- Email: [email protected]
Fixed intercaval block in the setting of atrial fibrillation promotes the development of atrial flutter
Fixed intercaval block in the setting of atrial fibrillation promotes the development of atrial flutter Hanh M. Bui, MD, MPH, Celeen M. Khrestian, BS, Kyungmoo Ryu, PhD, Jayakumar Sahadevan, MD, Albert L. Waldo, MD, FHRS From the Department of Medicine, Division of Cardiology, Case Western Reserve University/University Hospitals Case Medical Center, Cleveland, Ohio. BACKGROUND In the canine sterile pericarditis model, typically only atrial fibrillation (AF) is inducible on postoperative day 2. OBJECTIVE In this model, we tested the hypothesis that on postoperative day 2, placing a fixed line of block (LoB) between the vena cavae critically alters the atrial substrate, favoring the induction of sustained atrial flutter (AFL) instead of AF.
[CL] 110 ⫾ 10 ms), regular, left atrial reentrant driver, which caused fibrillatory conduction. After creation of the LoB, in 5 dogs, rapid atrial pacing now induced sustained AFL (mean CL 167 ⫾ 13 ms). In the 6th dog, AFL failed to develop because the left atrial driver that was induced before the LoB was still reproducibly induced despite the LoB.
METHODS In 6 sterile pericarditis dogs, sustained AF was induced by rapid atrial pacing. After terminating AF, a fixed LoB between the vena cavae was created (cryoablation), and AF reinduction was attempted. Simultaneous mapping from 400 to 420 electrodes on the right and left atrial epicardium and the interatrial atrial septum was performed during all studies.
CONCLUSION In this model of sustained AF, altering the substrate to create a fixed LoB between the vena cavae creates a substrate favoring the induction of AFL.
RESULTS Before creation of the LoB, in all 6 dogs, rapid atrial pacing induced sustained AF because of a rapid (mean cycle length
(Heart Rhythm 2008;5:1745–1752) © 2008 Heart Rhythm Society. All rights reserved.
Introduction
ing the atrial substrate to favor development of a LoB between the vena cavae. Supportive of the importance of atrial substrate changes to the development of AF and then AFL are data from the sterile pericarditis model that show that following creation of the model, virtually only AF is inducible on postoperative days 1 and 2,10,11 but on postoperative days 3 and 4, AFL is primarily inducible despite using the same pacing induction protocols.1,11–13 Noteworthy, the development of spontaneous AF and AFL in patients after open heart surgery follows a similar time course.14,15 Additionally, preventing or minimizing substrate changes in the canine sterile pericarditis model with steroids virtually abolishes induction of AFL, the latter associated with failure to develop a LoB between the vena cavae.11 Despite these associations indicating the importance of the LoB in the region between the vena cavae, it has never been shown that simply altering the atrial substrate by creating a fixed LoB in the region between the vena cavae will convert the induction of AF to AFL. In this study, we explore further the role of an LoB in the region between the vena cavae, testing the hypothesis that in the canine sterile pericarditis model, when only sustained AF is reproducibly inducible, placing a fixed LoB in the region between the vena cavae critically alters the atrial substrate, enabling preferential induction of sustained AFL over AF.
Previous work from our laboratory has shown that atrial fibrillation (AF) of variable, but usually short, duration precedes the development of atrial flutter (AFL).1,2 Key to the development of AFL is that during the preceding AF, a line of block (LoB) must form or already be present in the region between the vena cavae. This LoB is usually functional, but may be fixed.3 The necessary presence of an abnormal atrial substrate for the formation of sustained AFL (and, in fact, for AF) is also well established, beginning with the early work of Lewis et al.4 Creation of an LoB between the vena cavae or anterior in the right atrial free wall permits induction of AFL in otherwise normal atria.5– 8 Furthermore, it also has been recognized that AF can often be converted to AFL by some antiarrhythmic agents, for example, those with sodium channel blocking effects, especially the IC agents.9 Presumably the latter occurs by chang-
Supported in part by Grant R01 HL38408 from the United States Public Health Service, National Institutes of Health, National Heart, Lung and Blood Institute, Bethesda, Maryland, and Blue Dot, Glenstone, and Jennie Zoline Foundations, Washington, DC. Address reprint requests and correspondence: Dr. Albert L. Waldo, University Hospitals Case Medical Center, Division of Cardiovascular Medicine, MS LKS 5038, 11100 Euclid Avenue, Cleveland, Ohio 44106. E-mail address: [email protected]. (Received August 19, 2008; accepted August 28, 2008.)
KEYWORDS arrhythmia; atrial fibrillation; atrial flutter; mapping; electrophysiology
1547-5271/$ -see front matter © 2008 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2008.08.036
1746
Heart Rhythm, Vol 5, No 12, December 2008
Methods
Data acquisition and analysis
All studies were performed in accordance with guidelines specified by our Institutional Animal Care and Use Committee, the American Heart Association Policy on Research Animal Use, and the United States Public Health Service Policy on Use of Laboratory Animals. The study subjects were 6 mongrel 18-kg to 23-kg dogs.
During each episode of AF/AFL, data were simultaneously recorded from the RA free wall, the LA free wall, BB, and the interatrial septum or region between the PVs.10,16 For each episode of sustained AF, 1.2 seconds of data (12 consecutive 100-ms windows) from a midportion of an episode were analyzed to compute isochrone activation maps. Previous studies in this laboratory showed that electrograms recorded from the midportion of a sustained AF episode are representative of the activation pattern for the entire episode.10 For each 100-ms window, activation times (maximum resolution of 1 ms) were determined at each recording site. For AFL, the activation window was determined by the AFL cycle length (CL). Frequency analysis was performed on a subset of electrogram recordings using Fast Fourier transform (FFT) techniques.17,18 The driver and areas that respond 1:1 at the drive CL give rise to a single dominant frequency peak, whereas irregular activation is indicated by the presence of multiple frequency peaks with a broadband.17,18
Creation of the sterile pericarditis model As previously described,10,12 (1) each dog first underwent surgery to create the sterile pericarditis model; (2) before closure of the chest, stainless steel wire electrodes for subsequent use in pacing and recording were sutured to the Bachmann bundle (BB), the right atrial appendage (RAA), the posterior-inferior left atrium, and the right ventricle (RV); (3) standard postoperative care was provided.
Studies in the open-chest state On postoperative day 2, using standard techniques during general anesthesia,10,12 His bundle ablation was performed to create complete atrioventricular block10; the RV was then paced at 60 to 100 beats/min, and the heart was exposed.
