circus movementatrial flutter in canine sterile...

1716

Circus Movement Atrial Flutter in CanineSterile Pericarditis Model

Activation Patterns During Entrainment andTermination of Single-Loop Reentry In Vivo

Wolfgang Schoels, MD; Mark Restivo, PhD; Edward B. Caref, MA;William B. Gough, PhD; and Nabil El-Sherif, MD

Background. Recently, we used a custom designed "jacket" electrode with 127 bipolarelectrodes in a flexible nylon matrix to map the total atrial epicardial surface in the in situcanine heart. Atrial flutter in dogs with sterile pericarditis was shown to be due to a single wavefront circulating around a combined functional/anatomic obstacle, with the arc of functionalconduction block contiguous with one or more of the atrial vessels.Methods and Results. In the present study, this model was used to analyze the activation

pattern during pacing-induced entrainment and termination of single reentrant loops in asyncytium without anatomically predetermined pathways. Sustained atrial flutter was inducedin five dogs with 3-5-day-old sterile pericarditis. Atrial pacing at a cycle length 5-30 msecshorter than the spontaneous cycle length entrained the arrhythmia and could result in a"classical" activation pattern, characterized by an antidromic stimulated wave that collidedwith the reentrant orthodromic wave front of the previous beat at a constant site. However, twovariations of this classical activation pattern were also observed: 1) Pacing at short cyclelengths could lead to localized conduction block in antidromic direction, forcing a change in thepathway of the antidromic wave front. This could prevent the expected shift of the site ofcollision in antidromic direction. 2) The stimulated orthodromic wave front could also use apathway different from that of the original reentrant impulse, so that a different circuit wasactive during the pacing period. Termination of atrial flutter by rapid atrial stimulation wasassociated with progressive slowing and finally blocking of the paced orthodromic wave frontand a progressive shift of the site of collision in antidromic direction. The occurrence ofconduction block was determined by the cycle length of stimulation and the number ofstimulated beats. A longer train at the critical cycle length or the critical number of beats at ashorter cycle length could reinduce the same reentrant circuit or a different reentrant circuit,respectively, during stimulated cycles following the beat that terminated reentry.

Conclusions. The epicardial activation sequence during entrainment of reentrant arrhythmiasdoes not necessarily follow a standard activation pattern. Instead, the stimulated orthodromicas well as the antidromic wave front might use a pathway different from that of the originalreentrant wave front. The mechanisms of termination, failure of termination, and reinitiationof single-loop reentry are similar to those in the "figure-eight" reentrant circuit. (Circulation1991;83:1716-1730)

I n 1977, Waldo et a1l demonstrated that failure to as "an increase in the rate of all tissue of the chamberinterrupt atrial flutter with rapid atrial pacing being paced to a faster rate than the tachycardia,at rates faster than the atrial flutter rate was with resumption of the intrinsic rate of the tachycar-

due to transient entrainment of the arrhythmia. They dia upon either abrupt cessation of pacing or slowingdefined entrainment of a tachycardia by rapid pacing of the pacing rate below the intrinsic rate of the

From the Cardiology Division, Department of Medicine, StateUniversity of New York, Health Science Center and Veterans fur Herz und Kreislaufforschung.Administration Medical Center, Brooklyn, N.Y. Address for correspondence: Wolfgang Schoels, MD, Cardiol-

Supported by National Institutes of Health grants HL-36680 and ogy Section, Veterans Administration Medical Center, 800 PolyHL-31341 and the Veterans Administration medical research Place, Brooklyn, NY 11209.funds. W.S. is a recipient of a stipend of the Deutsche Gesellschaft Received July 5, 1990; revision accepted December 11, 1990.

by guest on June 4, 2018http://circ.ahajournals.org/

Dow

nloaded from

http://circ.ahajournals.org/

Schoels et al Entrainment of Single-Loop Reentry 1717

tachycardia." They also described'-4 three criteria,any of which, if present, established transient en-trainment of a tachycardia and suggested reentry asthe most likely underlying mechanism: 1) the dem-onstration of constant fusion, except for the lasttransiently entrained beat, 2) progressive fusion, thatis, different degrees of constant fusion at differentpacing rates, and 3) interruption of a tachyarrhyth-mia by rapid pacing associated with localized conduc-tion block to a site, followed by activation of that sitefrom a different direction and with shorter conduc-tion time by the next pacing impulse. Clinical interestin the entrainment phenomenon and these criteriaarose from their implications for antitachycardiapacing,5 their value in identifying and understandingreentrant arrhythmias,3 and their possible role inidentifying the critical site for ablative interventionsin the nonpharmacological treatment of reentrantarrhythmias.6 However, the entrainment concept hasbeen developed largely by deductive analysis, basedon the surface electrocardiogram and a limited num-ber of intracardial recordings. Only a few attemptshave been made to actually document the activationsequence during entrainment and termination ofventricular7 or supraventricular8 tachyarrhythmias.In fact, the spread of activation during entrainmentand termination of circus movement atrial flutter in afunctional model of reentry has never been shown.We have recently described a technique for epicar-dial mapping of the total atrial surface in the in situcanine heart.9 Using this approach in dogs withsterile pericarditis, we could demonstrate the epicar-dial activation sequence during induction, spontane-ous termination, and sustenance of circus movementatrial flutter in a syncytium without anatomicallypredetermined pathways. The present study was con-ducted to analyze the epicardial activation sequenceduring pacing-induced entrainment, termination, andfailure of termination of reentrant atrial flutter inthis animal model.

