IntroductionRecently it has been shown that paroxysmal

atrial fibrillation (AF) can be initiated by atrial pre-mature beats (APBs) originating from ectopic foci,predominantly located within the pulmonaryveins.1–2 Abolition of the focal activity by radiofre-quency catheter ablation has been introduced as acurative approach.1–5 Mapping and ablation isguided mainly by the localization of ectopic beatsand thus by the identification of the arrhythmo-genic pulmonary vein. The ablation procedure istherefore often limited by the absence of APBs.6–7

The incidence of spontaneous and provoked APBsin patients with focally-initiated AF has been not

investigated. The aim of the present study was tononinvasively assess the incidence of APBs at base-line and to analyze the impact of various provoca-tive maneuvers in increasing the number of APBsin patients with documented focally-initiated AF.

MethodsPatients

Fifty consecutive patients with documentedAF initiated by APBs were included in the presentstudy (Fig. 1). The mean duration of the history ofparoxysmal AF was 4 years (range 1–12). Antiar-rhythmic treatment was attempted with a medianof three (range 1–5) drugs; amiodarone was admin-istered in nine patients. All antiarrythmic drugswere withdrawn at least five half-life amiodarone 6months before entering the study. A single electri-cal cardioversion (from 1 to 3) was previously re-quired in 16 patients. Coronary artery diseasewithout remote myocardial infarction was knownin five patients. Arterial hypertension without left

Atrial Premature Beats in Patients with FocalAtrial Fibrillation: Incidence at Baseline andImpact of Provocative ManeuversRODOLFO VENTURA, CHRISTIAN WEISS, STEPHAN WILLEMS, NINA STURM, HANNO KLEMM, and THOMAS MEINERTZFrom the Department of Cardiology, University Hospital Eppendorf, Hamburg, Germany

VENTURA, R., ET AL.: Atrial Premature Beats in Patients with Focal Atrial Fibrillation: Incidence atBaseline and Impact of Provocative Maneuvers. This study evaluated the incidence of atrial prematurebeats (APBs) and the impact of various provocative maneuvers in patients with focally initiated AF. Fiftypatients (39 men, 57 6 0.4 years) with focally initiated, paroxysmal AF underwent Holter recording anda standardized protocol of provocative maneuvers: vagomechanical stimulation, adenosine 12 mg IV, es-molol 500 ng/kg IV, orciprenaline IV 5 mg/50 mL saline 0.9%, and atropine 0.01 mg/kg IV A surface ECGwas recorded for 20 minutes at baseline and following each part of the protocol. High focal activity wasdefined as $ 1 APB/minute. During Holter ECG, 29 (58%) patients had an amount of , 200 APBs, 12(24%) patients , 700 $ 200, and 9 (18%) patients $ 700 APBs. Less than 1 hour of high focal activity wasobserved in 34 (68%) of the 50 patients. Before starting provocative maneuvers 15 (30%) patients had highfocal activity whereas 35 (70%) presented , 1 APB/minute. In 29 (58%) patients APBs increased by $1/min during provocative maneuvers: by vagomechanical stimulation in 11 patients, after adenosine in15, esmolol in 12, orciprenaline in 15, and after atropine in 9 patients. In all patients with $ 1 APB/minat baseline, focal activity decreased or disappeared during some single provocative maneuvers. AF oc-curred in eight patients under provocative maneuvers. No predictive factors of a successful provocativemaneuver were detected with regard to the baseline patients characteristics and Holter results. In conclu-sion, patients with focally initiated AF have a low incidence of spontaneous APBs. Various provocativemaneuvers successfully increase APBs in more than half of the patients; orciprenaline had the highest ef-ficacy. Some provocative maneuvers may suppress APBs in the setting of high focal activity at baseline.(PACE 2002; 25:1467–1473)

ablation catheter, atrial fibrillation, Holter ECG

PACE, Vol. 25, No. 10 October 2002 1467

Address for reprints: Rodolfo Ventura, M.D., Dept. of Cardiol-ogy, University Hospital Eppendorf, Martinistr. 52, 20246Hamburg, Germany. Fax: 149-40-42803-4125; e-mail: ven-tura@uke.uni-hamburg.de

Received July 16, 2001; revised September 3, 2001; acceptedNovember 15, 2001.