Simultaneous multisite mapping Before and after creation of an LoB in the region between the vena cavae, induction of AF was attempted using a programmable stimulator to deliver bursts of rapid atrial pacing.10 Sustained AF was deemed reproducible when a specific pacing protocol induced sustained AF at least twice before creating the LoB. After creating the LoB, reinduction was attempted using the same pacing induction protocol to see whether either AF or AFL were induced. If induction of either was unsuccessful, the pacing protocol was repeated from the beginning. During episodes of induced, sustained AF, and subsequently AFL, simultaneous multisite mapping of atrial activation was performed by recording electrograms from 400 to 420 sites. In all dogs, recordings were made from electrode arrays overlying BB (24 electrodes), the right atrial (RA) free wall (188 electrodes), and the left atrial (LA) free wall (168 electrodes) as previously described.10 In all dogs except one (#4), an additional 20-pole electrode catheter was inserted into the RA via the right internal jugular vein and advanced to record from the interatrial septum.10 In Dog #4, electrograms were recorded from the epicardial surface between the pulmonary veins (PVs) using two 20-pole electrode catheters. One catheter was inserted between the right and left PV, and the other between the superior and inferior PV.
Creation of the LoB After reproducible induction of sustained (ⱖ2 min) AF and its termination (spontaneous or paced), a fixed, transmural LoB approximately 25 mm in length was created with a series of contiguous epicardial cryolesions (– 80°C) to the RA free wall between the vena cavae and slightly anterior to the crista terminalis. After creating the LoB, AF induction was reattempted.
Results Activation patterns during induced, sustained AF before creation of the LoB Before cryoablation, burst rapid atrial pacing induced sustained AF in all 6 dogs. Each episode of induced AF resulted from a stable, rapid, regular LA driver of very short CL (104 to 128 ms, mean 112 ⫾ 6 ms) that was caused by reentry around the inferior PVs in one dog (#4), and appeared to be caused by reentry circulating around 1 or more PVs in the other 5 dogs. We say “appeared to be” because we never mapped the entire reentry circuit. However, in 1 dog (#6), we were able to show clockwise and counterclockwise activation of the putative reentrant circuit, further supporting reentry as the mechanism of the driver. In all AF episodes, the driver served as a stable generator of daughter waves that spread to the rest of the atria via several routes, and produced fibrillatory conduction to variable portions of the LA and much or all of the RA. Maps of these areas showed a changing pattern of activation that included irregular beat-to-beat CLs, variable electrogram morphologies, variable wave fronts, and unstable LoBs. The recorded electrocardiogram (ECG) showed classic AF. In all instances, without a driver, there was no sustained tachyarrhythmia. Thus, although the initial AF was induced by burst rapid atrial pacing (the initial driver), if no new driver formed during or after the atrial pacing, there was no sustained AF. Figure 1 shows data from a representative example of induced, sustained AF before the creation of the LoB in Dog #4. Panel A shows maps of 2 consecutive 100-ms activation windows. A reentrant circuit (orange arrows) courses around the inferior PVs, and sends off daughter waves (gray arrows). Panel B shows an ECG simultaneously recorded with electrograms from sites A-L, denoted in Panel A, over a period of about 900 ms. Electrograms from sites A to E, recorded from within the reentrant circuit, show a regular
Bui et al
Converting AF to AFL
1747
Figure 1 A: Two consecutive 100-ms activation maps from a representative example of AF in Dog #4 before creation of an LoB between the vena cavae. A reentrant driver (orange) circulates around the inferior PVs. In this and subsequent figures, isochrones are at 10 ms intervals, lines with arrows indicate direction of activation, daughter waves are denoted by gray arrows, an asterisk (*) represents an epicardial breakthrough point, dashed lines represent functional lines of block, and gray areas are not activated during the illustrated activation window. A through L are electrogram recording sites shown in Panel B. B: Electrograms simultaneously recorded from sites A through L (shown in Panel A) over a period of approximately 900 ms. Beat-to-beat CL is indicated in milliseconds (ms). See text for discussion. BB ⫽ Bachmann bundle; IVC ⫽ inferior vena cava; LAA ⫽ left atrial appendage; PV ⫽ pulmonary vein; QRS ⫽ ventricular complex; RAA ⫽ right atrial appendage; SVC ⫽ superior vena cava.
beat-to-beat CL of 117 ⫾ 6 ms with electrograms of constant morphology. In contrast, electrograms from sites F, G, and I to L, recorded from outside the reentrant circuit, show a slower rate with an irregular CL and morphology. The regular rhythm of short CL around the PVs serves as the driver, and the irregular rhythm in parts of the LA and most of the RA indicates fibrillatory conduction caused by the driver. Site H recorded from BB is also outside the reentrant circuit, but shows a regular CL identical to that of the driver despite being a daughter wave. Located just superior to the upper PVs, the midportion of BB was usually a part of or very close to the reentrant circuit of the AF driver in these studies. Its activation was usually remarkably regular during AF, even when not part of the reentrant circuit, that is, it usually did not manifest fibrillatory conduction. Figure 2 shows 12 consecutive 100-ms windows from that same episode of AF (Dog #4). In each window, the reentrant circuit travels counterclockwise around the inferior left and right PVs. Daughter waves from this reentrant circuit travel superiorly between the left and right PVs to activate BB. After exiting BB, these daughter waves activate other portions of the LA and the RA, producing variable LoBs (dashed lines) and CLs. Also, a daughter wave travels from the reentrant circuit toward and usually across the intercaval region to activate a portion of the RA. In some activation windows, portions of the LA and RA are not activated (gray areas). The heterogeneity of activation seen
is the result of fibrillatory conduction. This activation pattern continued until the reentrant rhythm terminated, either spontaneously or because of atrial pacing. Figure 3 shows FFT analysis of electrograms recorded from selected atrial sites (Dog #4). Sites within the reentrant circuit show a single, narrow, dominant frequency peak at the driver’s CL (8.54 Hz). Sites not in the reentrant circuit, but activated by daughter waves at the driver’s CL (‡), also have a single dominant frequency peak at the driver’s CL (8.54 Hz). The inferior LA and all the RA sites, however, show a broadband with multiple frequency peaks consistent with fibrillatory conduction. Broadband RA sites between the vena cavae that are near the LA also include a frequency peak at the driver’s CL (8.54 Hz), indicating LA-to-RA conduction from the driver.16,19
Activation patterns during induced, sustained atrial tachyarrhythmias after creation of the LoB Using the same pacing protocol as before creation of the LoB, sustained AFL was now reproducibly induced in 5 of 6 six dogs. These AFL reentrant circuits showed significantly longer CLs (mean 165 ⫾ 13 ms; P ⬍ .05) compared with those during AF (Table 1). Figure 4 summarizes the course of all 5 postablation AFL reentrant circuits. Figure 5 illustrates activation during induced AFL in a representative example (Dog #4) after creation of the LoB between the vena cavae. The activation map (left panel)
1748
Heart Rhythm, Vol 5, No 12, December 2008
Figure 2 Twelve consecutive 100-ms activation windows of AF in Dog #4. The reentrant circuit (orange) travels around the inferior PVs, driving the atria via daughter waves (gray). See text for discussion. Abbreviations as in Figure 1.
