Protected Circumferential Conduction in the Posterior Atrioventricular Vestibule of the Left Atrium:...

Protected Circumferential Conduction in the PosteriorAtrioventricular Vestibule of the Left Atrium:Electrophysiologic and Anatomic CorrelatesJIE CHENG,* YANFEI YANG, PHILIP C. URSELL,† RANDALL J. LEE,PARVIN C. DOROSTKAR, KWABENA A. BOAHENE,* and MELVIN M. SCHEINMANFrom the Section of Cardiac Electrophysiology, University of California, San Francisco, California, *CardiacElectrophysiology Laboratory, SUNY Health Science Center, Syracuse, New York, and †Department of Pathology,University of California, San Francisco, California

CHENG, J., ET AL.: Protected Circumferential Conduction in the Posterior Atrioventricular Vestibule ofthe Left Atrium: Electrophysiologic and Anatomic Correlates. Background: The anatomic substrate forprotected isthmus conduction in the right atrium has been well defined. Little is known of similar substratesin the left atrium (LA).

Methods: Patients (pts) with reentrant tachycardia (AVRT) supported by a single left-sided accessorypathway were studied retrospectively (n = 64) and prospectively (n = 31). Intracardiac electrograms wererecorded from the His bundle position and coronary sinus (CS). The LA was mapped with a steerablecatheter using the transseptal approach. LA anatomy was examined grossly and histologically in sixcadaver hearts after removal of endocardium.

Results: A distal-to-proximal CS activation sequence during AVRT was seen in all patients with a leftlateral accessory pathway before ablation. After one to three radiofrequency (RF) energy deliveries thatdid not interrupt accessory pathway conduction, the CS activation sequence was reversed in three patientsin the retrospective group and bidirectional conduction block in the posterior atrioventricular vestibule ofthe LA (PAVV) was demonstrated in nine patients in the prospective group. Four of the six cadaver heartsshowed a distinct circumferential inferoposterior myocardial bundle that coursed parallel to the CS in thePAVV.

Conclusions: We described evidence of bidirectional intraatrial block in the PAVV after application ofRF energy during accessory pathway ablation. Such conduction block may mimic the presence of a secondaccessory pathway. Our data suggest that circumferential conduction in the PAVV may be poorly coupledto the rest of the LA and may be involved in the macro-reentrant circuit around the mitral annulus. Thecircumferential inferoposterior myocardial bundle may serve as the underlying anatomic substrate (PACE2005; 28:692–701)

electrophysiology, accessory pathway, reentry

IntroductionProtected isthmus conduction is critical in

typical right atrial flutter and its anatomic sub-strate has been well described.1–10 However, thepresence of a similar substrate in the left atrium isnot known. Recently, left atrial flutter supported byreentrant circuits around the mitral annulus hasbeen reported.11 In addition, prior reports haveprovided evidence of protected circumferentialconduction along the atrioventricular vestibule ofthe left atrium.12,13 Yet the underlying anatomic

Jie Cheng is now at Texas Heart Institute/St. Luke’s EpiscopalHospital, Houston, Texas.

Address for reprints: Melvin M. Scheinman, M.D., CardiacElectrophysiology, University of California, San Francisco,500 Parnassus Avenue, MU East 4S, Box 1354, San Fran-cisco, CA 94143-1354. Fax: (415)476-6260; e-mail: [email protected]

Received January 19, 2005; revised March 2, 2005; acceptedMarch 27, 2005.

structure has not been identified. In preliminarystudies, we found a distinct muscle bundle (infer-oposterior bundle) in the posterior atrioventricu-lar vestibule of the left atrium (Fig. 1). The purposeof our study is to describe both the electrophysio-logical characteristics and the anatomic correlatesof the protected circumferential conduction alongthe atrioventricular vestibule of the left atrium.

MethodsPatient Selection

A total of 95 patients who were referred for ab-lative therapy of a single left-sided accessory path-way were studied. Of them, 64 patients underwentradiofrequency ablation at University of CaliforniaSan Francisco (UCSF) between 1995 and Decem-ber 1997 and were studied retrospectively (Retro-spective Group); 31 patients were studied prospec-tively from December 1997 to 1998 at UCSF andfrom July 1998 to February 2000 at the State Uni-versity of New York Upstate Medical University

692 July 2005 PACE, Vol. 28

PROTECTED CONDUCTION IN THE LEFT ATRIUM

MA

PVsLAA base

AS

CS

Inferoposterior bundle

Figure 1. This schema shows the inferoposterior musclebundle in relationship to surrounding left atrial struc-ture. This muscle bundle is found parallel to the coro-nary sinus (CS) and divides into two parts, one ante-rior and one posterior to the base of the left atrial ap-pendage (LAA). The atrial septum (AS) and pulmonaryveins (PVs) are also shown.

