This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Roithinger, F. X. Articles by Lesh, M. D. Search for Related Content PubMed PubMed Citation Articles by Roithinger, F. X. Articles by Lesh, M. D.

(Circulation. 1997;96:3484-3491.)
© 1997 American Heart Association, Inc.

Articles

Relationship Between Atrial Fibrillation and Typical Atrial Flutter in Humans

Activation Sequence Changes During Spontaneous Conversion

Franz X. Roithinger, MD; Martin R. Karch, MD; Paul R. Steiner, MD; Arne SippensGroenewegen, MD; ; Michael D. Lesh, MD

From the Section of Cardiac Electrophysiology, Department of Medicine and Cardiovascular Research Institute, University of California San Francisco.

Correspondence to Michael D. Lesh, MD, Section of Cardiac Electrophysiology, Department of Medicine and the Cardiovascular Research Institute, University of California, San Francisco, 500 Parnassus Ave, Room MU 428, Box 1354, San Francisco, CA 94143-1354. E-mail lesh{at}ep4.ucsf.edu

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Background A transitional rhythm precedes the spontaneous onset of atrial flutter in an animal model, but few data are available in man.

Methods and Results In 10 patients, 16 episodes of atrial fibrillation (166±236 seconds) converting into atrial flutter during electrophysiological evaluation were analyzed. A 20-pole catheter was used for mapping the right atrial free wall. Preceding the conversion was a characteristic sequence of events: (1) a gradual increase in atrial fibrillation cycle length (150±25 ms after onset, 166±28 ms before conversion, P<.01); (2) an electrically silent period (267±45 ms); (3) "organized atrial fibrillation" (cycle length, 184±24 ms) with the same right atrial free wall activation direction as during atrial flutter; (4) another delay on the lateral right atrium (283±52 ms); and (5) typical atrial flutter (cycle length, 245±38 ms). The coronary sinus generally had a different rate than the right atrial free wall until the beat that initiated flutter, when right atrium and coronary sinus were activated in sequence.

During 1313 seconds of fibrillation, there were 171 episodes of "organized atrial fibrillation." An additional activation delay at least 30 ms longer than the mean organized atrial fibrillation cycle length was sensitive (100%) and specific (99%) for impending organization into atrial flutter. During organized atrial fibrillation, right atrial free wall activation was craniocaudal in 70% and caudocranial in 30%, which may explain why counterclockwise flutter is a more common clinical rhythm than clockwise flutter. Atrial flutter never degenerated into fibrillation, even after adenosine infusion.

Conclusions Anatomic barriers, along with statistical properties of conduction and refractoriness during atrial fibrillation, may explain the remarkably stereotypical pattern of endocardial activation during the initiation of atrial flutter via fibrillation and the rarity of degeneration of flutter to fibrillation once it stabilizes.

Key Words: fibrillation • atrial flutter • mapping

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
In typical human atrial flutter, the barriers of the right atrial reentrant circuit are reasonably well established, and curative ablation with a high rate of success is available for this arrhythmia.^1–5 Atrial fibrillation has long been described as a disorganized or "random" phenomenon.⁶ Recent studies, however, found evidence that activation during atrial fibrillation is not entirely random,⁷ suggesting similarities among these common arrhythmias. The conversion of atrial fibrillation into atrial flutter and vice versa has been shown to occur spontaneously in an animal model by use of high-density electrode mapping.⁸ In humans, however, the conversion of one arrhythmia to the other has not been extensively analyzed, although recent data using a limited number of recording sites support the notion that a transitional arrhythmia is critical for the onset of atrial flutter.⁹ Thus, in the present study, the mode of organization of atrial fibrillation into atrial flutter has been studied from multisite endocardial recordings in patients undergoing electrophysiological evaluation before atrial flutter ablation. Our hypothesis was that atrial fibrillation becomes more organized just before conversion of atrial flutter and that the activation sequence along the right atrial free wall during organized atrial fibrillation determines the ultimate direction of rotation of the stable atrial flutter reentrant wave. Finally, we hypothesized that the conversion of flutter to fibrillation would be a much rarer event because of the stabilizing nature of the anatomical barriers that define the atrial flutter substrate.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Patients
Of 89 consecutive patients referred to the University of California, San Francisco for curative atrial flutter ablation, spontaneous and induced atrial fibrillation lasting >15 seconds and converting into typical atrial flutter during the electrophysiological study before attempted catheter ablation occurred in 10 patients (1 woman and 9 men; age, 58 to 80 years; mean age, 66±6 years). Patients with atrial fibrillation converting into atypical atrial flutter (not involving the subeustachian isthmus), as recently defined,¹⁰ were excluded from the study. Of the 10 patients, 7 had no structural heart disease with echocardiographically normal left ventricular function and atrial size. One patient had moderately reduced left ventricular function (ejection fraction, 45%) without evidence of coronary artery disease and moderate biatrial enlargement. Two patients had coronary artery disease with a history of myocardial infarction, bypass surgery, compromised left ventricular function (ejection fraction, 25% and 35%, respectively), and mild biatrial enlargement. In 6 patients, all antiarrhythmic drugs, except for digoxin, were stopped at least 24 hours before the study. Four patients were receiving amiodarone treatment because of atrial arrhythmias (3 patients) or ventricular tachycardia (1 patient). Informed consent was obtained from all patients before the study according to the protocol approved by the Committee on Human Research.