Bui et al
Converting AF to AFL
1749
Figure 3 Fast Fourier transform (FFT) analysis of selected electrograms from Dog #4. The reentrant circuit (driver) is orange. ‡Sites outside the reentry circuit activated by daughter waves at the driver’s CL. See text for discussion. Abbreviations as in Figure 1.
shows a figure-of-8 reentrant circuit in the RA that drives the rest of the atria at a CL of 161 ms. The right panel shows an ECG recorded simultaneously with electrograms from the same sites as in Figure 1, except that site H is now recorded from the intercaval region. The reentrant circuit consists of 2 limbs sharing a common pathway in the RA free wall. One limb travels down the RA free wall and up the atrial septum, and the other travels clockwise around the lesion. This reentrant circuit drove the atria at a CL of 161 ms. The electrogram morphology and CL at all sites were constant from beat to beat, and electrograms at sites A to E, previously (prelesion) in the AF reentrant circuit (the driver), now were activated by daughter waves originating from the new driver, the AFL reentrant circuit.
Figure 4 Examples of the AFL reentrant circuits induced after creation of an LoB between the vena cavae. See text for discussion. CRYO ⫽ cryolesion between the vena cavae; other abbreviations as in Figure 1.
In 1 of 6 dogs (Dog #6), despite the intercaval LoB, AFL could not be induced. Figure 6, upper left panel, illustrates prelesion activation for a representative 100-ms window during induced AF, and shows an apparent reentrant circuit, that is, a driver (mean CL 128 ms) traveling clockwise around the PVs. The right panel illustrates electrograms recorded during this AF episode from selected sites (A to K) in both atria. Sites A to G are part of the reentrant circuit (left upper panel), and were activated sequentially at the driver’s CL. Daughter waves from this driver produced variable activation (H to K), resulting in AF. Per protocol, pacing was performed several times in each dog to verify the reproducibility and consistency of the arrhythmia induction. Figure 6, lower left panel, shows representative activation along with an ECG rhythm strip during another prelesion episode of AF (Dog #6). Once again, a driver was found circulating around the four PVs, but this time, in a counterclockwise direction (CL 120 ms). In each example, activation of the portion of the putative reentrant circuit between right PVs was assumed, as it was not mapped. Ability to map in this region from our electrode arrays is quite difficult, largely because of the presence of the reflection of the visceral pericardium. Additionally, activation in this region may be via an endocardial pathway. Nevertheless, the reversal of the direction of the putative reentrant circuit lends support to a reentrant mechanism as the AF driver. In dogs #1, 2, 3, and 5, a portion of the reentrant circuit also was assumed because it was not mapped, in all instances because we did not have electrodes placed between the PVs. After creating the LoB with a cryolesion, burst rapid atrial pacing in this dog (#6) still resulted in AF rather than
1750
Heart Rhythm, Vol 5, No 12, December 2008
Figure 5 Left: Activation window (161 ms) showing a figure-of-8 AFL reentrant circuit (blue) induced in Dog #4 after creating the line of block (CRYO) between the vena cavae. A through K denote sites from which electrograms displayed in the right panel were recorded. Right: ECG lead recorded simultaneously with electrograms from sites A through K. These are the same sites as in Figure 1. In contrast to Figure 1, now the driver is the AFL reentrant circuit (CL 161 ms), activation at all atrial sites is at the AFL (driver’s) CL, and electrogram morphology is constant from beat-to- beat. See text for discussion. Abbreviations as in Figure 4.
AFL (Figure 7). Figure 7, upper left panel, shows a representative activation map and an ECG rhythm strip during AF after the cryolesion. AF was caused by a stable LA driver traveling counterclockwise around the PVs (mean CL 122 ms). The right panel shows simultaneously recorded electrograms from sites A to K. The reentrant circuit (A to G), the driver, generated fibrillatory conduction to most of the rest of the atria (H to K). In effect, the cryolesion was an innocent bystander. As shown in Figure 7, lower left panel, FFT analysis of prelesion and postlesion data showed a narrow dominant frequency peak from areas around the PVs, consistent with a driver. In the RA and parts of the LA, frequency spectra were broadband with multiple peaks, consistent with irregular activation resulting from fibrillatory conduction.
reason. Because on postoperative day 2 the atria of the canine sterile pericarditis model are vulnerable to the induction of AF, it probably should not be surprising that creating an LoB between the vena cavae will not always change the pacing-induced rhythm from AF to AFL. This may be a paradigm for why patients who manifest AFL may, at other times, also manifest sustained AF. Another reason why only AF was induced in this dog may have been because the LoB was not quite long enough to permit a stable AFL reentrant circuit to become established. Nevertheless, this study does establish yet more clearly that when AF is exclusively inducible in a vulnerable atrial substrate, development of an LoB by whatever means between the vena cavae will favor the subsequent development of AFL instead of AF.
Discussion
Implications for the AF–AFL interrelationship
Changing inducible AF to inducible AFL
The data from this study further reinforce the interrelationship of AF and AFL and the role of an atrial driver as a part of their mechanism. The data show that both sustained AF and AFL required a driver, which formed during a period of preceding AF, the latter, in this case, induced by burst rapid atrial pacing. Thus, burst rapid atrial pacing first produced AF, the driver being the external pacemaker. With the cessation of the burst rapid atrial pacing, the AF that it produced no longer had a driver unless one formed during pacing or formed after pacing. If a driver of very short CL did not form, there was no sustained AF. In our studies, on postoperative day 2, only a left atrial driver formed, and it produced fibrillatory conduction and manifest AF. By altering the substrate with the creation of a fixed LoB between the vena cavae, burst rapid atrial pacing (the driver) still caused fibrillatory conduction, but with cessation of pacing, now AFL was most often induced because a critical component needed for development of its stable reentrant cir-
In the presence of the vulnerable substrate created by the sterile pericarditis in our canine model, AF, but not AFL, is typically induced with burst rapid atrial pacing on postoperative day 2.10 –13 Thus, as expected, in this postoperative day 2 study, initially only AF was induced in all 6 dogs. It was caused by a rapid, regular rhythm circulating around 1 or more PVs. This reentrant rhythm of very short CL served as a driver, giving rise to fibrillatory conduction and manifest AF. However, after creating a fixed LoB between the vena cavae, AFL was preferentially induced over AF. These data provide additional support for the concept that the difference between whether sustained AF or AFL forms (either spontaneously or induced) depends principally on the presence or development of a stable LoB in the region between the vena cavae. Why creation of the LoB did not result in induction of AFL in 1 dog is of interest, but we can only speculate on the
Bui et al
Converting AF to AFL
1751
Figure 6 Activation patterns in Dog #6 during induced, sustained AF before creating an LoB between the vena cavae. A: Two representative 100-ms activation windows during 2 different episodes of AF. In the upper activation window, a reentrant circuit (orange), the driver, travels around the 4 PVs in a clockwise direction (mean CL 128 ms). Sites A through G and sites H through K are outside the reentrant circuit. B: Electrograms recorded from A through K. Sites in the reentrant circuit (A through G) show a regular CL (the driver’s CL) and constant electrogram morphology. Sites outside the reentrant circuit (H through K) display irregular CLs and electrogram morphology. The lower activation window of Panel A shows another episode of AF, but the LA driver (orange) now travels counterclockwise (mean CL 120 ms) around the 4 PVs. See text for discussion. S ⫽ stimulus artifact; other abbreviations as in Figure 4.
cuit, a LoB between the vena cavae, was in place. Thus, for classic AFL to develop, a LoB between the vena cavae must form or be present. This LoB provides a critical substrate that preferentially leads to development of sustained AFL because of a right atrial reentrant driver.