(prospective group). Antiarrhythmic drugs werediscontinued at least three half-lives prior to thestudy.

Electrophysiologic Testing and Ablation

All patients were admitted to the Electrophys-iology Laboratory in a post-absorptive and unse-dated state after informed consent was obtained.Venous access was obtained with 6–8 F sheaths in-serted in the femoral veins. In some cases, the coro-nary sinus catheter was inserted through the rightinternal jugular veins. Atrial activation at the ante-rior septum (His bundle position) and coronary si-nus was also recorded. A quadripolar catheter wasplaced in the right ventricle. The left atrium wasmapped with a steerable catheter via a trans-septalapproach (Fig. 2). Pulse oximetery and vital signswere closely monitored throughout the study.

The 12-lead surface electrocardiograms (ECG)and intracardiac signals were recorded with a com-puterized multi-channel data acquisition system

RAO VIEW LAO VIEW

RVRV

HISHIS

CS CS

ABL

ABL

successfulsite

successfulsite

Figure 2. Typical catheter posi-tioning in patients with left lateralaccessory pathway. ABL = abla-tion catheter; HIS = His catheter;RV = right ventricular catheterplaced in the apex; CS = coronarysinus catheter; RAO = right ante-rior oblique view; LAO = left ante-rior oblique view. After successfulablation, the ablation catheter wasmoved slightly more lateral to thesite of accessory pathway (arrow).

(CardioLabTM by Prucka Engineering Inc.). All en-docardial signals were filtered with a low cut-offfrequency of 30 Hz and a high cut-off frequency of500 Hz. A notch filter was used occasionally to fil-ter out the 60 Hz noise. All signals were sampledat a rate of 1,000 Hz and stored digitally on op-tic disks. Programmed stimulation was deliveredthrough a multi-channel programmable stimulator(Bloom Associates Ltd., Reading, PA, USA).

The presence of a left-sided accessory path-way was confirmed by standard criteria. Allpatients received adequate anticoagulation withintravenous heparin after successful transseptalpuncture. RF ablation was typically performedwith steerable ablation catheters (EP Technology,Inc., San Jose, CA, USA) during entrainment ofthe tachycardia from the right ventricle using atemperature-controlled RF energy generator (EPTechnology, Inc.). The temperature measured atthe tip of the ablation catheter was typically setat 50–65◦C. The duration of each RF energy appli-cation varied from 30 seconds to 2 minutes.

Demonstration of Bidirectional ConductionBlock and Determination of Left Atrial EffectiveRefractory Period

In the prospective group, atrial overdrive pac-ing during sinus rhythm was performed either me-dial or lateral to the RF lesion to demonstrate thepresence of intraatrial conduction block in the pos-terior atrioventricular vestibule of the left atrium.The left atrial effective refractory period (ERP) wasalso determined in 9 patients studied prospec-tively, using either the coronary sinus catheter orthe steerable ablation catheter. Single atrial ex-trastimuli were delivered after a drive train of eightbeats at a basic cycle length of 250 ms. The pacingoutput was set at twice the diastolic pacing thresh-old with a pulse width of 2 ms. The atrial ERP atthe site of pacing was defined as the longest cou-pling interval that failed to capture the atrium.

PACE, Vol. 28 July 2005 693

CHENG, ET AL.

Gross Anatomy and Histologic Examinationof Autopsy Specimens

Six cadaver hearts were studied. All heartscame from patients with no cardiac history and au-topsy revealed no evidence of structural heart dis-ease. Age at death ranged from 42 to 54 years. Threewere from females. All hearts had been preservedin 10% buffered formalin. After the endocardiumwas carefully dissected away,14,15 the fiber ori-entation of the subjacent myocardium in the re-gion of the atrioventricular vestibule of the leftatrium was examined. Under gross examination,circumferential bundles were defined as myocar-dial fibers that course parallel to the plane of themitral annulus and longitudinal bundles to thosethat course perpendicular to the plane of mitralannulus. In selected areas, full thickness blocks ofmyocardium were taken for histology. Tissue sec-tions were stained with hematoxylin and eosin forlight microscopy.

