lpu�ww

CHHBD2

C

MppcDDNwt

0d

Meta-Analysis of Ablation of Atrial Flutter andSupraventricular Tachycardia†

Peter Spector, MDa,*, Matthew R. Reynolds, MD, MScb, Hugh Calkins, MDc,Manu Sondhi, MD, MBAd, Yingxin Xu, PhDd, Amber Martin, BSd, Catherine J. Williams, MPHd,

and Isabella Sledge, MD, MPHd

The purpose of this study was to perform a systematic review and meta-analysis to evaluatethe safety and efficacy of radiofrequency ablation (RFA) of typical atrial flutter (AFL) andatrioventricular node–dependent supraventricular tachycardia (SVT) in adult patients.Medline and EMBASE were searched (1990 to 2007) for all study design trials of RFA.Data relating to single- and multiple-procedure success, arrhythmia recurrence, repeatablation, adverse events, and death were extracted. For RFA in AFL, 18 primary studieswith 22 treatment arms and 1,323 patients were identified. Single-procedure success forAFL was 91.7% (95% confidence interval [CI] 88.4% to 94.9%). Multiple-procedure successwas 97.0% (95% CI 94.7% to 99.4%). Postablation arrhythmia was noted in 13.2% ofpatients (95% CI 7.5% to 18.9%), while repeat ablation was reported in 8% (95% CI 4.5%to 11.4%). For RFA of SVT, 39 primary studies with 49 treatment arms in 7,693 patientswith accessory pathways and atrioventricular nodal reentrant tachycardia were identified.Single-procedure success for SVT was 93.2% (95% CI 90.8% to 95.5%). Multiple-proceduresuccess was 94.6% (95% CI 92.4% to 96.9). Postablation arrhythmia was noted in 5.6%patients (95% CI 4.1% to 7.2%). Repeat ablation occurred in 6.5% (95% CI 4.7% to 8.3%).For AFL studies, all-cause mortality was 0.6%, and adverse events were reported in 0.5%of patients. For SVT studies, all-cause mortality was 0.1%, and adverse events werereported in 2.9% of patients. In conclusion, studies of RFA for the treatment of patientswith AFL and SVT report high efficacy rates and low rates of complications. © 2009

Elsevier Inc. All rights reserved. (Am J Cardiol 2009;104:671–677)

etryrtt

M

dAPsaafitdathcssr

Despite high reported success rates, radiofrequency ab-ation (RFA) in patients with atrial flutter (AFL) and su-raventricular tachycardia (SVT) in clinical practice may benderused. For example, in 2006, RFA was performed in25% of United States hospital admissions in which AFLas the first-listed diagnosis.1 The objective of this studyas to identify, evaluate, and summarize the best available

aUniversity of Vermont College of Medicine and Fletcher Allen Healthare, Burlington, Vermont; bBeth Israel Deaconess Medical Center andarvard Clinical Research Institute, Boston, Massachusetts; cThe Johnsopkins University School of Medicine, Baltimore, Maryland; and dUnitedioSource Corporation, Lexington, Massachusetts. Manuscript receivedecember 16, 2008; revised manuscript received and accepted April 21,009.

This study was funded by Biosense Webster, Inc., Diamond Bar,alifornia.

*Corresponding author: Tel: 802-847-4539; fax: 802-847-5176.E-mail address: Peter.Spector@vtmednet.org (P. Spector).

† Conflicts of interest: Dr. Spector is a consultant for Medtronic, Inc.,inneapolis, Minnesota; Biosense Webster, Inc.; and Boston Scientific Cor-

oration, Natick, Massachusetts. Dr. Spector has received research grant sup-ort from Medtronic, Inc., and Biosense Webster, Inc. Dr. Reynolds is aonsultant for Biosense Webster, Inc., and Sanofi-Aventis, Paris, France.r. Calkins is a consultant for Biosense Webster, Inc.; CryoCor, Saniego, California; Medtronic, Inc.; and ProRhythm, Inc., Ronkonkoma,ew York. Drs. Sondhi, Xu, and Sledge and Ms. Martin and Ms. Williamsere employees of United BioSource Corporation during the conduct of

lhis study.

002-9149/09/$ – see front matter © 2009 Elsevier Inc. All rights reserved.oi:10.1016/j.amjcard.2009.04.040

vidence regarding the safety and efficacy of RFA in pa-ients with AFL and SVT. To accomplish this, a systematiceview of published studies was conducted, and meta-anal-ses of outcomes were performed. For the purpose of thiseport, “SVT” refers only to atrioventricular nodal reentrantachycardia (AVNRT) and accessory pathway–mediatedachycardia, and “AFL” refers only to typical AFL.

ethods

To identify and retrieve all potentially relevant researchescribing cardiac RFA in the treatment of patients withFL and SVT, we conducted a search in Medline (viaubMed), EMBASE, and Current Contents. EMBASE wasearched for studies published from January 1, 1990, to Janu-ry 1, 2007, using the following search strategy: (1) heartrrhythmia (with synonym search) OR tachycardia OR atrialbrillation OR atrial flutter OR atrial tachycardia OR sinus

achycardia OR WPW OR “Wolff-Parkinson-White syn-rome” OR ventricular (with synonym search); (2) “catheterblation” (with synonym search) OR “radiofrequency abla-ion”; and (3) 1 AND 2. Results were limited to studies inumans published from 1990 to January 1, 2007, and ex-luded reviews, letters, comments, and editorials. Thisearch strategy identified studies of RFA in all arrhythmias;tudies in AFL and SVT were chosen from among theseesults.

