Canine model of paroxysmal atrial fibrillation and paroxysmal atrial ...

Am J Physiol Heart Circ Physiol 289: H1851–H1857, 2005. First published July 8, 2005; doi:10.1152/ajpheart.00083.2005.

Canine model of paroxysmal atrial fibrillation and paroxysmal atrial tachycardia Moshe Swissa,1 Shengmei Zhou,1 Offir Paz,1 Michael C. Fishbein,2 Lan S. Chen,3 and Peng-Sheng Chen1 1

Division of Cardiology, Department of Medicine, Cedars-Sinai Medical Center; 2Department of Pathology and Laboratory Medicine, David Geffen School of Medicine, University of California, Los Angeles, California; and 3Department of Neurology, Childrens Hospital Los Angeles and Keck School of Medicine of the University of Southern California, Los Angeles, California Submitted 27 January 2005; accepted in final form 21 June 2005

Swissa, Moshe, Shengmei Zhou, Offir Paz, Michael C. Fishbein, Lan S. Chen, and Peng-Sheng Chen. Canine model of paroxysmal atrial fibrillation and paroxysmal atrial tachycardia. Am J Physiol Heart Circ Physiol 289: H1851–H1857, 2005. First published July 8, 2005; doi:10.1152/ajpheart.00083.2005.—Both autonomic nerve activity and electrical remodeling are important in atrial arrhythmogenesis. Therefore, dogs with sympathetic hyperinnervation, myocardial infarction (MI), and complete atrioventricular block (CAVB) may have a high incidence of atrial arrhythmias. We studied eight dogs (experimental group) with MI, CAVB, and sympathetic hyperinnervation induced either by nerve growth factor infusion (n ⫽ 4 dogs) or subthreshold electrical stimulation (n ⫽ 4 dogs) of the left stellate ganglion. Cardiac rhythm was continuously monitored by a Data Sciences International transmitter for 48 (SD 27) days. Three normal control dogs were also monitored. Six additional normal dogs were used for histology control. Paroxysmal atrial fibrillation (PAF) and paroxysmal atrial tachycardia (PAT) were documented in all dogs in the experimental group, with an average of 3.8 (SD 3) episodes/day, including 1.3 (SD 1.6) episodes of PAF and 2.5 (SD 2.2) episodes of PAT. The duration averaged 298 (SD 745) s (range, 7– 4,000 s). There was a circadian pattern of arrhythmia onset (P ⬍ 0.01). Of 576 episodes of PAF and PAT, 236 (41%) episodes occurred during either sustained or nonsustained ventricular tachycardia (VT). Among these 236 episodes, 53% started before VT, whereas 47% started after the onset of VT. Normal dogs did not have either PAF or PAT. The hearts from the experimental group had a higher density of nerve structures immunopositive (P ⬍ 0.01) for three different nerve specific markers in both right and left atria than those of the control dogs. We conclude that the induction of nerve sprouting and sympathetic hyperinnervation in dogs with CAVB and MI creates a high yield model of PAF and PAT.

sympathetic hyperinnervation in dogs with complete atrioventricular block (CAVB) and myocardial infarction (MI). In that study, cardiac nerve sprouting was induced by chronic (4 –5 wk) infusion of NGF to the left stellate ganglion (LSG). The onset of VT was preceded by an increased ventricular rate, suggesting sympathetic activation. However, atrial activation cycle lengths lengthened after the termination of VT, suggesting that VT induced vagal activation (3). Because a combined sympathetic and vagal activation could facilitate the development of atrial fibrillation (AF) (13), it is possible that spontaneous AF episodes could occur during VT in this animal model. Chen’s laboratory (15) subsequently showed that subthreshold electrical stimulation of the LSG can induce even more nerve sprouting and a significantly higher incidence of spontaneous paroxysmal ventricular arrhythmias than NGF infusion to the LSG. Because these dogs had continuous atrial and ventricular recording through implanted Data Sciences International (DSI) transmitters, we performed a retrospective analysis of the atrial rhythms recorded in these dogs. The purpose was to test the hypothesis that atrial sympathetic hyperinnervation in dogs with electrical and structural remodeling can lead to PAF and PAT and provide an experimental model for these arrhythmias. METHODS

