441

Arrhythmogenic Potential of Pulmonary Venous Tissue: Triggersfor Atrial Fibrillation Identified Within the Remnant of a Vein

SANJAY DIXIT, M.D., F.H.R.S., WILLIAM H. SAUER, M.D., DAVID J. CALLANS, M.D., F.H.R.S.,and FRANCIS E. MARCHLINSKI, M.D., F.H.R.S.

From the Cardiovascular Division, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania; and the University of Colorado,Denver, Colorado, USA

Arrhythmogenic Potential of Pulmonary Venous Tissue. Background: Pulmonary veins (PVs)have frequently been identified as triggers for atrial fibrillation (AF), and higher arrhythmogenic potentialof superior PVs has been attributed to their larger size, which can more rigorously support abnormalitiesof impulse formation and/or conduction.

Case Report: Contrary to this belief, we report our observations in a 63-year-old patient with history oflung cancer, S/P left upper lobectomy, undergoing ablation for paroxysmal AF. Circular mapping (Lasso)and ablation (ABL; 8-mm) catheters were deployed in left atrium (LA). Intracardiac ultrasound revealedseparate right superior (RS) and inferior (RI) PVs and a single left PV. Segmented LA anatomy from theCT angiogram images corroborated this, although on the latter there appeared to be a “stump” at superioraspect of the left PV. This stump likely was the remnant of the left superior (LS) PV. Thus, the patent leftvein was likely the dilated left inferior (LI) PV. With the Lasso and ABL deployed at the LIPV ostium andLSPV remnant, respectively, AF was reproducibly seen to initiate with earliest activity in the latter. Singleradio-frequency ablation (RFA) lesion within the LSPV remnant abolished AF triggers. Additional RFAwas done to isolate LI, RS, and RI PVs. Over a follow-up period of 24 months, this patient has remainedfree from AF off any drugs.

Conclusions: Our observations suggest that even very proximal remnants of PVs can serve as triggersfor AF. Recognition of this phenomenon was facilitated by the use of advanced imaging technique and thedeployment of multiple catheters. (J Cardiovasc Electrophysiol, Vol. 20, pp. 441-444, April 2009)

atrial fibrillation, catheter ablation, pulmonary vein, electroanatomical mapping, trigger, lung cancer

Introduction

Pulmonary veins (PVs) have frequently been identifiedas triggers for atrial fibrillation (AF).1,2 In the seminal stud-ies from the Bordeaux group as well as the observationsof other investigators, PV triggers for AF were much morecommonly observed in the superior as compared with theinferior veins.1-3 Although the reason for this discrepancyis not clear, it has been speculated that the higher “arrhyth-mogenic” potential of superior PVs is on account of theirlarger dimensions and/or longer extensions of atrial mus-culature within their confines resulting in a complex archi-tecture, which may support abnormalities of impulse for-mation/conduction.4 This would imply that smaller veins,especially short PV remnants that lack such complex archi-tecture, may be unable to initiate AF. Contrary to this belief,we report our observations in a patient where triggers for AFwere identified in the proximal remnant of the left superior(LS) PV.

Dr. Dixit, Dr. Callans, and Dr. Marchlinski report receiving honoraria fromBiosense Webster for participation in a speaker’s bureau.

Address for correspondence: Sanjay Dixit, M.D., F.R.H.S., 9 FoundersPavilion, Hospital of the University of Pennsylvania, 3400 SpruceStreet, Philadelphia, PA 19104, USA. Fax: +1-215-615-4350; E-mail:Sanjay.Dixit@uphs.upenn.edu

Manuscript received 5 August 2008; Revised manuscript received 21 August2008; Accepted for publication 26 August 2008.

doi: 10.1111/j.1540-8167.2008.01338.x

Case ReportA 63-year-old man was referred to us for ablation of symp-

tomatic paroxysmal AF resistant to antiarrhythmic drugs (fle-cainide and sotalol). The patient had a remote history of lungcancer for which he had undergone left upper lobectomy andwas subsequently deemed cured. The patient was admitted theday before and all antiarrhythmic medications were stopped 3days prior to admission. Our ablation protocol has been pre-viously described.7 Briefly, two decapolar catheters with 5-mmelectrodes and 2-mm interelectrode spacing were placed in thecoronary sinus (CS) and posterolateral right atrium (RA). AnAcuNav diagnostic ultrasound catheter (5.5–10 MHz, 10 F, Acu-son, Sunnyvale, CA, USA) was advanced to the level of fossaovalis in RA. Using long introducer sheaths (USCI Mullins),two serial transseptal punctures were made through whichthe ablation catheter (CARTO-Biosense, Inc., Diamond Bar,CA, USA; 8-mm tip) and decapolar circular mapping catheter(Lasso, Webster, Inc., CA, USA; 20 mm circumference; 1- to–2-mm interelectrode spacing) were advanced into left atrium(LA). A weight-adjusted bolus of unfractionated heparin wasadministered before the first transseptal puncture and subse-quent infusion was titrated to maintain activated clotting time≥ 300 seconds for the duration of the procedure. A detailedelectroanatomic shell of LA was created in sinus rhythm andindividual PVs were tagged and tubed. This electroanatomicshell was “merged” with the LA geometry that was segmentedfrom a CT angiogram of the heart that was acquired the eveningbefore. Validity of merge was confirmed by alignment of keystructures, including LA appendage, roof, midposterior wall,and septum. On the electroanatomic shell, the right superior(RS) and right inferior (RI) veins were widely separated andthere appeared to be two closely approximated left-sided veins(Fig. 1, panels A and B). Intracardiac ultrasound revealed