Statistical Analysis

All variables were reported as mean ± SD.Comparison between means was performed withthe two-tailed t-test. A P value of <0.05 was con-sidered statistically significant.

ResultsAll patients ultimately had successful abla-

tion of the left-sided accessory pathway except onepatient in whom the conduction over the acces-sory pathway persisted after repeated ablation at-tempts. There were 66 cases of single left lateralaccessory pathways and 29 cases of single left pos-terior pathways.

Changes in the Left Atrial Activation SequenceAfter Discrete RF Lesions

After a limited number of RF energy appli-cation (mean: 1.7 ± 0.9, range: 1–3 applications)which did not interrupt accessory pathway con-duction, reversal of left atrial activation sequencerecorded along the coronary sinus catheter eitherduring pacing or during tachycardia was seen inthree patients in the retrospective group (5%) andnine in the prospective group (29%). All the 12 pa-tients had a structurally normal heart. Of the 12 pa-tients, there were 2 females (age: 23.0 ± 9.7). Nonetook antiarrhythmic drugs, β- or calcium channelblockers or digitalis. Among them, 11 had a left lat-eral accessory pathway and one had a left posteriorpathway.

The reversal of the activation sequence alongthe coronary sinus was observed during RF ap-plication in five patients with a single left lateralpathway. Figure 3 illustrates such reversal in a pa-tient in the retrospective group with an accessory

pathway at the 3:30 o’clock position on the mitralannulus (left anterior oblique view). The catheterpositions were similar to those shown in Figure2 except that the ablation catheter was positionedclose to but not right on the site of accessory path-way during the first RF energy application. RF en-ergy was applied during entrainment of AVRT withpacing from the right ventricle. During RF abla-tion, there was a sudden reversal of the activationsequence of the CS recordings, from a distal-to-proximal to a proximal-to-distal pattern (Fig. 3).However, the VA intervals at the proximal ablationcatheter and at the His bundle position remainedthe same despite the change of the activation se-quence in the CS, indicating the persistence ofretrograde conduction over the original accessorypathway. In addition, the dominant componentsof the coronary sinus electrograms also reversedthe polarity (Fig. 3). Since bipolar recordings arenot sensitive to changes in direction of propagationunless it is parallel to the recording catheter, such

Figure 3. Simultaneous recordings from surface ECGlead I, and intracardiac electrograms from the proxi-mal ablation catheter (pABL), the His catheter (His), thecoronary sinus catheter (CS) with CS 1–2 being the mostdistal site and CS 7–8 the most proximal site, and theright ventricle catheter (RV) shows a reversal of activa-tion sequence recorded in the coronary sinus during RFenergy application in a patient with left lateral acces-sory pathway. RF energy was applied during entrain-ment of orthodromic tachycardia from the right ventri-cle. The VA interval recorded on the proximal ablationcatheter and the His position remained identical, indi-cating the persistence of retrograde conduction over theleft lateral accessory pathway before and after the CSactivation sequence reversal. In addition, activation ofCS followed activation of anterior septum (His) after thereversal, consistent with a counterclockwise activationsequence due to block at the site. Also, note the reversalof the polarity of the atrial electrograms recorded alongthe CS that suggests the activation wavefront was paral-lel to the CS catheter (see the text).



complete reversal of the electrogram polarity sup-ports the hypothesis that the activation wavefrontwas traveling parallel to the coronary sinus beforeand after reversal of activation sequence. This pa-tient showed no evidence of retrograde conduc-tion after the second RF energy application dur-ing which the ablation catheter was repositionedslightly lateral to the site of the first attempt.

In the prospective group, we performed atrialoverdrive pacing to document the presence of bidi-rectional block after RF ablation. Figure 4 demon-strates bidirectional block after a single RF energyapplication in a patient who underwent success-ful ablation of a left lateral accessory pathway at4:00 o’clock on the mitral annulus (LAO view).The catheter position is shown in Figure 2. Afterthe successful ablation of the accessory pathway,the ablation catheter was moved slightly lateralto the site of RF lesion. Pacing from distal coro-nary sinus uncovered a counterclockwise (fromthe proximal-to-distal CS) conduction block at thesite of RF lesion (Fig. 4A). Similarly, pacing fromthe ablation catheter positioned lateral to the RFlesion revealed a clockwise (from distal to proxi-mal) conduction block (Fig. 4B).