Reprinted with permission fromJOURNAL OF PACING AND CLINICAL ELECTROPHYSIOLOGY, Volume 25, No. 10, October 2002

Copyright © 2002 by Futura Publishing Company, Inc., Armonk, NY 10504-0418.

ventricular hypertrophy in the echocardiogramwas present in 11 patients. Increased levels of thy-roidal hormones were not found in any patient.Distinction between vagally- and adrenergicallymediated AF was made following criteria pro-posed by Coumel.8 Basically, AF was consideredas vagally mediated if arrhythmia occurred exclu-sively at rest, favored by digestion or alcohol, orduring a relaxed period after physical or emotionalstress in a setting of bradycardia and in patientswith no structural heart disease. Conversely, AFwas considered as adrenergically mediated if it oc-curred under stress or exercise following an in-crease in heart rate. Since all data in this regardwere anamnestic, patients were assumed as havingpresumably vagally or adrenergically mediatedAF. However, a nonempiric AF classification isquite impossible because strict criteria are lacking,due to the complex interaction between the sym-pathetic and parasympathetic nervous systems.

Thus, 18 (36%) of the 50 patients probablyhad vagally, and 15 (30%) adrenergically, medi-ated AF. The remaining 17 (34%) patients showedevidences of both AF forms. Patient characteris-tics are summarized in Table I. The study com-

VENTURA, ET AL.

1468 October 2002 PACE, Vol. 25, No. 10

Figure 1. Holter electrocardiogram showing an example of atrial fibrillation initiated by atrialpremature beats. The onset is preceded by typical P on T ectopy.

Table I.

Baseline Characteristics of the 50 Patients

Patients (*n) 50Male (n) 39Age (mean 6 SD†) 57 6 0.4 yearsHistory of paroxysmal ‡AF (median) 4 (1–12) yearsVagally mediated AF (n) 18 (36%)Adrenergically mediated AF (n) 15 (30%)Mixed type (n) 17 (34%)Ineffective §AA-drugs (median) 3 (1–5)Episodes of persisent AF (n of patients) 16

1 episode 82 episodes 53 episodes 3

CAD (n) 5Arterial hypertension (n) 11Left atrium diameter (mean 6 SD) 41 6 7 mm

*Number of patients; †standard deviation; ‡atrial fibrillation;§antiarrhythmic.Episodes of persistent AF required electrical cardioversion. Allpatients with CAD had no history of myocardial infarction and allpatients with arterial hypertension had no left ventriclehypertrophy on the echocardiogram.

plied with the Declaration of Helsinski, the proto-col was approved by the ethics committee of theInstitutional Board, and informed consent was ob-tained from all patients.

24-Hour Holter Electrocardiogram (ECG)

During sinus rhythm a Holter ECG (at least 22hours) was performed in each patient. Analysis ofHolter recordings was conducted to determine thetotal number of APBs during the day and night-time. Based on the total number of APBs recordedin 24 hours, patients were arbitrarily subdividedinto three groups: patients with a high ($ 700),medium ($ 200 , 700), and a low (, 200) APBnumber. In addition, the number of hours withmore than one APB/min was calculated. Episodesof AF were evaluated with respect to initiation,time of onset, and duration. Nonconducted APBsand P wave morphologies were not evaluated.

