nadvertent electrical isolation of the left atrial appendageuring catheter ablation of persistent atrial fibrillation
hin Pang Chan, MRCP, Wai Shun Wong, MD, Satchana Pumprueg, MD, Srikar Veerareddy, MD,reedhar Billakanty, MD, Christopher Ellis, MD, Sanders Chae, MD, Daniel Buerkel, MD,ohan Aasbo, DO, Thomas Crawford, MD, Eric Good, DO, Krit Jongnarangsin, MD,atthew Ebinger, DO, Frank Bogun, MD, Frank Pelosi, MD, Hakan Oral, MD, Fred Morady, MD,man Chugh, MD
rom the Division of Cardiovascular Medicine, Cardiovascular Center, University of Michigan, Ann Arbor, Michigan.
amm
CarBtii
Kt
ACa
(
ACKGROUND Left atrial appendage (LAA) isolation is rare anday be associated with impaired transport function and thrombo-mbolism.
BJECTIVE The purpose of this study was to determine the mech-nisms of inadvertent isolation of the LAA during atrial fibrillationAF) ablation.
ETHODS This study consisted of 11 patients (ejection fraction.43 � 0.18, left atrial diameter 51 � 8 mm) with persistent AFho had LAA conduction block during a procedure for AF (n � 8)r atrial tachycardia (AT) (n � 3).
ESULTS LAA conduction block occurred during ablation at theachmann bundle region in 6 patients, mitral isthmus in 3, LAAase in 2, and coronary sinus in 1. The mean distance from theblation site to the LAA base was 5.0 � 1.9 cm. LAA isolation wasransient in all 6 patients in whom LAA conduction was monitorednd was permanent in the 4 patients in whom conduction was notonitored during energy delivery. The remaining patient was
blation. Nine of (82%) the 11 patients have remained arrhyth-ia-free without antiarrhythmic drugs at mean follow-up of 6 � 7onths, and all have continued taking warfarin.
ONCLUSION Electrical isolation of the LAA may occur duringblation of persistent AF and AT even when the ablation site isemote from the LAA. This likely is due to disruption of theachmann bundle and its leftward extension, which courses alonghe anterior left atrium and bifurcates to surround the LAA. Mon-toring of LAA conduction during ablation of persistent AF or AT ismportant in avoiding permanent LAA isolation.
oted to have LAA isolation during a redo procedure before any rights reserved.
Patients with persistent atrial fibrillation (AF) frequentlyequire ablation of the left atrial (LA) substrate, consistingf ablation of complex atrial electrograms and/or linearblation. Although this approach has been shown to beighly effective,1 there is a concern that such extensiveblation may be associated with electrical and mechanicalmpairment of the LA in some patients. Indeed, a prior caseeport showed that the left atrial appendage (LAA) may belectrically isolated during extensive LA ablation for per-istent AF.2 This has important implications for LA trans-ort function and the potential for thromboembolic compli-ations despite maintenance of sinus rhythm. The goal ofhis study was to determine the mechanisms of LAA isola-
Drs. Oral and Morady are co-founders of Ablation Frontiers. Addresseprint requests and correspondence: Dr. Aman Chugh, Division ofardiovascular Medicine, University of Michigan Hospital, Cardiovascularenter, SPC 5853, 1500 East Medical Center Drive, Ann Arbor, Michigan8109-5853. E-mail address: [email protected]. (Received August 10,
ion and to discuss strategies to prevent this complication inatients undergoing catheter ablation of persistent AF.
ethodshe subjects of this study were 11 patients who experiencedAA conduction block/isolation as a result of catheter ab-
ation of persistent AF or postablation atrial tachycardiaAT). All antiarrhythmic medications except for amioda-one were discontinued at least five half-lives prior to ab-ation. Amiodarone was discontinued at least 8 weeks prioro the ablation procedure. All patients were anticoagulatedith warfarin for at least 1 month before the procedure andere bridged with low-molecular-weight heparin followingarfarin discontinuation 3 days prior to the procedure.ransesophageal echocardiography was performed within4 hours of the procedure to exclude LA thrombus. Thelinical characteristics of the patients are listed in Table 1.ine (82%) of the 11 patients had evidence of structuraleart disease (congestive heart failure, ischemic and non-
efect, left ventricular hypertrophy), and all had evidence ofA enlargement. The indication for the procedure duringhich LAA conduction block was observed was AF in 8atients and AT in 3 patients. Eight patients had previouslyndergone a procedure for persistent AF; the other 3 pre-ented for their first ablation procedure for persistent AF.