Clinical implications By extrapolation, the data further add to our understanding of the development of recurrent AFL in many patients after open heart surgical repair of some forms of congenital heart disease (e.g., atrial septal defects, transposition of the great vessels, some ventricular septal defects, etc.) in which an incision is made in the RA in the region between the vena cavae. Should AFL subsequently develop, it is attributable to cavotricuspid isthmus-dependent AFL in about twothirds and to lesion reentry in about one-third.20 Clearly, it is reasonable to suggest that these surgical procedures be modified, when possible, to avoid creating an arrhythmogenic incision (LoB) in the region between the vena cavae. These data also provide further support for the concept that a driver of very short CL is an important mechanism of sustained AF. Also, the importance of the substrate in development of the driver is also further shown. And finally, the demonstration of a reentrant driver as a mechanism of AF in
this model suggests there may be or likely is a clinical counterpart.21,22
Study limitations This study recorded simultaneously from a large number of electrodes, which provided considerable recording resolution within both atria. Nevertheless, no activation data from the LA endocardium and only limited activation data from the interatrial septum were obtained. Also, no recordings were made from within the PVs, and in 5 of 6 dogs, a portion of the putative reentrant circuit was not mapped.
Conclusion In these studies in the canine sterile pericarditis model, rapid atrial pacing-induced sustained AF resulted from a stable, regular reentrant LA rhythm of very short CL that drove the atria, causing fibrillatory conduction. Creating a fixed LoB between the vena cavae provided an atrial substrate favorable for the subsequent rapid atrial pacing-induction of AFL rather than AF. However, when the prelesion reentrant circuit was reengaged during rapid atrial pacing-induction, AF recurred despite the presence of a fixed LoB between the vena cavae.
1752
Heart Rhythm, Vol 5, No 12, December 2008
Figure 7 Data from Dog #6 during AF after creation of the LoB. A, Upper Activation Window: representative activation map with a counterclockwise reentrant circuit (orange), mean CL 122 ms, driving the atria, causing fibrillatory conduction. An ECG lead II rhythm strip is shown below. B: Electrograms recorded from the reentrant circuit (A through G) again show a regular CL (the driver’s CL) and constant electrogram morphology; recordings from outside the reentrant circuit (H through K) display irregular CLs and electrogram morphology. A, lower portion: FFT analysis of electrograms from selected sites provides further support for the above. Thus, sites within the reentrant circuit show a narrow peak at the frequency of the reentrant circuit (the driver). Right atrial sites show a broadband with multiple peaks. Some LA sites not in the reentrant circuit are activated at the driver’s CL. RA sites near the LA contain a peak at the driver’s CL, indicating LA-to-RA conduction from the driver to those sites. See text for discussion. Abbreviations as in Figure 6.
References 1. Shimizu A, Nozaki A, Rudy Y, et al. Onset of induced atrial flutter in the canine pericarditis model. J Am Coll Cardiol 1991;17:1223–1234. 2. Waldo AL, Cooper TB. Spontaneous onset of type I atrial flutter in patients. J Am Coll Cardiol 1996;28:707–712. 3. Waldo AL, Feld GK. Inter-relationships of atrial fibrillation and atrial flutter. Mechanisms and clinical implications. J Am Coll Cardiol 2008;51:779 –786. 4. Lewis T, Feil HS, Stroud WD. Observations upon flutter and fibrillation. Part II. The nature of auricular flutter. Heart 1920;7:191–242. 5. Rosenblueth A, Garcia-Ramos J. Studies on flutter and fibrillation. II. The influence of artificial obstacles on experimental auricular flutter. Am Heart J 1947;33:677– 684. 6. Frame LH, Page RL, Boyden PA, et al. Circus movement in the canine atrium around the tricuspid ring during experimental atrial flutter and during reentry in vivo. Circulation 1987;76:1155–1175. 7. Feld GK, Shahandeh-Rad F. Mechanism of double potentials recorded during sustained atrial flutter in the canine right atrial crush-injury model. Circulation 1992;86:628 – 641. 8. Tomita Y, Matsuo K, Sahadevan J, et al. Role of functional block extension in lesion related atrial flutter. Circulation 2001;103:1025–1030. 9. Huang DT, Monohan KM, Zimetbaum P, et al. Hybrid pharmacologic and ablative therapy: a novel and effective approach for the management of atrial fibrillation. J Cardiovasc Electrophysiol 1998;9:462– 467. 10. Kumagai K, Khrestian C, Waldo AL. Simultaneous multisite mapping studies during induced atrial fibrillation in the sterile pericarditis model—insights into the mechanism of its maintenance. Circulation 1997;95:511–521. 11. Goldstein RN, Ryu K, Khrestian C. Prednisone prevents inducible atrial flutter in the canine sterile pericarditis model. J Cardiovasc Electrophysiol 2008;19: 74 – 81. 12. Pagé P, Plumb VJ, Okumura K, et al. A new model of atrial flutter. J Am Coll Cardiol 1986;8:872– 879.
13. Okumura K, Plumb VJ, Pagé PL, et al. Atrial activation sequence during atrial flutter in the canine pericarditis model and its effects on the polarity of the flutter wave in the electrocardiogram. J Am Coll Cardiol 1991;17:509 – 518. 14. Aranki SF, Shaw DP, Adams DH, et al. Predictors of atrial fibrillation after coronary artery surgery: current trends and impact on hospital resources. Circulation 1996;94:390 –397. 15. Mitchell LB. Incidence, timing and outcome of atrial tachyarrhythmias after cardiac surgery. In: Atrial Fibrillation After Cardiac Surgery. Steinberg J, editor. Norwell, MA: Kluwer, 2000:37–50. 16. Matsuo K, Uno K, Khrestian CM, et al. Conduction from left-to-right and right-to-left across the crista terminalis. Am J Physiol Heart Circ Physiol 2001;280:H1683–H1691. 17. Sahadevan J, Ryu K, Peltz L, et al. Epicardial mapping of chronic atrial fibrillation in patients: preliminary observations. Circulation 2004;110: 3293–3299. 18. Ryu K, Sahadevan J, Khrestian CM, et al. Use of Fast Fourier transform analysis of atrial electrograms for rapid characterization of atrial activation—implications for delineating possible mechanisms of atrial tachyarrhythmias. J Cardiovassc Electrophysiol 2006;17:198 –206. 19. Ryu K, Sahadevan J, Khrestian CM, et al. Frequency analysis of atrial electrograms identifies conduction pathways from the left to the right atrium during atrial fibrillation. Studies in two canine models. J Cardiovasc Electrophysiol 2008; In press. 20. Chan D, Van Hare GF, Mackall JA, et al. Importance of the typical atrial flutter isthmus in post-operative intra-atrial reentry. Circulation 1998;98 Suppl I:I– 616. 21. Jalife J. Experimental and clinical atrial fibrillation mechanisms. Bridging the divide. Intervent Cardiac Electrophysiol 2003;9:85–92. 22. Waldo AL. Mechanisms of atrial fibrillation. J Cardiovasc Electrophysiol 2003; 14:5267–5274.
