763

Right Atrial Compartmentaiization Using RadiofrequencyCatheter Ablation for Management of Patients with

Refractory Atrial Fihrillation

ASHOK GARG, M.D.. WILLIAM FINNERAN, M.D., MICHAEL MOLLERUS, M.D.,ULRIKA BIRGERSDOTTER-GREEN, M.D., OSAMU FUJIMURA, M.D.,

LINDA TONE, R.N.. and GREGORY K. FELD, M.D.

From the Electrophysiology Program, Division of Cardiology, Department ol" Medicine,Utiiversity of California, San Diego, California

Ablation of Atrial Fibrillation, introduction: Atrial fibrillation (AF) is often refractory toantinrrhythmic drugs, and patients who are intolerant of AF may require the maze operationfor cure. As a less Invasive alternative, a catbeter-based, right atrial compartmentaiizationprocedure was evaluated.

Methods and Results: Twelve patients with AF refractory to Class I and III antiarrhythmicdrugs were studied. Four linear ri}̂ ht atrial radiofrequency ablations were performed, fromsuperior to inferior vena cava in the posterior wall and interatrial septum, anteriorly from thesuperior vena cava to the tricuspid annulus through the appendage, and across the tricuspidvalve-inferior vena cava isthmus. The radiofrequency catheter was dragged along each linethree to four times, until the atrial electrogram amplitude decreased hy 75% and there wasbidirectional condnction block in the tricuspid valve-inferior vena cava isthmus. One compli-cation occurred: sinus node dysfunction requiring a pacemaker. Eight patients were dischargedfrom the hospital on no antiarrhythmic drugs, and four were discharged on previously ineffec-tive antiarrhythmic drugs. Total duration of follow-up was 21.3 ± 11.2 months. Four patientsdischarged on previously ineffective antiarrhythmic drugs had no recurrence of AF. One patientdischarged off antiarrhythmic drugs had no recurrence of AF. Seven patients discharged offantiarrhythmic drugs had recurrent AF by 12.6 ± 13.0 months (median 6, range 1 to 39); 3 ofthese 7 responded to previously ineffective antiarrhythmic drugs without further AF and 4 didnot. Thus, 8 of 12 patients (67%) had suppression of AF after ahlation on previously ineffectivemedication or no medication.

Conclusion: Right atrial compartmentaiization may alter the substrate for AF, thus improv-ing the efficacy of previously ineffective antiarrhythmic drugs. Because it is relatively safe, itmay he a reasonable adjunctive intervention to maintain sinus rhythm in patients with drug-refractory AF. (J Cardiovasc Eleclrophysiol, Vol. 10. pp. 763-771. June 1999)

arrhythmia, atrial fibrillation, catheter ablation

fntroduction

Atrial tibrillation (AF) is an increasingly prev-alent cardiac arrhythmia.'- When uncontrolled.

Address for corrcspiindence: Gregory K. Feld, M.D., Electrophys-iology Program, Division ol" Ciirdiology, Department of Medicine.University of California. 2()0 W. Arbor Dr., San Diego, CA 92103.Fax: 619-543-7418.

Manuscript received 21 December 1998; Accepted for publication26 February 1999,

AF is associated with significant morbidity, es-pecially tbe risk of stroke secondary to thrombo-embolism. Although unproven. suppression ofAF is still widely accepted as tbe preferred treat-ment, in contrast to rate control and long-termanticoagulation with warfarin.'-^ However, be-cause antiarrhythmic drugs often are ineffectivein preventing recurrence of AF and may causeserious side effects or increase mortality due toprourrhythmia.,'^'' alternative treatments for pa-

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

tients with AF bave been developed, includingAV nodal ablation or modification for ventricularrate control only,^ *" and tbe maze and otber com-partment operations for cure of AF."*" Altbougbtbe maze operation is curative, tbe fact that it isan open beart operation makes it inappropriate orunacceptable for many patients witb AF."*"