A decapolar catheter (EZ Steer, Biosense-Webster, Dia-ond Bar, CA, USA) was positioned within the coronary sinus
CS) or recording electrograms and atrial pacing. Intracardiacchocardiography was used to guide a double transseptaluncture. Heparin was infused to maintain an activatedlotting time of 300 to 350 seconds. An open-irrigation,.5-mm-tip deflectable-tip ablation catheter (ThermoCool,iosense-Webster) was used for mapping and ablation. Bi-olar electrograms were recorded at a bandpass of 30 to 500z (EPMedSystems, West Berlin, NJ, USA). A three-di-ensional replica of the LA was created with an electro-
natomic mapping system (CARTO, Biosense-Webster). Tovoid collateral injury to the esophagus, the esophagus wasisualized by a radiopaque probe or barium administration.3
adiofrequency (RF) energy was delivered at a maximumower output of 35 W at a flow rate of 30 mL/min andaximum temperature of 45°C. Power was reduced to 20 to
5 W at a flow rate of 17 mL/min during applications ofnergy near the pulmonary vein (PV) ostia, in the posteriorA, and in the CS.
In 9 of the 11 patients, the procedural end-point wasonversion of AF to sinus rhythm during RF ablation. Theblation strategy included antral PV isolation followed byatheter ablation of sites exhibiting complex electrograms.he acute end-point during the latter step was organizationf the local atrial electrogram. The next step was linearblation at the roof and mitral isthmus, with demonstrationf conduction block as previously described.4,5 If AF didot terminate after these steps, complex electrograms in theight atrium were targeted. Two patients underwent theblation procedure using a combination of these steps with-ut the end-point of acute termination of AF. After we notedhat the LAA may be isolated even when ablation waserformed several centimeters away from the structure,2 weoutinely monitored conduction by placing a ring catheter
able 1 Clinical characteristics of the study patients
AF � atrial fibrillation; AT � atrial tachycardia.
Lasso, Biosense-Webster) in the LAA. w
LAA isolation was defined as the complete eliminationr dissociation of LAA potentials. If LAA conduction waseing monitored during ongoing AF, LA conduction im-airment was defined as reproducible slowing of the LAAycle length by at least 50% while documenting no changen the global AF cycle length. If conduction slowing orlock into the LAA was observed during energy delivery,F current was immediately discontinued.
apping and ablation of ATn patients who underwent a procedure for postablation AT,he arrhythmia was mapped by a combination of activationnd entrainment mapping.6 In patients who presented ininus rhythm, isoproterenol (5–20 �g/min) was infused toacilitate arrhythmia induction. During ablation of AT, theing catheter was placed in the LAA to monitor conductionnto the appendage. RF current was discontinued if energyelivery resulted in LAA conduction slowing or block.fter termination of AT, rapid atrial pacing and pro-rammed atrial stimulation were performed with and with-ut isoproterenol infusion.
ostablation careatients were monitored overnight and discharged the nextay. They were prescribed warfarin and low-molecular-eight heparin until the international normalized ratio was2.0. Patients were also treated with rate-control medica-
ions, but none received rhythm-control medications. Pa-ients were seen in an outpatient clinic at 3 and 6–9 monthsfter the procedure. Rhythm status was assessed with anuto-triggered monitor worn for 30 days or by interrogationf the existing pacemaker/defibrillator.
tatistical analysisontinuous variables are given as mean � SD.
esultslectrophysiologic findingshe presenting rhythm was AF in 6 patients, AT in 2, andinus in 3 (Table 2). Among the 3 patients who presented ininus rhythm, isoproterenol was required to induce AF in 1atient and AT in another. In the third patient (no. 9), norrhythmias could be induced (see later). In patients no. 3nd 5, who presented for their first ablation procedure andad no evidence of structural heart disease, mapping re-ealed confluent areas of very low voltage (�0.5 mV),onsistent with an atrial myopathic process.
