Br HeartJ 1995;74:671-675

TECHNIQUE

Long term results of fast pathway ablation inatrioventricular nodal reentry tachycardia using a

modified technique

Davendra Mehta, J Anthony Gomes

AbstractObjective-To assess immediate and longterm success of "fast" pathway catheterablation with graded use of radiofre-quency energy in patients with classicatrioventricular nodal reentrant tachy-cardia (AVNRT) and evaluate clinical,procedure related, and electrophysiologi-cal features affecting long term results.Design-31 consecutive patients withclassic AVNRT at electrophysiologicalstudy, who were candidates for radiofre-quency ablation. Patients were followedfor an average of 24 months after abla-tion.Setting-All studies and ablations wereperformed in an electrophysiological lab-oratory under fluoroscopic guidanceusing standard electrophysiological tech-niques.Intervention-Radiofrequency applica-tion was performed at the site of proxi-mal His bundle electrogram with A:Vratio of > 1. It was started at 10W withincrement of 5 W to a maximum of 25 Wat 60s. With the onset of junctionalrhythm, atrial pacing was begun in orderto monitor the PR interval. Applicationwas terminated prematurely with a non-conducted P wave, continued prolonga-tion of the PR interval beyond 50% of thebaseline, or a threefold rise in imped-ance.

Results-Successful ablation was possiblein 30/31 patients (97%) with an average ofseven applications (range 1-10). It was

associated with significant prolongationof PR interval (P < 0.001) and AVWenckebach cycle length (P = 0-01).Ventriculo-atrial conduction was abol-ished in 24/30 patients (82%) with suc-cessful ablation. Two patients developedtransient complete heart block (3 and 12min) and one persistent right branchblock. Four patients had late recurrence.

Presence of ventriculo-atrial block was

the only electrophysiological index pre-dictive oflong term success (P = 0-01).Conclusions-Graded use of radiofre-quency energy and atrial pacing to moni-tor PR interval decreases the risk ofatrioventricular block in patients under-going fast pathway ablation for AVNRT.Ventriculo-atrial block is predictive of

long term success and should be a pre-ferred end point for fast pathway abla-tion.

(Br Heart _J 1995;74:671-675)

Keywords: atrioventricular nodal reentry tachycardia;fast pathway ablation

Atrioventricular nodal reentrant tachycardia(AVNRT) is the commonest cause ofparoxysmal supraventricular tachycardia. '

Radiofrequency (RF) catheter ablation is nowthe treatment of choice for management ofthese patients as it is associated with a successrate in the range of 9095%.26 Successfulablation in these patients can be performed byablation of either the "fast" or the "slow"limb of the reentrant circuit.4 The slow path-way is ablated at a relatively posterior septalsite of the tricuspid valve annulus. It is guidedeither by slow pathway potentials, which are

difficult to identify with confidence, or morefrequently anatomically, in which case abla-tion is performed in the mid and posteriorseptal portion of the tricuspid valve annulusuntil tachycardia is not inducible. Slow path-way ablation is preferred over fast pathwayablation as it is associated with a lower risk ofcomplete heart block.2-6 However, there is a

wide area over which slow pathway potentialscan be recorded. Fast pathway ablation isachieved in the region of the proximal Hisbundle with His:atrium electrogram ratio of0-5 or less.34 It results in prolongation of thePR interval. Although relatively easy, as

guided by a specific electrogram in a smallanatomical area, fast pathway ablation hasbeen reported to be associated with a higherincidence of inadvertent complete heartblock. It has been suggested that graded use

of RF energy decreases the incidence of heartblock.4 In the present study factors affectinglong term success of fast pathway ablationwere assessed. RF energy applications were

performed in a graded manner and with onsetof junctional rhythm, atrial pacing was per-formed to monitor the PR interval.