442 Journal of Cardiovascular Electrophysiology Vol. 20, No. 4, April 2009

Figure 1. Panels A, B, C and D repre-sent the electroanatomic shell and seg-mented CT angiogram images of the leftatrium with pulmonary veins (PVs) in su-perior and posterior projections, respec-tively. Red dot represents remnant of theleft superior (LS) PV. LI = left inferior;RI = right inferior; RS = right superior.

a single left vein ostium that measured 25 mm. On the CTangiogram, there appeared to be a “stump” located superiorand anterior to the proximal aspect of a single left PV (Fig. 1,panels C and D), which was not evident on the electroanatomicLA shell. We felt that this stump represented the proximal rem-nant of the surgically “ligated” LSPV as a result of the left upperlobectomy. Thus, the patent left vein was likely the left inferior(LI) PV that had dilated to compensate for lack of LSPV flow.At baseline (prior to any stimulation), the patient demonstratedfrequent spontaneous atrial premature complexes (APCs) andshort runs of self-terminating AF. The morphology of the APCsand activation pattern of decapolar catheter in the CS (distalto proximal) suggested that these were likely originating fromthe left-sided veins. Thus, the Lasso catheter was deployed atthe LIPV ostium and the ablation catheter was carefully ad-vanced into the LSPV remnant. Frequent APCs (both isolatedand initiating AF) persisted, and during these earliest electri-cal depolarizations were recorded in the distal bipole of theablation catheter followed by sequential activation of the itsproximal bipole and then the Lasso catheter bipoles coveringthe superior aspect of the LIPV ostium (Fig. 2, panels A and B).It was also interesting to note that during sinus rhythm thereappeared to be delayed conduction within the LIPV. A singleradiofrequency ablation lesion at the site of earliest activationin the LSPV remnant abolished the APCs. Additional circum-ferential lesions were delivered to achieve isolation (entry/exitblock) of the LI, RS, and RI veins. Following PV isolation, usinga stimulation protocol consisting of incremental isoproterenolinfusion (up to 20 mcg) and cardioversion of pacing induced AF,no other triggers were identified. The patient had an uneventfulpostprocedure recovery, and, over a follow-up period exceeding2 years, he has not experienced any further AF episodes andremains off all antiarrhythmic drugs.

Discussion

To the best of our knowledge, this is the first instancewhere triggers for AF have been shown to occur in the prox-imal remnant of a surgically ligated LSPV. Another uniqueobservation in this case was the compensatory dilatation of

the LIPV so that it mimicked a common left vein. Thesestructures were correctly identified on the CT angiogram im-ages, and recognition of the electrophysiologic phenomenonwas facilitated by deploying a circular mapping catheter atthe LIPV ostium and an ablation catheter within the LSPVremnant. Thus, we were able to appropriately categorize andsuccessfully target the source of AF.

The unique anatomy and electrophysiologic properties ofPVs have been cited as the reason why they are the most com-mon sites for AF triggers.1-3,5 It has been proposed that AFtriggers are likely the result of microreentry and so are morecommonly observed in the superior veins, which tend to bebigger with longer extensions of atrial musculature resultingin more complex tissue anisotropy that favors abnormalitiesof impulse conduction.5,6 Contrary to this belief, our obser-vations show that even proximal remnants of PVs can bearrhythmogenic. This has implications for the mechanismsunderlying AF triggers. As seen in this case, the consistentinitiation of APCs, both isolated and degenerating into AF bysingle, similarly coupled electrical depolarization recordedon the distal bipole of the catheter in the LSPV remnant,argues against sustained reentry.7 Although we cannot ruleout single reentrant events, the more likely explanation forthis observation may be abnormal automaticity and/or trig-gered activity.8 More importantly, our observations show thatAF triggers can exist fairly close to the PV ostium, and thismakes a strong case for targeting ablation lesions proximallyduring PV isolation. Our observations also highlight the util-ity of advanced imaging techniques such as CT angiogram orMRI in defining unanticipated anatomical variations. With-out the segmented CT images, we would have likely classifiedthese AF triggers as originating from the LA roof. While thiswould not have impacted our ability to successfully targetthem, it might have prevented recognition of this uniquephenomenon. It is interesting to note that intracardiac ultra-sound was unable to identify the PV remnant. This is likelybecause the lack of blood flow in this structure prevented