Among the 7 patients with AF, including 1 patient inhom AF was induced, RF ablation resulted in AF termi-ation in 6 patients. AF converted to AT in all 6 patients.he mechanisms of AT in these patients were macroreentry
nvolving the left and right atrial septum in 2 patients; smalleentrant circuit near the right upper PV, typical flutter, anderi-mitral reentry in 2 patients each; focal AT from thease of the LAA in one patient; and small reentrant circuitt the anterior base of the LAA in 1 patient. All of these ATs
ere successfully eliminated during RF energy delivery.
wifcwfsI1ep
MIifiswaLpeiptt
iL(3crtippp
cwRcosmpLpdcpar
r
T
P
11
l*†‡
Frab
175Chan et al Left Atrial Appendage Isolation After Catheter Ablation
Among the 3 patients with AT, including 1 patient inhom it was induced, the mechanism was macroreentry
nvolving the LA roof (patient no. 10), focal tachycardiarom the LA septum (patient no. 4), and a small reentrantircuit at the LA septum (patient no. 3). All of these ATsere successfully ablated. After termination of the roof and
ocal septal AT in 1 patient each, both patients underwentuccessful ablation of inducible macroreentrant septal AT.n these 2 patients (no. 4 and 10) and 2 other patients (no.and 11) with macroreentrant septal AT described earlier,
nergy delivery at the right atrial septum was required in 3atients for tachycardia termination.
echanism of LAA conduction impairmentn patients in whom LAA conduction was not acutely mon-tored (n � 4), the rhythm during which LAA block wasrst observed was septal AT in 2 patients (no. 4 and 10) andinus in 2 (no. 1 and 8; Table 2). In the former, LAA blockas noted immediately after linear ablation from the septal
spect of the mitral valve to the anterosuperior aspect of theA, the region of the Bachmann bundle. In the latter 2atients, LAA isolation was observed immediately after RFnergy delivery at the base of the LAA for AF in 1 and ATn the other patient. LAA isolation was permanent in theseatients. Subsequent mapping revealed that the area of elec-rical silence extended beyond the appendage and includedhe anterior wall, portion of the septum, and mitral isthmus.
In patients in whom LAA conduction was acutely mon-tored with a ring catheter (n � 6), the rhythm during whichAA conduction block was observed was AF in 3 patients
no. 2, 6, and 7), septal AT in 1 (no. 11), and sinus in 2 (no.and 5). The sites at which RF energy delivery resulted in
onduction block (Figure 1) were the Bachmann bundleegion in 4 patients either during electrogram-guided abla-ion of AF (Figure 2) or linear ablation of AT, mitralsthmus in 3 patients (Figure 3), and the distal CS in 1atient during an attempt to obtain isthmus block in aatient with peri-mitral reentry (Figures 4A and 4B). In 2
able 2 Electrophysiologic findings and mechanisms of LAA iso
atient no. Prior procedures Rhythm Prior MI line
1 2 AF Yes2 1 AF Yes3 0 SR No4 3 AT Yes5 0 AF Yes6 0 AF Yes7 2 AF No8 1 AF Yes9* 3 SR Yes0 2 AT Yes1 2 SR Yes
AF � atrial fibrillation; AT � atrial tachycardia; BB � Bachmann bundeft atrial appendage; MI � mitral isthmus; NA � not applicable; SR � sPatient no. 9 was noted to have LAA isolation at the time of the redo pCulprit site is site at which radiofrequency energy delivery resulted in coDistance is distance from culprit site to base of LAA.