MethodsPATIENT CHARACTERISTICS

Over the period of this study, fast pathway

The ElectrophysiologyandElectrocardiographySection, Division ofCardiology,Department ofMedicine, The MountSinai Medical Center,New York, New York,USAD MehtaJ A GomesCorrespondence to:Dr D Mehta, AssistantProfessor in Medicine,Division of Cardiology,Mount Sinai Hospital, 1Gustave L Levy Place, NewYork NY 10029, USA.Accepted for publication30 June 1995

671

Mehta, Gomes

ablation was attempted in 31 consecutivepatients with the classic variety of AVNRT."7

Mean age of the patients was 58 years (range18-80). There were nine men and 22 women.Average duration of symptoms was sevenyears (<1-35 years). All patients had palpita-tions and seven also had syncope. Threepatients had coronary artery disease withoutsignificant ischaemia. Standard electrophysio-logical techniques were used. During thestudy surface electrocardiographic leads 1, 2,and Vl, and high right atrial, proximal, anddistal His bundle, and proximal and distalcoronary sinus electrograms were monitoredcontinuously. Atrial and ventricular stimula-tion protocols were used to identify the ante-grade and retrograde atrioventricularconduction characteristics. If sustainedAVNRT could not be initiated at the baselinestudy despite two extrastimuli, the stimula-tion protocol was repeated with an isopre-naline infusion. Standard criteria were usedto establish the diagnosis of dual atrioventric-ular nodal pathways and atrioventricularnodal reentry.7

ABLATIVE PROCEDUREA 7F deflectable quadripolar catheter with a4 mm bulbous tip electrode (MansfieldScientific) introduced through the rightfemoral vein was used for ablation. Thecatheter was positioned to obtain a bipolarrecording of the His bundle deflection. It wasthen withdrawn and rotated anteriorly until itrecorded a ventricular deflection that wassmaller than the atrial deflection, withatrial:ventricular electrogram ratio of >1 andHis bundle potential of <100,uV, and therewas no beat to beat variation, the latter indi-cating a stable catheter tip position. If initialapplication of RF failed, the above procedurewas repeated in an attempt to obtain bettertissue contact. If a satisfactory electrogramcould not be obtained from the anterior sep-tal site of the tricuspid annulus, the ablationcatheter was gradually rotated posteriorly tosee if the target electrogram (A:V ratio of >1with relatively small His bundle deflection) orthe earliest atrial activation during tachycar-dia could be recorded from a relatively poste-rior septal site. At the target site RFapplication was started at 10W and wasincreased by 5 W every 15 s to a maximum of25 W. Atrial pacing at 100-110 beats/minwas begun with the onset of junctionalrhythm so as to consistently capture theatrium. The PR interval was closelymonitored. RF energy application was termi-nated with continuing PR prolongation

PR 180ms

Figure 1 Rhythm strip (electrocardiographic lead II) recording during radiofrequency(RF) application showing prolongation ofPR interval. RF application was started 10 sbefore the recording of the above strip. Atrial pacing (as identified by the pacing spike) wasbegun with the onset ofjunctional rhythm. There was progressive prolongation ofPRinterval during 15 s ofRF application.

beyond 50% (fig 1), occurrence of a singlenon-conducted P wave, two- to threefold risein impedance, or at 60 s. Inducibility oftachycardia was evaluated after each RFapplication. Ablation was terminated whenno tachycardia could be induced with atrialand ventricular premature stimulation withtwo extrastimuli before and during isopre-naline infusion. Ablation was repeated if twoor more echo beats were inducible. If notachycardia was induced, complete stimula-tion protocol was repeated to document theatrioventricular and ventriculo-atrial conduc-tion characteristics. Postablative electrophysi-ological assessment was performed beforeisoprenaline infusion. Patients were followedone month after the ablation and were sub-sequently asked to report if they hadrecurrence of symptoms. Up to date tele-phone follow up was available in all patients.

STATISTICAL METHODSElectrophysiological variables are expressedas mean (SD). Statistical comparisons of preand post ablation electrophysiological vari-ables were performed by Student's t test forpaired and unpaired values. Clinical vari-ables, and pre and post ablation electrophysi-ological indices in predicting long termsuccess of fast pathway ablation were assessedin the Cox stepwise regression analysis.Association of ventriculo-atrial conductionand late recurrence of AVNRT was evaluatedin addition by Fisher's exact test. A probabil-ity value of <0 05 was considered statisticallysignificant.