Catheters
For mapping the right atrium, a 7F, 20-pole catheter (Cordis-Webster) with alternating 2- and 10-mm interelectrode distance (2-mm interpolar distance) was situated in the trabeculated portion of the right atrium in 9 patients, as previously described.¹ In 1 patient, two 7F octapolar catheters (EP Technologies) with 10- and 5-mm interelectrode distance were positioned along the trabeculated right atrium. In all patients, a multielectrode catheter was placed in the coronary sinus via the right internal jugular vein. In 4 patients, a quadripolar catheter was positioned along the septum in the region of the His bundle, with adequate signal amplitude.

Induction of Atrial Flutter and Fibrillation
Specific attempts to induce atrial fibrillation were not made. However, in our laboratory, attempts are made during sinus rhythm to use rapid pacing and programmed stimulation to induce atrial flutter. During induced or spontaneous typical atrial flutter, attempts are also made to entrain atrial flutter from multiple sites and multiple pace cycle lengths and to pace terminate or internally cardiovert atrial flutter to sinus rhythm. If atrial fibrillation resulted during attempts at induction, entrainment, or termination of atrial flutter, such an event was carefully logged. Episodes of atrial fibrillation that lasted at least 15 seconds and subsequently organized into typical atrial flutter (as defined below) form the material for the present study. During induced typical atrial flutter in 7 patients, adenosine (6 to 18 mg) was infused. The effect of adenosine was assessed in terms of the cycle length of atrial flutter, and any degeneration back to fibrillation was noted.

Electrogram Recordings and Calculations
Bipolar intracardiac electrograms filtered between 30 and 500 Hz were recorded and stored digitally on a Cardiolab system (Prucka Engineering) simultaneously with the 12-lead surface ECG. Calculation of atrial fibrillation cycle length was performed on the Cardiolab system with digital calipers. Twenty-five cycles of atrial fibrillation in two different bipoles along the trabeculated right atrium were measured in every episode of conversion of atrial fibrillation to atrial flutter: (1) 5 seconds after the onset of atrial fibrillation, (2) just before a more organized pattern of atrial fibrillation (see definitions below), and (3) during the period of a more organized pattern of atrial fibrillation immediately before conversion to typical atrial flutter. Atrial flutter cycle length at baseline and after adenosine administration was determined by calculating a mean of 10 cycles. Additionally, all recorded atrial fibrillation was scanned for all periods of "organized atrial fibrillation," as defined below, and notes were made of whether those periods resulted in further organization into overt typical atrial flutter or whether they once again became a more disorganized form of atrial fibrillation.

Definitions
Typical counterclockwise atrial flutter was defined as a macroreentrant right atrial arrhythmia, having a classic appearance of negative flutter waves in the inferior leads of the 12-lead ECG and a constant counterclockwise endocardial activation sequence around the tricuspid annulus. Typical clockwise atrial flutter was defined as a macroreentrant right atrial arrhythmia having predominantly positive flutter waves in the inferior leads and a constant clockwise endocardial activation sequence around the tricuspid annulus.² For any induced atrial flutter, entrainment was performed from multiple sites, with a comparison of the postpacing interval to the flutter cycle length, particularly during entrainment from the low right atrium, to be certain that the subeustachian isthmus was critical to the maintenance of the arrhythmia.^1,2 Atrial fibrillation was defined as a rapid atrial rhythm (rate >260 bpm) with a characteristic surface ECG morphology and a variability of the beat-to-beat cycle length, morphology, and/or amplitude of recorded bipolar atrial electrograms.¹¹ Along the trabeculated right atrium, organized atrial fibrillation was considered present if discrete atrial complexes, separated by an isoelectric baseline, with a constant activation sequence, either caudocranial or craniocaudal, were seen during three or more cycles over at least 3 cm along the right atrial free wall. Despite such organization, the rhythm was still consistent with atrial fibrillation because of a variability in beat-to-beat cycle length of >30 ms, a variability in electrogram morphology, a rate of >260 bpm, and a characteristic surface ECG morphology. Disorganized atrial fibrillation was considered present if atrial electrograms failed to demonstrate discrete complexes or isoelectric intervals over at least 3 cm along the right atrial free wall.

Statistical Analysis
Values are expressed as mean±SD. Statistical comparisons were performed by use of Student's t test, paired and unpaired when appropriate, and ² test. A value of P<.05 was considered statistically significant.

	Results

Top Abstract Introduction Methods Results Discussion References

 
In 10 patients, a total of 16 episodes (range, 1 to 4) of atrial fibrillation organizing into typical atrial flutter were observed. The duration of atrial fibrillation varied between 15 and 862 seconds (mean, 166±236; median, 70 seconds; a total of 2656 seconds or 44 minutes). Atrial fibrillation converted into typical counterclockwise atrial flutter in 13 episodes and into typical clockwise atrial flutter in 3 episodes. The mean atrial flutter cycle length was 245±38 ms (range, 200 to 315 ms). Atrial fibrillation was induced by overdrive pacing during atrial flutter in 8 episodes (7 patients), by low-energy internal cardioversion in 3 episodes (2 patients), and by burst pacing during sinus rhythm in 3 episodes (3 patients). In 2 episodes (2 patients), spontaneous atrial fibrillation during sinus rhythm occurred. During typical atrial flutter, increasing doses of adenosine were administered in 7 patients titrated to achieve significant AV block (3 and 6 mg in 1 patient; 6 and 12 mg in 5 patients; and 6, 12, and 18 mg in 1 patient). None of the 15 adenosine administrations caused a degeneration into atrial fibrillation (Fig 1). In these 7 patients, the mean atrial flutter cycle length at baseline (242±34 ms) was identical to the mean atrial flutter cycle length during maximum adenosine effect (242±35 ms, P=NS).