atients, RF ablation at more than 1 site resulted in LAA C
onduction slowing. Slowing of LAA conduction was notedithin 1 to 9 seconds of initiation of RF current. During theF application that resulted in LAA conduction delay, nooncomitant slowing of the sinus or atrioventricular node,r hypotension, was observed. When LAA conductionlowing was observed during ablation, RF current was im-ediately discontinued. As a result, LAA conduction im-
roved, and none of these patients experienced permanentAA disconnection (Figure 4A). However, in 3 of theseatients (no. 3, 5, and 11), conduction was significantlyelayed such that LAA activation occurred after the QRSomplex (Figure 5). Also, mitral isthmus block was notursued further in the 2 patients (no. 3 and 5) in whom RFblation performed endocardially or epicardially (in the CS)esulted in transient LAA isolation.
In the remaining patient (no. 9) who presented in sinushythm and was noninducible despite isoproterenol infu-
in the study patients
it site† Distance‡ (cm) LAA monitored LAA isolation
of LAA NA No Permanent5.4 Yes Transient3.0 Yes Transient8.4 No Permanent
S 3.1 Yes Transient6.6 Yes Transient
I 5.5 Yes Transientof LAA NA No Permanent
NA NA Permanent5 No Permanent3.2 Yes Transient
the anterosuperior aspect of the left atrium; CS � coronary sinus; LAA �thm.e prior to any radiofrequency ablation.on slowing/block into the LAA.
igure 1 Sites (red circles) at which radiofrequency energy deliveryesulted in left atrial appendage (LAA) conduction block. In two patients,blation at more than one site caused LAA conduction block. Bachmannundle region is located at the anterosuperior left atrium (dashed oval).
lation
Culpr
BaseBBMIBBMI/CBBBB/MBaseNABBBB
le, nearinus rhyrocedurnducti
S � coronary sinus; RSPV � right superior pulmonary vein.
sLRmtrar
etp
mtic
Fis s (CS),
Fd(p
176 Heart Rhythm, Vol 7, No 2, February 2010
ion, mapping revealed dissociated potentials in the LAA,A roof, and left PV antrum prior to any ablation (Figure 6).eview of electrograms from a procedure performed 32onths earlier showed clear evidence of 1:1 conduction into
he LAA during sinus rhythm. Five months before the cur-ent procedure, the patient experienced a transient ischemicttack despite sinus rhythm and an international normalizedatio of 3.2. Interrogation of her pacemaker revealed no
igure 2 Slowing of conduction into the left atrial appendage during ranitiation of RF current (arrow), the LAA cycle length (CL) slows fromimultaneously prolong globally in lead V1 or locally in the coronary sinu
igure 3 Isolation of the left atrial appendage (LAA) during ablation atelivery of radiofrequency energy, conduction into the LAA ceased (asternot shown). LAA conduction was being monitored during distal coronary
roximal bipoles of the ablation catheter (Abl). V � far-field ventricular electrog
pisodes of atrial arrhythmias after she experienced theransient ischemic attack, whereas 4 months earlier, she hadersistent AT/AF requiring cardioversion.
The mean distance (measured on the electroanatomicap) from the site where RF ablation caused LAA conduc-
ion block to the base of the LAA, excluding those patientsn whom the culprit site was the LAA base, was 5.0 � 1.9m. In 2 of the 11 patients, mapping during sinus rhythm
uency (RF) energy delivery at the Bachmann bundle region. Shortly afteran of 150 ms to 300 ms. Note that the atrial fibrillation CL does notas evidenced by periods of continuous electrical activity (red bracket).
ral isthmus in a patient with an atrial myopathy. Within a few seconds ofA conduction resumed shortly after discontinuing radiofrequency energyS) pacing. Also shown are bipolar electrograms recorded by the distal and
diofreqa me
the mitisk). LAsinus (C
ram recorded by ring catheter placed at base of LAA.