Figure 5 shows the presence of bidirectionalblock in the posterior atrioventricular vestibule ofthe left atrium in a patient with a left posterior ac-cessory pathway that was created by the initial un-successful RF energy application in the proximalcoronary sinus. The catheter positions are shownin Figure 5A. After the first unsuccessful ablationat a site more distal to the final successful site, adouble potential developed at the proximal coro-nary sinus that bracketed the site of the RF lesionduring tachycardia along with the reversal of ac-tivation sequence in the left atrium such that theelectrogram at the distal coronary sinus was laterthan the atrial electrogram at the anterior septum(Fig. 5B), indicating counterclockwise conductionblock in the septal aspect of the posterior atrioven-tricular vestibule of the left atrium. Similarly, pac-ing from distal CS revealed double potential at thesame site, consistent with clockwise conductionblock created by the RF lesion in the spetal aspectof the posterior atrioventricular vestibule of the leftatrium (Fig. 5C).

Effective Refractory Periods and ConductionTime Along the Mitral Annulus

In nine patients in the prospective group, ERPwas determined at the distal coronary sinus epicar-dially and lateral left atrium endocardially withthe steerable ablation catheter. The ERP deter-mined epicardially at the distal coronary sinus andthat determined endocardially at the lateral leftatrium were similar (197 ± 24 ms and 193 ± 20 ms,respectively, P = 0.38). We defined the mitral an-

nular conduction time in 8 patients with left lateralaccessory pathway and evidence of bidirectionalconduction block after RF ablation as the time in-terval between the pacing artifact at the coronarysinus immediately proximal to the site of block andthe electrogram recorded at the ablation catheterpositioned immediately lateral to the RF lesion (asshown in Fig. 4). In one patient with a left posterioraccessory pathway, we failed to capture the prox-imal coronary sinus. In this case, the mitral annu-lar conduction time was calculated by adding thetime interval from the pacing artifact at the distalcoronary sinus to the first component of the doublepotential (E1) and the time interval from the pac-ing artifact to E2 (Fig. 5C). The measured mitralannular conduction time was 118 ± 23 ms whichwas significantly shorter than the atrial ERP deter-mined epicardially at the distal coronary sinus orendocardially at the lateral left atrium (P < 0.001).

Anatomic Evidence of a Muscle Bundle in thePosterior Atrioventricular Vestibule of the LeftAtrium Parallel to the Mitral Annulus

After carefully removing the endocardium,the fiber orientation in the posterior atrioventric-ular vestibule of the left atrium in the six normalcadaver hearts was examined grossly. In four of thesix specimens, there was a circumferential bun-dle of myocardial fibers that could be identifiedcoursing parallel to the mitral annulus in the pos-terior atrioventricular vestibule of the left atrium(Fig. 6). We termed this bundle as inferoposteriorbundle according to its anatomic location in theinferoposterior left atrium. As it approached thebase of the left atrial appendage, the inferoposte-rior bundle divided to form two branches, eachcoursing along either side of the base of left atrialappendage (Fig. 6A). In the remaining specimens,however, the bundle seems to simply deviate frommitral annulus (Fig. 6B). In these specimens, theinferoposterior bundle was continuous with theseptoatrial bundle.16 Histologically, the inferopos-terior bundle appeared to occupy most, if not theentire thickness, of the posterior atrioventricularvestibule of the left atrium (Fig. 6A). The mini-mum width of the bundle was 1.1 ± 0.3 cm (rangefrom 0.8 to 1.5 cm) and its length before it eitherdivided into two branches or deviated from the mi-tral annulus was 3.4 ± 1.2 cm (range: 1.5–5.0 cm).

DiscussionEvidence of Protected CircumferentialConduction in the Posterior AtrioventricularVestibule of the Left Atrium

The major observation of our study is thereversal of left atrial activation sequence after


CHENG, ET AL.

Figure 4. Bidirectional block in the posterior atrioventricular vestibule of left atrium (PAVV) aftera single successful RF energy application in a patient with a concealed left lateral accessory path-way. The catheter positions were shown in Figure 2. (A) Prior to RF energy application, pacingat distal CS led to simultaneous clockwise (distal to proximal) activation along the CS catheter(posterior mitral annulus) and counterclockwise activation along the anterior mitral annulus withcollision in the vicinity of the septum (left panel). After the ablation, the ablation catheter waspositioned immediately lateral to the RF lesion. Pacing from distal CS revealed only a clockwiseactivation along the mitral annulus (right panel), consistent with counterclockwise conductionblock created by the RF lesion in the lateral aspect of PAVV. (B) During orthodromic tachycar-dia prior to RF energy application, the left atrium was activated simultaneously in a clockwisedirection along the posterior mitral annulus (from dCS to pCS) and in the counterclockwise di-rection along the anterior mitral annulus with wave front collision at the septum (left panel). Theablation catheter was positioned at the site of the successful ablation. After the ablation, pacingfrom the ablation catheter positioned immediately lateral to the site of RF lesion revealed only acounterclockwise activation in the left atrium (right panel), consistent with clockwise (distal toproximal) conduction block at the RF lesion. The mitral annulus conduction time in this patientwas determined by the time interval between pacing artifact at the distal CS and local activationat the distal ablation catheter (108 ms). Note the conduction time from the distal CS and theablation catheter increased more than 80 ms after the RF energy application.