View larger version (82K): [in this window] [in a new window]   Figure 1. Surface ECG lead III in 7 patients during infusion of adenosine (arrow). High-grade AV block develops, but none of the patients shows degeneration into atrial fibrillation. Furthermore, the atrial flutter cycle length does not change significantly before and during adenosine infusion. AS, JR, HA, HE, SR, RJ, and SJ are the initials of the patients receiving adenosine. CL indicates the mean duration of 10 flutter cycles in milliseconds of every patient before (left column) and during adenosine infusion (right column).

Sequence of Changes From Atrial Fibrillation to Atrial Flutter
Before conversion of atrial fibrillation to atrial flutter, a characteristic sequence of events was present in all 16 episodes (Figs 2 and 3). First, a gradual increase in atrial fibrillation cycle length was found. The mean atrial fibrillation cycle length increased from 150±25 ms (range, 104 to 182 ms) 5 seconds after the onset of atrial fibrillation to 166±28 ms (range, 142 to 202 ms) immediately before the occurrence of a more organized pattern of atrial fibrillation (P<.01). Second, a sudden change in activation sequence was noted such that the lateral right atrial free wall was not activated for 267±45 ms (range, 194 to 345 ms). This electrically silent period, or pause, that heralded the onset of organized atrial fibrillation was an average of 157±22% (range, 117% to 187%) of the mean cycle length of the preceding 25 cycles of disorganized atrial fibrillation. The mean cycle length of organized atrial fibrillation was 184±24 ms (range, 153 to 225 ms). This more organized atrial fibrillation lasted a mean of 9±8 cycles (median, 7 cycles). Finally, another sudden change in cycle length with a mean electrically silent period on the right atrial free wall of 283±52 ms, significantly longer than the mean atrial flutter cycle length of 245±38 ms (P<.01), preceded the onset of typical atrial flutter. The mean percentage of this delay on the right atrial free wall compared with the preceding organized atrial fibrillation was 147±20% (range, 115% to 187%). During the whole period of organized atrial fibrillation directly preceding atrial flutter, distinct potentials on the coronary sinus catheter were present in all 16 episodes, although atrial activation in the coronary sinus was not synchronized to the activation sequence on the right atrial free wall. With the onset of typical atrial flutter, however, a stable 1:1 relation between the activation sequence along the trabeculated right atrium and the atrial activation recorded in the coronary sinus was established. In 5 episodes (4 patients), septal recordings were available. With the onset of typical atrial flutter, the septal activation showed discrete atrial signals, and a stable 1:1 relation to the activation along the trabeculated right atrium became established. Before this, septal recordings showed a disorganized pattern in 1 episode and an organized pattern with discrete atrial signals in 4 episodes, but the rate was different from the rate of the right atrial free wall and coronary sinus atrial activation.

View larger version (57K): [in this window] [in a new window]   Figure 2. A characteristic sequence of events in the conversion of atrial fibrillation to typical counterclockwise atrial flutter. First, disorganized atrial fibrillation is present. Then, a sudden change in the activation sequence occurs such that the right atrial free wall is not activated for 335 ms (bipole TA 1-2). A period of 6 cycles of organized atrial fibrillation follows, with an activation sequence characteristic for typical counterclockwise atrial flutter. After another activation delay of 354 ms, typical stable counterclockwise atrial flutter with a cycle length of 300 ms resumes, with negative flutter waves in the inferior leads of the surface ECG and intermittent 1:1 AV conduction. Note the recording in the coronary sinus showing distinct atrial depolarizations even during disorganized atrial fibrillation, although a stable 1:1 relation of the coronary sinus atrial signal preceding the activation sequence along the trabeculated right atrium does not occur until the onset of typical atrial flutter. III, AVF, V1 indicate the surface ECG leads; TA, bipolar recordings along the tricuspid annulus (TA 1–2 is the low lateral right atrium and TA 15–16 is the roof of the right atrium); CS mid, middle coronary sinus catheter recording; CS prox, proximal coronary sinus catheter recording; numbers, cycle lengths in milliseconds; bold numbers, change in activation sequence (335), activation delay before flutter onset (354), cycle length of typical atrial flutter (300); and V, ventricular depolarization on the coronary sinus catheter.

View larger version (54K): [in this window] [in a new window]   Figure 3. A characteristic sequence of events in the conversion of atrial fibrillation to typical clockwise atrial flutter. After 5 cycles of counterclockwise organized atrial fibrillation and a change in activation sequence along the middle right atrial free wall (TA 11–12, 245 ms), clockwise organized atrial fibrillation starts, which, after an activation delay of 286 ms along the middle right atrial free wall, converts into typical clockwise atrial flutter. During counterclockwise organized atrial fibrillation, the atrial recording in the coronary sinus shows a disorganized pattern without clearly discernible atrial activation. During clockwise organized atrial fibrillation, however, distinct atrial signals are present, but a stable 1:1 relation with the atrial activation recorded in the coronary sinus related to the activation along the trabeculated right atrium is first established with the onset of typical clockwise atrial flutter. Abbreviations as in Fig 2.

Fig 2 demonstrates that atrial activation in the coronary sinus first preceded the activation of the right atrial free wall after a silent period of 354 ms and remained related to it subsequently during typical counterclockwise atrial flutter. Fig 3 shows that the atrial activation sequence along the right atrial free wall preceded the atrial coronary sinus recording in a stable relation, beginning with the onset of typical clockwise atrial flutter.