Tbere bave been a few recent reports in botbexperimental animal models of AF and in bu-mans'- '̂ on tbe use of radiofrequency catheterablation (RFCA) for making multiple linear en-docardial lesions to emulate tbe incision lines intbe surgical maze procedure. In clinical studies,variable success bas been reported, with com-plete suppression of AF or a reduction in thefrequency of AF episodes in 30% to 60% ofpatients.''*-'"' However, some investigators baveproposed that linear lesions in tbe left atrium areessential for cure of AF.'"" Unfortunately, tbere isa greater potential for complications witb linearRFCA in tbe left atrium, including tbe proceduralrisks associated witb transseptal catbeterizationand tbe risk of tbromboembolic stroke.''*'*

To test tbe bypotbesis that tbe right atriummay be important in initiating or sustaining AF,we performed linear RFCA in tbe rigbt atrium ina series of patients witb AF refractory to Class Iand/or III antiarrhythmic drugs.'^ Our aim was tomodify tbe substrate for AF to prevent its recur-rence, witb <ir wilbout adjunctive antiarrbytbmicdrug therapy., while minimizing tbe risks of pro-cedural complications such as tbromboembolicstroke.

Methods

Patient Characteristics

Tbe study population consisted of 12 patients(10 men and 2 women; mean age 56 ± 11 years)referred from September 1995 to June 1998 formanagement of refractory paroxysmal, persis-tent, or chronic AF. Patient characteristics arelisted in Table I. Tbe time since first onset ofsymptomatic AF was 5.3 ± 3.4 years. Tbreepatients bad underlying coronary artery disease:two bad reduced leit ventricular function (ejec-tion fraction 30% to 40%) and one bad byper-tensive beart disease witb left ventricular byper-tropby but normal left ventricular systolicfunction. Tbe remaining eight patients bad nostructural beart disease. Left atrial si/e was nor-mal (< 4 cm) in eigbt patients and mildly en-larged (> 4 but < 5 cm) in four patients. Allpatients had recurrent AF despite having re-ceived a mean of 3.2 ± 0.9 Class I and/or illantiarrhythmic drugs, including amiodarone insix patients. All patients had symptomatic AFdocumented by ECG, 24-bour ambulatory ECG,or transtelepbonic event monitor witbin tbe 6montbs prior to entry into tbe study. Prior toablation. I patient bad daily episodes, 2 badweekly episodes, and 5 bad montbly episodes ofparoxysmal AF (i.e., self-terminating AF. last-ing < 24 bours per episode). Tbree patients badone or more episodes of persistent AF (sustainedAF lasting 24 to 72 bours and requiring direct

Pt.No.

123456789

101112

Patient Demographics and Results of Right

Age(years)

754450576451516463424670

56 ± 1 1

Sex

MMMMMMFMMMFM

CVI)

CADNoneNoneCADCADNoneNoneHTNNoneNoneNoneNone

Time sinceFirst AF(years)

12253334

10633

105.3 ± 3.4

Type ofAF

ChronicParoxysmalPersistentParoxysmalPersistentParoxysmalParoxysmalParoxysmalParoxysmalPersistentParoxysmalParoxysmal

TABLE 1Atrial Linear

AFFre<|Ut'ncv

Chronic> 1/months 1/6 month> I/week2^1/6 month> 1 /week>l/month>l/month>l/month> 1/6 month>i/day>!/month

Radiofrequency Catheter Ablation during

No. A ADFailed

233433434144

3.2 i 0.9

Amio(Y/N)

NNYNYYNYYNYN

Posl-RFFollow-Up(nionth.s)

29343231302823

7201065

21.3 ± 1 1 . 2

AFRecur(Y/N)

NYYNYNYNYYYN7/5

Long-Term

Time toAF Rtvur(months)

—

618—

3—

I—

661

—12.6 ± 13.0

FoUow-Up

AF unA AD*(Y/N)

NYNNNNYNYNYN4/8

A AD*Post-RF

SotalolSotalolAmioSotalolAmioAmioAmioAmioSotalolSoialolAmio/PANone

AAD = Ciuss I i)r ill antiarrhythmic drug: AF on AAD = recurrence or lack of recurrence of atrial librillation while taking a previouslyineffective antiarrhythmic drug; Amio = amiodarone: CAD = coronary artery disease: CVD = cardiovascular disease: HTN = hypertension;PA = pri>cainamide: Persistent = AF requiring cardioversion to convert to normal sinus rhythm; Post-RF follow-up = duration oi" follow-upafter ablation: Time since first AF = lime since first documented episode of atrial fibrillation: Time to AF Recur = time after ablation thatpatients discharged on no antiarrhyihmic drugs had recurrence of atrial fibrillation.* Previously ineffective AAD.