Additional strategies were used to identify recently pub-

ished reports that may not yet be indexed in Medline. The

www.AJConline.org

PmfCroa

rwlrmiwdfhjatltt

pwtocbiiDa

co

gram of

TSs

V

TP

S

C

I

ndt

a

w

672 The American Journal of Cardiology (www.AJConline.org)

ubMed search included a keyword search for the past 6onths with no limits, and Current Contents was searched

or the past 6 months, using similar search terms. Theochrane Library was searched for any recent systematic

eview of the subject. A manual check of the reference listsf all accepted studies and of recent reviews and meta-nalyses was performed to supplement these searches.

Eligible studies included those with all study designseporting efficacy and safety data after patients under-ent RFA for AFL and SVT. Study selection included 2

evels of screening. At initial screening, abstracts wereeviewed for the following exclusion criteria: reviews oreta-analyses; animal or in vitro studies; pediatric stud-

es (subjects aged �18 years) or mixed populations ofhich �15% were pediatric patients; follow-up of �7ays; not studies of RFA; alternative energy sources usedor ablation (i.e., cryoablation, microwave ablation, origh-frequency ultrasound ablation); atrioventricularunction ablation with pacemaker implantation (ablatend pace strategy); �40 patients per arrhythmia or abla-ion technique; published only in abstract form; pub-ished before 1990; and published in languages otherhan English, Spanish, French, Italian, German, and Por-uguese.

The full reports of accepted abstracts and titles thatassed the initial screening underwent a second screening,hich required the consensus of 2 independent investiga-

ors. Only studies reporting �1 of the efficacy or safetyutcomes for �1 group of patients were accepted. All ac-epted studies were included for data extraction and data-ase development. Data extraction involved capturing var-ous data elements from each study and was performed by 1nvestigator and then reviewed by a second investigator.iscrepancies were resolved by consensus or by consulting

Figure 1. Schematic dia

third investigator. Study characteristics, baseline patient a

haracteristics, study intervention, clinical outcomes, safetyutcomes, and cost and utilization outcomes per treatment

the screening process.

able 1tudy characteristics for catheter ablation studies for atrial flutter andupraventricular tachycardias

ariable AFL SVT*

k t N k t N

otal 18 22 1,323 39 49 7,693ublication year1990–1999 8 10 578 32 38 4,6672000–2007 10 12 745 7 11 3,026tudy designRCT 4 7 336 1 2 100nRCT, prospective 2 3 223 3 4 368Comparative, retrospective — — — 1 2 87UCS, prospective 9 9 499 18 20 2,877Single arm, retrospective 3 3 265 16 21 4,261EBM level of evidence†

I 4 7 336 — — —II 2 3 223 3 5 340III 9 9 499 20 23 3,092IV 3 3 265 16 21 4,261

nstitutional settingSingle center 11 13 638 29 34 5,543Multicenter 7 9 685 10 15 2,150

CEBM � Centre for Evidence-Based Medicine; k � number of studies; N �umber of patients evaluated in studies reporting characteristic; nRCT � nonran-omized controlled trial; RCT � randomized controlled trial; t � number ofreatment groups reporting characteristic; UCS � uncontrolled case series.

* Includes Wolff-Parkinson-White syndrome, accessory pathways,trioventricular reentrant tachycardia, and AVNRT.

† Following CEBM rules for level of evidence (available at http://ww.cebm.net).

rm were collected.

psotpfri

rilswttegwCC

R

pt

stitsttreSd

TomAiscscB

ts8f

TC

B

T

G

C

M

aventric

673Arrhythmias and Conduction Disturbances/Ablation for Atrial Flutter and SVT

Studies with treatment arms having �40 study partici-ants per technique or treatment group were rejected. Iftudies included some treatment arms with �40 patients andthers with fewer, only the treatment arms with �40 pa-ients were extracted. For the safety data, commonly re-orted adverse events were collected. Zeroes were imputedor events only when it was clear that there was a completeeporting of complications or adverse events and the eventsn question did not occur.