discrete RF applications in the posterior atrioven-triuclar vestibule of the left atrium. As shown inFigure 3, the constant VA interval recorded at theablation catheter speaks against the possibility of asecond septal or right-sided pathway being respon-sible for such reversal of atrial activation sequence.In all of these patients, the absence of a second ac-cessory pathway was carefully confirmed. Further-more, bidirectional conduction block in the pos-terior atrioventricular vestibule of the left atriumwas demonstrated with overdrive pacing mediallyand laterally to the RF lesion site in the prospec-tive group. This suggests that atrial myocardium inthe posterior atrioventricular vestibule of the left

RAO VIEW LAO VIEW

RV

HIS

CS

ABL

successfulsite

RV

CS

HIS

ABL

successfulsite

A

Figure 5. (A) Catheter positioning in a patient with left posterior accessory pathway immediatelyafter the first unsuccessful RF energy application. The abbreviations are the same as in Figure 2.The arrow marks the site of subsequent successful ablation. (B) Bidirectional block in the septalaspect of the posterior atrioventricular vestibule of the left atrium (PAVV) after a single unsuc-cessful RF energy application in the same patient of (A). Prior to the ablation, the activationalong the posterior mitral annulus proceeded in a counterclockwise direction (left panel) duringorthodromic AVRT (TCL = 246 ms). The tracings on the right were recorded after the initial un-successful ablation of the pathway and showed the activation along the posterior mitral annulusoccurred after the activation of the anterior mitral annulus and became clockwise during thesame orthodromic AVRT. This is consistent with counterclockwise conduction block created bythe RF lesion in the septal aspect of the PAVV. Note the tachycardia cycle length increased from246 to 338 ms due to the withdrawal of Isuprel. In addition, the ablation catheter was moved tothe inferior tricuspid annulus where the atrial activation was much later than the proximal CS,excluding a second right-sided accessory pathway. A double potential developed at the proximalCS recording site. Note the timing of the two components, E1 and E2, of the double potential withE2 being the latest atrial activation electrogram, indicating the proximal CS recording site brack-eted the site of block. (C) The tracings on the left were recorded prior to ablation during pacingfrom the distal CS, showing continuous conduction in the left atrium along the posterior mitralannulus. Pacing at distal CS after the initial unsuccessful ablation resulted in double potential atthe proximal CS with the second component, E2, occurring after the activation of anterior septum(His), consistent with a clockwise conduction block in the septal aspect of the PAVV (right panel).The mitral annulus conduction time was calculated by adding the time interval from pacing ar-tifact at the distal coronary sinus to the first component of the double potential (E1) and the timeinterval from the pacing artifact to E2 (120 ms).

atrium may be poorly coupled to the reminder ofthe left atrium in some patients.

Our initial finding12 is consistent with theelegant study by Luria et al. who found similarfindings in 11 patients with a left free wallaccessory pathway.13 They postulated that thearea between the left inferior pulmonary veinand the inferolateral mitral annulus might serveas an isthmus for a pathway around the mitralannulus. However, anatomic correlates are lackingto substantiate the hypothesis. Our study sug-gests that inferoposterior bundle in the posterioratrioventricular vestibule of the left atrium mayserve as the anatomic substrate for such protected


CHENG, ET AL.

Figure 5. Continued.