Predictor of Flutter Onset and Direction of Flutter Rotation
During the period of organized atrial fibrillation that immediately preceded the onset of typical atrial flutter, the sequence of activation on the right atrial free wall was analyzed in comparison to the sequence during atrial flutter. Of 16 episodes of typical atrial flutter, 13 were counterclockwise, and in all, the sequence of activation during the preceding period of organized atrial fibrillation was craniocaudal (see Fig 2). In the 3 episodes of clockwise atrial flutter, the sequence of activation during the preceding period of organized atrial fibrillation was caudocranial (see Fig 3). In other words, the activation sequence on the right atrial free wall during those episodes of organized atrial fibrillation that heralded the onset of typical atrial flutter was always predictive of the direction of rotation of the flutter wave.

A total of 1313 seconds of atrial fibrillation (22 minutes) were completely scanned for imbedded periods of organized atrial fibrillation, as defined above. Organized atrial fibrillation with a craniocaudal activation sequence along the right atrial free wall (counterclockwise) occurred 120 times (mean, 9±7; median, 6; range, 1 to 30 per epoch of atrial fibrillation) and lasted from 3 to 89 cycles (mean, 14±12 cycles; median, 9 cycles), or from 0.5 to 19.5 seconds (mean, 2.6±2.6 seconds). Organized atrial fibrillation with a caudocranial activation sequence along the right atrial free wall (clockwise) was observed 51 times (mean, 3.6±3 times; median, 2 times; range, 0 to 8 times per epoch of atrial fibrillation). The range of duration was from 3 to 41 cycles (mean, 9±7 cycles; median, 7 cycles) or from 0.5 to 6.6 seconds (mean, 1.6±1.2 seconds). Therefore, a craniocaudal activation sequence of organized atrial fibrillation occurred more frequently (P<.01) and lasted longer (P<.01) than a caudocranial organized activation sequence. Even in 2 patients who each had 2 different episodes of atrial fibrillation that converted to either counterclockwise or clockwise typical atrial flutter, the craniocaudal activation sequence occurred more frequently (78 versus 28 episodes, P<.01) and lasted longer (16±12 versus 9±7 cycles, P<.01) than the caudocranial activation sequence.

The simple presence of atrial fibrillation with an organized pattern of activation on the right atrial free wall was a common finding and did not, in itself, predict impending conversion to typical atrial flutter. Although, as Fig 4 shows, an organized pattern of activation along the right atrial free wall may be present, disorganized atrial fibrillation usually resumes. However, the presence of an organized atrial fibrillation activation sequence on the right atrial free wall that was then followed by an electrical silence longer than the atrial flutter cycle length along the lateral right atrium that did not result in conversion to typical atrial flutter occurred only once during the 171 episodes of organized atrial fibrillation. Thus, the presence of organized atrial fibrillation followed by a pause longer than the typical clinical atrial flutter cycle length, or at least 15% (or 30 ms) longer than the preceding organized pattern of atrial fibrillation, was highly sensitive (100%) and specific (99%) in predicting conversion of atrial fibrillation into stable typical atrial flutter.

View larger version (60K): [in this window] [in a new window]   Figure 4. Organized atrial fibrillation with craniocaudal activation sequence, which does not convert to atrial flutter (left). Note the presence of distinct atrial signals recorded with the coronary sinus catheter. Disorganized atrial fibrillation resumes (right), because no significant activation delay on the right atrial free wall is present and no stable relationship between the atrial activation recorded in the coronary sinus and along the right atrial free wall is established. Abbreviations as in Fig 2.

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
In this study, right atrial multisite endocardial recordings during the conversion of atrial fibrillation to flutter were obtained in patients before routine radiofrequency catheter ablation of typical atrial flutter. A remarkably stereotypical pattern of endocardial organization was observed in 16 conversion episodes obtained in 10 patients. After a significant increase in atrial fibrillation cycle length, a sudden change in organization of the activation sequence along the trabeculated right atrium was noted, converting the rhythm to what we have termed "organized atrial fibrillation." Another local activation delay on the right atrial free wall, significantly longer than the atrial flutter cycle length, together with the establishment of a 1:1 synchronization of the trabeculated right atrial activation sequence and the coronary sinus os, spontaneously initiated the onset of stable typical clockwise or counterclockwise atrial flutter. The direction of rotation (counterclockwise versus clockwise) was always predicted by the direction of activation on the right atrial free wall during the organized phase of atrial fibrillation. The sudden change in activation sequence and the subsequent organization of atrial fibrillation were common findings in flutter patients with atrial fibrillation. However, the additional activation delay at least 15% (or 30 ms) longer than the mean preceding organized atrial fibrillation cycle length and longer than the subsequent flutter cycle length was highly sensitive and specific to predict conversion to typical atrial flutter. Furthermore, a craniocaudal right atrial free wall activation sequence was much more common and lasted longer than a caudocranial activation sequence when organized atrial fibrillation did occur. Unlike animal studies of the conversion of atrial fibrillation to flutter and vice versa,⁸ adenosine administration never resulted in degeneration of atrial flutter to fibrillation, despite its known effect to shorten atrial refractory periods.