Provocative ManeuversA stepwise approach of various provocative

maneuvers was performed in each patient withthe aim to induce or to increase focal activity. A12-lead surface ECG was recorded for 20 minutesat baseline. Tests were started with vagomechani-cal maneuvers: carotid sinus massage and Val-salva test. Then, patients underwent four differentpharmacologic tests using fixed dosages: (1) intra-venous (IV) bolus of 12 mg adenosine, (2) IV bolusof esmolol at the dose of 500ng/kg bodyweight, (3)infusion of orciprenaline (5mg/500 mL NaCl0.9%) by 20% increment of heart rate, and (4) IVbolus of atropine at the dose of 0.01 mg/kg afterthe heart rate reached the baseline level. Duringand after each single provocative maneuver, a 20-minute ECG was recorded. A 10-minute baselineECG between the single provocative maneuverwas recorded when the APB rate returned to theinitial value. To avoid interactions between thepharmacologic agents used, provocative maneu-vers were performed in the above mentioned orderin all patients. Since the present study was de-signed as noninvasive, atrial pacing and electricalcardioversion were not included in the PM proto-col.

The individual provocative test was consid-ered as effective following an increase of at least 1APB/min. In case of AF induction, provocativemaneuvers were continued following electricalcardioversion. In case of persistent AF furtherprovocative maneuvers were ceased.

Statistical Analysis

All values of APBs are expressed as mean 6SD or median with minimal and maximal value.The incidence of APB in different subgroups wascompared using the two-tailed Student’s t-test. A

value of P , 0.05 was assumed as statistically sig-nificant.

Results24-Hour Holter ECG

During the Holter ECG a median overall num-ber of 134 APBs was recorded (range 2–3,006).There were 29 (58%) patients showing , 200APBs/24 hours 12 (24%) patients between 200and 700/24 hours and 9 (18%) patients . 700APBs/24 hours (Fig. 2). A median of 63 APBs(range 0–1,224) was recorded during the day (be-tween 8:00 AM and 8:00 PM) and a median of 57APBs (range 0–1,782) during nighttime (between8:00 PM and 8:00 AM). In the complete Holter ECGrecording any differences in the overall incidenceof APBs between the day and nighttime were de-tected. The incidence of APBs, during 24 hours ofHolter recording, in all 50 patients appeared ho-mogeneous. Between patients with presumed va-gally and adrenergically mediated AF, no signifi-cant differences with regard to the occurrence ofAPBs could be detected comparing the day andnighttime. At least 1 APB/min was recorded in 97(8.6%) of the overall 1,125 hours of Holter ECGrecording. During the complete Holter ECG therewere more than 8 hours with $ 1 APB/min in four(8%) patients, between 5 and 8 hours in 3 (6%) pa-tients, between 1 and 4 hours in 9 (18%) patients,and less than 1 hour in 34 (68%) patients (Fig. 3).The onset of 27 AF episodes could be recorded in14 patients. All episodes terminated sponta-neously after 1 6 0.6 hours. The initiation of AFwas observed during the day in 13 espisodes and

APBS IN AF

PACE, Vol. 25, No. 10 October 2002 1469

Figure 2. Diagram showing the number of spontaneousatrial premature beats (APBs) detected during a 24-hourHolter electrocardiogram. The majority of patientspresent a low APB number.

during the nighttime in 14 episodes. The focal ac-tivity during the 60 minutes preceding AF onsetwas low (, 1 APB/min) prior to 20 AF episodesand high ($ 1 APB/min) before the remaining sixepisodes. The heart rate did not increase or de-crease significantly before AF onset. Patients withdocumented AF had the following characteristics:a clinical history suggestive of adrenergically me-diated AF (n 5 4), vagally mediated AF (n 5 3)and both AF forms (n 5 7). In the four patientswith presumed adrenergically mediated AF, fourAF episodes were recorded during the day andtwo during the nighttime. In the three patientswith probably vagally mediated AF, threeepisodes were detected during the day and five inthe nighttime. In the seven patients with a proba-bly mixed AF form, six events of AF occurred dur-ing the day and seven during the nighttime. In ad-dition, of the 14 patients who had AF duringHolter ECG, the APB number in the same record-ing was low in 9 (64%) patients, medium in 3(21%), and high in 2 (14%) patients.