MethodsDetails of the model preparation, the recording

techniques, and the methods for constructing iso-chronal maps have previously been described.9-12Studies were performed in five mongrel dogs(15-20 kg) with sterile pericarditis and induciblesustained atrial flutter. Sterile pericarditis was cre-ated by generously dusting the epicardial surface ofboth atria with talcum powder.13 After recovery,induction of atrial flutter by rapid atrial pacing wasattempted daily with the dogs in the conscious,nonsedated state. Mapping studies were performedon the day following induction of sustained atrialflutter (3-5 days after the initial surgery). The chestwas reopened, and a specially designed electrodearray was placed on the epicardial surface of bothatria. This atrial "jacket" contained 111 bipolarrecording electrodes with an interpolar distance of1-2 mm in a flexible nylon mesh. The interelectrodedistance ranged from 3 mm to 5 mm but could reach

5-8 mm in certain areas (e.g., the atrial append-ages) when the nylon mesh was stretched. Fivesmall Teflon patches containing two to five elec-trode pairs provided 16 additional bipolar record-ings from the area between the pulmonary veins,the area below the inferior vena cava, and theanterior aspect of the atrial appendages. Data werestored and analyzed using a 128-channel computer-ized multiplexer recording system. Activation timeswere marked manually after review of each individ-ual recording. In electrograms showing a sharpintrinsic deflection, the maximum first derivativewas taken as the moment of activation. In multipha-sic electrograms, a contribution of the ventricularactivation or the stimulus artifact to the recordedsignal was excluded by comparing the timing of theelectrogram components with the surface electro-cardiogram and other epicardial recordings. Thepeak of the major deflection was then chosen as themoment of activation. From these activation times,isochronal epicardial activation maps were con-structed manually at 10-msec intervals. During themapping experiments, electrocardiographic lead II,a left or right atrial electrogram, and aortic bloodpressure were continuously monitored on a VR 12recorder (Electronics for Medicine, Lenexa, Kan.).

Study ProtocolAfter placement of the jacket electrode, several

spontaneous beats were recorded to confirm anormal sinus activation. If necessary, the electrodearray was repositioned to assure adequate record-ings from all 16 patch electrodes and at least 90% ofthe electrodes on the jacket. Induction of atrialflutter was then attempted either by rapid burstpacing or by critically timed premature stimuli,provided by a DTU-101 digital stimulator (Bloom,Reading, Pa.). Once sustained atrial flutter wasinduced, recordings from all 127 electrode siteswere obtained. The cycle length and electrocardio-graphic morphology were observed over a timecourse of at least 30 minutes to confirm the stabilityof the arrhythmia. Atrial pacing was then appliedthrough an electrode pair on the jacket electrode.Rectangular impulses of four times diastolic thresh-old strength assured atrial capture at rapid rates.The stimulation protocol started with a single beatat a cycle length 5 msec shorter than the averagecycle length of the induced arrhythmia. The numberof stimuli was increased in steps of one to amaximum of 15 beats without a change in cyclelength. The above sequence was then repeated atprogressively shorter cycle lengths (in steps of 5msec) until 1:1 atrial capture was no longerachieved. If a stimulation protocol terminated thearrhythmia, atrial flutter was reinduced. The stim-ulation sequence was completed if the reinducedepisode of atrial flutter had an identical cycle lengthand activation sequence. Otherwise, the sequencewas started again.


Dow

nloaded from


1718 Circulation Vol 83, No 5 May 1991

SUSTAINED ATRIAL FLUTTER 140 140 140 140

A

B

CD

E

F

G

H

J

K

A

FIGURE 1. Epicardial activation map (left panel) and selected electrograms (right panel) during an episode of sustained atrialflutter. Heavy black dots in map indicateposition ofselected electrode sites (A-K). Representative examplesfor each isochronal areaand site ofstimulation were chosen for demonstration. Heavy solid lines denote arcs offunctional conduction block. Arrows outlinegeneral direction ofmajor activation wave front. Numbers in electrogram recordings reflect interval between consecutive activationsin milliseconds.

Statistical MethodsData are expressed as mean±SD.

ResultsAll five dogs had inducible sustained atrial flutter

with a mean cycle length of 139±21 msec (110-170msec) in the open-chest state. During the arrhythmia,the atrial electrogram morphology was uniform, andthe beat-to-beat interval variation was less than 10msec. Epicardial activation maps revealed single reen-trant loops in the left (n=1) or right (n=4) atrium,with the circumference of one or more of the atrialvessels and an arc of functional conduction blockforming a combined functional/anatomic obstacle.Subsequent episodes in the same dog had the sameactivation sequence and a similar average cycle length(±5 msec). In one dog, however, the reentrant wavefront proceeded either clockwise or counterclockwisearound a similar functional/anatomic obstacle in theleft atrium. Respective episodes differed in cyclelength by 15 msec and were analyzed separately.