Prior Studies
In humans, little is known about the mechanism and significance of the conversion of atrial fibrillation to atrial flutter. Watson and Josephson¹² described a brief period of irregular atrial activity in one or more intracardiac leads in most of the patients with inducible atrial flutter during programmed atrial stimulation. Tunick et al¹³ reported on 96 patients with atrial fibrillation undergoing Holter monitoring, 25 of whom also had atrial flutter. However, most of these patients were studied shortly after cardiac surgery. Careful observations recently reported by Waldo and Cooper⁹ featured the presence of a transitional rhythm, usually atrial fibrillation, preceding the spontaneous onset of atrial flutter in patients after open heart surgery. A single intracardiac recording site was available in these patients, so any change in activation sequence, such as that observed in our study, would not have been discerned. However, these authors noted that atrial fibrillation often slowed and became more regular in cycle length before conversion to typical atrial flutter. Our observation of a more organized atrial fibrillation preceding typical atrial flutter may well correspond to that period of stability described by Waldo and Cooper.⁹

In the canine model of sterile pericarditis, Shimizu et al¹⁴ demonstrated that the onset of atrial flutter is preceded by a short period of atrial fibrillation. By use of multielectrode mapping in the open-chest state, development of an area of slow conduction and unidirectional conduction block was demonstrated. Unidirectional block occurred at the area where the wave front crossed perpendicular to the orientation of the atrial muscle fibers, mostly along the sulcus terminalis, suggesting anisotropic conduction along anatomic obstacles. Using the same model,⁸ the aforementioned group performed multisite mapping of the conversion of atrial flutter to atrial fibrillation and atrial fibrillation to atrial flutter. It was shown that the length of a line of functional block along the right atrial free wall was critical for the maintenance of stable atrial flutter.

Hypothetical Mechanism of Atrial Fibrillation Organization in Humans
The characteristic sequence of events present during right atrial endocardial mapping of the conversion of atrial fibrillation to atrial flutter in humans and the similarities to prior human and animal data^8,9,14 allow us to speculate on the following mechanism of organization of atrial fibrillation to atrial flutter (Fig 5): During disorganized atrial fibrillation, it may be assumed that multiple reentrant wavelets are activating the right atrium. The significant increase in atrial fibrillation cycle length and the subsequent activation delay heralding the onset of organized atrial fibrillation could be due to the coalescence and annihilation of wavelets (Fig 5A and 5B). Subsequently, perhaps because of anisotropic conduction properties¹⁵ or the anatomic barriers of crista terminalis and eustachian ridge,^1–3 the trabeculated right atrium is activated by a single wavelet by part of the counterclockwise flutter circuit (Fig 5C). Most commonly, the orthodromic craniocaudal wavelet along the right atrial free wall can be expected to coalesce with an antidromic wavelet from the septum and then block in or near the subeustachian isthmus. After several organized cycles, disorganized atrial fibrillation resumes (Figs 4 and 5C). If the craniocaudal wavelet finds the subeustachian isthmus and the adjacent septum excitable, however (statistically, a rare event), this descending wave is able to activate the septum and the left atrium (Fig 5D), and atrial flutter occurs (Fig 5E). This would explain the synchronization of the trabeculated right atrial activation with the coronary sinus os and the high septum with the onset of typical atrial flutter. Thus, the excitability of the narrow isthmus between the inferior vena cava and the tricuspid annulus and of the adjacent septum and left atrium may be the ultimate prerequisites for the occurrence of stable typical counterclockwise atrial flutter.

View larger version (70K): [in this window] [in a new window]   Figure 5. Hypothetical mechanism of atrial fibrillation organizing into typical atrial flutter. Top, Circular schematics of the right atrial anatomy, with the trabeculated right atrium on the left and the smooth right atrium on the right from the crista terminalis (CT) and the outer border representing the tricuspid annulus (TA). The five stages during the conversion from atrial fibrillation to atrial flutter are represented by A through E. An increase in atrial fibrillation cycle length is the result of a decrease in the number of reentrant wavelets (A). Sudden change in the activation sequence results from block of reentrant wavelets along the right atrial free wall and the subeustachian isthmus (B). Organized atrial fibrillation with a craniocaudal activation of the right atrial free wall (C). Antidromic activation of the subeustachian isthmus and block of the craniocaudal wavelet prevents the onset of typical atrial flutter. Only if the craniocaudal activation finds the subeustachian isthmus and the adjacent septum excitable, a single wavelet may establish. Consequently, activation of the subeustachian isthmus and the further coalescence of wavelets cause another significant activation delay along the right atrial free wall (D). As a stable relation between right atrial free wall activation and coronary sinus atrial activation is established, typical counterclockwise atrial flutter initiates (E). With the onset of typical atrial flutter (D, E), discrete septal atrial signals with a 1:1 relation to the activation along the trabeculated right atrium become established. Bottom, Simultaneous intracardiac and surface ECG tracing divided into five stages (A through E). SVC indicates superior vena cava; CT, crista terminalis; IVC, inferior vena cava; CS, coronary sinus os; ER, eustachian ridge; and SEPT, bipolar septal recordings during conversion from atrial fibrillation to typical atrial flutter.