PAROXYSMAL ATRIAL FIBRILLATION (PAF) and paroxysmal atrial tachycardia (PAT) are characterized by spontaneous onset and offset of the arrhythmias without any apparent outside intervention. It is difficult to determine the mechanisms of paroxysmal atrial arrhythmias in part because a clinically relevant animal model is not available. Many clinical studies (1, 4) suggested that autonomic nervous system hyperactivity, electrical, and structural remodeling are important factors for the development of atrial arrhythmias in humans. Chen’s laboratory (3) has developed an animal model of paroxysmal ventricular tachycardia (VT), ventricular fibrillation, and sudden cardiac death by the induction of cardiac nerve sprouting and

The research protocol was approved by the Institutional Animal Care and Use Committee of the Cedars-Sinai Medical Center (Los Angeles, CA) and followed the guidelines of the American Heart Association. The studies included data from a total of 17 dogs, including 3 normal dogs with chronic cardiac rhythm monitoring using a DSI transmitter and 6 normal dogs for histological control. In addition, we studied 8 dogs with nerve sprouting, sympathetic hyperinnervation, MI, and CAVB. Six of the latter 8 dogs were included in a previous study (15) on ventricular arrhythmias. We did not realize until the study was published that there were spontaneous atrial arrhythmia episodes in the same dogs. We have now performed retrospective analyses of the data acquired by the DSI transmitters for atrial arrhythmias and performed atrial nerve density analyses in the same dogs. These data on atrial arrhythmia and atrial nerve densities have not been previously reported. Experimental group: dogs with MI, CAVB, and nerve sprouting. The experimental group included eight dogs [23.1 (SD 2.3) kg] that underwent general anesthesia for sterile surgery. The details of the surgical preparation have been previously reported (15). Briefly, CAVB was induced by radiofrequency ablation of the atrioventricular (AV) junction. MI was induced by ligation of the left anterior descending

Address for reprint requests and other correspondence: P.-S. Chen, Rm. 5342, CSMC, 8700 Beverly Blvd., Los Angeles, CA 90048-1865 (e-mail: [email protected]).

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

electrophysiology; stellate ganglion; atrial arrhythmia; pathology

http://www.ajpheart.org

0363-6135/05 $8.00 Copyright © 2005 the American Physiological Society

H1851

H1852

CANINE MODEL OF PAROXYSMAL AF

coronary artery below the first diagonal branch. Subthreshold electrical stimulation was given to the LSG in four dogs, and NGF infusion was given to the LSG in four dogs. We sutured bipolar electrical wires to the atria and to the ventricles. Both atrial and ventricular leads were connected to a DSI transmitter in a submuscular chest pocket for continuous recording with a sampling rate of 1,000 per second. A pacemaker was used for continuous backup pacing at 40 beats/min. Cardiac rhythm was continuously monitored through a DSI transmitter for 48 (SD 27) days. Two additional dogs of that study (15) were monitored only with an implantable cardioverter defibrillator. These latter two dogs were not included in the present study. Control groups. We performed sterile surgery under isoflurane general anesthesia in three normal dogs. Through a skin incision at the level of fourth left intercostal space, a subcutaneous pocket was created, and a DSI transmitter was implanted. Two bipolar electrodes from the DSI transmitter were tunneled through the subcutaneous tissues to the midsternum and to the back of the chest. The wound was closed, and the ambulatory ECG was recorded over a 2-wk period. The data were analyzed to determine the cardiac rhythms of normal dogs. The hearts from an additional six normal dogs were used for histological control. Immunocytochemistry. The hearts from the experimental group (n ⫽ 6 dogs) and from the normal control group (n ⫽ 6 dogs) were removed and fixed in 4% buffered formalin for 1 h and then preserved in 70% ethanol (15). Tissues were obtained from multiple sites of the left and right atria. The nerve markers tyrosine hydroxylase, synaptophysin, and growth associated protein-43 were stained using 5-␮m transmural sections. The immunostaining methods and the sources of antibodies have been reported elsewhere (3). The nerve density was measured by using Image-Pro Plus 4.0 software (15). The nerve density was expressed as the total area of nerve fibers per unit area (in ␮m2/mm2). Statistical analysis. We analyzed the cardiac rhythm manually. For normal control dogs, we analyzed all 14 days of recordings. For dogs in the experimental group, we analyzed a total of 11–28 days of the data (see Table 1 for details on each dog). A minimal of 11 days and at least 2 days/wk were analyzed for each dog. Days with poor data quality were excluded from analyses. The actual days analyzed in each dog are presented in Table 1. We also divided the events into 4-h periods and computed the percentage of events for each dog in each time period. We then used a Friedman nonparametric rank test for