Garg. et al. Ablation of Atrial Fihrillation 765

current [DC] cardioversion) in the 6 months priorto ablation, and one patient had heen in chronicAF for 12 years (sustained AF lasting more than72 hours and/or unresponsive to cardioversion).Patients were excluded if they had evidence ofleft atrial thrombus on transthoracic or trans-esophageal echocardiogram, marked left atrialLMilargement (> 5 cm), pulmonary arterial hyper-tension, severe tricuspid regurgitation, severemitral or aortic valve disease, or any other car-diac condition requiring open heart surgery in thenear future. The right atrial compartmentaliza-tion procedure was offered as an alternative toAV nodal ablation with pacemaker implantation,AV node modification., or the surgical maze op-eration. Informed consent was obtained under aprotocol approved by the institutional reviewboard of the University of California at SanDiego.

Electrophysiologic Study

All Class I or III antiarrhythmic drugs werediscontinued 5 to 7 days prior to ablation, withthe exception of amiodarone, which was beingtaken by six patients prior to ablation. All pa-tients had been taking warfarin with an interna-tional noimalized ratio (INR) in the 2 to 3 rangefor at least 4 weeks prior to the study. Warfarinwas stopped 3 days prior to the study to ensure anormal INR at the time of ablation. If heparinwas being administered, it was stopped 6 hoursprior to the ablation. Conscious sedation wasadministered by continuous infusion of propofolor intermittent doses of intravenous fentanyl andmidazolam. Intravenous sheaths (6 to 8 French)were placed using Seldinger percutaneous tech-nique in the right internal jugular vein and theright and left femoral veins. An electrophysi-ologic study with atrial and ventricular pro-grammed stimulation was performed in sinusrhythm in all 11 patients with paroxysmal orpersistent AF, except in one patient whose refer-ring physician had already performed an electro-physiologic study. The purpose of the baselineelectrophysiologic study was to rule out any cur-able supraventricular tachycardia (AV nodal re-entrant tachycardia, AV reentrant tachycardia, orfocal atrial tachycardia) that might be triggeringparoxysmal AF. In the one patient presenting inchronic AF. baseline electrophysiologic testingwas not performed. For the electrophysiologicstudy, a 20-pole Halo"̂ "̂ catheter (Cordis-Web-ster. Inc., Baldwin Park, CA. USA) was posi-

tioned in the right atrium around the tricuspidvalve annulus, a decapolar catheter (Daig, Inc.,Minnetonka, MN, USA) in the coronary sinus,and two quadripolar catheters (Daig, Inc.) at theHis bundle and right ventricular apex. The sur-face 12-lead ECG and al! endocardial bipolarelectrograms were recorded simultaneously us-ing 30- to 500-Hz bandpass filters on a mul-tichannel digital recorder and stored to opticaldisk drive (Bard Electrophysiology, Inc., Tewks-bury, MA, USA).