Two relevant differences seem to be present in humans and in the canine sterile pericarditis model. First, in dogs, the change from atrial fibrillation to atrial flutter was gradual, as was the documented length of the line of block, whereas in humans, a sudden change in activation sequence and a sudden and significantly longer activation delay preceded stable atrial flutter. Second, in the canine model, the conversion from atrial flutter to atrial fibrillation and from atrial fibrillation to atrial flutter occurred in approximately equal proportions. In our patients, however, the spontaneous conversion of typical atrial flutter to atrial fibrillation never occurred, even after administration of multiple doses of adenosine. Likewise, Stambler et al¹⁶ noted that adenosine administration during type I atrial flutter did not cause degeneration into atrial fibrillation in 41 patients. In contrast, adenosine has been reported to consistently cause degeneration of atrial flutter into atrial fibrillation in the canine model.⁸ One possible explanation for the disparity noted in the effect of adenosine in the canine model and in humans is that the line of block along the crista terminalis and eustachian valve^1–3 may be fixed in patients susceptible to typical atrial flutter and is therefore not affected by the adenosine-induced shortening of the refractory period, whereas in the canine model, the line of block around which flutter circulates has a functional component that can be abolished by adenosine-induced shortening of refractory period. Once atrial fibrillation organizes into atrial flutter in humans, as a result of the coalescence of circulating fibrillatory wavelets, stable atrial flutter around anatomic obstacles results and reversion back to atrial fibrillation is unlikely. This is also in keeping with the clinical observation that atrial flutter is a remarkably stable arrhythmia, which may occasionally be present for decades once it is initiated.

Counterclockwise Versus Clockwise Atrial Flutter
Recent studies provide evidence that the anatomic barriers that define the atrial flutter circuit in typical clockwise and in typical counterclockwise atrial flutter are the same.^10,17,18 However, the reason for counterclockwise atrial flutter to be much more commonly observed clinically remains unknown. For the initiation of atrial flutter by programmed stimulation in the electrophysiology laboratory, it has been shown that the site of pacing determines the direction of rotation of the flutter wave¹⁸: clockwise atrial flutter is induced after pacing from the trabeculated right atrium, whereas counterclockwise atrial flutter is induced from the smooth right atrium. Our study provides evidence that atrial fibrillation is much more likely to organize in a craniocaudal (counterclockwise) than in a caudocranial (clockwise) direction along the trabeculated right atrium. Possibly, tissue-specific determinants of anisotropic conduction¹⁵ or an anatomic "funneling" effect statistically favor impulse propagation on the trabeculated right atrium in a craniocaudal rather than a caudocranial direction. In any case, because the direction of activation along the right atrial free wall during organized atrial fibrillation predicts the direction of rotation of subsequent atrial flutter and because craniocaudal organized atrial fibrillation occurs more commonly than caudocranial organization, counterclockwise flutter is statistically more likely and therefore is the predominant clinical arrhythmia.

Study Limitations
In the present study, endocardial recordings with sufficient resolution have been obtained only from the trabeculated right atrium. Left atrial activation has been represented only by coronary sinus recordings, and septal recordings with limited resolution were available in 4 patients (5 episodes). Therefore, the contribution of the intraatrial septum and the left atrium in the conversion from atrial fibrillation to atrial flutter cannot be fully elucidated from our data. In addition, higher-density recordings will be necessary to detail the exact site(s) of block that occur when atrial fibrillation becomes more organized and when organized atrial fibrillation initiates atrial flutter. Six episodes of conversion of atrial fibrillation to atrial flutter in 4 patients were documented during amiodarone treatment. Although the mean flutter cycle length was significantly longer in these patients (276±35 versus 220±15 ms, P<.01) and amiodarone might increase the propensity of atrial fibrillation to organize, the characteristic activation sequence was not different in patients with (Fig 1) or without (Fig 2) amiodarone therapy. All but two of the epochs of atrial fibrillation in the present study were electrically induced by pacing or shock delivery. Therefore, we cannot be sure if the changes in the activation sequence during these electrically induced periods of atrial fibrillation are the same as those occurring with spontaneously elicited clinical atrial fibrillation. However, in the two episodes of the spontaneous onset of atrial fibrillation that occurred in the electrophysiology laboratory, the ultimate organization of atrial fibrillation to atrial flutter followed the same stereotypical pattern as the electrically induced episodes. In addition, we analyzed only episodes of atrial fibrillation >15 seconds.

Most of the patients in our study had relatively normal right atrial size, and only 3 patients had significant biatrial enlargement. Therefore, patients with substantial atrial enlargement caused by valvular heart disease might exhibit a different pattern of organization and a different response to adenosine infusion. Likewise, patients with atrial fibrillation organizing into atypical atrial flutter have not been included in the present study. In these patients, a frequent transition from and to atrial fibrillation has been reported,¹⁰ but the circuit and barriers of atypical atrial flutter have yet to be defined.

Conclusions
A remarkably stereotypical pattern of endocardial organization during conversion from atrial fibrillation to atrial flutter was documented in humans. The characteristic activation sequence may shed light on the mechanism of the onset of clinical atrial flutter. The more common appearance of counterclockwise as opposed to clockwise atrial flutter is explained by the predominant craniocaudal activation of the right atrial free wall during organized atrial fibrillation in patients with the anatomic substrate for typical atrial flutter. That spontaneous degeneration of typical atrial flutter back to fibrillation never occurred implies that stable, probably anatomic, barriers provide the substrate for reentry in typical atrial flutter. The significance of atrial fibrillation during electrophysiological testing in patients with atrial flutter relative to the long-term risk of clinical atrial fibrillation after successful ablation of atrial flutter requires further studies.