repeated measures data to determine whether the event occurred randomly throughout the day. The same statistical methods were used to analyze the occurrence of ventricular arrhythmia throughout the day. Data were presented as means (SD). Nonpaired t-tests were used to compare the means of nerve densities. ANOVA with Dunn’s (Bonferroni) correction was used to compare the means of ⱖ3 groups. P ⱕ 0.05 was considered significant. RESULTS

PATs in experimental group. Atrial tachyarrhythmia was diagnosed when there was an abrupt (⬎50 beats䡠min⫺1 䡠s⫺1) increase in the atrial rate to ⬎210 beats/min and persisted for at least 5 s. The rate of 210 beats/min was selected because it was ⬎3 SD from the mean of the normal heart rate. We found in each dog multiple episodes of atrial tachyarrhythmias (⬎210 beats/min) with abrupt onset and offset. AF is defined by a rapid and irregular rhythm, whereas atrial tachycardia (AT) is defined by a rapid and regular rhythm. We used the term AT rather than atrial flutter to be more inclusive. PAF and PAT were documented in all dogs of the experimental group, with an average of 3.8 (SD 3) episodes/day, including 1.3 (SD 1.6) episodes of PAF and 2.5 (SD 2.2) episodes of PAT. The duration averaged 298 (SD 745) s (range, 7– 4,000 s). Table 1 summarizes the characteristics of the atrial arrhythmia documented by the DSI transmitters in each dog studied. Figure 1A shows an example of PAF that occurred during sustained VT. Figure 1B shows an abrupt onset of PAF (arrow), which spontaneously converted to AT. Atrial activations were irregular before the onset of this episode of arrhythmia. Figure 1C shows another spontaneous conversion between AF and AT in the same episode. Figure 1D shows the spontaneous AF termination. When PAF converted to PAT in the same episodes, we counted it as a single episode of PAF. Figure 2A shows an episode of PAT during a period of intermittent VT. Because of complete heart block in both directions, the atrial and ventricular arrhythmias did not interfere with each other. The PAF and PAT did not always occur

Table 1. Cardiac rhythm in experimental and control group Atrial Rhythm, no. of episodes/day Nerve Sprouting Induction

Monitor Duration, days

Analyzed Period, days (%)

SCD Due to VF

Sinus, beats/min

PAF mean (SD)

PAT mean (SD)

Total atrial arrhythmias

Study Dogs LSGS Dog Dog Dog Dog NGF Dog Dog Dog Dog

1 2 3 4

77 25 17 11

26 (37%) 15 (60%) 13 (76%) 11 (100%)

Yes No Yes Yes

104 (SD 21) 99 (SD 25) 109 (SD 26) 114 (SD 21)

2.5 (SD 2.6) 1.5 (SD 1.5) 2.7 (SD 2.4) 2.3 (SD 1.9)