Right Atrial Compartmentalization

Following placement of diagnostic cathetersand baseline electrophysiofogic study, the pa-tients underwent right atrial compartmentaliza-tion using a 7-French. steerable ablation catheterwith a 4- or 5-mm ablation electrode with ther-mistor for temperature monitoring (EP Technol-ogies, Inc., Mountain View, CA, USA). The ab-lation catheter was coupled to an EPT 1000 (EPTechnologies, Inc.) radiofrequency (RF) energygenerator. RF energy up to 50 W wa.s applied inan automatic temperature-controlled manner toachieve a constant tissue temperature of 50° to60°C. The ablation catheter was positioned in theright atrium through a standard I5-cm. 8-Frenchvenous sheath from the right femoral vein. Underfiuoroscopic guidance, four linear RF lesionswere created by dragging the ablation catheter 3to 5 mm every 30 to 60 seconds during contin-uous RF energy application (Fig. I). The ablationcatheter was dragged 3 to 4 times over each ofthe four predefined lines to produce a continuouslinear lesion. Linear lesion continuity was con-firmed following ablation by a reduction of themaximum observed atrial electrogram amplitude(bipolar), recorded at the distal ablation elec-trode, by > 75% along the entire length of thelesion. Following ahlation, the one patient withchronic AF underwent DC cardioversion to re-store sinus rhythm. In addition, bidirectionalconduction block across the tricuspid valve-infe-rior vena cava (TV-IVC) isthmus was confirmedby pacing at the low right atrium and coronarysinus ostium, as previously described for ablationof type 1 atrial flutter.''' Attempts were made toinduce atrial flutter with burst pacing cyclelengths down to 2:1 atrial capture from the lowright atrium and proximal coronary sinus. How-ever, no systematic attempt was made to induceAF. Patients then were managed in a standardmanner, including observation overnight under

766 Juurnal of Cardiovascular Electrophysiology Vot. 10, No. 6, June 1999

Septal

Figure 1. Schematic diagram of the right atrium, with thefree wall opened to depict the locations of the four linearradiofrequency ahlation lesions in the right atrium, includ-ing the posterior, septat. anterior, and isthmus tesions. Notethat the superior portion of the anterior tesion (dotted line)extetiding from the right atrial appendage to the superiorvena cava was performed only in the first seven patients andthen abandoned due to sinus node injury in one patient.

continuous telemetry monitoring. Patients weredischarged from the hospital when they werestable, usually within 24 hours following abla-tion.

Follow-Up

Four patients (1. 4. 6. 8) were discharged on aClass I or III antiarrhythmic drug that previouslywas ineffective in suppressing AF. Eight patients(2, 3. 5. 7. 9. 10.. 11, 12) were discharged on noantiarrhythmic drug,, at the discretion of the pa-tient's primary physician or the investigator.However, patients were asked to continue war-farin anticoagulation for at least 1 month follow-

ing discharge. Patients were instructed to reportimmediately any adverse events following dis-charge and to retum to the clinic or emergencydepartment for any symptoms suggesting recur-rence of AF. No attempts were made to docu-ment asymptomatic AF during follow-up, be-cause all patients had markedly symptomaticepisodes prior to ablation. All patients were seenin the arrhythmia clinic 1 month after ablationand at 3- to 6-month intervals thereafter in thearrhythmia clinic or at the office of their referringphysician. At each clinic visit, a history of ar-rhythmia symptoms was taken, a physical exam-ination was performed, and an ECG was ob-tained.

Results

Acute Results of Ablation

Right atrial compartmentalization was at-tempted in 12 patients. In all patients except theone patient with chronic AF. linear RFCA wasperformed in sinus rhythm. The posterior, septal.and anterior linear lesions were performed in all12 patients, although the right atrial appendage tosuperior vena cava portion of the anterior lesionwas only performed in tbe first seven patientsand subsequently abandoned due to sinus nodedamage in one patient (Table 2). The TV-IVCisthmus lesioti was performed in only ten pa-tients, because two previously had successfulatrial flutter ablation with persistent bidirectionalisthmus block. Total procedure time was 283 ±104 minutes (median 285, range 120 to 525),with a mean fluoroscopy time of 137 ± 48 min-utes (median I 10. range 67 to 226). The averageduration of RF energy application required tocomplete all linear ablation lines in each patient

Procedure durationFluoroscopy durationDuration of radiofrequency energy applicationSVC-IVC (posterior) lesionSVC-IVC (septLiI) lesionRAA-TVA/RAA-SVC (atiterior) lesionIVC-TVA isthmus lesionImpedance rises per patientComplications: sinus lurest

TABI.K 2Procedural Parameters (N = 12)

Mean ± 1 SI)

283 ± 104 mill137 ± 48 min53 ± 17niio

1212

12/710*

0.50 ± 0.66!