ResultsIMMEDIATE RESULTSAll 31 patients had evidence of dual atrioven-tricular nodal pathway physiology. AVNRTwas induced by premature atrial stimulation(25 patients) or rapid ventricular pacing (twopatients), and required isoprenaline infusionin the other four. Using the above technique,"fast" pathway ablation was successful in 30of the 31 patients (96%). In 26 patients suc-cessful ablation of the fast pathway was possi-ble at the anterior-superior septal site of thetricuspid annulus with the delivery of an aver-age of seven applications of RF energy (range1-11). In five patients in whom ablation atthe anterior septal site failed, the mid-septalarea was mapped. In four of these patients,the earliest atrial activation during tachycar-dia was recorded at a site between the Hisbundle location and os of the coronary sinus.RF application was performed at these sites;however, in only two of these four patientscould His bundle potentials or potentials sim-ilar to the timing of His-bundle electrogrambe recorded from the site (posterior) of suc-cessful ablation (fig 2). As seen with ablationat the anterior site, ablation at these posteriorsites in all four patients was associated withprolongation of the PR interval, loss of ven-ticulo-atrial conduction, and abolition ofAVNRT.

Following successful ablation there wassignificant prolongation of PR and atrium-

PR lOOms

672

Long term results offast pathway ablation in atrioventnicular nodal reentry tachycardia using a modified technique

Figure 2 Intracardiacelectrocardiograms recordedin sinus rhythm beforedelivery ofa successfulradiofrequency pulse. Thetop 3 traces representsurface electrocardiographicleads I, II, and Vl.Intracardiac electrogramsshownfrom top to bottomare at sites of ablationcatheter (AC), distal andproximal His bundleelectrograms (HBI,HB2), and distal andproximal coronary sinuselectrograms (CSJ, CS2).Ablation catheterrecordings, in addition tothe atrial and ventricularelectrograms, show arelatively low frequencypotential occuring slightlybefore the rapid His bundledeflection on HB2. Thiswas recorded at a relativelyposterior site of tricuspidannulus where delivery ofradiofrequency energyresulted in "fast" pathwayablation.

I. I

HB1 A.mmmm -,_-

HB2-

AH 4

I II*

cs1iv

CS2

A

His intervals and of atrioventricularWenckebach cycle length and no significantchange in atrioventricular nodal effectiverefractory period (table). Six patients (19%)had evidence of retrograde conductiondespite non-inducibility of AVNRT andseven patients had single reentrant echobeats. Retrograde conduction was by the slowpathway in five and probably by the fast path-way in one (short ventriculo-atrial interval).Two of the 30 patients who had successful

ablation developed complete heart block dur-ing RF application. In one it occurred at 15 sof RF application and the other at 35 s. Inboth patients heart block developed with RFapplication at an anterior site and was tran-sient. In the first patient it recovered at 3 minof termination of RF application whereas inthe other patient atrioventricular conductionrecovered after 12 min. Surprisingly in boththese patients the PR interval returned tonormal within the next half hour.Immediately after ablation neither patient hadinducible arrhythmia but one had ventriculo-atrial conduction. The latter patient had laterecurrence of AVNRT, seven months afterablation. Two other patients had right bundlebranch block, pobably related to cathetermovement to a distal position during RFapplication. Right bundle branch block haspersisted in one patient and has recovered inthe other. Two patients, who had PR interval

Electrophysiological indices before and after ablation in 30patients who undervent successfulfast pathway ablation.Values are means (SD)

Index Pre-ablation Post-ablation P

PR (ms) 137 (19) 208 (32) <0 001AH (ms) 69 (20) 164 (48) <0 001HV (ms) 42 (11) 46 (14) NSAVERP (ms) 277 (42) 287 (50) NSAVWCL (ms) 313 (46) 356 (62) 0 01VA block 0 24 (82%)

AH, low right atrium to His conduction time as recorded atthe site of His bundle electrogram; AVERP, AV effectiverefractory period; AVWCL, AV Wenckebach cycle length;HV, His bundle to ventricular conduction time.

of more than 300 ms after ablation of fastpathway, complained of a heavy feeling in thethroat for two and three weeks after ablationrespectively. These symptoms were not asso-ciated with any physical finding or pericardialeffusion and were thought to be related totransient haemodynamic consequences ofprolonged PR interval. In both these patientssymptoms subsided spontaneously withoutany change in the PR interval and at threemonths follow up both were totally asympto-matic.