[CL] 110 ⫾ 10 ms), regular, left atrial reentrant driver, which caused fibrillatory conduction. After creation of the LoB, in 5 dogs, rapid atrial pacing now induced sustained AFL (mean CL 167 ⫾ 13 ms). In the 6th dog, AFL failed to develop because the left atrial driver that was induced before the LoB was still reproducibly induced despite the LoB.
METHODS In 6 sterile pericarditis dogs, sustained AF was induced by rapid atrial pacing. After terminating AF, a fixed LoB between the vena cavae was created (cryoablation), and AF reinduction was attempted. Simultaneous mapping from 400 to 420 electrodes on the right and left atrial epicardium and the interatrial atrial septum was performed during all studies.
CONCLUSION In this model of sustained AF, altering the substrate to create a fixed LoB between the vena cavae creates a substrate favoring the induction of AFL.
RESULTS Before creation of the LoB, in all 6 dogs, rapid atrial pacing induced sustained AF because of a rapid (mean cycle length
(Heart Rhythm 2008;5:1745–1752) © 2008 Heart Rhythm Society. All rights reserved.
Introduction
ing the atrial substrate to favor development of a LoB between the vena cavae. Supportive of the importance of atrial substrate changes to the development of AF and then AFL are data from the sterile pericarditis model that show that following creation of the model, virtually only AF is inducible on postoperative days 1 and 2,10,11 but on postoperative days 3 and 4, AFL is primarily inducible despite using the same pacing induction protocols.1,11–13 Noteworthy, the development of spontaneous AF and AFL in patients after open heart surgery follows a similar time course.14,15 Additionally, preventing or minimizing substrate changes in the canine sterile pericarditis model with steroids virtually abolishes induction of AFL, the latter associated with failure to develop a LoB between the vena cavae.11 Despite these associations indicating the importance of the LoB in the region between the vena cavae, it has never been shown that simply altering the atrial substrate by creating a fixed LoB in the region between the vena cavae will convert the induction of AF to AFL. In this study, we explore further the role of an LoB in the region between the vena cavae, testing the hypothesis that in the canine sterile pericarditis model, when only sustained AF is reproducibly inducible, placing a fixed LoB in the region between the vena cavae critically alters the atrial substrate, enabling preferential induction of sustained AFL over AF.
Previous work from our laboratory has shown that atrial fibrillation (AF) of variable, but usually short, duration precedes the development of atrial flutter (AFL).1,2 Key to the development of AFL is that during the preceding AF, a line of block (LoB) must form or already be present in the region between the vena cavae. This LoB is usually functional, but may be fixed.3 The necessary presence of an abnormal atrial substrate for the formation of sustained AFL (and, in fact, for AF) is also well established, beginning with the early work of Lewis et al.4 Creation of an LoB between the vena cavae or anterior in the right atrial free wall permits induction of AFL in otherwise normal atria.5– 8 Furthermore, it also has been recognized that AF can often be converted to AFL by some antiarrhythmic agents, for example, those with sodium channel blocking effects, especially the IC agents.9 Presumably the latter occurs by chang-
Supported in part by Grant R01 HL38408 from the United States Public Health Service, National Institutes of Health, National Heart, Lung and Blood Institute, Bethesda, Maryland, and Blue Dot, Glenstone, and Jennie Zoline Foundations, Washington, DC. Address reprint requests and correspondence: Dr. Albert L. Waldo, University Hospitals Case Medical Center, Division of Cardiovascular Medicine, MS LKS 5038, 11100 Euclid Avenue, Cleveland, Ohio 44106. E-mail address: [email protected]. (Received August 19, 2008; accepted August 28, 2008.)
KEYWORDS arrhythmia; atrial fibrillation; atrial flutter; mapping; electrophysiology
1547-5271/$ -see front matter © 2008 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2008.08.036
1746
Heart Rhythm, Vol 5, No 12, December 2008
Methods
Data acquisition and analysis
All studies were performed in accordance with guidelines specified by our Institutional Animal Care and Use Committee, the American Heart Association Policy on Research Animal Use, and the United States Public Health Service Policy on Use of Laboratory Animals. The study subjects were 6 mongrel 18-kg to 23-kg dogs.
During each episode of AF/AFL, data were simultaneously recorded from the RA free wall, the LA free wall, BB, and the interatrial septum or region between the PVs.10,16 For each episode of sustained AF, 1.2 seconds of data (12 consecutive 100-ms windows) from a midportion of an episode were analyzed to compute isochrone activation maps. Previous studies in this laboratory showed that electrograms recorded from the midportion of a sustained AF episode are representative of the activation pattern for the entire episode.10 For each 100-ms window, activation times (maximum resolution of 1 ms) were determined at each recording site. For AFL, the activation window was determined by the AFL cycle length (CL). Frequency analysis was performed on a subset of electrogram recordings using Fast Fourier transform (FFT) techniques.17,18 The driver and areas that respond 1:1 at the drive CL give rise to a single dominant frequency peak, whereas irregular activation is indicated by the presence of multiple frequency peaks with a broadband.17,18
Creation of the sterile pericarditis model As previously described,10,12 (1) each dog first underwent surgery to create the sterile pericarditis model; (2) before closure of the chest, stainless steel wire electrodes for subsequent use in pacing and recording were sutured to the Bachmann bundle (BB), the right atrial appendage (RAA), the posterior-inferior left atrium, and the right ventricle (RV); (3) standard postoperative care was provided.
Studies in the open-chest state On postoperative day 2, using standard techniques during general anesthesia,10,12 His bundle ablation was performed to create complete atrioventricular block10; the RV was then paced at 60 to 100 beats/min, and the heart was exposed.
Simultaneous multisite mapping Before and after creation of an LoB in the region between the vena cavae, induction of AF was attempted using a programmable stimulator to deliver bursts of rapid atrial pacing.10 Sustained AF was deemed reproducible when a specific pacing protocol induced sustained AF at least twice before creating the LoB. After creating the LoB, reinduction was attempted using the same pacing induction protocol to see whether either AF or AFL were induced. If induction of either was unsuccessful, the pacing protocol was repeated from the beginning. During episodes of induced, sustained AF, and subsequently AFL, simultaneous multisite mapping of atrial activation was performed by recording electrograms from 400 to 420 sites. In all dogs, recordings were made from electrode arrays overlying BB (24 electrodes), the right atrial (RA) free wall (188 electrodes), and the left atrial (LA) free wall (168 electrodes) as previously described.10 In all dogs except one (#4), an additional 20-pole electrode catheter was inserted into the RA via the right internal jugular vein and advanced to record from the interatrial septum.10 In Dog #4, electrograms were recorded from the epicardial surface between the pulmonary veins (PVs) using two 20-pole electrode catheters. One catheter was inserted between the right and left PV, and the other between the superior and inferior PV.