Mfdmn

2,S5 nun110 min54 min

———.———

Kanj;e

l2U-525iiiin67-226 min22-89 min

——————

IVC = inferior vena cava: RAA = right atriat appendage: SVC = superior vena cava: TVA = tricuspid valve annulus.* Two patietils had previously undergone IVC-TVA isthmus ahlation tor atrial flutter.

Garg. et al. Abtation of Atrial Fibrillation 767

was 53 ± 17 minutes (median 54, range 22 to89). By automatically maintaining temperature inthe 50" to 60°C range at all times, impedancerises with coagulum formation requiring catheterremoval and cleaning occurred in only five pa-tients: two episodes in one patient and one epi-sode in four patients, for an average of only0.50 ± 0.67 episodes per patient. Following ab-lation, there was marked (> 75%) reduction inamplitude and fractionation of electrogramsalong the entire length of the posterior, septal,and anterior linear lesions. In addition, bidirec-tional conduction block at tbe TV-IVC isthmuswas documented in all 12 patients, AF did notterminate in the patient with chronic AF duringablation, and DC cardioversion was required torestore sinus rhythm. However, an episode of AFinduced during catheter manipulation in one pa-tient with a history of paroxysmal AF convertedto type I atrial flutter during completion of tbeposterior, septal, and anterior linear lesions, andthen to sinus rhythm during completion of theTV-IVC isthmus lesion (Fig. 2),

permanent pacemaker implantation developedacutely in one patient (8). This complicationoccurred while perfonning the appendage to su-perior vena cava portion of the anterior lesion.Therefore, the superior portion of the anteriorlesion was abandoned, while the appendage totricuspid annulus portion of the anterior lesioncontinued to be performed, with no further inci-dence of sinus node damage. There were no otherprocedural complications or any long-term com-plications during follow-up in any patient.

Discussion

This study suggests that right atrial compart-mentaiization by linear RFCA improves mainte-nance of sinus rhythm in some patients with AFpreviously refractory to Class I and/or III antiar-rhythmic drugs. During long-tenn follow-up. AFwas suppressed in 67% of patients followingright atrial compartmentaiization, alone or incombination with previously ineffective antiar-rhytbmic drugs.

Effects of Ablation on AF Recurrence DuringLong-Term Follow-Up

The mean duration of follow-up for all 12patients was 21.3 ± 11.2 months (Table 1). Fourpatients (I.. 4, 6, 8) discharged from tbe hospitalon previously ineffective antiarrhythmic drugsand one patient (12) discharged without antiar-rhythmic drugs had no recurrence of AF duringfollow-up. Seven patients (2, 3, 5, 6, 7, 10, 11)initially discharged from the hospital withoutantiarrhythmic drugs had recurrent symptomaticAF. The mean time to recurrence was 12.6 ±13.0 months (median 6, range I to 39). Of these.seven patients, three (3, 5, 10) responded to aClass I or III antiarrhythmic drug that previouslyhad been ineffective without further episodes ofAF, and four patients (2, 7, 9, II) had recurrentAF despite antiarrhythmic drug tberapy. Overall,8 of 12 patients (67%) ultimately bad completesuppression of symptomatic AF after ablation. 7on a previously ineffective antiarrhythmic drugand I on no antiarrhythmic drugs. There was nocorrelation between recurrence of AF and leftatrial size or left ventricular function.

Complications

During the rigbt atria! compartmentaiizationprocedure, sinus node dysfunction requiring

Right Atrial CompartmentaiizationProcedure Design

Although previous studies used I to 7 linearlesions for right atrial ablation of AF,'"^ '"̂ the 4linear lesions used in this study were designed tointerrupt all potential reentrant circuits aroundanatomic or functional obstacles, including supe-rior and inferior venae cavae, crista tcrminalis,fossa ovalis. atrial appendage, and tricuspidvalve annulus. It was hoped that these four linearlesions would prevent the propagation of trigger-ing arrhythmias and reentrant wavelets in theright atrium tbat might initiate and/or sustain AF,while having no effect on conduction from thesinus node to tbe AV node. However, it also wasexpected that right atrial lesions alone might notprevent recurrence of AF, unless combined withClass I or III antiarrhythmic drugs.