	Acknowledgments

 
Dr Roithinger is funded by a grant from the Max Kade Foundation Inc. Dr Lesh is supported in part by a grant from the NIH (No. 1-RO-1-HL-55227)

Received April 21, 1997; revision received July 28, 1997; accepted August 5, 1997.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Kalman JM, Olgin JE, Saxon LA, Fisher WG, Lee RJ, Lesh MD. Activation and entrainment mapping defines the tricuspid annulus as the anterior barrier in typical atrial flutter. Circulation. 1996;94:398–406.[Abstract/Free Full Text]
   
2. Olgin JE, Kalman JM, Fitzpatrick AP, Lesh MD. Role of right atrial endocardial structures as barriers to conduction during human type I atrial flutter: activation and entrainment mapping guided by intracardiac echocardiography. Circulation. 1995;92:1839–1848.[Abstract/Free Full Text]
   
3. Nakagawa H, Lazarra R, Khastgir T, Beckman KJ, McClelland JH, Imai S, Pitha JV, Becker AE, Arruda M, Gonzalez MD, Widman LE, Rome M, Neuhauser J, Wang X, Calame JD, Goudeau MD, Jackman WM. Role of the tricuspid annulus and the eustachian valve/ridge on atrial flutter. Circulation. 1996;94:407–424.[Abstract/Free Full Text]
   
4. Feld GK, Fleck P, Chen PS, Boyce K, Bahnson TD, Stein JB, Calisi CM, Ibarra M. Radiofrequency catheter ablation for the treatment of human type I atrial flutter. Circulation. 1992;86:1233–1240.[Abstract/Free Full Text]
   
5. Kirkorian G, Moncada E, Chevalier P, Canu G, Claudel JP, Bellon C, Lyon L, Touboul P. Radiofrequency ablation of atrial flutter. Circulation. 1994;90:2804–2814.[Abstract/Free Full Text]
   
6. Moe GK. On, the multiple wavelet hypothesis of atrial fibrillation. Arch Int Pharmacodyn Ther. 1962;140:183–188.
   
7. Gerstenfeld EP, Sahakian AV, Swiryn S. Evidence for transient linking of atrial excitation during atrial fibrillation in humans. Circulation. 1992;86:375–382.[Abstract/Free Full Text]
   
8. Ortiz J, Niwano S, Abe H, Rudy Y, Johnson NJ, Waldo AL. Mapping the conversion of atrial flutter to atrial fibrillation and atrial fibrillation to atrial flutter: insights into mechanisms. Circ Res. 1994;74:882–894.[Abstract/Free Full Text]
   
9. Waldo AL, Cooper TB. Spontaneous onset of type I atrial flutter in patients. J Am Coll Cardiol. 1996;28:707–712.[Abstract]
   
10. Kalman JM, Olgin JE, Saxon LA, Lee RJ, Scheinman MM, Lesh MD. Electrocardiographic and electrophysiologic characterization of atypical atrial flutter in man: use of activation and entrainment mapping and implications for catheter ablation. J Cardiovasc Electrophysiol. 1997;8:121–144.[Medline] [Order article via Infotrieve]
    
11. Wells JL Jr, Karp RB, Kouchoukos NT, MacLean WAH, James TN, Waldo AL. Characterization of atrial fibrillation in man: studies following open heart surgery. PACE Pacing Clin Electrophysiol. 1978;1:426–438.[Medline] [Order article via Infotrieve]
    
12. Watson RM, Josephson ME. Atrial flutter, I: electrophysiologic substrates and modes of initiation and termination. Am J Cardiol. 1980;45:732–740.[Medline] [Order article via Infotrieve]
    
13. Tunick PA, McElhinney L, Mitchell T, Kronzon I. The alternation between atrial flutter and atrial fibrillation. Chest. 1992;101:34–36.[Abstract/Free Full Text]
    
14. Shimizu A, Nozaki A, Rudy Y, Waldo AL. Onset of induced atrial flutter in the canine pericarditis model. J Am Coll Cardiol. 1991;17:1223–1234.[Abstract]
    
15. Saffitz JE, Kanter HL, Green KG, Tolley TK, Beyer EC. Tissue-specific determinants of anisotropic conduction velocity in canine atrial and ventricular myocardium. Circ Res. 1994;74:1065–1070.[Abstract/Free Full Text]
    
16. Stambler BS, Wood MA, Ellenbogen KA. Pharmacologic alterations in human type I atrial flutter cycle length and monophasic action potential duration: evidence of a fully excitable gap in the reentrant circuit. J Am Coll Cardiol. 1996;27:453–461.[Abstract]
    
17. Saoudi N, Nair M, Abdelazziz A, Poty H, Daou A, Anselme F, Letac B. Electrocardiographic patterns and results of radiofrequency catheter ablation of clockwise type I atrial flutter. J Cardiovasc Electrophysiol. 1996;7:931–942.[Medline] [Order article via Infotrieve]
    
18. Olgin JE, Kalman JM, Saxon LA, Lee RJ, Lesh MD. The mechanism of initiation of atrial flutter in man: site of unidirectional block and direction of rotation. J Am Coll Cardiol. 1997;29:376–384.[Abstract]
    

This article has been cited by other articles:

				A. L. Waldo and G. K. Feld Inter-Relationships of Atrial Fibrillation and Atrial Flutter: Mechanisms and Clinical Implications J. Am. Coll. Cardiol., February 26, 2008; 51(8): 779 - 786. [Abstract] [Full Text] [PDF]

				Z. Qu Critical mass hypothesis revisited: role of dynamical wave stability in spontaneous termination of cardiac fibrillation Am J Physiol Heart Circ Physiol, January 1, 2006; 290(1): H255 - H263. [Abstract] [Full Text] [PDF]