Provocative ManeuversImmediately before starting provocative ma-

neuvers, a focal activity of at least 1 APB/min waspresent in 15 (30%) patients while in the remain-ing 35 (70%) a focal activity of , 1 APB/min couldbe detected. The individual provocative maneu-vers showed the following results: carotid sinusmassage was successful in 5 (10%) patients, Val-salva test in 6 (12%), adenosine in 15 (30%), es-

molol in 12 (24%), orciprenaline in 15 (30%), andatropine in 9 (18%) patients. In all 15 patientswith presents focal activity at baseline, at least oneprovocative maneuver caused a decrease (reduc-tion of at least 1 APB/min) of APBs: carotid sinusmessage in 12 (80%) patients, Valsalva test in 12(80%), adenosine in 9 (60%), esmolol in 8 (53%),orciprenaline in 13 (87%), and atropine in 14(93%) patients. In summary, at least a singleprovocative maneuver was effective to increasethe number of APBs in 29 (58%) patients. Suc-cessful provocative maneuvers in patients withand without focal activity at baseline are shown inFigure 4. There was no significant difference ob-served with respect to a positive provocative ma-neuver result comparing patients presenting withpresumed vagally and adrenergically mediatedAF (Table II). Due to various provocative maneu-vers, AF occurred in 8 patients: in 2 patients afteradenosine, in 4 under orciprenaline, and in 2 pa-tients after atropine. Electrical cardioversion wasnecessary in seven patients. All provocative ma-neuvers could be performed in each patient. In ad-dition, each patient returned to the baseline statusfollowing each provocative maneuver after 7 6 6minutes, and a 10-minute baseline ECG was

VENTURA, ET AL.

1470 October 2002 PACE, Vol. 25, No. 10

Figure 3. Diagram showing the number of hours with $1 atrial premature beat (APB) per minute during a 24-hour Holter electrocardiogram. The majority of patientspresent , 1 hour with $ 1 APB/min.

Figure 4. The two diagrams are showing the positive(increase of at least 1 atrial premature beat (APB) perminute) results of all provocative maneuvers. The upperpart is showing successful provocative maneuvers inpatients presenting high focal activity ($ 1 APB/min) atbaseline (before beginning of the provocativemaneuvers). The lower part is showing successfulprovocative maneuvers in patients presenting no focalactivity (, 1 APB/min) at baseline. In the same patientsmore than one provocative maneuver can be successful.

recorded every time. Comparing the results ofHolter ECG and provocative maneuvers no corre-lation could be detected since spontaneous occur-rence of APBs during Holter ECG did not predictthe results of provocative maneuvers. In otherwords, a high incidence of APBs during Holterrecording was not predictive for successfulprovocative maneuvers and vice versa. Table IIIshows the lack of correlation between Holter re-sults and successful provocative maneuvers in allstudy patients.

DiscussionMain Findings

The present study demonstrates that most pa-tients with documented, focally initiated AF havea low number of APBs at baseline. Clusters ofAPBs are observed for only 1 hour a day in the ma-jority of these patients. Provocative maneuvers areeffective in inducing or increasing APBs in morethan half of the patients. However, the number ofAPBs at baseline does not correlate with the resultof provocative maneuvers. No differences in inci-dence of spontaneous APBs and impact ofprovocative maneuvers are detectable between pa-tients with a clinical history suggestive of vagallyand adrenergically mediated AF.