Comparison of Results of Present Studies withPrevious Studies on Linear RFCA for TreatmentofAF

Left atrial ablation was not performed in thisstudy because of the limitations of currenl cath-eter technology and the risk of serious compli-cations such as thromboembolic stroke and pul-monary vein stenosis.''' "'*̂ *' A small number ofstudies of right and/or lefl atrial ablation have

768 Juurnal of Cardiovascular Electrophysiology VoL 10, No. 6, June 1999

Figure 2. (A) Twelve-lead ECGs from one patient who developed AF during the diagnostic electrophysiologic study. (B)During comptetion of the posterior, septal. and anterior at?tation lesions. AF was converted to type 1 atrial fiutter. The atrialfiutter subxequentty amverted to normat sinus rhythm during comptetion of the tricuspid valve-inferior vena cava isthmustesion (not recorded).

been performed, however, demonstrating the fea-sibility of linear RFCA to reduce the frequency,or even cure. AF.'-* '̂ However, in the.se previousstudies, right atrial linear RFCA alone was foundto prevent recurrence of AF in < 50% of pa-

tients.'4-''' These studies enrolled similar patientsand used similar success criteria, but the durationof follow-up was shoner than in the presentstudy.i^-'^ Therefore, the results of this studyappear to be slightly better than those of previous

Garg, et at. Ablation of Atrial Fibrillation 769

studies, with long-term suppression of AF in67% of patients undergoing right atrial compart-mentaiization alone or in combination with pre-viously ineffective Class I or III antiarrhythmicdrugs. Although no .statistical comparison can bemade between the results of this study and thoseof previous studies, it is possible that the slightlybetter results observed in this study may be re-lated to the placement of the linear lesions toincorporate all potential anatomic and functionalobstacles in the right atrium that might form areentrant circuit.

Possihie Mechanisms of the Antiarrhythmic Effectof the Right Atrial CompartmentaiizationProcedure

The niechanism(s) of improvement in AF con-trol by right atrial compartmentaiization alone is(are) uncertain. However, the present under-standing of the electrophysiologic basis of AFsuggests several possible mechanisms based onablation-induced changes in atrial electrophysi-ologic substrate. AF has been shown to be due tomultiple reentrant wavelets-' and to require ashort tissue wavelength and critical mass of tis-sue for its initiation and maintenance.^^ Initiationand perpetuation of AF also may be partly de-pendent on the many anatomic and functionalobstacles that may exist in the right atrium.-^ Inthe presence of the necessary functional eiectro-physiologic abnormalities and anatomic obsta-cles, a triggering arrhythmia such as prematureatrial depolarizations, foca! atrial tachycardia, oratrial flutter may then initiate and maintain AF.The right atrium also may be a site for such atriggering arrhythmia.-* Therefore, it is possiblethat right atriai compartmentaiization alone, or incombination with antiarrhythmic drugs, may pre-vent triggering arrhythtnias from initiating ormaintaining AF by preventing propagation of theinitiating wavefront, preventing reentry aroundanatomic obstacles, and/or reducing the criticalmass of tissue required for multiple wavelet re-entry.

Vagal denervation, which would prevent va-gaily mediated bradycardia and shortening ofatrial refractory period, has been proposed as amechanism of prevention of recurrent AF fol-lowing right atrial ablation procedures.'^-^^ It ispossible that the linear ablations perfonned inthis study could have affected vagal innervationto the right atrium, but this was not evaluateddirectly. However, indirect measures of vagal

innervation, such as sinus rate, were unalteredduring or following ablation, suggesting that thismechanism of prevention of recurrent AF is un-likely.

Another possible mechanism that has beenproposed for the suppression of AF by atrialablation is a reduction of connexin43, the inter-cellular gap junction protein, in areas of atria thathave been ablated.-'' A reduction of connexin43by ahlation may produce antiaiThythmic elTectsby altering atrial myocardial conduction.-'' aswas shown in one canine study in which rightatrial lesions alone prevented induction of AFin > 50% of the dogs studied. However, confir-mation of this mechanism in vivo currently is notpossible in humans.