				F. Kilicaslan, A. Verma, H. Yamaji, N. F. Marrouche, O. Wazni, J. E. Cummings, S. Hao, M. W. Andrews, S. Beheiry, A. Abdul-Karim, et al. The need for atrial flutter ablation following pulmonary vein antrum isolation in patients with and without previous cardiac surgery J. Am. Coll. Cardiol., March 1, 2005; 45(5): 690 - 696. [Abstract] [Full Text] [PDF]

				B. Olshansky Combining ablation of atrial fibrillation with ablation of atrial flutter: are we there yet? J. Am. Coll. Cardiol., June 2, 2004; 43(11): 2063 - 2065. [Full Text] [PDF]

				O. Wazni, N. F. Marrouche, D. O. Martin, A. M. Gillinov, W. Saliba, E. Saad, A. Klein, M. Bhargava, D. Bash, R. Schweikert, et al. Randomized Study Comparing Combined Pulmonary Vein-Left Atrial Junction Disconnection and Cavotricuspid Isthmus Ablation Versus Pulmonary Vein-Left Atrial Junction Disconnection Alone in Patients Presenting With Typical Atrial Flutter and Atrial Fibrillation Circulation, November 18, 2003; 108(20): 2479 - 2483. [Abstract] [Full Text] [PDF]

				S. Schmieder, G. Ndrepepa, J. Dong, B. Zrenner, J. Schreieck, M. A.E. Schneider, M. R. Karch, and C. Schmitt Acute and long-term results of radiofrequency ablation of common atrial flutter and the influence of the right atrial isthmus ablation on the occurrence of atrial fibrillation Eur. Heart J., May 2, 2003; 24(10): 956 - 962. [Abstract] [Full Text] [PDF]

				O. Berenfeld, A. V. Zaitsev, S. F. Mironov, A. M. Pertsov, and J. Jalife Frequency-Dependent Breakdown of Wave Propagation Into Fibrillatory Conduction Across the Pectinate Muscle Network in the Isolated Sheep Right Atrium Circ. Res., June 14, 2002; 90(11): 1173 - 1180. [Abstract] [Full Text] [PDF]

				A. L. Waldo Mechanisms of atrial flutter and atrial fibrillation: distinct entities or two sides of a coin? Cardiovasc Res, May 1, 2002; 54(2): 217 - 229. [Full Text] [PDF]

				B. Brembilla-Perrot Clinical course after radiofrequency ablation of type I atrial flutter. Identification of patients who risk atrial arrhythmia recurrences Eur. Heart J., March 2, 2002; 23(6): 441 - 443. [Full Text] [PDF]

				J. B. Morton, M. J. Byrne, J. M. Power, J. Raman, and J. M. Kalman Electrical Remodeling of the Atrium in an Anatomic Model of Atrial Flutter: Relationship Between Substrate and Triggers for Conversion to Atrial Fibrillation Circulation, January 15, 2002; 105(2): 258 - 264. [Abstract] [Full Text] [PDF]

				P. D. Bella, A. Fraticelli, C. Tondo, S. Riva, G. Fassini, and C. Carbucicchio Atypical atrial flutter: clinical features, electrophysiological characteristics and response to radiofrequency catheter ablation Europace, January 1, 2002; 4(3): 241 - 253. [Abstract] [PDF]

				P. Delta Bella and C. Tondo Stratification and outcomes in patients with paroxysmal atrial fibrillation: pulmonary vein ablation versus linear lesions Eur. Heart J. Suppl., November 1, 2001; 3(suppl_P): P33 - P40. [Abstract] [PDF]

				P. B. Sparks and J. M. Kalman Is atrial flutter a risk factor for stroke? J. Am. Coll. Cardiol., September 1, 2001; 38(3): 785 - 788. [Full Text] [PDF]

				C. W. Israel, J. R. Ehrlich, G. Gronefeld, A. Klesius, T. Lawo, B. Lemke, and S. H. Hohnloser Prevalence, characteristics and clinical implications of regular atrial tachyarrhythmias in patients with atrial fibrillation: insights from a study using a new implantable device J. Am. Coll. Cardiol., August 1, 2001; 38(2): 355 - 363. [Abstract] [Full Text] [PDF]

				G. Stabile, A. De Simone, P. Turco, V. La Rocca, P. Nocerino, C. Astarita, F. Maresca, C. De Matteis, T. Di Napoli, E. Stabile, et al. Response to flecainide infusion predicts long-term success of hybrid pharmacologic and ablation therapy in patients with atrial fibrillation J. Am. Coll. Cardiol., May 1, 2001; 37(6): 1639 - 1644. [Abstract] [Full Text] [PDF]

				H. J. Sih, D. P. Zipes, E. J. Berbari, D. E. Adams, and J. E. Olgin Differences in organization between acute and chronic atrial fibrillation in dogs J. Am. Coll. Cardiol., September 1, 2000; 36(3): 924 - 931. [Abstract] [Full Text] [PDF]

				H. Paydak, J. G. Kall, M. C. Burke, D. Rubenstein, D. E. Kopp, R. J. Verdino, and D. J. Wilber Atrial Fibrillation After Radiofrequency Ablation of Type I Atrial Flutter : Time to Onset, Determinants, and Clinical Course Circulation, July 28, 1998; 98(4): 315 - 322. [Abstract] [Full Text] [PDF]

This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Roithinger, F. X. Articles by Lesh, M. D. Search for Related Content PubMed