Study, patient, and clinical outcomes data were summa-ized using basic descriptive statistics. Efficacy outcomes ofnterest were synthesized using meta-analytic pooling ofike treatment group (arrhythmia type) across studies. Re-tricted maximum-likelihood random-effects meta-analysesere conducted and used to construct 95% confidence in-

ervals (CIs) for each of the RFA success rates. This methodakes into account study sample size and incorporates het-rogeneity of effects estimates across the studies. Hetero-eneity was measured using Q statistics. All calculationsere performed using SAS version 8.1 (SAS Institute Inc.,ary, North Carolina) and SPSS version 14.0 (SPSS, Inc.,hicago, Illinois).

esults

Figure 1 shows a schematic diagram of the screeningrocess. The entire publication search yielded 1,870 cita-

able 2haracteristics of patients who underwent catheter ablation for atrial flutt

aseline Characteristic AF

t n

otal patientsMean age (years) 22 1,307Mean no. drugs refractory 7 462Mean duration of arrhythmia (years) 6 371

t n/

enderMale 22 1,011Female 22 296o-morbid conditionsPrevious ablation 3 5Ischemic heart disease 9 97Valvular heart disease 11 104Structural heart disease 11 402Cardiomyopathy 6 52Congestive heart failure 2 6Congenital heart disease 5 92Implantable cardioverter-defibrillator 2 16Previous cardiac surgery (PCI/CABG) 2 21Stroke 1 2Diabetes 3 11Hypertension 8 141edication history

Antiarrhythmic agents 4 242

CABG � coronary artery bypass grafting; n � number of patients with ts in Table 1.

* Includes Wolff-Parkinson-White Syndrome, accessory pathways, atrio

ions, excluding duplicate citations. Of these, 1,333 ab- y

tracts were rejected during initial screening. Full reports ofhe remaining 537 citations were obtained and screened fornclusion and exclusion criteria. This resulted in the rejec-ion of 356 additional reports, most commonly because ofample sizes �40 per treatment arm (135 studies). Addi-ional reasons for study rejection are detailed in Figure 1. Ofhe 537 full reports reviewed, 181 (142 primary plus 39elated) passed the screening criteria. Of these, 19 studiesvaluated AFL (18 primary plus 1 related) and 48 evaluatedVT (39 primary plus 9 related), with the remainder ad-ressing other arrhythmias.

Key characteristics of the AFL studies are presented inable 1. The studies varied widely in terms of the numberf patients (40 to 185) and follow-up duration (6 to 48onths). Of the 18 studies, 10 were published after 1999.bout 2/3 of the studies (k � 11, n � 747) were conducted

n Europe. Eleven of the studies (61%) were single-centertudies. Most of these studies were either prospective, un-ontrolled case series (k � 9) or retrospective, single-armtudies (k � 3), although we did identify 4 randomizedontrolled trials. We rated 12 of 18 as Centre for Evidence-ased Medicine level of evidence III or IV.

The overall mean age was 61 years (range of means 52o 79), and 77% of patients were men (Table 2). Amongtudies reporting it, previous antiarrhythmic use occurred in9% of patients, with a mean of 2.9 antiarrhythmic drugailures per subject. The mean duration of AFL was 2.2

upraventricular tachycardias

SVT*

Mean (Range) t n Mean (Range)

61 (52–79) 41 6,800 47 (27–80)2.9 (2.0–3.3) 19 2,676 1.5 (0.4–3.0)2.2 (0.8–4.7) 7 621 16 (11.0–19.6)

% t n/N %

77.4 43 2,994/6,690 44.822.6 43 3,696/6,690 55.2

2.4 4 38/705 5.421.2 14 137/2,694 5.116.7 9 83/1,680 4.951.9 19 203/2,462 8.215.6 2 4/981 0.4

4.8 4 16/759 2.134.5 — — —10.7 2 1/168 0.620.6 1 0/128 0.0

4.7 6 11/904 1.27.7 1 9/128 7.0

33.7 2 4/205 2.0

89.0 12 1,088/1,155 94.2

acteristic; PCI � percutaneous coronary intervention. Other abbreviations

ular reentrant tachycardia, and AVNRT.

er and s

L

N

/1,307/1,307

/210/457/621/774/334/124/267/150/102/43/143/419

/272

his char

ears. Commonly reported co-morbidities included isch-

TE

S

SMPR

TS

S

M

V

P

TO

Y

674 The American Journal of Cardiology (www.AJConline.org)

able 3fficacy outcomes for catheter ablation for atrial flutter and supraventricular tachycardias

uccess Rate AFL Accessory Pathways* AVNRT

t (n) M-A % (95% CI) t (n) M-A % (95% CI) t (n) M-A % (95% CI)

ingle-procedure success total 21 (1,251) 91.66 (88.40–94.93)† 13 (1,763) 90.91 (87.71–94.11)† 23 (4,249) 94.33 (91.24–97.41)†

ultiple-procedure success total 11 (600) 97.04 (94.72–99.36)† 11 (1,448) 93.32 (89.41–97.22)† 12 (1,942) 95.97 (93.40–98.53)†

ostablation arrhythmia 21 (1,251) 13.20 (7.54–18.86)† 14 (1,696) 7.16 (4.68–9.64)† 27 (4,786) 4.91 (2.92–6.91)†

epeat ablation procedure 18 (1,111) 7.97 (4.53–11.42)† 12 (1,618) 8.04 (4.70–11.37)† 21 (4,474) 5.57 (3.56–7.58)†

M-A � meta-analyzed. Other abbreviations as in Table 1.* Includes Wolff-Parkinson-White Syndrome and atrioventricular reentrant tachycardia.