Safety of the Right Atrial CompartmentaiizationProcedure

By using careful technique, including temper-ature feedback control during ablatit)n, no seri-ous complications occurred during this study,other than a potentially avoidable sinus nodeinjury in one patient. However, the average pro-cedure and fluoroscopy times in this study werelong, although they were shorter than those re-quired for linear ablation of" both the right andleft atria.'**'̂ The risk of such long fluoroscopyexposure times to patients and staff is unceriain.but it may be significant. Thus, future researchneeds to be done using new catheter ablationsystems capable of creating long linear lesionsmore quickly and more effectively or anatomicmapping systems that can precisely localize cath-eter position without using fluoroscopy. to re-duce both total procedure and fluoroscopy times.

Limitations of the Study

One possible limitation of this study is thatonly symptomatic episodes of AF were docu-mented. It is a well-known phenomenon thatsome patients with AF are asymptomatic duringa significant percentage of episodes documentedby ambulatory ECG monitoring. However, thepatients in this study were all highly symptom-atic during AF. so it is likely that they would, andin fact did, recognize when they had a recunenceof AF, as documented by ECG in all instances atthe time of symptomatic recurrence.

Another limitation of this study is that therewas no parallel control population, such as amedically or surgically treated group of patients

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

with refractory AF. Instead, eacb patient wasconsidered his or her own historic control. Aparallel, drug-treated control group would havebeen desirable, particularly considering the smallnumber of patients entering this study, but thoseenrolled bad previously failed most available an-tiarrhythmic drugs, and they would not havemade appropriate comparative controls. It can beargued that there may be random fluctuation inthe severity and frequency of AF over time, andwithout a contro! group we cannot rule out purechance in tbe observed improvement in AF fre-quency in these patients. However, because im-provement was seen over a long follow-up pe-riod in a significant number of patients, most ofwhom had frequent episodes of AF prior to ab-lation, the observed benefits of right atrial linearRFCA are likely to be real and not due to chancevariation in AF frequency or a placebo effect.

Clinical hnplieations of the Study

The results of the present study suggestthat right atrial compartmentalization by linearRFCA may provide an adjunctive therapeuticoption for the treatment of AF refractory to an-tiarrbythmic drugs and thus may deserve furtherlarge-scale study. However, the present data alsosuggest that right atrial linear RFCA alone is notsufficient to cure AF in most patients.

References

1. Wcilcns HJJ: Alrial librillalion. the last big hurdle intreating supraventricular tachycardia. N Engl J MedI984-.331:944-945.

2. Kannel WB. Abbol RD. Savage DD. et ai: Epidemio-logic features of chronic atrial fibrillation: The Fra-mingham study. N Engl J Med 1982;.1()6:IOI8-1O22.

3. Stn)ke Prevention in Atrial Fibrillalion Study GroupInvestigators: Wartarin versus aspirin for prevention ofthroniboembulism in atrial fibrillation. Stroke Preven-tion in alrial fibriiiation 2 study. Lancet 1994:343:687-691.

4. NHLBi AFFIRM Investigators: Atrial fibdllation fol-low-up investigation of rhythm management—THEAFFIRM study design. Am J Cardiol i997;79:li98-1120.

5. Cardiac Arrhythmia Suppression Trial (CAST) Inves-tigators: Preliminary report: Effect of encainide andflecainide on mortality in a randomized trial of arrhyth-mia .suppression alier myocardial infarction. N EngiJ Med 1989;32i:406-4i2.

6. Fiaicer GC. Biackshear JL. McBride R. et ai: Antiar-rhythmic drug therapy and cardiac mortaiity in atrial

fibrillation. The Stroke Prevention in Atrial FibrillationInvestigators. J Am Coll Cardiol 1992:20:527-532.