rcoBt
CApatwrParacn
DMTiAeswtcrlmmtvi
FIhbatA
177Chan et al Left Atrial Appendage Isolation After Catheter Ablation
evealed asynchronous activation of the right and left atria,onsistent with Bachmann bundle block (Figure 5). Only 1f the 11 patients (no. 11) demonstrated ECG evidence ofachmann bundle block (Figure 6), that is, terminal nega-
ivity of the P wave in the inferior leads.7
linical outcomet mean follow-up of 6 � 7 months after the last ablationrocedure, 9 (82%) of the 11 patients have remained free oftrial arrhythmias in the absence of antiarrhythmic medica-ions. One patient in whom complete mitral isthmus blockas not obtained for fear of LAA isolation experienced
ecurrence of peri-mitral flutter (based on comparison of-wave morphology and cycle length) 6 weeks after herblation procedure. However, after treatment of hyperthy-oidism, she has experienced no further arrhythmias in thebsence of rhythm-control medications. All patients haveontinued taking warfarin, and none has experienced any
igure 4 A: Example of transient isolation of the left atrial appendage (Lmmediately after LAA conduction block was noted, RF energy was discoad easily inducible peri-mitral flutter, and endocardial ablation terminated tlock. The rhythm is sinus tachycardia with first-degree atrioventricular bblation catheter in the distal CS and a ring catheter in the LAA. Image wehat led to transient LAA isolation. Note the distance between the tip of thP � anteroposterior; LAO � left anterior oblique.
eurologic complications after their last procedure. q
iscussionain findingshe principal finding of this study is that the LAA may be
nadvertently isolated during catheter ablation of persistentF or postablation AT. LAA conduction may be impaired
ven when ablation is performed more than 5 cm from thetructure. In a prior case report, 25 minutes of RF energyas required to intentionally isolate the LAA.8 However, in
he current study, only a few seconds of RF energy, in theontext of extensive prior ablation or an atrial myopathy,esulted in LAA conduction block. Isolation of the LAA isikely to be associated with impaired transport function anday predispose to thromboembolic complications despiteaintenance of sinus rhythm. The observation that none of
he patients in whom LAA conduction was monitored de-eloped permanent LAA isolation suggests that monitorings important in avoiding this complication and its conse-
ring radiofrequency (RF) energy delivery in the distal coronary sinus (CS).(arrow), resulting in prompt resumption of LAA conduction. This patientycardia. Additional ablation was being performed in the distal CS for linear: Cinefluoroscopic views of the same patient showing the position of therded during RF energy delivery in the distal CS (for mitral isthmus block)ion catheter and the base of the LAA as demarcated by the ring catheter.
AA) duntinuedhe tachlock. Bre recoe ablat
uences.
MTetlbcrtntma
blcwl“tcl
rpa
Fow mann b
F(rb
178 Heart Rhythm, Vol 7, No 2, February 2010
echanism of LAA disconnectionhe fact that the LAA may be electrically disconnectedven when RF ablation is performed at a distant site implieshe presence of long intra-atrial myocardial connectionsinking the culprit site to the appendage. The Bachmannundle and its relationship to other interatrial and intra-atrialonnections help explain these observations. The septalaphe is the region where the two atria appear to be “knit-ed” together and is the origin of the major interatrial con-ections. The raphe is bounded by the superior vena cava onhe right, the right superior PV on the left, and the Bach-ann bundle anteriorly.9 The Bachmann bundle itself is notdiscrete structure; it resembles a wide band connecting the
igure 5 Example of Bachmann bundle block resulting in asynchronouf the left atrial appendage (LAA). As a result, the LA is activated caudocras positioned at the anterosuperior left atrium, in the region of the Bach
igure 6 Example of delayed isolation of the left atrial appendage (LAAsolid arrows), consistent with complete LAA isolation. Review of the elevealed preserved LAA conduction. Note terminal negativity of the P w
undle block.
ases of the right and left appendages. Subepicardially, theeftward extension of the Bachmann bundle is joined byircular fibers from the anterior and anterosuperior LA,hich then bifurcates to encircle the LAA.9 Deeper to the
eftward extension of the Bachmann bundle, fibers of theseptopulmonary bundle” combine with bands of the “sep-oatrial bundle,” which are endocardial in location, furtherontributing to the myocardium at the base of the LAA andeft lateral ridge.10
Based on the LA myoarchitecture, it is possible to inter-upt LAA conduction at upstream areas, such as endocardialrojections of the Bachmann bundle, downstream sites suchs the anterior/posterior base of the LAA and the mitral
tion of the right atrium (RA) and left atrium (LA) and delayed activationvia the musculature of the coronary sinus (solid arrows). The LA catheterundle. Abbreviations as in previous figures.