† p �.01 for heterogeneity.

able 4afety outcomes for catheter ablation procedures by indication

afety Outcome AFL Accessory Pathways* AVNRT

t n/N % t n/N % t n/N %

ortalityDeath overall 13 4/723 0.6 18 7/2,267 0.3 27 2/4,748 0.0Procedure related 12 0/671 0.0 18 2/2,267 0.1 26 0/4,296 0.0Not procedure related 13 6/723 0.8 18 5/2,267 0.2 27 2/4,748 0.0ascular access complicationsArteriovenous fistula 8 0/459 0.0 11 4/1,474 0.3 12 2/2,530 0.1Hematoma 8 0/459 0.0 10 4/1,196 0.3 14 10/3,566 0.3Pneumothorax 8 1/459 0.2 10 1/1,228 0.1 12 2/2,911 0.1eriprocedural eventsStroke, ischemic 12 0/673 0.0 12 2/1,687 0.1 16 0/1,992 0.0Transient ischemic accident 12 0/673 0.0 13 3/1,856 0.2 16 0/1,992 0.0Cardiac tamponade 12 0/673 0.0 13 8/1,856 0.4 17 2/2,893 0.1Pulmonary embolus 12 1/673 0.1 13 0/1,856 0.0 19 5/3,297 0.2Deep-vein thrombosis 12 0/673 0.0 11 0/1,443 0.0 16 2/2,648 0.1Other embolism 12 0/673 0.0 11 1/1,443 0.1 15 1/1,747 0.1Pericardial effusion 12 2/673 0.3 13 3/1,856 0.2 18 4/3,169 0.1Atrioventricular block 12 3/673 0.4 15 16/2,044 0.8 25 78/4,661 1.7CHF exacerbation 16 0/926 0.0 13 0/1,856 0.0 16 1/1,992 0.1Need for a pacemaker 11 1/550 0.2 9 3/1,095 0.3 24 34/4,625 0.7Total no. of patients with events 8 2/409 0.5 11 39/1,417 2.8 12 61/2,013 3.0

Abbreviations as in Tables 1 and 2.

* Includes Wolff-Parkinson-White syndrome and atrioventricular reentrant tachycardia.

able 5utcomes for atrial flutter and supraventricular tachycardia over time

ear† AFL Accessory Pathways* AVNRT

Outcome t n/N % Year† t n/N % Year† t n/N %

1995–1997 Single-procedure success 2 93/103 90.3 1990–1995 8 1,153/1,285 89.7 1990–1995 8 710/772 92.0Multiple-procedure success 1 43/44 97.7 8 1,086/1,170 92.8 5 513/524 97.9Repeat ablation 2 10/103 9.7 9 114/1,318 8.6 6 48/580 8.3

1998–2000 Single-procedure success 7 350/409 85.6 1996–2000 5 447/478 93.5 1996–2000 7 1,145/1,203 95.2Multiple-procedure success 6 348/368 94.6 3 269/278 96.8 6 1,102/1,144 96.3Repeat ablation 7 51/409 12.5 3 15/300 5.0 6 51/1,144 4.5

2001–2003 Single-procedure success 5 315/342 92.1 2001–2007 — — — 2001–2007 8 2,214/2,276 97.3Multiple-procedure success 2 99/100 99.0 — — — 1 273/276 98.9Repeat ablation 5 18/336 5.4 — — — 9 87/2,750 3.2

2004–2007 Single-procedure success 7 381/403 94.5Multiple-procedure success 2 83/88 94.3Repeat ablation 4 10/263 3.8

Abbreviations as in Tables 1 and 2.* Includes Wolff-Parkinson-White Syndrome and atrioventricular nodal tachycardia.

† Year of publication.

ea

a9st7(

wpmuwa

sa

RAtaAststtcc(B

oe5di9tr

pR9a9wR(Rwtl

wvft

7r

oibp

abc(sn

ailoihtclpdna

D

mroatwdbwstA

ctmrapaapp

tdpd

675Arrhythmias and Conduction Disturbances/Ablation for Atrial Flutter and SVT

mic heart disease (21%), congenital heart disease (35%),nd valvular heart disease (17%).

Efficacy outcomes are listed in Table 3. The meta-nalyzed single-procedure success for AFL ablation was1.7% (95% CI 88.4% to 94.9%). Multiple procedureuccess was 97.0% (95% CI 94.7% to 99.4%). Postabla-ion arrhythmia was noted in 13.2% of patients (95% CI.5% to 18.9%), and repeat ablation was reported in 8%95% CI 4.5% to 11.4%).

All-cause mortality was �0.6% (4 of 723), of which thereas no reported procedure-related mortality (Table 4). Theercentage of patients with adverse events was 0.5%. Theost commonly reported adverse events were atrioventric-

lar block (0.4%) and pericardial effusion (0.3%). Thereere no reports of transient ischemic attacks or strokes

mong the studies reviewed.The single-procedure success rate for AFL has improved

ince 1995 (Table 5). The development of new arrhythmiasnd the need for repeat ablation have decreased over time.