1.7 (SD 1.2) 0.9 (SD 1) 1.9 (SD 1.6) 2.2 (SD 1.9)

4.2 (SD 3.1) 2.4 (SD 1.9) 4.6 (SD 3.4) 4.5 (SD 3.6)

5 6 7 8

82 44 65 62

28 (34%) 19 (43%) 20 (31%) 19 (31%)

No No No No

119 (SD 22) 100 (SD 23) 109 (SD 28) 103 (SD 25)

2.6 (SD 2.1) 1.7 (SD 1.3) 4.3 (SD 2.8) 2.1 (SD 0.8)

2.2 (SD 2.1) 1.1 (SD 0.8) 0 (SD 0) 0 (SD 0)

4.8 (SD 3.6) 2.8 (SD 1.9) 4.3 (SD 2.8) 2.1 (SD 0.8)

0 0 0

0 0 0

Control Dogs None Dog 1 Dog 2 Dog 3

14 14 14

14 14 14

NA NA NA

119 (SD 38) 124 (SD 42) 112 (SD 37)

0 0 0

Paroxysmal atrial fibrillation (PAF) and paroxysmal atrial tachycardia (PAT) values are means (SD); SCD, sudden cardiac death; VF, ventricular fibrillation; LSGS, subthreshold electrical stimulation of the left stellate ganglion stimulation; NGF, nerve growth factor. AJP-Heart Circ Physiol • VOL

289 • NOVEMBER 2005 •

www.ajpheart.org


during VT. For example, Fig. 2B shows AT at an atrial rate of 270 beats/min during a ventricular escape rhythm at 56 beats/ min. Of 576 episodes of PAF and PAT detected in this study, 236 (41%) episodes occurred during either sustained or nonsustained VT. Among these 236 episodes, 53% started before the onset of VT, whereas 47% started after the onset of VT. The remaining episodes occurred during either spontaneous escape rhythm (n ⫽ 133, 23% for PAT) or during ventricular pacing at 40 beats/min (n ⫽ 209, 36% for PAF). We determined the exact time of onset in 432 episodes of PAF or PAT. These episodes represent 61% to 100% of the episodes in each dog [54 (SD 20) episodes/dog]. There is an obvious circadian variation of the incidence of PAF and PAT (P ⬍ 0.01) in all dogs studied (Fig. 3). The arrhythmias occurred most often in the morning and early afternoon and

H1853

less frequently in the evening. Figure 4 shows the occurrence of atrial arrhythmias (AF or AT) over time in five dogs with the longest follow-up period. The results show that paroxysmal atrial arrhythmias started in week 1 and persisted for up to 14 wk after the first surgery. We noted that weeks 3 and 4 in all five dogs were associated with a lower incidence of arrhythmia than weeks 1 and 2 (or weeks 5, 6, and 7) (P ⬍ 0.06 by sign test). Atrial rhythms of normal dogs. The DSI recordings of three normal dogs showed no episodes of PAF or PAT. However, all dogs had significant sinus arrhythmias. The sinus rate ranged from 50 to 180 beats/min, with significant sinus arrhythmias characterized by an abrupt increase or decrease of the heart rate. The rate changed from 55 to 120 beats/min and back to baseline within 5 s in this episode.

Fig. 1. Paroxysmal atrial fibrillation (PAF) and paroxysmal atrial tachycardia (PAT) during ventricular tachycardia (VT) in dog 4 of the experimental group. A: simultaneous ECG and bipolar atrial electrogram (EGM) recording of PAF that lasted 140 s. Arrow points to onset of atrial fibrillation (AF). Ventricular rhythm during this episode of PAF was sustained VT at a rate of 130 beats/min. B–D: selected 10-s segments of same PAF episode. B: initial rhythm was a short period of AF that lasted 0.5 s (arrow), followed by spontaneous conversion of AF to atrial tachycardia (AT). C: (not continuous) EGM documented conversion between AF and AT in same episode. Abrupt termination of AF was documented in EGM (D), whereas VT continued. Sinus node recovery time (SNRT) was 3 s. Irregular atrial activation was seen both before and after PAF episode. A, atrial EGM; V, distant ventricular EGM. AJP-Heart Circ Physiol • VOL