Spontaneous, Atrial, Premature Beats

The finding by Haissaguerre et al.2 that parox-ysmal AF can be initiated from a focal source/shas modified the therapeutic approach to patientswith paroxysmal AF. Radiofrequency catheter ab-lation of the initiating foci has been introduced asa therapeutic option to cure paroxysmal AF.1–3,7

Thus, the presence of APBs is a prerequisite to lo-calize the focal origin of AF, since foci can be lo-cated outside the pulmonary veins in 16%.9,10

Furthermore, each pulmonary vein has the poten-tial to carry the site of origin of AF and thus to bearrhythmogenic.8 However, no data were avail-able about the spontaneous incidence of APBs in

patients with evidence of focally initiated AF.Prior to radiofrequency ablation of AF, the num-ber of spontaneous APBs is known to be low, oftenresulting in long mapping procedures.7 In the pre-sent study, clusters of APBs were limited to a fewminutes during 24 hours of Holter recording in68% of patients. A distinct influence of the auto-nomic nervous system has been hypothesized forthe onset of AF.11,12 It has long been known thatAF could be triggered by vagal or sympatheticstimuli. Clinically, many patients describe the on-

APBS IN AF

PACE, Vol. 25, No. 10 October 2002 1471

Table III.

Distribution of Successful Provocative Maneuversin Various Patient Subgroups, Selected withRegard to Spontaneous Incidence of APBs

During a Holter Recording

EffectiveProvocative

24-hour ManeuversHolter ECG Patients Percentage

Total APBs numberHigh ($ 700)9 patients 6 67%Medium (, 700 $ 200)12 patients 7 58%Low (, 200)29 patients 16 55%Hours with $ 1 APB/min. 8 hours4 patients 2 50%5–8 hours3 patients 2 67%1–4 hours9 patients 5 55%, 1 hour34 patients 20 59%

*Provocative maneuvers.

Table II.

Result of Provocative Maneuvers in Patients with a Clinical History Suggestive of Vagally-,Adrenergically-Mediated, and Mixed Atrial Fibrillation

Vagal †AF Sympathetic AF Mixed AF*PMs 18 (36%) 15 (30%) 17 (34%)

EffectiveVagal provocative maneuvers 3 (17%) 4 (27%) 4 (24%)Sympathetic provocative maneuvers 1 (5%) 2 (13%) 4 (24%)Both 3 (17%) 4 (27%) 3 (18%)Ineffective 11 (61%) 5 (33%) 6 (35%)

*Provocative maneuvers; †atrial fibrillation.

set of AF at rest or during mental or physicalstress. However, in the present study, no differ-ences were detected regarding incidence and cir-cadian variation of APBs between these patientsubgroups. In addition, a lack of correlation be-tween APB incidence and number or duration ofAF episodes was observed. Of the 14 patients whoshowed AF during the Holter recording, 9 (64%)had a low number, 3 (21%) a medium number,and 2 (14%) patients a high number of APBs. Thefocal activity was low (, 1 APB/min) during 60minutes preceding the onset of 20 AF episodesand high ($ 1 APB/min) prior to the remaining 6episodes.

Provocative ManeuversVarious, arbitrarily chosen maneuvers have

been used in previous reports to induce or to in-crease the number of atrial premature beats.6–7,13 Itwas demonstrated that provocative maneuverswere successful in inducing APBs in more thanhalf of the patients. In the setting of absent APBsat baseline, infusion of orciprenaline was the mosteffective drug in inducing APBs. In patients withpreexisting APBs, vagal maneuvers, like adeno-sine and esmolol, were more effective than or-ciprenaline or atropine.

In the present study, single provocative ma-neuvers were frequently associated with a de-crease or disappearance of APBs if focal activitywas present at baseline. Following orciprenalineand atropine, a decrease in the number of APBswas observed in almost 90% of this patient co-hort. These data are partly supported by previouspublished studies. Tai et al.14 reported the effectsof phenylephrine infused in patients with focalAF. Phenylephrine completely or partially sup-pressed ectopic activity originating from the pul-monary veins, indicating a reflex increase in va-gal tone following the adrenergic effects ofphenylephrine as an underlying mechanism. In-terestingly, phenylephrine could not suppressectopic activity originating from the superiorvena cava. Cheung15 demonstrated an increase inspontaneous activity from the pulmonary veinsof the guinea pig after administration of nora-drenaline and a decrease after acetylcholine. Inan experimental study Chen et al.16 could pro-voke ectopic activity from the pulmonary veinswith isoproterenol and the suppressed it withadenosine, propranolol, nifedipin, and acetyl-choline. In addition they demonstrated the pres-ence of various myocardial cells in the pul-monary veins characterized by differentelecrophysiological properties; some of themhaving spontaneous activity. Conversely, in theclinical setting Hsieh et al.9 used a high doseadenosine bolus to induce initiation of AF by ec-