LATE RESULTSDuring a follow up period of 12 to 30 months(average 24), four patients (13%) had symp-tomatic recurrences two weeks to ninemonths after apparently successful ablation.There was no difference in postablation elec-trophysiological features in patients with andwithout late recurrences. The role of numberof RF lesions, changes in PR and atrium-Hisintervals, prolongation of atrioventricularWenckebach cycle length, atrioventricularnodal effective refractory period, andventriculo-atrial block in predicting long termsuccess was assessed in a multivariate analy-sis. Absence of ventriculo-atrial conductionwas the only electrophysiological feature pre-dictive of long term success (P < 0 01). Threeof the six patients with preserved ventriculo-atrial conduction developed late recurrencesas compared with one of the 26 patients withno ventriculo-atrial conduction (P = 0 01,Fisher's exact test). Three of these fourpatients have undergone a repeat electrophys-iological study and successful ablation of thefast pathway with loss of ventriculo-atrialconduction. In all three patients the recurrentarrhythmia was also classic AVNRT.

DiscussionIn patients with AVNRT, catheter ablation ofthe "slow" or "fast" AV nodal pathways hasbeen reported to have a success rate in therange of 55-94%, higher success rates beingseen with increasing operator experience.34The present study confirns the high successrate for "fast" pathway ablation. The majorimmediate complication associated with fastpathway ablation is a high incidence of inad-vertent atrioventricular block, reported to bein the range of 8-21 %.2 4 This could berelated to the learning curve of the operator,the aggressiveness with which a successfulresult was pursued, the onset of junctional

673

I

Mehta, Gomes

rhythm making monitoring of the atrioven-tricular interval difficult, variable location ofthe fast pathway, mode of delivery of energy,and total energy delivered at the site of abla-tion. In large proportion of patients in thepresent study, adequate results were obtainedat a relatively low energy level. Graded RFapplication allows energy to be titrated toeach patient. As the energy used is relativelysmall, lesions are smaller and localised to thefast pathway. The use of high energy is morelikely to involve both the slow and the fastpathways and thus more likely to lead toheart block. Secondly, application of RFenergy in the region of the atrioventricularnode leads to junctional rhythm making itdifficult to monitor the PR interval, therebyincreasing the risk of atrioventricular block.Atrial pacing with onset of junctional rhythm,at a rate faster than the junctional rate, helpsin monitoring the PR interval; thus with anyearly evidence of atrioventricular block, RFapplication can be promptly stopped. In fourof the 30 patients with successful ablation(13%), "fast" pathway ablation was achievedat a relatively posterior location along the tri-cuspid valve ring. Location of the fast path-way was confirmed by earliest activation atthese sites during AVNRT. Furthermore postablation electrophysiological features wereconsistent with those reported with fast path-way ablation, that is, prolongation of the PRinterval and loss or marked prolongation ofventriculo-atrial conduction. These sites areconventionally targeted for "slow" pathwayablation. Ablation at these sites could explaincomplete atrioventricular block seen in somepatients during slow pathway ablation.4

Limited data are available about long termfollow up patients with AVNRT undergoingRF ablation.34 In the present study late recur-rence rate was 12%, which is similar to thatreported in previous studies. Electro-physiological features seen after ablation weresimilar to those reported by Jazayeri et al withsuccessful fast pathway ablation.3 Sixty sevenpercent of their patients lost ventriculo-atrialconduction following successful fast pathwayablation. However, loss of ventriculo-atrialconduction was not correlated with long termsuccess. Our study shows that loss ofventriculo-atrial conduction is the onlyfeature predictive of long term success. It islikely that in patients who had persistentventriculo-atrial conduction and were non-inducible, non-inducibility was related totransient effects of vagal stimulation as aresult of RF energy application. A recentstudy by Kocovic et al has shown that RFapplication in the region of atrioventricularnode leads to features suggestive of auto-nomic denervation of the conduction systemand associated changes in heart rate variabil-ity which recover in due course of time andcould explain transient non-inducibility inpatients with persistent ventriculo-atrialconduction.8 It can also be hypothesised thatin these patients RF application resulted inrevesible tissue damage and not actual coagu-lation necrosis, while loss of ventriculo-atrial

conduction was related to coagulation necro-sis and thus long term "cure".

It is suggested that prolonged PR intervalassociated with fast pathway ablation is non-physiological and could result in symptomsakin to pacemaker syndrome. None of thepatients in the present series or in the previ-ous studies has reported long term symptomsrelated to prolonged PR interval. It might,however, be a concern in patients withimpaired left ventricular function. Secondly,as a result of a prolonged PR interval thereappears to be a theoretical possibility of lateatrioventricular block. However, none of ourpatients (including the two patients who hadtransient atrioventricular block) has had elec-trocardiographic evidence of atrioventricularblock at a mean follow up of 24 months.