289 • NOVEMBER 2005 •

www.ajpheart.org

H1854


Fig. 2. PAT in experimental group. A: AT during intermittent VT (dog 1). EGM (top) shows 60-s segment during which there is continuous atrial tachyarrhythmia consistent with AT and intermittent VT, and EGM (bottom) shows 10-s segment of same episode. B: EGM shows AT during spontaneous ventricular escape rhythm (dog 5, experimental group).

Cardiac nerve density. The nerve densities at the left and right atria were all significantly (P ⬍ 0.01) higher in the experimental group than in normal control dogs. Figure 5 shows typical examples of atrial nerve distribution in the experimental group as determined by immunostaining using three different markers. Abundant nerves are present in both the right and the left atria. Figure 6 shows the nerve densities of the experimental and the control groups. The experimental group showed significantly higher nerve densities than the control group for all three nerve specific markers. DISCUSSION

Continuous subthreshold electrical stimulation of the LSG or NGF infusion to the LSG induces atrial nerve sprouting and sympathetic hyperinnervation in dogs with MI and CAVB. These dogs developed multiple episodes of spontaneous PAF AJP-Heart Circ Physiol • VOL

and PAT within 1–3 mo after the first surgery. The arrhythmias often occurred in association with VT and were more frequent in the morning than in the evening. In comparison, normal dogs did not have any PAF or PAT episodes. These results show that nerve sprouting and sympathetic hyperinnervation in dogs with electrical and structural remodeling provide an animal model of spontaneous PAF and PAT. Animal models of PAF. The incidence of AF in healthy domestic animals is low (2). In a series of 3,000 dogs brought to a large veterinary clinic for examination, 124 abnormal arrhythmia or conduction disturbances ware found in 95 animals. These 124 episodes included 14 atrial premature contractions, 13 AF, 2 atrial flutter, and 3 AT, giving an incidence of 1% (9). Machida et al. (7) evaluated 285 apparently healthy Holstein dairy cows over an 18-mo period. AF was noted in seven cows with an incidence of 2.5%. The authors noted no

289 • NOVEMBER 2005 •

www.ajpheart.org


H1855

Fig. 3. Frequencies of PAF and PAT over 24-h period in 6 dogs. We divided the events into 4-h periods and computed percentage of events for each dog in each time period. There is a circadian pattern of arrhythmia frequencies (P ⬍ 0.01), with the peak incidence in early morning.

particular disease conditions in any case at the time of onset of AF and thereafter. Because there has been no high-yield and clinically relevant animal model of PAF, it has been difficult to study the mechanism by which AF occurs spontaneously. Mechanisms of paroxysmal atrial tachyarrhythmias. Without multiple control groups, it is not possible to determine the relative importance of MI, CAVB, and nerve sprouting in creating this animal model. However, it is known that autonomic nerve activity is important in atrial arrhythmogenesis (4). There is a circadian trend for the onset of PAF in humans (16, 17). Sharifov et al. (13) reported that direct infusion of isoproterenol and adrenalin into the sinus node artery can induce AF in 21% of open-chest dogs. Acetylcholine infusion induced AF in 100% of the dogs. Acetylcholine-mediated AF was facilitated by isoproterenol, which decreased the threshold acetylcholine concentration for AF induction and increased the AF duration. These data suggest that both autonomic systems play a role in AF. Scherlag et al. (12) reported that endovascular stimulation of the autonomic nerves within the left pulmonary artery can induce PAF. Similarly, high-frequency electrical stimulation of the pulmonary veins and superior vena cava during atrial refractoriness can induce atrial arrhythmias