Creation of the LoB After reproducible induction of sustained (ⱖ2 min) AF and its termination (spontaneous or paced), a fixed, transmural LoB approximately 25 mm in length was created with a series of contiguous epicardial cryolesions (– 80°C) to the RA free wall between the vena cavae and slightly anterior to the crista terminalis. After creating the LoB, AF induction was reattempted.
Results Activation patterns during induced, sustained AF before creation of the LoB Before cryoablation, burst rapid atrial pacing induced sustained AF in all 6 dogs. Each episode of induced AF resulted from a stable, rapid, regular LA driver of very short CL (104 to 128 ms, mean 112 ⫾ 6 ms) that was caused by reentry around the inferior PVs in one dog (#4), and appeared to be caused by reentry circulating around 1 or more PVs in the other 5 dogs. We say “appeared to be” because we never mapped the entire reentry circuit. However, in 1 dog (#6), we were able to show clockwise and counterclockwise activation of the putative reentrant circuit, further supporting reentry as the mechanism of the driver. In all AF episodes, the driver served as a stable generator of daughter waves that spread to the rest of the atria via several routes, and produced fibrillatory conduction to variable portions of the LA and much or all of the RA. Maps of these areas showed a changing pattern of activation that included irregular beat-to-beat CLs, variable electrogram morphologies, variable wave fronts, and unstable LoBs. The recorded electrocardiogram (ECG) showed classic AF. In all instances, without a driver, there was no sustained tachyarrhythmia. Thus, although the initial AF was induced by burst rapid atrial pacing (the initial driver), if no new driver formed during or after the atrial pacing, there was no sustained AF. Figure 1 shows data from a representative example of induced, sustained AF before the creation of the LoB in Dog #4. Panel A shows maps of 2 consecutive 100-ms activation windows. A reentrant circuit (orange arrows) courses around the inferior PVs, and sends off daughter waves (gray arrows). Panel B shows an ECG simultaneously recorded with electrograms from sites A-L, denoted in Panel A, over a period of about 900 ms. Electrograms from sites A to E, recorded from within the reentrant circuit, show a regular
Bui et al
Converting AF to AFL
1747
Figure 1 A: Two consecutive 100-ms activation maps from a representative example of AF in Dog #4 before creation of an LoB between the vena cavae. A reentrant driver (orange) circulates around the inferior PVs. In this and subsequent figures, isochrones are at 10 ms intervals, lines with arrows indicate direction of activation, daughter waves are denoted by gray arrows, an asterisk (*) represents an epicardial breakthrough point, dashed lines represent functional lines of block, and gray areas are not activated during the illustrated activation window. A through L are electrogram recording sites shown in Panel B. B: Electrograms simultaneously recorded from sites A through L (shown in Panel A) over a period of approximately 900 ms. Beat-to-beat CL is indicated in milliseconds (ms). See text for discussion. BB ⫽ Bachmann bundle; IVC ⫽ inferior vena cava; LAA ⫽ left atrial appendage; PV ⫽ pulmonary vein; QRS ⫽ ventricular complex; RAA ⫽ right atrial appendage; SVC ⫽ superior vena cava.
beat-to-beat CL of 117 ⫾ 6 ms with electrograms of constant morphology. In contrast, electrograms from sites F, G, and I to L, recorded from outside the reentrant circuit, show a slower rate with an irregular CL and morphology. The regular rhythm of short CL around the PVs serves as the driver, and the irregular rhythm in parts of the LA and most of the RA indicates fibrillatory conduction caused by the driver. Site H recorded from BB is also outside the reentrant circuit, but shows a regular CL identical to that of the driver despite being a daughter wave. Located just superior to the upper PVs, the midportion of BB was usually a part of or very close to the reentrant circuit of the AF driver in these studies. Its activation was usually remarkably regular during AF, even when not part of the reentrant circuit, that is, it usually did not manifest fibrillatory conduction. Figure 2 shows 12 consecutive 100-ms windows from that same episode of AF (Dog #4). In each window, the reentrant circuit travels counterclockwise around the inferior left and right PVs. Daughter waves from this reentrant circuit travel superiorly between the left and right PVs to activate BB. After exiting BB, these daughter waves activate other portions of the LA and the RA, producing variable LoBs (dashed lines) and CLs. Also, a daughter wave travels from the reentrant circuit toward and usually across the intercaval region to activate a portion of the RA. In some activation windows, portions of the LA and RA are not activated (gray areas). The heterogeneity of activation seen
is the result of fibrillatory conduction. This activation pattern continued until the reentrant rhythm terminated, either spontaneously or because of atrial pacing. Figure 3 shows FFT analysis of electrograms recorded from selected atrial sites (Dog #4). Sites within the reentrant circuit show a single, narrow, dominant frequency peak at the driver’s CL (8.54 Hz). Sites not in the reentrant circuit, but activated by daughter waves at the driver’s CL (‡), also have a single dominant frequency peak at the driver’s CL (8.54 Hz). The inferior LA and all the RA sites, however, show a broadband with multiple frequency peaks consistent with fibrillatory conduction. Broadband RA sites between the vena cavae that are near the LA also include a frequency peak at the driver’s CL (8.54 Hz), indicating LA-to-RA conduction from the driver.16,19
Activation patterns during induced, sustained atrial tachyarrhythmias after creation of the LoB Using the same pacing protocol as before creation of the LoB, sustained AFL was now reproducibly induced in 5 of 6 six dogs. These AFL reentrant circuits showed significantly longer CLs (mean 165 ⫾ 13 ms; P ⬍ .05) compared with those during AF (Table 1). Figure 4 summarizes the course of all 5 postablation AFL reentrant circuits. Figure 5 illustrates activation during induced AFL in a representative example (Dog #4) after creation of the LoB between the vena cavae. The activation map (left panel)
1748
Heart Rhythm, Vol 5, No 12, December 2008
Figure 2 Twelve consecutive 100-ms activation windows of AF in Dog #4. The reentrant circuit (orange) travels around the inferior PVs, driving the atria via daughter waves (gray). See text for discussion. Abbreviations as in Figure 1.
Bui et al
Converting AF to AFL
1749
Figure 3 Fast Fourier transform (FFT) analysis of selected electrograms from Dog #4. The reentrant circuit (driver) is orange. ‡Sites outside the reentry circuit activated by daughter waves at the driver’s CL. See text for discussion. Abbreviations as in Figure 1.
shows a figure-of-8 reentrant circuit in the RA that drives the rest of the atria at a CL of 161 ms. The right panel shows an ECG recorded simultaneously with electrograms from the same sites as in Figure 1, except that site H is now recorded from the intercaval region. The reentrant circuit consists of 2 limbs sharing a common pathway in the RA free wall. One limb travels down the RA free wall and up the atrial septum, and the other travels clockwise around the lesion. This reentrant circuit drove the atria at a CL of 161 ms. The electrogram morphology and CL at all sites were constant from beat to beat, and electrograms at sites A to E, previously (prelesion) in the AF reentrant circuit (the driver), now were activated by daughter waves originating from the new driver, the AFL reentrant circuit.