ping of the LAA with a ring catheter revealed dissociated atrial potentialsysiologic study from the ablation procedure performed 32 months earlierlead II (dashed arrows), the electrocardiographic hallmark of Bachmann
s activaanially,
). Mapectrophave in
iiplaats
tw2apmirlcig
PTtaaipwb
ppiLdctpalsc
Lgtobpcp
DIu
pgiaitAittd
LAiiridBttiv
tlcmoCmIdiw
tmdAttMwab
SFmsftim
179Chan et al Left Atrial Appendage Isolation After Catheter Ablation
sthmus, and at any point in between. In most of the patientsn this study, the cumulative effect of multiple ablationrocedures likely contributed to conduction slowing or iso-ation of the LAA. Specifically, ablation at the anteriorspect of the left antrum during PV isolation, LA septum,nterior wall, and LAA base while targeting complex elec-rograms, plus linear ablation at the mitral isthmus, in es-ence transected these major intra-atrial connections.
Extensive ablation cannot be invoked to explain the facthat LAA conduction block was observed in 3 patients whoere undergoing their first ablation procedure. Mapping inof these patients revealed extensive areas of low voltage
nd scar, particularly at the anterosuperior and anterior LA,recisely the route of the leftward extension of the Bach-ann bundle. At the outset, these patients likely had poor
nteratrial and intra-atrial coupling, and RF energy deliveryeadily led to further conduction impairment and LAA iso-ation. Real-time monitoring of LAA conduction may berucial in preventing LAA isolation in patients with preex-sting atrial scarring, even in those patients who are under-oing their first ablation procedure.
revalence and prevention of LAA isolationhe epicardial location of the Bachmann bundle and the fact
hat it usually is composed of several bundles of myofibers,s opposed to being a discrete cable-like structure,9,11 prob-bly help explain why Bachmann bundle block and LAAsolation are relatively uncommon. Also, the redundancyrovided by the septopulmonary and septoatrial bundles,ith respect to the myoarchitecture of the lateral LA, maye protective.
Although acute monitoring of LAA conduction was im-lemented relatively recently, it is unlikely that patientsrior to this change left the laboratory with undetected LAAsolation. Because linear block at the mitral isthmus and theA roof is assessed (in patients with persistent AF or AT)uring LAA pacing in our laboratory, LAA conduction isonfirmed in every such patient. Finally, when LAA isola-ion does occur, it is not a subtle finding as the loss of atrialotentials was documented at the anterior wall, the base andpex of the appendage, and the mitral isthmus, that is, aarge portion of the LA itself. Such widespread electricalilence also speaks to the long course of critical intra-atrialonnections.
One could argue that in an effort to minimize the risk ofAA isolation, extensive ablation of complex atrial electro-rams should be avoided during ablation of AF. Bypassinghese sites may prevent injury to the major intra-atrial routesf conduction. Although complex atrial electrograms maye nonspecific,12 they also may be critical in the fibrillatoryrocess.13,14 Thus, it is possible that avoiding these sitesompletely may result in decreased efficacy of the ablationrocedure.
elayed LAA isolationn one patient (no. 9) in this study, LAA isolation was
nrelated to acute RF energy delivery. It is likely that the w
rior procedure performed almost 3 years ago, which tar-eted the PVs, LA septum and anterior wall, and mitralsthmus, resulted in conduction slowing along critical intra-trial connections. Ongoing structural remodeling in thenterim may have resulted in further myocardial loss alonghese routes and widespread scarring and LAA isolation.lthough LAA isolation late after ablation probably is rare,
t may be reasonable to prescribe indefinite anticoagulationo patients who may be at risk of this complication, such ashose with atrial myopathy or very delayed LAA activation,espite maintaining sinus rhythm.