(11). The response to the high-frequency stimulation is blunted or prevented after ␤-receptor blockade and abolished by atropine. These findings support the importance of autonomic tone in the induction of atrial tachyarrhythmia in this model. We noted that the incidence of atrial arrhythmia was lower in weeks 3 and 4 than in earlier weeks 1 and 2 and later weeks 6 and 7. The mechanisms by which this biphasic changes occurred are unclear. Jardine et al. (6) recently documented increased cardiac sympathetic nerve activity after acute MI in a sheep model. Their results showed significantly increased sympathetic discharges for at least 7 days after infarction. An elevated sympathetic discharge as well as acute surgical trauma could be responsible for a high incidence of arrhythmia in the first 2 wk of monitoring in the present study. The increased incidence of atrial arrhythmia in weeks 6 and 7 might be related to nerve sprouting and increased cardiac sympathetic activation. However, because we did not directly monitor sympathetic nerve activity in this study, these hypotheses remain untested. Association of paroxysmal atrial tachyarrhythmias and VT. We observed in this animal model a high temporal association between PAF and PAT with the development of VT. Because

Fig. 4. Occurrence of atrial arrhythmias (AA) over time. We plotted incidences of AA (both AF and AT), documented by Data Sciences International transmitters in five dogs that have longest duration of monitoring. Graph shows that AA can occur at any time during recording. AJP-Heart Circ Physiol • VOL

289 • NOVEMBER 2005 •

www.ajpheart.org

H1856


Fig. 5. Growth associated protein-43 (GAP-43), synaptophysin (Syn), and tyrosine hydroxylase (TH) immunolabeled cardiac nerves (brown twigs) in dog 2 of the experimental group (original magnification ⫻40). RA, right atrium; LA, left atrium.

all dogs have CAVB, it was impossible for VT to have induced secondary AF through rapid retrograde AV nodal conduction and for AF to have induced VT through rapid anterograde AV nodal conduction. Rather, the atrial arrhythmia and ventricular arrhythmia in this model might be separate events that shared the same trigger. Because of a statistically significant circadian distribution of PAF and PAT and because the same circadian distribution was observed in VT in the same animal model (3), we propose that there is a common trigger for both atrial and ventricular arrhythmias in this model. Chen’s laboratory (3) has previously analyzed atrial and ventricular rate before the onset of paroxysmal VT in this canine model of sudden death. The onset of VT was preceded by accelerated idioventricular escape rhythm, suggesting sympathetic activation leading to

the onset of VT. However, after the onset of VT, the atrial rate decreased, suggesting increased vagal tone. These findings suggest that increased sympathetic tone before VT and/or increased vagal tone during and after VT might have played a role in the generation of PAF or PAT in at least some of the PAF and PAT episodes in this model. An alternative explanation for a temporal association between atrial and ventricular arrhythmias is a reflex-induced change in the cardiac nervous system. The onset of atrial arrhythmia could activate autonomic reflex and induce ventricular arrhythmia, whereas the onset of ventricular arrhythmia could activate autonomic reflex and induce atrial arrhythmia. Because we did not measure sympathetic nerve activity, whether or not there is arrhythmia-induced autonomic activation is unclear.

Fig. 6. Atrial nerve densities. Filled bars show nerve densities from dogs with either subthreshold left stellate ganglion stimulation (LSGS, n ⫽ 4 dogs) or NGF infusion (n ⫽ 2 dogs). Opened bars show nerve densities of control (CNT) dogs. Top: number of nerves per mm2. Bottom: area of nerves (␮m2/mm2). Difference between experimental group and control group is statistically significant (*P ⬍ 0.05).