CS

PVsAS Appendage

anterior mitral leaflet

posterior mitral leaflet

Posterior LA

Anterior LA

LAAbase

epicardium

A

CS Appendage

anterior mitral

leaflet

posterior mitral leaflet

AS

Posterior LA Anterior LACS

LAAbase

PVsAS

epicardium

B

Figure 6. Fiber orientation in the posterior atrioventricular vestibule (PAVV) of the left atrium. After removal of theendocardium, the underlying fiber orientation was examined. Four of the six hearts showed a distinct inferoposteriorbundle in the PAVV that coursed parallel to the coronary sinus but of various lengths and widths (see text). Cir-cumferential bundles refer to the myocardial fibers that course parallel to the plane of mitral annulus under grossexamination. Longitudinal bundles refer to those coursing perpendicular to the plane of mitral annulus. (A) In four ofthe hearts, the inferoposterior bundle divided into two branches as it approached the base of the left atrial appendage(curved solid arrows): one branch coursing between the appendage and mitral annulus and a posterior branch skirtingposteriorly and superiorly to the appendage. The histological section taken from the inferoposterior bundle (indicatedby the dashed arrow) is shown at the left lower panel. Note the circumferential fibers occupy the entire thickness ofthe myocardium at this point. The carton (left upper panel) depicts the fiber orientation of the inferoposterior bundleand its anatomic relationship with other structures in the left atrium. (B) In two specimens, the inferoposterior bundledeviated away from the annulus as it approached the appendage (solid curved arrows) and the annular branch wasabsent, being replaced by longitudinal fibers (short straight arrows). Histological slide made parallel to the mitral an-nulus at the site between the circumferential fibers and longitudinal fibers (indicated by the dashed arrow) shows theinterlaced circumferential fibers and longitudinal fibers. The carton (left upper panel) represents the fiber orientationof the inferoposterior bundle in this specimen. LAA = left atrial appendage; AS = atrial septum; PVs = pulmonaryveins; CS = coronary sinus.


CHENG, ET AL.

circumferential conduction. It is not surprising toexpect complete interruption of conduction alongthe inferoposterior bundle with discrete RF le-sions since the minimum width of the inferopos-terior bundle was in the same order of magni-tude as the typical RF lesion size.17,18 Wang etal. have described in a great detail of the orienta-tion of the atrial musculature.14 They found that,epicardially, the lower branch of the Bachmann’sbundle extended to the posterior atrioventricularvestibule of the left atrium in a circumferential ori-entation along the mitral annulus. They also notedthat, endocardially, portion of the septoatrial bun-dle coursed circumferentially along the atrioven-tricular vestibule of the left atrium. Ho et al. fromthe same laboratory also reported there were my-ocardial fibers that form “a broad circumferentialband around the inferior part of the posterior wall”of the left atrium.15 However, the transmural dis-tribution of the circumferential bundle in the pos-terior atrioventricular vestibule of the left atriumhas not been previously defined. Our histologicaldata indicate that, at least in some patients, there isa distinct circumferential bundle, inferoposteriorbundle, that courses along the mitral annulus andextends the entire thickness of the atrial wall inthe posterior atrioventricular vestibule of the leftatrium. Moreover, there is a substantial variationin the width of this bundle along its course as wellas in its length. It is possible that this inferoposte-rior bundle may consist of intermingling fibers thatextend from the Bachmann’s bundle and the cir-cumferential branch of the septoatrial bundle.14–16

Recently, Antz et al. reported that the mus-culature of the coronary sinus forms an electricalconnection between the right and left atria in thecanine heart.19 These connections extend as muchas a few centimeters into the coronary sinus. Olginet al. recently reported a case of reentrant tachycar-dia that was interrupted by circumferential RF le-sions in the coronary sinus.20 Because of the prox-imity to the posterior atrioventricular vestibule ofthe left atrium, the RF application in the poste-rior mitral vestibule may affect the musculatureof the coronary sinus. However, the limited num-ber of discrete RF energy applications applied inour study is unlikely to have resulted in completeinterruption of such musculature, which requireseither surgical incision or circumferential contin-uous RF lesions19,20 Furthermore, interruption ofthe coronary sinus musculature could not explainthe complete reversal of the left atrial activationsequence. In addition, one would expect doublepotentials along the coronary sinus since the ac-tivation sequence in the posterior atrioventricularvestibule of the left atrium and in the coronary si-nus would be in the opposite direction which wasnot observed in our study.20

Jaıs et al. described electrically silent areas inthe left atrium in patients with left atrial flutter,presumably due to scars or fibrosis.11 When suchscarring is extensive, it is conceivable that a nar-row isthmus may be formed in the posterior atri-oventricular vestibule of the left atrium and mayexplain the left atrial flutter around the mitral an-nulus. However, all of the patients in our study hada structurally normal heart, and electrically silentareas or scars were not identified.