Thirty-nine studies with 49 treatment arms evaluatingFA in 7,693 patients with accessory pathways andVNRT were identified. Key characteristics of the 39

otal (16 accessory pathway and 24 AVNRT) primaryccepted studies are listed in Table 1. One study reportedVNRT and accessory pathway treatment arms.2 The

tudies varied widely in terms of the number of patients (40o 901) and follow-up duration (4 to 98.4 months). Twenty-even of these studies had follow-up durations of 6 monthso 2 years. Of the 39 studies, 32 were published from 1990o 1999. Twenty-nine of the studies (74%) were single-enter studies. Most were either prospective, uncontrolledase series (k � 18) or retrospective, single-arm studiesk � 16), and 36 of 39 were rated as Centre for Evidence-ased Medicine level of evidence III or IV.

The mean age of patients with SVT was 47 years (rangef means 27 to 80; Table 2) but was likely skewed by ourxclusion of pediatric studies. In contrast to AFL studies,5% of patients in SVT studies were women. The meanuration of SVT was about 16 years, but this was reportednfrequently. Previous antiarrhythmic use was reported in4% of these patients at baseline, and patients were refrac-ory to an average of 1.5 drugs. SVT studies reported lowates of baseline co-morbid conditions.

SVT ablation outcomes for all treatment arms wereooled and meta-analyzed. Single-procedure success ofFA for all SVTs combined was 93.2% (95% CI 90.8% to5.5%) on the basis of 6,012 patients from 36 treatmentrms. Multiple-procedure success was 94.6% (95% CI2.4% to 96.9%) of 3,390 patients. Postablation arrhythmiaas noted in 5.6% of 6,482 patients (95% CI 4.1% to 7.2%).epeat ablation was required in 6.5% of 6,092 patients

95% CI 4.7% to 8.3%). As listed in Table 3, the efficacy ofFA was slightly higher for AVNRT than accessory path-ay tachycardias (single procedure 94.3% vs 90.9%, mul-

iple procedure 96.0% vs 93.3%), with correspondinglyower rates of recurrent SVT and repeat ablation.

The total proportion of patients with adverse eventsas 2.9% (Table 4). The most commonly reported ad-erse events included atrioventricular block (1.4%), needor pacemaker (0.65%), hematoma (0.3%), and cardiac

amponade (0.2%). All-cause mortality was 0.13% (9 of p

,015), of which there were 2 procedure-related deathseported among 6,563 patients.

Single- and multiple-procedure success rates improvedver time, while recurrent SVT, need for repeat ablation,ncidence of cardiac tamponade, and stroke decreased (Ta-le 5). Similar trends were seen for RFA of accessoryathways and AVNRT.

There was significant heterogeneity of efficacy outcomesmong studies of AFL and SVT. Such heterogeneity woulde expected with the diversity of patient populations (in-luding variability in SVT subtype) and treatment settingsincluding variability of techniques and catheter types). De-pite the heterogeneity, the 95% CI for efficacy was fairlyarrow (Table 3).

Outlier studies were examined to understand which char-cteristics might be contributing to the heterogeneity. Stud-es that included patients refractory to 3 � 1 drugs hadower success rates and higher recurrence rates than studiesf patients refractory to fewer drugs.3,4 The SVT data setncluded 1 large study (901 patients) that reported a veryigh initial success rate of 97%. Studies that included pa-ients without previous electrophysiologic studies5,6 or in-luded subsets of patients with fast pathway ablation5 hadower initial success rates. Postablation recurrence and re-eat ablation rates were higher in studies with protocol-riven, postprocedure electrophysiologic studies,7,8 largerumbers of pediatric patients, or patients with Ebstein’snomaly.9

iscussion

To our knowledge, this is the first systematic review andeta-analysis of RFA for AFL and SVT. Our analysis

evealed several noteworthy points relevant to formulationf evidence-based treatment plans for patients with theserrhythmias. Specifically, our analysis confirms that abla-ion is safe and highly effective. Procedure-related mortalityas extremely rare (2 of 7,286), with a low (�0.5%) inci-ence of most adverse events, except for atrioventricularlock (1.4%) and need for pacemaker (0.6%) in patientsith SVT, generally because the successful ablation of

ome SVTs requires placing lesions very near the atrioven-ricular node. In recent years, single-procedure efficacy forFL and SVT has been near or �95%.The largest previously published series on RFA for SVT

onsisted of voluntary registries and surveys conducted inhe 1990s.10–12 We did not include these studies in our

eta-analysis, because many of the individual studies weeviewed had been incorporated into those pooled analyses,nd we wished to avoid “double-counting” them. Theserevious studies reported very similar outcomes to ours:bout 95% efficacy for AVNRT and accessory pathwayblation, 2% to 4% rates of significant complications, and arocedural mortality rate of about 0.1% with accessoryathway ablation.