AJP-Heart Circ Physiol • VOL

289 • NOVEMBER 2005 •

www.ajpheart.org


Interaction between sympathetic tone and electrophysiological remodeling. In addition to nerve sprouting, concomitant atrial structural and electrophysiological remodeling is also known to occur after the creation of CAVB and/or MI in animals and in humans (8, 10, 14). These remodeling processes might facilitate the development of cardiac arrhythmia. Yamashita et al. (18) have demonstrated a circadian variation of potassium channel gene expression in rat hearts, further supporting the idea that ion channel function could also play a role in the circadian variation of cardiac arrhythmias. We propose that the mechanisms of PAF and PAT in this canine model most likely were due to an active interaction between autonomic tone and electrophysiological alterations. Limitations. We did not directly record sympathetic or vagal nerve activity in this study. Therefore, whether or not sympathetic or vagal discharges are causally related to the onset of atrial and ventricular arrhythmias in this model remain undetermined. We have not tested the atrium for electrophysiological or structural remodeling. Although others have reported that the remodeling processes can occur after the creation of MI and/or CAVB in humans and in dogs (8, 10, 14), the contribution of the remodeling processes in atrial tachyarrhythmias of this model remain unclear. The AT was diagnosed based on the electrical activity recorded from a single pair of bipolar electrodes. It is possible that the regular activation is present only at the recording site, whereas the other atrial regions are in fact fibrillating. Therefore, the PAT episodes might be overestimated, whereas the PAF episodes might be underestimated in this study. Clinical implications. Feinberg et al. (5) reported that the overall incidence of AF in the United States is 0.89%, affecting ⬎2 million people. Among this large number of AF patients, roughly two-thirds have sustained AF and one-third have intermittent PAF. The exact trigger of the PAF is unclear, in part because an animal model of spontaneous PAF was not available. In this study we reported a model of PAF and PAT by creating sympathetic hyperinnervation in dogs with MI and complete heart block. There were frequent PAF and PAT episodes in each dog, and a circadian pattern of onset of these atrial arrhythmias. These PAF and PAT episodes often occur simultaneously with VT, which in this model clearly showed a diurnal variation of onset (3, 15). Most importantly, these atrial arrhythmias occurred in an ambulatory state without investigator intervention. This animal model suggests that elevated autonomic nerve activity may be an important trigger of PAF and PAT. This model provides an opportunity to study the mechanisms associated with spontaneously onset of PAF and PAT in large animals. ACKNOWLEDGMENTS We thank Hongmei Li, Avile McCullen, and Elaine Lebowitz for assistance. GRANTS This study was supported by fellowship grants from the Cedars-Sinai Save A Heart Foundation and the Israel Pacing Foundation (to M. Swissa and O. Paz), a North American Society of Pacing and Electrophysiology Michel Mirowski International Fellowship, an American Heart Association Western