7. Feld GK, Fleck RP, Fujimura O. et al: Control of rapidventricuiar response by radiofrequency catheter modi-fication of the atrioventricular node in patients withmedicaily refractory atrial fibrillation. Circulation1994:90:2299-2307.

8. Feld GK: Radiofrequency catheter ablation versu.s mod-ification of the A V node tor control of rapid ventricularresponse in atrial fibrillation. J Cardiovasc Electro-phy.siol 1995:6:217-228.

9. Morady F. Hasse C, Strickberger SA. et al: Long-termfollow-up after radiofrcquenL'y modification of theatrioventricular node in patients with atrial fibrillation.J Am Coll Cardioi i997;29:l 13-121.

10. Cox JL. Boineau JP, Schuessler RB, et al: Five yearexperience with Maze procedure for atrial fibrillation.Ann Thorac Surg 1993:56:814-824.

11. Shyu KG. Cheng JJ. Chen JJ. et al: Recovery of atrialfunction after compartment operation for chronic atrialfibrillation in mitral valve disease. J Am Coll Cardioi1994:24:392-398.

12. Eivan A. Pride HP, Ebie JN, et al: Radiofrequencycatheter ablation of the atria reduces inducibliity andduration of atrial Hbriiiation in dogs. Circulati(»n 1995:91:2235-2244.

13. Mitcheii MA, McRury ID. Haines DE: Linear atrialablations in a canine model of chronic atrial fibriltation.Circulation i 998;97:1176-1 i 85.

14. Swanz JF. Pellerscls G, Silvers J. et al: A catheterba.sed curative approach to atriai fibriiiation in humans.(Abstract) Circuiation 1994;90:i-335.

15. Man KC. Daoud F., Knight B. et ai: Rigiit atriai radio-frequency catheter ablation of paroxysmal atriai fibrii-iation. (Abstract) J Am Colt Cardioi i996:27:188A.

16. Haissaguerre M, Jais P. Shah DC, et ai: Right and leftatrial radiofrequency catheter therapy of paroxysmaiatrial libriiiation. J Cardiovasc Eiectrophysioi i 996:7:Ii32-li44.

17. Gaita F. Riccardi R. Calo L. et al: Atrial mapping andradiofrequency catheter ablation in patients witii idio-pathic atriai fibriiiation. Circuiation i998;97:2136-2i45.

18. Feid GK. Birgersdotter-Green U. Fujimura O. et al:Radiofrequency catheter ablation of the right atrium forcontroi of athai libriiiation refractory to anti-arrhythmicdrugs. (Abstract) J Am Coll Cardiol 1997:29:176A.

19. Poty H, Saoudi N, Aziz A A, et al: Radiofrequencycatheter abiation of type 1 atriai flutter. Prediction oflate success by electrophysiologic criteria. CircuiationI995;92:i389-1392.

20. Robbins IM, Colvin EV, Doyle TP, et al: Pulmonaryvein stenosis after catheter ablation of atrial fibriiiation.Circulation i99K;98:1769-1 775.

21. Moe GK: On the niuitiple wavelet hypothesis of atriaifibrillation. Arch int Pharmacodyn 1962: i40:183-i88.

22. Garrcy WE: Auricular hbriiiation. Physiol Rev i924;4:215-250.

Garg, et al. Ablation of Atrial Fibrillation 77 i

23. Aiiessie MA, Konings K, Kirchhof CJ. et ai: Electro- iological effects of maze abiation procedure on caninephysioiogic mechanisms of perpeluation of atrial fibrii- sinoatriai node iunction. PACE i992:i5:2084-209Llation. Am J Cardiol I9%;77:IOA-23A. 26. Eivan A. Huang XD. Pressier ML. et ai: Radiofre-

24. Jais P. Haissaguerre M, Shiih DC. et ai: A focai source quency catheter abiation of the atria eliminates pac-of atriai fibrillation treated by discrete radiofrequency ing induced sustained atriai fibrillation and reducesablation. Circulation 1997:95:572-576. connexin 43 in dogs. Circuiation 1997:96:1675-

25. McLoughiin DE, Blitz A, Simmons J. et al: Eiectrophys- 1685.