RFA was initially offered only to patients who had life-hreatening arrhythmias (e.g., Wolff-Parkinson-White syn-rome with syncope or other high-risk features) and/oratients with frequent and highly symptomatic tachycardiaespite drug therapy. Previous therapeutic options for such

atients included fairly drastic measures, such as arrhythmia

smdfo

lrndvmblmp

tsAa3simRsdsD2

iictRdocSsocpa

2emmpctrsndcsm

tbdabp�dpgf

rodWocAitarae

nScefp

rawdprwsqawt

ssiafirsaemw

676 The American Journal of Cardiology (www.AJConline.org)

urgery13 or atrioventricular junction ablation and pace-aker implantation.14 For clinical and ethical reasons, ran-

omized studies comparing RFA with such options were noteasible, and the early evidence base for RFA grew mainlyut of single-center studies and cooperative registries.

As these procedures matured, their demonstrated successed to application in broader patient groups. Because largeandomized comparisons of RFA with medical therapy wereever conducted, clinicians today must base managementecisions on comparisons of ablation outcomes data withery limited information from separately conducted phar-acologic studies. Although such comparisons are limited

y potential differences between treatment groups, we be-ieve that briefly reviewing the published research on phar-acologic treatment for AFL and SVT will provide appro-

riate context for the results of our study.There is a remarkable paucity of data regarding drug

herapy for AFL, in large part because most studies have noteparated outcomes for patients with atrial fibrillation andFL. Although class I drugs are known to have some

ctivity for AFL, only a single study separated outcomes in6 patients with AFL, 50% of whom achieved long-termuccess with oral flecainide.15 The class III drug dofetilides also believed to have some efficacy for AFL, basedainly on the Symptomatic Atrial Fibrillation Investigativeesearch on Dofetilide (SAFIRE-D) study, which reported

eparate results for 30 patients with AFL who receivedofetilide and found that roughly 50% to 70% maintainedinus rhythm over 12 months, depending on the dose.16

ofetilide is associated with QT prolongation and a 1% to% risk for Torsades de Pointes.

We know of 2 prospective randomized studies compar-ng antiarrhythmic drugs with RFA for AFL; 1 of these wasncluded in our study database,17 and the other was ex-luded because its 2 study arms included �40 patients.18 Inhe study by Natale et al,18 61 patients were randomized toFA or a variety of class I and class III antiarrhythmicrugs, including amiodarone. After a mean follow-up periodf 21 � 11 months, AFL recurred in 6.4% of RFA patients,ompared to 60% of patients in the drug therapy group.imilarly, da Costa et al17 randomized 104 patients afteringle episodes of AFL to RFA or amiodarone. After a meanf 13 months, flutter had recurred in 3.8% of RFA patients,ompared to 30% in the amiodarone group. Ten percent ofatients in the amiodarone group experienced drug-relateddverse events.

On the basis of much of the aforementioned data, the003 consensus guidelines for SVT management19 stronglyndorse RFA over antiarrhythmic drugs for the long-termanagement of AFL. In those guidelines, RFA was recom-ended as a class I intervention for all scenarios, except in

atients with first episodes of well-tolerated flutter (RFAlass IIa), whereas the highest recommendation for drugherapy was for dofetilide (class IIa) for patients withecurrent and well-tolerated AFL. Given the apparentuperiority of RFA in safety and efficacy, we see little too role for repeated cardioversion or antiarrhythmicrugs in the long-term management of AFL when rhythmontrol is desired, and we believe that the guidelineshould more strongly discourage the use of antiarrhyth-

ic drugs for AFL. e

As with AFL, the evidence defining the role of drugherapy in treating SVT is limited. Atrioventricular nodallocking agents (� blockers, calcium channel blockers, andigoxin) are frequently used in the management of AVNRTnd tachycardias involving concealed accessory pathwaysecause of their mechanistic appeal and excellent safetyrofile. Very small studies from the early 1980s, involving20 patients each, showed that verapamil, propranolol, and

igoxin reduced the frequency and duration of episodes inatients with frequent paroxysmal SVT by �50% but sug-ested that few patients became completely arrhythmiaree.20,21

Before the advent of RFA, many patients with SVTefractory to nodal blocking drugs were treated with class Ir class III antiarrhythmic drugs. Unlike nodal agents, theserugs are thought to be safe when used in patients witholff-Parkinson-White syndrome because they act directly

n accessory pathway conduction. Randomized, placebo-ontrolled studies of flecainide (systematically reviewed bynderson et al22), propafenone,23,24 sotalol,25 and dofetil-

de26 all showed significant reductions in SVT frequency,ypically near 70% to 80% compared to placebo. However,

smaller proportion of patients, likely in the 50% to 60%ange, were rendered arrhythmia free. The trade-off in usingntiarrhythmic drugs, of course, is toxicity, with adversevents ranging from 7% to about 25% in these studies.