AJP-Heart Circ Physiol • VOL

H1857

Region Fellowship and an American Heart Association Scientist Development Award 0435135N (to S. Zhou), a Piansky endowment (to M. C. Fishbein), a Pauline and Harold Price Endowment (to P. S. Chen), and by National Heart, Lung, and Blood Institute Grants R01-HL-78932, R01-HL-71140, P50-HL52319, and R01-HL-66389. REFERENCES 1. Allessie M, Ausma J, and Schotten U. Electrical, contractile and structural remodeling during atrial fibrillation. Cardiovasc Res 54: 230 –246, 2002. 2. Buchanan JW. Spontaneous arrhythmias and conduction disturbances in domestic animals. Ann NY Acad Sci 127: 224 –238, 1965. 3. Cao JM, Chen LS, KenKnight BH, Ohara T, Lee MH, Tsai J, Lai WW, Karagueuzian HS, Wolf PL, Fishbein MC, and Chen PS. Nerve sprouting and sudden cardiac death. Circ Res 86: 816 – 821, 2000. 4. Coumel P. Autonomic influences in atrial tachyarrhythmias. J Cardiovasc Electrophysiol 7: 999 –1007, 1996. 5. Feinberg WM, Blackshear JL, Laupacis A, Kronmal R, and Hart RG. Prevalence, age distribution, and gender of patients with atrial fibrillation. Analysis and implications. Arch Intern Med 155: 469 – 473, 1995. 6. Jardine DL, Charles CJ, Ashton RK, Bennett SI, Whitehead M, Frampton CM, and Nicholls GM. Increased cardiac sympathetic nerve activity following acute myocardial infarction in a sheep model. J Physiol 565: 325–333, 2005. 7. Machida N, Nakamura T, Kiryu K, and Kagota K. Electrocardiographic features and incidence of atrial fibrillation in apparently healthy dairy cows. Zentralbl Veterinarmed A 40: 233–239, 1993. 8. Miyauchi Y, Zhou S, Okuyama Y, Miyauchi M, Hayashi H, Hamabe A, Fishbein MC, Mandel WJ, Chen LS, Chen PS, and Karagueuzian HS. Altered atrial electrical restitution and heterogeneous sympathetic hyperinnervation in hearts with chronic left ventricular myocardial infarction: implications for atrial fibrillation. Circulation 108: 360 –366, 2003. 9. Patterson DF, Detweiler DK, Hubben K, and Botts RP. Spontaneous abnormal cardiac arrhythmias and conduction disturbances in the dog. A clinical and pathologic study of 3,000 dogs. Am J Vet Res 22: 355–369, 1961. 10. Popescu BA, Macor F, Antonini-Canterin F, Giannuzzi P, Temporelli PL, Bosimini E, Gentile F, Maggioni AP, Tavazzi L, Piazza R, Ascione L, Stoian I, Cervesato E, and Nicolosi GL. Left atrium remodeling after acute myocardial infarction (results of the GISSI-3 Echo Substudy). Am J Cardiol 93: 1156 –1159, 2004. 11. Schauerte P, Scherlag BJ, Patterson E, Scherlag MA, Matsudaria K, Nakagawa H, Lazzara R, and Jackman WM. Focal atrial fibrillation: experimental evidence for a pathophysiologic role of the autonomic nervous system. J Cardiovasc Electrophysiol 12: 592–599, 2001. 12. Scherlag BJ, Yamanashi WS, Schauerte P, Scherlag M, Sun YX, Hou Y, Jackman WM, and Lazzara R. Endovascular stimulation within the left pulmonary artery to induce slowing of heart rate and paroxysmal atrial fibrillation. Cardiovasc Res 54: 470 – 475, 2002. 13. Sharifov OF, Fedorov VV, Beloshapko GG, Glukhov AV, Yushmanova AV, and Rosenshtraukh LV. Roles of adrenergic and cholinergic stimulation in spontaneous atrial fibrillation in dogs. J Am Coll Cardiol 43: 483– 490, 2004. 14. Sparks PB, Mond HG, Vohra JK, Jayaprakash S, and Kalman JM. Electrical remodeling of the atria following loss of atrioventricular synchrony: a long-term study in humans. Circulation 100: 1894 –1900, 1999. 15. Swissa M, Zhou S, Gonzalez-Gomez I, Chang CM, Lai AC, Cates AW, Fishbein MC, Karagueuzian HS, Chen PS, and Chen LS. Long-term subthreshold electrical stimulation of the left stellate ganglion and a canine model of sudden cardiac death. J Am Coll Cardiol 43: 858 – 864, 2004. 16. Viskin S, Golovner M, Malov N, Fish R, Alroy I, Vila Y, Laniado S, Kaplinsky E, and Roth A. Circadian variation of symptomatic paroxysmal atrial fibrillation. Data from almost 10,000 episodes. Eur Heart J 20: 1429 –1434, 1999. 17. Yamashita T, Murakawa Y, Sezaki K, Inoue M, Hayami N, Shuzui Y, and Omata M. Circadian variation of paroxysmal atrial fibrillation. Circulation 96: 1537–1541, 1997. 18. Yamashita T, Sekiguchi A, Iwasaki YK, Sagara K, Iinuma H, Hatano S, Fu LT, and Watanabe H. Circadian variation of cardiac K⫹ channel gene expression. Circulation 107: 1917–1922, 2003.

289 • NOVEMBER 2005 •

www.ajpheart.org