ofablation success. Circulation 1996:94:407-424.

13. Kalman JM. Van Hare GF, Olgin JE. et al: Ablation of"incisional" reentrant airial tachycardia complicatingsurgery for congenital heart disease: Use of entrainmentto define a critical isthmus of conduction. Circulation1996:93:302-312.

14. Triedman JK, Jenkins KJ, Colan SD. el al: Inlra-alrialreentrant tachycardia after palliation of congenital heartdisease: Characterization of multiple maeroreentrantcircuits using fluoroscopically based three-dimensionalendocardial mapping. J Cardiovasc Electrophysiol1997:8:239-270.

15. SwarU JF, Pellersels G, Silvers J, ct al: A catheter-based curative approach to atrial fibrillation in humans.(Abstract) Circulation l994;90:I-335.

16. Haissaguerre M, Gencel L. Fischer B, et al: Successfulcatheter ablation of atrial fibrillation. J CardiovascElectrophysiol 1994:3:1045-1032.

17. Haissaguerre M, Jais P. Shah DC, et al: Right and left

atrial radiofrequency catheter therapy of paroxysmalatrial fibrillation. J Cardiovasc Electrophysiol I996;7:1132-1144.

18. Cox JL, Schuessler RB. D'Agostino HJ. et al: Thesurgical treatment of atrial fibrillation: 111. Develop-ment of a definite surgical procedure. J Thorac Cardio-vasc Surg 1991; 101:569-583.

19. Cox JL, Boineau JP, Schuessler RB, ct at: Five-yearexperience with the maze procedure for atrial tibrilla-tion. Ann Thorac Surg 1993:56:814-824.

20. Kottkamp H. Hindricks G. Borggrefe M. et al: Intraop-erative catheter-guided radiofrequency ablation of atrialfibrillation: Experimental studies on methodological as-pects, biophysical parameters and feasibility. (Abstract)Eur Heart J 1997:18(Suppl):320.

21. Cox JL, Canavan TE, Schuessler RB, et al: The surgicaltreatment of atdal fibrillation. II. Intraoperative eiectro-physiologic mapping and description of the electro-physiologic basis of atdal flutter and atdal fibrillation.J Thorac Cardiovasc Surg 1991; 101:40(v426.

22. Konings KTS. Kirchhof CJHJ, Smeets JRLM, el al:High-density mapping of electrically induced atdal fi-bdllation in humans, Cireulation I994;89:1665-168O.

23. Moe GK. Rheinboldt WC, Abildskov JA: A computermodel of atdal fibdllation. Am Heart J I964:67:2(K)-22O,

24. AUessie MA, Lammers WJEP, Bonke FIM, et al: Ex-pedmental evaluation of Moe's multiple wavelet hy-pothesis of atrial fibrillation. In Zipes DP. Jalife J, eds:Cantiac Etcctrophysiotogy ami Arrhythmias. Gruneand Stratton. New York. 1985, pp. 265-273,

23, Gaita F, Riccardi R. Calo L. et al: Atrial mapping andradiofrequency catheter ablation in patients with idio-pathic atrial librillation. Electrt)physiological findingsand ablaiion results. Circulation 1998:97:2136-2145.

26, Wijffels MCEF. Kirchhof CJHJ. Doriand R. et al:Atdal flbdllation begets atdal fibdllation, A study inawake chronically instrumented goats. Circulation1995;92:t954-1968.

27. Zarse M. Deharo J, AUessie MA: Radiofrequency ab-lation of anatomical atria! circuits in a rabbit model ofatdal fibditation. (Abstract) Circulation l996;94:l-676.

28, Eivan A. Pride HP, Eble JN, et al: Radiofrequencycatheter ablation of the atda reduces inducibility andduration of atdal fibrillation in dogs. Circulation 1993;91:2233-2244.

29, Williams JM, Ungerleider RM, Lofland GK. et al: Leftatrial isolation: A new technique for the treatment ofsupraventricular arrhythmias. J Thorac Cardiovasc Surg1980:80:373-380.

30. Sueda T. Nagata H. Odhashi K, et al: Efficacy of asimple left atdal procedure for chronic atrial flbdllationin mitral valve operations. Ann Thorac Surg 1997;63:1070-1073.

31. Kawaguchi AT, Koskai Y, Sasako Y. et al: Risks andbenefits of combined maze procedure for airial flbdlla-tion associated wilh organic heari disease. J Am Ct)llCardioi 1996:28:985-990.