Figure 4 Examples of the AFL reentrant circuits induced after creation of an LoB between the vena cavae. See text for discussion. CRYO ⫽ cryolesion between the vena cavae; other abbreviations as in Figure 1.
In 1 of 6 dogs (Dog #6), despite the intercaval LoB, AFL could not be induced. Figure 6, upper left panel, illustrates prelesion activation for a representative 100-ms window during induced AF, and shows an apparent reentrant circuit, that is, a driver (mean CL 128 ms) traveling clockwise around the PVs. The right panel illustrates electrograms recorded during this AF episode from selected sites (A to K) in both atria. Sites A to G are part of the reentrant circuit (left upper panel), and were activated sequentially at the driver’s CL. Daughter waves from this driver produced variable activation (H to K), resulting in AF. Per protocol, pacing was performed several times in each dog to verify the reproducibility and consistency of the arrhythmia induction. Figure 6, lower left panel, shows representative activation along with an ECG rhythm strip during another prelesion episode of AF (Dog #6). Once again, a driver was found circulating around the four PVs, but this time, in a counterclockwise direction (CL 120 ms). In each example, activation of the portion of the putative reentrant circuit between right PVs was assumed, as it was not mapped. Ability to map in this region from our electrode arrays is quite difficult, largely because of the presence of the reflection of the visceral pericardium. Additionally, activation in this region may be via an endocardial pathway. Nevertheless, the reversal of the direction of the putative reentrant circuit lends support to a reentrant mechanism as the AF driver. In dogs #1, 2, 3, and 5, a portion of the reentrant circuit also was assumed because it was not mapped, in all instances because we did not have electrodes placed between the PVs. After creating the LoB with a cryolesion, burst rapid atrial pacing in this dog (#6) still resulted in AF rather than
1750
Heart Rhythm, Vol 5, No 12, December 2008
Figure 5 Left: Activation window (161 ms) showing a figure-of-8 AFL reentrant circuit (blue) induced in Dog #4 after creating the line of block (CRYO) between the vena cavae. A through K denote sites from which electrograms displayed in the right panel were recorded. Right: ECG lead recorded simultaneously with electrograms from sites A through K. These are the same sites as in Figure 1. In contrast to Figure 1, now the driver is the AFL reentrant circuit (CL 161 ms), activation at all atrial sites is at the AFL (driver’s) CL, and electrogram morphology is constant from beat-to- beat. See text for discussion. Abbreviations as in Figure 4.
AFL (Figure 7). Figure 7, upper left panel, shows a representative activation map and an ECG rhythm strip during AF after the cryolesion. AF was caused by a stable LA driver traveling counterclockwise around the PVs (mean CL 122 ms). The right panel shows simultaneously recorded electrograms from sites A to K. The reentrant circuit (A to G), the driver, generated fibrillatory conduction to most of the rest of the atria (H to K). In effect, the cryolesion was an innocent bystander. As shown in Figure 7, lower left panel, FFT analysis of prelesion and postlesion data showed a narrow dominant frequency peak from areas around the PVs, consistent with a driver. In the RA and parts of the LA, frequency spectra were broadband with multiple peaks, consistent with irregular activation resulting from fibrillatory conduction.
reason. Because on postoperative day 2 the atria of the canine sterile pericarditis model are vulnerable to the induction of AF, it probably should not be surprising that creating an LoB between the vena cavae will not always change the pacing-induced rhythm from AF to AFL. This may be a paradigm for why patients who manifest AFL may, at other times, also manifest sustained AF. Another reason why only AF was induced in this dog may have been because the LoB was not quite long enough to permit a stable AFL reentrant circuit to become established. Nevertheless, this study does establish yet more clearly that when AF is exclusively inducible in a vulnerable atrial substrate, development of an LoB by whatever means between the vena cavae will favor the subsequent development of AFL instead of AF.
Discussion
Implications for the AF–AFL interrelationship
Changing inducible AF to inducible AFL
The data from this study further reinforce the interrelationship of AF and AFL and the role of an atrial driver as a part of their mechanism. The data show that both sustained AF and AFL required a driver, which formed during a period of preceding AF, the latter, in this case, induced by burst rapid atrial pacing. Thus, burst rapid atrial pacing first produced AF, the driver being the external pacemaker. With the cessation of the burst rapid atrial pacing, the AF that it produced no longer had a driver unless one formed during pacing or formed after pacing. If a driver of very short CL did not form, there was no sustained AF. In our studies, on postoperative day 2, only a left atrial driver formed, and it produced fibrillatory conduction and manifest AF. By altering the substrate with the creation of a fixed LoB between the vena cavae, burst rapid atrial pacing (the driver) still caused fibrillatory conduction, but with cessation of pacing, now AFL was most often induced because a critical component needed for development of its stable reentrant cir-
In the presence of the vulnerable substrate created by the sterile pericarditis in our canine model, AF, but not AFL, is typically induced with burst rapid atrial pacing on postoperative day 2.10 –13 Thus, as expected, in this postoperative day 2 study, initially only AF was induced in all 6 dogs. It was caused by a rapid, regular rhythm circulating around 1 or more PVs. This reentrant rhythm of very short CL served as a driver, giving rise to fibrillatory conduction and manifest AF. However, after creating a fixed LoB between the vena cavae, AFL was preferentially induced over AF. These data provide additional support for the concept that the difference between whether sustained AF or AFL forms (either spontaneously or induced) depends principally on the presence or development of a stable LoB in the region between the vena cavae. Why creation of the LoB did not result in induction of AFL in 1 dog is of interest, but we can only speculate on the
Bui et al
Converting AF to AFL
1751
Figure 6 Activation patterns in Dog #6 during induced, sustained AF before creating an LoB between the vena cavae. A: Two representative 100-ms activation windows during 2 different episodes of AF. In the upper activation window, a reentrant circuit (orange), the driver, travels around the 4 PVs in a clockwise direction (mean CL 128 ms). Sites A through G and sites H through K are outside the reentrant circuit. B: Electrograms recorded from A through K. Sites in the reentrant circuit (A through G) show a regular CL (the driver’s CL) and constant electrogram morphology. Sites outside the reentrant circuit (H through K) display irregular CLs and electrogram morphology. The lower activation window of Panel A shows another episode of AF, but the LA driver (orange) now travels counterclockwise (mean CL 120 ms) around the 4 PVs. See text for discussion. S ⫽ stimulus artifact; other abbreviations as in Figure 4.
cuit, a LoB between the vena cavae, was in place. Thus, for classic AFL to develop, a LoB between the vena cavae must form or be present. This LoB provides a critical substrate that preferentially leads to development of sustained AFL because of a right atrial reentrant driver.