Overall, limited-quality data suggest that atrioventricularodal blocking agents have only a modest effect in reducingVT frequency and duration. Better-quality evidence indi-ates that class IC or class III antiarrhythmic drugs improvefficacy in treating patients with SVT, but with a higher riskor adverse events, and still without levels of efficacy ap-roaching that observed with RFA.

The 2003 consensus guidelines for SVT managementeflect these sentiments.19 In general, RFA is recommendeds a class I intervention for almost all categories of patientsith SVT, the only exceptions being patients who do notesire ablation after discussion of its risks and benefits andatients with asymptomatic preexcitation on electrocardiog-aphy (class IIa). It is recognized, however, that patientsith SVT have a wide range of arrhythmia burden and

ymptom severity, such that ablation is certainly not re-uired for all patients, and medical therapy may be moreppropriate for many. Patient preferences, risk tolerance,illingness or aversion to daily medical therapy, occupa-

ion, and other factors may influence these decisions.The quality of all research synthesis is dependent on the

tudies included in the data set, and such analyses areubject to the limitations imposed by the criteria set fornclusion in the review. Certain outcomes of interest, suchs the long-term incidence of atrial fibrillation after success-ul AFL ablation, were reported variably in the source stud-es, and data on this were not systematically extracted. Thiseview included a broad range of study designs, treatmentettings, and inclusion criteria. Conducting meta-analysescross such disparate study settings could be of concern. How-ver, including different settings and study designs in theeta-analysis reflected the disease in the population as ahole. Thus, this review provides an estimate of “average

xpectations” across treatment settings.

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

677Arrhythmias and Conduction Disturbances/Ablation for Atrial Flutter and SVT

1. Agency for Healthcare Research and Quality. HCUPnet. Available at:http://hcupnet.ahrq.gov. Accessed June 12, 2008.

2. Kay GN, Epstein AE, Dailey SM, Plumb VJ. Role of radiofrequencyablation in the management of supraventricular arrhythmias: experi-ence in 760 consecutive patients. J Cardiovasc Electrophysiol 1993;4;:371–389.

3. Schumacher B, Pfeiffer D, Tebbenjohanns J, Lewalter T, Jung W,Luderitz B. Acute and long-term effects of consecutive radiofrequencyapplications on conduction properties of the subeustachian isthmus intype I atrial flutter. J Cardiovasc Electrophysiol 1998;9:152–163.

4. Tai CT, Chen SA, Chiang CE, Lee SH, Wen ZC, Huang JL, Chen YJ,Yu WC, Feng AN, Lin YJ, Ding YA, Chang MS. Long-term outcomeof radiofrequency catheter ablation for typical atrial flutter: risk pre-diction of recurrent arrhythmias. J Cardiovasc Electrophysiol 1998;9:115–121.

5. Calkins H, Epstein A, Packer D, Arria AM, Hummel J, Gilligan DM,Trusso J, Carlson M, Luceri R, Kopelman H, Wilber D, Wharton JM,Stevenson W. Catheter ablation of ventricular tachycardia in patientswith structural heart disease using cooled radiofrequency energy. J AmColl Cardiol 2000;35:1905–1914.

6. Kuck KH, Schluter M, Geiger M, Siebels J, Duckeck W. Radiofre-quency current catheter ablation of accessory atrioventricular path-ways. Lancet 1991;337:1557–1561.

7. Kimman GP, van Hemel NM, Jessurun ER, van Dessel PF, Kelder JC,Defauw JJ, Guiraudon GM. Comparison of late results of surgical orradiofrequency catheter modification of the atrioventricular node foratrioventricular nodal reentrant tachycardia. Eur Heart J 1999;20:527–534.

8. Salerno JA, Storti C, De Ponti R, Zardini M, Longobardi M, Ferrari A,Massacci E. Transcatheter ablation by radiofrequency in paroxysmalatrioventricular junctional reentrant tachycardia. The role of mapping.New Trends Arrhythm 1991;7:367–378.

9. Lesh MD, Van Hare GF, Schamp DJ, Chien W, Lee MA, Griffin JC,Langberg JJ, Cohen TJ, Lurie KG, Scheinman MM. Curative percu-taneous catheter ablation using radiofrequency energy for accessorypathways in all locations: results in 100 consecutive patients. J AmColl Cardiol 1992;19:1303–1309.

0. Hindricks G; The Multicentre European Radiofrequency Survey(MERFS) Investigators of the Working Group on Arrhythmias of theEuropean Society of Cardiology. The Multicentre European Radiofre-quency Survey (MERFS): complications of radiofrequency catheterablation of arrhythmias. Eur Heart J 1993;14:1644–1653.

1. Scheinman MM. Nonpharmacologic management of supraventriculartachycardia. Am J Geriatr Cardiol 2000;9:159–161.

2. Scheinman MM, Huang S. The 1998 NASPE prospective catheterablation registry. Pacing Clin Electrophysiol 2000;23:1020–1028.

3. Gallagher JJ, Gilbert M, Svenson RH, Sealy WC, Kasell J, WallaceAG. Wolff-Parkinson-White syndrome. The problem, evaluation, andsurgical correction. Circulation 1975;51:767–785.