topic beats. In the present study the sameprovocative maneuver could induce APBs in asetting of low focal activity and suppress APBs inpresence of high focal activity. Therefore, the au-thors postulated a different substrate sensitivityto the autonomic nervous system with respect tothe basal condition. Furthermore, various kindsof ectopic foci may respond differently to a cer-tain autonomic influence.

Occurrence of AF was observed in 16% of thepatients; mostly following adrenergic provocativemaneuvers as orciprenaline and atropine admin-istration. The rate of AF occurrence was lowerthan that reported by other authors.4,9 It may bedue to the fact that invasive provocative maneu-vers were not performed like atrial pacing or elec-trical cardioversion, known to induce AF or topredispose the atrium to AF. In addition, the aimof the provocative maneuvers was to induce APBsavoiding AF onset. Thus, drugs were adminis-tered only in standard dose and orciprenalinewas tritrated to obtain an increase in heart rate ofabout 20%. Although the left atrium dimensionwas comparable with that reported in previousstudies, all patients had no structural heart dis-ease in terms of myocardial damage or valve dis-ease.

It is already known that different drug regi-mens for patients with vagally or adrenergicallymediated AF might show different efficacy. Thus,drug treatment with b-blockers was shown to bemore successful in patients with adrenergicallymediated than vagally mediated AF.8 Contrarily,flecainid was reported to be superior compared toother drugs in patients with vagally mediated AF.8The results showed that in patients with pre-sumed vagally mediated AF, vagal maneuverswere not more effective than sympathetic maneu-vers in inducing APBs. Likewise, in patients withpresumed adrenergically mediated AF, sympa-thetic maneuvers were not more successful thanvagal maneuvers in inducing APBs. According tothis finding, focal AF episodes might be influ-enced by both branches of the autonomic nervoussystem. This is supported by the fact that vagallyand adrenergically mediated AF can coexist in thesame patient.17 In addition, many patients reportthe onset of AF related to rest or stress changingduring the years of follow-up.17 In the presentstudy, positive results of PMs could be obtained inthe same patient following vagal or sympathetictests. Therefore, the choice of the PM can not bedirectly deduced from the presumed pattern ofAF.

The results showed no correlation betweenspontaneous, ectopic activity and the impact ofspecific provocative maneuvers, since in patientswith a low, medium, and high number of APBs

VENTURA, ET AL.

1472 October 2002 PACE, Vol. 25, No. 10

during the Holter recording PMs were successfulin 67%, 58%, and 55% of the cases, respectively(Table III). Thus, a low incidence of APBs duringHolter recording is not predictive for unsuccessfulprovocative maneuvers.

Study Limitations

Due to the difficulty to evaluate the P wavemorphology and to recognize nonconducted APBson the surface ECG, data in this setting were notcollected. Further investigations using intracar-diac ECG are needed for evaluation of the P wavemorphology and the incidence of nonconductedAPBs in patients with focally mediated AF. HolterECG and provocative maneuvers were not re-peated in the same patient. Therefore, day-by-dayvariation in incidence of APBs was not evaluated

and no consideration about the reproducibility ofresults could be made.

ConclusionsMost patients with focally initiated, AF have

a low incidence of spontaneous APBs. Provoca-tive maneuvers are successful in inducing APBs inmore than half of the patients; orciprenalineseems to be the most efficacious. In the presenceof focal activity at baseline, provocative maneu-vers, particularly adrenergic agents, can suppressAPBs. Holter recording has no predictive value forthe impact of provocative maneuvers. Therefore,Holter recording of a low, focal activity should notpreclude patients from further preliminary evalu-ation before radiofrequency ablation of focal atrialfibrillation.