With increasing experience, slow pathwayablation is becoming the treatment of choicefor patients with classic AVNRT and is usu-ally attempted first. Our findings in consecu-tive patients suggest that fast pathwayablation can be safely undertaken in allpatients with classic AVNRT and is bestsuited for patients with unsuccessful slowpathway ablation. Modification of the tech-nique so as to closely monitor the PR inter-val, and graded application of RF energy, islikely to reduce or abolish the incidence ofinadvertent heart block. Although there is atheoretical possibility of increased incidenceof heart block when fast pathway ablation isperformed after a failed attempt at slow path-way ablation, this was not the experience in aprevious crossover study by Langberg et al,when fast pathway ablation was undertakenin all patients with failed slow pathway abla-tion.4

CONCLUSIONSFast pathway ablation is feasible in the major-ity of patients with classical AVNRT andshould be attempted following failed slowpathway ablation. Graded RF energy applica-tion and atrial pacing with the onset of junc-tional rhythm significantly decrease the risk ofcomplete heart block. In some patients fastpathway seems to have a relative posteriorseptal location. The only electrophysiologicalfeature predictive of absence of recurrenceappears to be loss of ventriculo-atrial conduc-tion at ablation. A prolonged PR interval as aresult of successful fast pathway ablation doesnot seem to be associated with any long termcomplication or delayed occurrence of AVblock.

1 Wu D, Denes P, Amat-y-Leon F, Dhingra R, WyndhamCRC, Bauernfeind R, et al. Clinical, electrocardio-graphic and electrophysiologic observations in patientswith paroxysmal supraventricular tachycardia. Am JCardiol 1978;41:1045-51.

2 Jackman WM, Beckman KJ, McClelland JH, Wang X,Friday KJ, Roman CA, et al. Treatment of supraventric-ular tachycardia due to atrioventricular nodal reeentryby radiofrequency ablation of slow-pathway conduction.NEngljMed 1992;327:313-8.

3 Jazayeri MR, Hempe SL, Sra JS, Dhala AA, Blanck Z,Deshpande SS, et al. Selective transcatheter ablation ofthe fast and slow pathways using radiofrequency energyin patients with atrioventricular nodal reentrant tachy-cardia. Circulation 1992;85:1318-28.

674

Long term results offast pathway ablation in atrioventricular nodal reentry tachycardia using a modified technique

4 Langber IJ, Leon A, Borganelli M, Kalbfleisch SJ, El-Atassi R, Calkins H, et al. A randomized, prospectivecomparison of anterior and posterior approaches toradiofrequency catheter ablation of atrioventricularnodal tachycardia. Circulation 1993;87: 1551-6.

5 Haissaguerre M, Gaita F, Fischer F, Commenges D,Montserrat P, d'Ivernois C, et al. Elimination of atrio-ventricular nodal reentrant tachycardia using discreteslow potentials to guide application of radiofrequencyenergy. Circulation 1992;85:2162-75.

6 Chen S-A, Chiang C, Tsang W-P, Hsia C-P, Wang D-C,Yeh C-T, et al. Selective radiofrequency catheter abla-

tion of fast and slow pathway in 100 patients with atrio-ventricular nodal reentrant tachycardia. Am Heart J1993;125: 1-10.

7 Denes P, Wu D, Dhingra RC, Chuquimia R, Rosen KM.Demonstration of dual AV nodal pathways in patientswith supraventricular tachycardia. Circulation 1973;48:549-55.

8 Kocovic DZ, Harada T, Shea JB, Soroff D, Freidman PL.Alterations of heart rate and of heart rate variability afterradiofrequency catheter ablation of supraventriculartachycardia. Delineation of parasympathetic pathways inthe human heart. Circulation 1993;88:1671-81.

IMAGES IN CARDIOLOGY

Pericardial cyst

This chest x ray shows an incidental findingof a smooth round opacity in the left cardio-phrenic angle. Pericardial (spring-water) cystsare a rare developmental anomaly most com-monly found in the cardiophrenic angles, andmore frequently on the right. Occasionallythey are located superiorly in the media-stinum. They usually do not cause symptomsor signs and are invariably benign. Computedtomographic scanning is a useful method ofdistinguishing between cysts and solidtumours.

M K DAVIES

675