4. Scheinman MM, Evans-Bell T. Catheter ablation of the atrioventric-ular junction: a report of the percutaneous mapping and ablationregistry. Circulation 1984;70:1024–1029.

5. Hohnloser SH, Zabel M. Short- and long-term efficacy and safety offlecainide acetate for supraventricular arrhythmias. Am J Cardiol1992;70(suppl):3A–10A.

6. Singh S, Zoble RG, Yellen L, Brodsky MA, Feld GK, Berk M, Billing CBJr. Efficacy and safety of oral dofetilide in converting to and maintainingsinus rhythm in patients with chronic atrial fibrillation or atrial flutter: theSymptomatic Atrial Fibrillation Investigative Research on Dofetilide

(SAFIRE-D) study. Circulation 2000;102:2385–2390.

7. Da Costa A, Thévenin J, Roche F, Romeyer-Bouchard C, AbdellaouiL, Messier M, Denis L, Faure E, Gonthier R, Kruszynski G, Pages JM,Bonijoly S, Lamaison D, Defaye P, Barthélemy JC, Gouttard T, IsaazK, for the Loire-Ardèche-Drôme-Isère-Puy-de-Dôme (LADIP) Trialof Atrial Flutter Investigators. Results from the Loire-Ardèche-Drôme-Isère-Puy-de-Dôme (LADIP) Trial on Atrial Flutter, a multicentricprospective randomized study comparing amiodarone and radiofre-quency ablation after the first episode of symptomatic atrial flutter.Circulation 2006;114:1676–1681.

8. Natale A, Newby KH, Pisano E, Leonelli F, Fanelli R, Potenza D,Beheiry S, Tomassoni G. Prospective randomized comparison of an-tiarrhythmic therapy versus first-line radiofrequency ablation in pa-tients with atrial flutter. J Am Coll Cardiol 2000;35:1898–1904.

9. Blomstrom-Lundqvist C, Scheinman MM, Aliot EM, Alpert JS,Calkins H, Camm AJ, Campbell WB, Haines DE, Kuck KH, LermanBB, Miller DD, Shaeffer CW Jr, Stevenson WG, Tomaselli GF,Antman EM, Smith SC Jr, Alpert JS, Faxon DP, Fuster V, Gibbons RJ,Gregoratos G, Hiratzka LF, Hunt SA, Jacobs AK, Russell RO Jr, PrioriSG, Blanc J-J, Budaj A, Burgos EF, Cowie M, Deckers JW, GarciaMAA, Klein WW, Lekakis J, Lindahl B, Mazzotta G, Morais JCA, OtoA, Smiseth O, Trappe H-J. ACC/AHA/ESC guidelines for the man-agement of patients with supraventricular arrhythmias—executivesummary: a report of the American College of Cardiology/AmericanHeart Association Task Force on Practice Guidelines and the EuropeanSociety of Cardiology Committee for Practice Guidelines (WritingCommittee to Develop Guidelines for the Management of PatientsWith Supraventricular Arrhythmias). Circulation 2003;108:1871–1909.

0. Mauritson DR, Winniford MD, Walker WS, Rude RE, Cary JR, HillisLD. Oral verapamil for paroxysmal supraventricular tachycardia: along-term, double-blind randomized trial. Ann Intern Med 1982;96:409–412.

1. Winniford MD, Fulton KL, Hillis LD. Long-term therapy of paroxys-mal supraventricular tachycardia: a randomized, double-blind compar-ison of digoxin, propranolol and verapamil. Am J Cardiol 1984;54:1138–1139.

2. Anderson JL, Jolivette DM, Fredell PA. Summary of efficacy andsafety of flecainide for supraventricular arrhythmias. Am J Cardiol1988;62(suppl):62D–66D.

3. UK Propafenone PSVT Study Group. A randomized, placebo-con-trolled trial of propafenone in the prophylaxis of paroxysmal supraven-tricular tachycardia and paroxysmal atrial fibrillation. Circulation1995;92:2550–2557.

4. Pritchett EL, McCarthy EA, Wilkinson WE. Propafenone treatment ofsymptomatic paroxysmal supraventricular arrhythmias. A randomized,placebo-controlled, crossover trial in patients tolerating oral therapy.Ann Intern Med 1991;114:539–544.

5. Wanless RS, Anderson K, Joy M, Joseph SP. Multicenter comparativestudy of the efficacy and safety of sotalol in the prophylactic treatmentof patients with paroxysmal supraventricular tachyarrhythmias. AmHeart J 1997;133:441–446.

6. Tendera M, Wnuk-Wojnar AM, Kulakowski P, Malolepszy J, Koz-lowski JW, Krzeminska-Pakula M, Szechinski J, Droszcz W,Kawecka-Jaszcz K, Swiatecka G, Ruzyllo W, Graff O. Efficacy andsafety of dofetilide in the prevention of symptomatic episodes ofparoxysmal supraventricular tachycardia: a 6-month double-blindcomparison with propafenone and placebo. Am Heart J 2001;142:

93–98.