Saffitz et al. showed that there is poor lat-eral cell-to-cell connection in canine atrial tis-sue.21 Anisotropy was proposed as the mecha-nism of preferential longitudinal conduction andpropensity for lateral conduction block in atrialmyocardium. We postulate that the protected cir-cumferential conduction in the posterior atrioven-tricular vestibule of the left atrium is due to theanisotropic conduction existing between the in-ferorposterior bundle and the surrounding atrialmyocardium (Fig. 6).

Left Atrial Effective Refractory Periods andConduction Time Around the Mitral Annulus

All of the patients in whom protected circum-ferential conduction was identified had no struc-tural heart disease or history of AFL. Moreover, inthe prospective group, none had inducible AFL inresponse to rapid atrial pacing. This could be ex-plained, at least in part, by the rapid conductiontime around the mitral annulus that is exceededby the left atrial ERP along the mitral annulus. Itappears that left atrial flutters typically occur inpatients with evidence of structurally abnormalleft atrium characterized by scars and regions ofprofound slow conduction.11 In addition, many ofthose patients had a history of atrial fibrillation. Itis also possible that, in those patients, the electricalremodeling process during atrial fibrillation mayalso lead to significant shortening of atrial ERP thatallows for a reduction in the wavelength and facil-itates the development of AFL.22 Thus, the pres-ence of a discrete inferoposterior bundle alone isnot a sufficient condition for the development ofatrial flutter around the mitral annulus.

Limitations

Although we demonstrated the bidirectionalconduction block in the posterior atrioventricularvestibule of the left atrium, the activation sequencealong the anterior mitral annulus was not mappedin detail. Since the left atrial electrograms were notmapped, we cannot completely exclude the pos-sibility that our findings were due to uncouplingof coronary sinus muscle to the left atrium. Fur-ther studies with other mapping techniques suchas the electromagnetic mapping system are needed



to more carefully delineate the activation sequencein the entire left atrium.

We determined the left atrial ERP at an arbi-trary cycle length of 250 ms and the ERP wouldbe shorter than reported if the pacing cycle lengthwas decreased. However, in a recent report of pa-tients with left atrial flutter, the average AFL CLwas reported to be approximately 300 ms.11

Our study alone does not prove that the pro-tected circumferential conduction in the posterioratrioventricular vestibule of the left atrium consti-tutes an integral portion of the reentrant circuit inatrial flutter around the mitral annulus. However, arecent clinical report with detailed mapping in theleft atrium during AFL around the mitral annulussupports our speculation.11 Natale et al. reportedsuccessful ablation of mitral AFL by linear drag le-

sions in the posterior atrioventricular vestibule ofthe left atrium from the mitral annulus toward theright lower pulmonary vein.23

Clinical Implications

There are two important clinical implica-tion of our study. First, acute reversal of theleft atrial activation sequence during ablationof the left-sided accessory pathways does notnecessarily indicate the presence of a secondpathway. This finding should prompt carefulevaluation of the retrograde conduction duringventricular pacing or orthodromic tachycardia.Second, our study provides the electrophysio-logic and anatomic correlates that may explainthe macro-reentrant tachycardia around the mitralannulus.

References1. Klein G, Guiraudon G, Sharma A, Milstein S. Demonstration of

macro-reentry and feasibility of operative therapy in the commontype of atrial flutter. Am J Cardiol 1986; 57:587–591.

2. Cosio FG, Lopez GM, Goicolea A, Arribas F. Electrophysiologicstudies in atrial flutter. Clin Cardiol 1992; 15:667–673.

3. Olgin JE, Kalman JM, Fitzpatrick AP, Lesh MD. Role of right atrialendocardial structures as barriers to conduction during human typeI atrial flutter: Activation and entrainment mapping guided by in-tracardiac echocardiography. Circulation 1995; 92:1839–1848.

4. Kalman J, Olgin J, Lee RJ, Saxon LA, Lesh MD. The anterior barrierin human atrial flutter: Role of the tricuspid annulus. Circulation1995; 92:I–406.

5. Saoudi N, Derumeaux G, Cribier A, Letac B. The role of catheterablation techniques in the treatment of classic (type I) atrial flutter.PACE 1991; 14:1022–2027.

6. Feld GK, Fleck RP, Chen PS, Boyce K, Bahnson TD, Stein JB, CalisiCM, Ibarra M. Radiofrequency catheter ablation for the treatmentof human type 1 atrial flutter. Identification of a critical zone in thereentrant circuit by endocardial mapping techniques. Circulation1992; 86:1233–1240.