APBS IN AF

PACE, Vol. 25, No. 10 October 2002 1473

References1. Jais P, Haissaguerre M, Shah DC, et al. A focal source of atrial fib-

rillation treated by discrete radiofrequency ablation. Circulation1997; 95:572–576.

2. Haissaguerre M, Jais P, Shah DC, et al. Spontaneous initiation ofatrial fibrillation by ectopic beats originating in the pulmonaryveins. N Engl J Med 1998; 339:659–666.

3. Hsieh MH, Chen SA, Tai CT, et al. Double multielectrode mappingcatheter facilitates radiofrequency catheter ablation of focal atrialfibrillation originating from pulmonary veins. J Cardiovasc Elec-trophysiol 1999; 10:136–144.

4. Chen SA, Hsieh MH, Tai CT, et al. Initiation of atrial fibrillation byectopic beats originating from the pulmonary veins: Electrophysi-ologic characteristics, pharmacologic responses, and effects of ra-diofrequency ablation. Circulation 1999; 100:1879–1886.

5. Chen SA, Tai CT, Yu WC, et al. Right atrial focal atrial fibrillation:Electrophysiologic characteristics and radiofrequency catheter ab-lation. J Cardiovasc Electrophysiol 1999; 10:328–335.

6. Shah DC, Haissaguerre M, Jais P, et al. Provocative maneuvres forinducing pulmonary vein ectopy. (abstract) PACE 1999; 22:153.

7. Haissaguerre M, Jais P, Shah DC, et al. Electrophysiological endpoint for catheter ablation of atrial fibrillation initiated from mul-tiple pulmonary venous foci. Circulation 2000; 101:1409–1417.

8. Coumel P. The role of the autonomic nervous system in atrial flut-ter and fibrillation. In N Saoudi, W Schoels, N El-Sherif, (ed.):Atrial Flutter and Fibrillation: From Basic to Clinical Applica-tions. Armonk, NY, Futura Publishing Company, Inc., 1998, pp.

89–102.9. Hsieh MH, Tai CT, Tsai CF, et al. Pulmonary vein electrogram

characteristics in patients with focal sources of paroxysmal atrialfibrillation. J Cardiovasc Electrophysiol 2000; 11:953–959.

10. Haissaguerre M, Shah DC, Jais P, et al. Incidence and topographyof foci triggering atrial fibrillation after successful pulmonary veindisconnection. (abstract) PACE 2001; 24:540.

11. Coumel P. Role of the autonomic nervous system in paroxysmalatrial fibrillation. Clin Cardiol 1990; 13:209–212.

12. Coumel P. Paroxysmal atrial fibrillation: A disorder of autonomictone? Eur Heart J 1994; 15(Suppl. A):9–16.

13. Haissaguerre M, Jais P, Shah DC, et al. Catheter ablation of chronicatrial fibrillation targeting the reinitiating triggers. J CardiovascElectrophysiol 2000; 11:2–10.

14. Tai CT, Chiou CW, Wen ZC, et al. Effect of phenylephrine on focalatrial fibrillation originating in the pulmonary veins and superiorvena cava. J Am Coll Cardiol 2000; 36:788–793.

15. Cheung DW. Electrical activity of the pulmonary vein and its in-teraction with the right atrium in the guinea-pig. J Physiol 1980;314:445–456.

16. Chen YJ, Chen SA, Chang MS, et al. Arrhythmogenic activity ofcardiac muscle in pulmonary veins of the dog: Implication for thegenesis of atrial fibrillation. Cardiovasc Res 2000; 48:265–273.

17. Fioranelli M, Piccoli M, Mileto GM, et al. Analysis of heart ratevariability five minutes before the onset of paroxysmal atrial fib-rillation. PACE 1999; 22:743–749.