AA isolation during mitral isthmus ablationblation of the mitral isthmus has been shown to be critical
n treating postablation ATs,6 preventing AF following PVsolation,4 and eliminating long-lasting persistent AF.1 Theesults of this study suggest that ablation at the mitralsthmus may be associated with impairment of LAA con-uction, likely as a result of injury to ramifications of theachmann bundle, which extend to the posterior base of
he appendage. Although this likely is a rare phenomenon,he fact that linear ablation is increasingly being performedn patients with persistent AF probably calls for increasedigilance.
Catheter ablation within the CS is frequently required forermination of peri-mitral atrial flutter15 and attainment ofinear block.4 However, it is important to keep in mind thatatheter ablation within the distal CS, even at low power,ay result in LAA isolation in the vulnerable patient. In one
f the study patients, it is likely that energy delivery in theS interrupted the distal leftward extension of the Bach-ann bundle at the epicardial aspect of the mitral isthmus.
n this patient with easily inducible peri-mitral flutter, aecision was made to forgo further mitral isthmus ablationn order to avoid LAA isolation even though she likelyould experience arrhythmia recurrence.Given that the ligament of Marshall is intimately related
o the myocardium of the mitral isthmus and the CS, oneight speculate as to whether its inadvertent elimination
uring catheter ablation could help explain LAA isolation.s mentioned earlier, the multiple layers of myocardium
hat constitute the lateral LA make it unlikely that elimina-ion of a narrow muscular tract such as the ligament of
arshall would lead to LAA isolation. Furthermore,hether the ligament of Marshall is interrupted endocardi-
lly16 or epicardially,17 subsequent LAA isolation has noteen reported.
eptal ATour of the 11 patients who experienced transient or per-anent LAA isolation required ablation of a macroreentrant
eptal AT. Specifically, linear ablation was commencedrom the septal aspect of the mitral valve and extended tohe right-sided PVs, near the Bachmann bundle region. Thiss precisely the route of the leftward extension of the Bach-ann bundle. Because ablation at the right atrial septum
as required for tachycardia termination in 3 of these pa-
tiit
SInttijbtv
t“(Glovsth
ssswoaL
CIatatb
Lp
R
1
1
1
1
1
1
1
1
180 Heart Rhythm, Vol 7, No 2, February 2010
ients, it may be reasonable to start with right atrial ablationn these patients in an attempt to minimize damage to thentra-atrial connections linking the LA septum/anterior wallo the appendage.
tudy limitationsn 4 of the study patients, real-time LAA monitoring wasot performed during RF ablation; hence, the precise abla-ion site that led to LAA isolation is not known with cer-ainty. It also can be speculated that damage to the longntra-atrial connections may be mediated by RF-related in-ury to the atrial vasculature. However, because conductionlock appears to occur along the known intra-atrial conduc-ion routes as opposed to the known atrial vascular routes, aascular mechanism is unlikely.
It has been previously reported that the cycle length ofhe right atrial appendage may be altered during stimulationat a distance” from the superior left ganglionated plexusGP).18 Importantly, such stimulation, or ablation, of thePs has not been shown to affect the amplitude (implying
oss of myocytes) or activation (implying slow conduction)f appendage potentials. Because LAA amplitude and acti-ation were altered as a result of ablation in the currenttudy without concomitant slowing of the sinus or atrioven-ricular node, or hypotension, a GP-related mechanism isighly unlikely.
Although LAA conduction slowing was observed withineconds of initiation of RF current, the results of this studyhould not imply that the appendage may be isolated with aingle lesion. Rather, it is more likely that LAA conductionas severely impaired either by prior ablation or as a resultf a myopathic process, such that further injury (related toblation) to the existing intra-atrial routes readily resulted inAA isolation.
onclusionnadvertent LAA isolation may occur in patients undergoingcatheter ablation procedure for persistent AF or postabla-
ion AT, even if the ablation site is far removed from theppendage. The mechanism most likely is due to injury tohe leftward extension of the Bachmann bundle and its
ranches that surround the base of the LAA. Monitoring
1
AA conduction is straightforward and appears to be im-ortant in preventing LAA isolation.
eferences1. O’Neill MD, Wright M, Knecht S, et al. Long-term follow-up of persistent atrial
fibrillation ablation using termination as a procedural endpoint. Eur Heart J2009;30:1105–1112.