Dixit et al. Arrhythmogenic Potential of Pulmonary Venous Tissue 443

Figure 2. Panel A: From top to bottomare ECG lead V1, 10 bipolar recordingsof circular mapping (Lasso) catheter lo-cated at the LIPV ostium, recordings fromdistal and proximal bipoles of the ablationcatheter located in the LSPV remnant,and five bipolar recordings from catheterspositioned in coronary sinus (CS) andright atrium (RA). Initial two beats rep-resent sinus rhythm with delayed conduc-tion within the LIPV (arrow heads). Thethird beat is an atrial premature complexduring which distal bipole of the abla-tion catheter (star) shows earliest elec-trical depolarization. Panel B: From topto bottom, tracings are arranged identi-cal to Figure 3, panel A. Initial two beatsrepresent sinus rhythm with delayed con-duction within the LIPV (arrow heads).The third beat is an atrial premature com-plex (APC), which degenerates into AF.Note earliest electrical depolarization inthe distal bipole of the ablation catheter(star) preceding APC. Similar electricalactivity is noticed in the distal bipole ofthe ablation catheter after the first sinusbeat (open arrow). This may represent lo-calized “depolarization” within the LSPVremnant that fails to propagate within orexit the vein.

us from visualizing it. We were also misled by the com-pensatory dilation of the LIPV that on ultrasound appearedconsistent with a left common vein. This case report sup-ports the practice of using two catheters (circular mappingand ablation) during PV isolation. In this instance, havingtwo separate catheters was very useful in correctly localizingand successfully targeting the AF trigger site.

Limitations

Although a single lesion within the LSPV remnant elim-inated the trigger, we still isolated all the veins. We adoptedthis approach because the patient had multiple risk fac-tors including long-standing hypertension, sleep apnea, andmodestly dilated LA, which might have predisposed him to

444 Journal of Cardiovascular Electrophysiology Vol. 20, No. 4, April 2009

additional AF triggers. From our observations we speculatethat the mechanism underlying these triggers was likely ab-normal automaticity and/or triggered activity. However, sincewe did not do any pacing maneuvers and/or pharmacologicintervention, we were unable to confirm this. Furthermore,following surgical truncation, the LSPV remnant could haveundergone fibrotic changes resulting in complex anisotropy,which could support any/all of the above arrhythmia mecha-nisms. We also point out that although we have categorizedthe structure as a proximal remnant of the ligated LSPV, itslocation is in proximity to the left atrial appendage and sowe cannot conclusively rule out the possibility that it mayrepresent atrial diverticulum and/or atrial outpouching, bothof which have been previously described.9,10

Conclusions

Our observations show that even the proximal remnant ofa PV can be the site for AF triggers. Recognition of this phe-nomenon was facilitated by the use of an advanced imagingtechnique and the deployment of multiple catheters.

References

1. Haissaguerre M, Jais P, Shah DC, Takahashi A, Hocini M, QuiniouG, Garrigues S, Le Moroux A, Le Metayer P, Clementy J: Sponta-

neous initiation of atrial fibrillation by ectopic beats originating in thepulmonary veins. N Engl J Med 1998;339:659-666.

2. Chen SA, Hsieh MH, Tai CT, Tsai CF, Prakash VS, Yu WC, Ding YA,Chang MS: Initiation of atrial fibrillation by ectopic beats originatingfrom the pulmonary veins. Circulation 1999;100:1879-1886.

3. Ashar MS, Pennington J, Callans DJ, Marchlinski FE: Localization ofarrhythmogenic triggers of atrial fibrillation. J Cardiovasc Electrophys-iol 2000;11:1300-1305.

4. Ho SY, Cabrera JA, Tran VH, Farre J, Anderson RH, Sanchez-QuintanaD: Architecture of pulmonary veins: Relevance to radiofrequency ab-lation. Heart 2002;86:265-270.

5. Jais P, Hocini M, Macle L, Choi K, Deisenhofer I, Weerasooriya R,Shah DC, Garrigue S, Raybaud F, Scavee C, Metayer PL, Clementy J,Haissaguerre: Distinctive electrophysiological properties of pulmonaryveins in patients with atrial fibrillation. Circulation 2002;106:2479-2485.

6. Mandapati R, Skane A, Chen J, Berenfeld O, Jalife J: Stable microreen-trant sources as a mechanism of atrial fibrillation in the isolated sheepheart. Circulation 2000;101:194-199.

7. Peters NS, Cabo C, Wit AL: In Zipes DP, Jalife J, eds. Cardiac Electro-physiology: From Cell to Bedside. Philadelphia: W.B. Saunders, 1999,pp. 345-349.

8. Dixit S, Gerstenfeld EP, Callans DJ, Marchlinski FE: Mechanisms un-derlying sustained firing from pulmonary veins: Evidence from pacingmaneuvers and pharmacologic manipulation. Pacing Clin Electrophys-iol 2004;27:1-11.

9. Terada H, Tanaka Y, Kashima K, Sannou K, Arima T: Left atrialdiverticulum associated with severe mitral regurgitation. Jpn Circ J2000;64:474-476.

10. Weerasooriya R, Murray C: Left atrial pouch. Europace 2007;9:1141.