Clinical implications By extrapolation, the data further add to our understanding of the development of recurrent AFL in many patients after open heart surgical repair of some forms of congenital heart disease (e.g., atrial septal defects, transposition of the great vessels, some ventricular septal defects, etc.) in which an incision is made in the RA in the region between the vena cavae. Should AFL subsequently develop, it is attributable to cavotricuspid isthmus-dependent AFL in about twothirds and to lesion reentry in about one-third.20 Clearly, it is reasonable to suggest that these surgical procedures be modified, when possible, to avoid creating an arrhythmogenic incision (LoB) in the region between the vena cavae. These data also provide further support for the concept that a driver of very short CL is an important mechanism of sustained AF. Also, the importance of the substrate in development of the driver is also further shown. And finally, the demonstration of a reentrant driver as a mechanism of AF in
this model suggests there may be or likely is a clinical counterpart.21,22
Study limitations This study recorded simultaneously from a large number of electrodes, which provided considerable recording resolution within both atria. Nevertheless, no activation data from the LA endocardium and only limited activation data from the interatrial septum were obtained. Also, no recordings were made from within the PVs, and in 5 of 6 dogs, a portion of the putative reentrant circuit was not mapped.
Conclusion In these studies in the canine sterile pericarditis model, rapid atrial pacing-induced sustained AF resulted from a stable, regular reentrant LA rhythm of very short CL that drove the atria, causing fibrillatory conduction. Creating a fixed LoB between the vena cavae provided an atrial substrate favorable for the subsequent rapid atrial pacing-induction of AFL rather than AF. However, when the prelesion reentrant circuit was reengaged during rapid atrial pacing-induction, AF recurred despite the presence of a fixed LoB between the vena cavae.
1752
Heart Rhythm, Vol 5, No 12, December 2008
Figure 7 Data from Dog #6 during AF after creation of the LoB. A, Upper Activation Window: representative activation map with a counterclockwise reentrant circuit (orange), mean CL 122 ms, driving the atria, causing fibrillatory conduction. An ECG lead II rhythm strip is shown below. B: Electrograms recorded from the reentrant circuit (A through G) again show a regular CL (the driver’s CL) and constant electrogram morphology; recordings from outside the reentrant circuit (H through K) display irregular CLs and electrogram morphology. A, lower portion: FFT analysis of electrograms from selected sites provides further support for the above. Thus, sites within the reentrant circuit show a narrow peak at the frequency of the reentrant circuit (the driver). Right atrial sites show a broadband with multiple peaks. Some LA sites not in the reentrant circuit are activated at the driver’s CL. RA sites near the LA contain a peak at the driver’s CL, indicating LA-to-RA conduction from the driver to those sites. See text for discussion. Abbreviations as in Figure 6.
References 1. Shimizu A, Nozaki A, Rudy Y, et al. Onset of induced atrial flutter in the canine pericarditis model. J Am Coll Cardiol 1991;17:1223–1234. 2. Waldo AL, Cooper TB. Spontaneous onset of type I atrial flutter in patients. J Am Coll Cardiol 1996;28:707–712. 3. Waldo AL, Feld GK. Inter-relationships of atrial fibrillation and atrial flutter. Mechanisms and clinical implications. J Am Coll Cardiol 2008;51:779 –786. 4. Lewis T, Feil HS, Stroud WD. Observations upon flutter and fibrillation. Part II. The nature of auricular flutter. Heart 1920;7:191–242. 5. Rosenblueth A, Garcia-Ramos J. Studies on flutter and fibrillation. II. The influence of artificial obstacles on experimental auricular flutter. Am Heart J 1947;33:677– 684. 6. Frame LH, Page RL, Boyden PA, et al. Circus movement in the canine atrium around the tricuspid ring during experimental atrial flutter and during reentry in vivo. Circulation 1987;76:1155–1175. 7. Feld GK, Shahandeh-Rad F. Mechanism of double potentials recorded during sustained atrial flutter in the canine right atrial crush-injury model. Circulation 1992;86:628 – 641. 8. Tomita Y, Matsuo K, Sahadevan J, et al. Role of functional block extension in lesion related atrial flutter. Circulation 2001;103:1025–1030. 9. Huang DT, Monohan KM, Zimetbaum P, et al. Hybrid pharmacologic and ablative therapy: a novel and effective approach for the management of atrial fibrillation. J Cardiovasc Electrophysiol 1998;9:462– 467. 10. Kumagai K, Khrestian C, Waldo AL. Simultaneous multisite mapping studies during induced atrial fibrillation in the sterile pericarditis model—insights into the mechanism of its maintenance. Circulation 1997;95:511–521. 11. Goldstein RN, Ryu K, Khrestian C. Prednisone prevents inducible atrial flutter in the canine sterile pericarditis model. J Cardiovasc Electrophysiol 2008;19: 74 – 81. 12. Pagé P, Plumb VJ, Okumura K, et al. A new model of atrial flutter. J Am Coll Cardiol 1986;8:872– 879.
13. Okumura K, Plumb VJ, Pagé PL, et al. Atrial activation sequence during atrial flutter in the canine pericarditis model and its effects on the polarity of the flutter wave in the electrocardiogram. J Am Coll Cardiol 1991;17:509 – 518. 14. Aranki SF, Shaw DP, Adams DH, et al. Predictors of atrial fibrillation after coronary artery surgery: current trends and impact on hospital resources. Circulation 1996;94:390 –397. 15. Mitchell LB. Incidence, timing and outcome of atrial tachyarrhythmias after cardiac surgery. In: Atrial Fibrillation After Cardiac Surgery. Steinberg J, editor. Norwell, MA: Kluwer, 2000:37–50. 16. Matsuo K, Uno K, Khrestian CM, et al. Conduction from left-to-right and right-to-left across the crista terminalis. Am J Physiol Heart Circ Physiol 2001;280:H1683–H1691. 17. Sahadevan J, Ryu K, Peltz L, et al. Epicardial mapping of chronic atrial fibrillation in patients: preliminary observations. Circulation 2004;110: 3293–3299. 18. Ryu K, Sahadevan J, Khrestian CM, et al. Use of Fast Fourier transform analysis of atrial electrograms for rapid characterization of atrial activation—implications for delineating possible mechanisms of atrial tachyarrhythmias. J Cardiovassc Electrophysiol 2006;17:198 –206. 19. Ryu K, Sahadevan J, Khrestian CM, et al. Frequency analysis of atrial electrograms identifies conduction pathways from the left to the right atrium during atrial fibrillation. Studies in two canine models. J Cardiovasc Electrophysiol 2008; In press. 20. Chan D, Van Hare GF, Mackall JA, et al. Importance of the typical atrial flutter isthmus in post-operative intra-atrial reentry. Circulation 1998;98 Suppl I:I– 616. 21. Jalife J. Experimental and clinical atrial fibrillation mechanisms. Bridging the divide. Intervent Cardiac Electrophysiol 2003;9:85–92. 22. Waldo AL. Mechanisms of atrial fibrillation. J Cardiovasc Electrophysiol 2003; 14:5267–5274.