7. O’Nunain S, Linker NJ, Sneddon JF, Debbas NM, Camm AJ, WardDE. Catheter ablation by low energy DC shocks for successful man-agement of atrial flutter. Br Heart J 1992; 67:67–71.

8. Cosio FG, Lopez-Gil M, Goicolea A, Arribas F, Barroso JL. Radiofre-quency ablation of the inferior vena cava-tricuspid valve isthmusin common atrial flutter. Am J Cardiol 1993; 71:705–709.

9. Nakagawa H, Lazzara R, Khastgir T, Beckman KJ, McClelland JH,Imai S, Pitha JV, Becker AE, Arruda M, Gonzalez MD, Widman LE,Rome M, Neuhauser J, Wang X, Calame JD, Goudeau MD, JackmanWM. Role of the tricuspid annulus and the eustachian valve/ridgeon atrial flutter. Relevance to catheter ablation of the septal isthmusand a new technique for rapid identification of ablation success.Circulation 1996; 94(3):407–424.

10. Shah DC, Jaıs P, Haıssaguerre M, Chouairi S, Takahashi A, HociniM, Garrigue S, Clementy J. Three-dimensional mapping of the com-mon atrial flutter circuit in the right atrium. Circulation 1997;96:3904–3912.

11. Jaıs P, Shah DC, Haıssaguerre M, Hocini M, Peng JT, Takahashi A,Garrigue S, Metayer PL, Clementy J. Mapping and ablation of leftatrial flutters. Circulation 2000; 101:2928–2934.

12. Cheng J, Dorostkar PC, Lee RJ, Lesh MD, Saxon LA, Scheinman MM.Protected mitral annular circumferential activation uncovered bydiscrete radiofrequency lesions. JACC 1998; 31(Suppl.):367A.

13. Luria DM, Nemec J, Etheridge SP, Compton SJ, Klein RC, ChughSS, Munger TM, Shen WK, Packer DL, Jahangir A, Rea RF, Ham-mill SC, Friedman PA. Intra-atrial conduction block along the mi-tral valve annulus during accessory pathway ablation: Evidence fora left atrial “isthmus.” J Cardiovasc Electrophysiol 2001; 12:744–749.

14. Wang K, Ho SY, Gibson DG, Anderson RH. Architecture of atrialmusculature in humans. Br Heart J 1995; 73:559–565.

15. Ho SY, Sanchez-Quintana D, Cabrera JA, Anderson RH. Anatomyof the left atrium: Implications for radiofrequency ablation of atrialfibrillation. J Cardiovasc Electrophysiol 1999; 10:1525–1533.

16. Papez JW. Heart musculature of the atria. Am J Anat 1920; 27:255–285.

17. Wittkampf FHM, Hauer RNW, Robles de Medina EO. Control ofradiofrequency lesion size by power regulation. Circulation 1989;80:962–968.

18. Langberg JJ, Lee MA, Chin MC, Rosenqvist M. Radiofrequencycatheter ablation: The effects of electrode size on lesion volumein vivo. Pacing Clin Electrophysiol 1990; 13:1242–1248.

19. Antz M, Otomo K, Arruda M, Scherlag BJ, Pitha J, Tondo C, LazzaraR, Jackman WM. Electrical conduction between the right atriumand the left atrium via the musculature of the coronary sinus. Cir-culation 1998; 98(17):1790–1795.

20. Olgin JE, Jayachandran JV, Engesstein E, Groh W, Zipes DP. Atrialmacroreentry involving the myocardium of the coronary sinus: Aunique mechanism for atypical flutter. J Cardiovasc Electrophysiol1998; 9(10):1094–1099.

21. Saffitz JE, Kanter HL, Green KG, Tolley TK, Beyer EC. Tissue-specific determinants of anisotropic conduction velocity in ca-nine atrial and ventricular myocardium. Circ Res 1994; 74:1065–1070.

22. Wijffels MC, Kirchhof CJ, Dorland R, Allessie MA. Atrial fibrilla-tion begets atrial fibrillation. A study in awake chronically instru-mented goats. Circulation 1995; 92(7):1954–1968.

23. Natale A, Richey M, Gery F, Tomassoni F, Beheiny S, RajkovichK, Wides B, Nickell D, Leonelli FM. Clinical characteristics andablation of left side atrial flutter. JACC 1999; 33:116A.


Protected Circumferential Conduction in the Posterior Atrioventricular Vestibule of the Left Atrium:...

Documents

Transcript of Protected Circumferential Conduction in the Posterior Atrioventricular Vestibule of the Left Atrium:...