2. Kuhne M, Ho SY, Morady F, et al. Elimination of left atrial appendage poten-tials during radiofrequency ablation near the right superior pulmonary vein.Heart Rhythm 2008;5:475–478.
3. Ruby RS, Wells D, Sankaran S, et al. Prevalence of fever in patients undergoingleft atrial ablation of atrial fibrillation guided by barium esophagraphy. J Car-diovasc Electrophysiol 2009;20:883–887.
4. Jais P, Hocini M, Hsu LF, et al. Technique and results of linear ablation at themitral isthmus. Circulation 2004;110:2996–3002.
5. Hocini M, Jais P, Sanders P, et al. Techniques, evaluation, and consequences oflinear block at the left atrial roof in paroxysmal atrial fibrillation: a prospectiverandomized study. Circulation 2005;112:3688–3696.
6. Chae S, Oral H, Good E, et al. Atrial tachycardia after circumferential pulmo-nary vein ablation of atrial fibrillation: mechanistic insights, results of catheterablation, and risk factors for recurrence. J Am Coll Cardiol 2007;50:1781–1787.
7. Cosio FG, Martin-Penato A, Pastor A, et al. Atrial activation mapping in sinusrhythm in the clinical electrophysiology laboratory: observations during Bach-mann’s bundle block. J Cardiovasc Electrophysiol 2004;15:524–531.
8. Takahashi Y, Sanders P, Rotter M, et al. Disconnection of the left atrialappendage for elimination of foci maintaining atrial fibrillation. J CardiovascElectrophysiol 2005;16:917–919.
9. Ho S, Anderson R, Sanchez-Quintana D. Atrial Structure and fibres: morpho-logical basis of atrial conduction. Cardiovasc Res 2002;54:325–336.
0. Cabrera JA, Ho SY, Climent V, et al. The architecture of the left lateral atrialwall: a particular anatomic region with implications for ablation of atrial fibril-lation. Eur Heart J 2008;29:356–362.
1. Lemery R, Guiraudon G, Veinot JP. Anatomic description of Bachmann’s bundleand its relation to the atrial septum. Am J Cardiol 2003;91:1482–1485, A8.
2. Oral H, Chugh A, Yoshida K, et al. A randomized assessment of the incrementalrole of ablation of complex fractionated atrial electrograms after antral pulmo-nary vein isolation for long-lasting persistent atrial fibrillation. J Am CollCardiol 2009;53:782–789.
3. Nademanee K, McKenzie J, Kosar E, et al. A new approach for catheter ablationof atrial fibrillation: mapping of the electrophysiologic substrate. J Am CollCardiol 2004;43:2044–2053.
4. Haissaguerre M, Sanders P, Hocini M, et al. Catheter ablation of long-lastingpersistent atrial fibrillation: critical structures for termination. J CardiovascElectrophysiol 2005;16:1125–1137.
5. Chugh A, Oral H, Good E, et al. Catheter ablation of atypical atrial flutter andatrial tachycardia within the coronary sinus after left atrial ablation for atrialfibrillation. J Am Coll Cardiol 2005;46:83–91.
6. Chugh A, Wimmer A, Morady F. Elimination of left superior pulmonary veinostial potentials during radiofrequency ablation at the mitral isthmus. HeartRhythm 2007;4:85–87.
7. Hwang C, Wu T-J, Doshi R, et al. Vein of Marshall cannulation for the analysisof electrical activity in patients with atrial fibrillation. Circulation 2000;101:1503–1505.
8. Hou Y, Scherlag BJ, Lin J, et al. Interactive atrial neural network: Determiningthe connections between ganglionated plexi. Heart Rhythm 2007;4:56–63.