Entrainment of the induced arrhythmia(s) could bedemonstrated in all five dogs during stimulation withup to three (2±0.9) different cycle lengths 5-30 msecshorter than the cycle length of the sustained atrialflutter. The term "entrainment" was only used if theobserved activation pattern was stable from beat tobeat even with long drive trains. The shortest pacingcycle that successfully entrained the arrhythmia was20±9 msec shorter than the cycle length of the spon-taneous tachycardia. Of the six different tachycardiasinduced in the five dogs, two tachycardias could beentrained within 25 msec, whereas entrainment waspossible within 5, 15, 20, and 30 msec in one episodefor the other four tachycardias. Repeated attempts toentrain a tachycardia at the same paced cycle length

resulted in the same activation pattern. Terminationwas always possible at a stimulated cycle length of10-35 msec (25+±9 msec) shorter than that of theinduced arrhythmia. The epicardial activation patternduring 22 different episodes of entrainment and ter-mination was critically analyzed, and representativeexamples are presented.Classical Entrainment of Single-Loop Reentry

Figure 1 shows the epicardial activation map (leftpanel) and selected electrograms (right panel) duringan undisturbed episode of sustained atrial flutter at acycle length of 140 msec. The epicardial atrial surfacein this as well as in all subsequent figures is displayedin a planar projection as if the atria were separatedfrom the ventricles along the atrioventricular (AV)ring and as if the inferior bodies of the atrial append-ages were incised from the AV ring to their tips andunfolded. The heavy black dots indicate the positionof the selected recording electrodes. A single reen-trant loop was oriented around an arc of block (theheavy solid line) extending from the proximity of thesuperior vena cava to one of the right pulmonaryveins. Conduction within the free right atrial wall wasvery slow, especially in the area between the centralobstacle and the superior vena cava. The right panelof Figure 1 depicts representative electrograms fromeach isochronal area within the reentrant circuit andfrom the future site of stimulation (site G). Here andin subsequent figures, electrograms are presented tounderline the stability and regularity of the describedactivation pattern, and respective sites were chosenon the basis of the quality of the recording. Thedouble potentials recorded at electrode sites I and Kwere interpreted as local activation and an electro-tonic potential reflecting activation at an adjacent


Dow

nloaded from



s15

A

D

E

H

K

A-

FIGURE 2. Top panel: Epicardial activation sequence during entrainment oftachycardia at cycle length of130 msec. A-Kare electrodesites. Two stimulated beats (S,, and 5,5) and first beat on cessation ofpacing (A,) are depicted. Duringpacing, moment ofstimulationis chosen as time refrrence. Indicates site ofstimulation. Shaded area represents tissue being activatedby reentrant orthodromic wavefront ofimpulses preceding stimulated beats S,, and S15 (S,O and S14, respectively); the common isochrone ofthis reentrant orthodromicand the stimulated antidromic wavefront, which contains the site ofcollision, is represented by dotted area. (See textfor details and Figure1 for explanation of activation maps.) Bottom panel: Selected electrogams during entrainment of tachycardia at cycle length of 130msec. Vertical lines represent stimulus artifact (S). Electrode site next to site ofstimulation is encircled. Anows outline activation sequencein orthodromic (-) and antidromic (---) direction. A, is first tachycardia beat on cessation ofpacing. Note activation sequence andelectrogram morphology at site F before, dunng, and afterpacing

recording site on the opposite side of the arc ofblock.9.11

Atrial stimulation at a cycle length of 130 msecincreased the rate of the tachycardia to the paced rate,with resumption of the spontaneous rate (140 msec)after cessation of pacing. Representative epicardialactivation maps and selected atrial electrograms aredemonstrated in Figure 2. Because the epicardial acti-vation pattern was constant during the pacing period,maps of only two stimulated beats (S,, and S15) areshown (Figure 2, top panel). Stimulation was applied to

a site in the posterioinferior aspect of the left atrium,outside the reentrant circuit. The time of stimulationwas chosen as the time reference. The stimulated wavefront proceeded within the original reentrant pathwayin orthodromic (i.e., in the direction of the originalreentrant wave front) as well as in antidromic (i.e.,opposite the original reentrant wave front) direction.The stimulated antidromic wave front collided after 50msec at a constant site with the reentrant orthodromicwave front of the beat preceding S,1 and S15 (SIO and S14,respectively). This collision blocked further spread of

S11

ATRIAL STIMULATION CL 130 miec


Dow

nloaded from



ATRIAL STIMULATION CL 120 msec

5 120 6 120 7 120

A

B

CD

E

F

H

K

A


Dow

nloaded from



FIGURE 3. Top panel: Epicardial activation maps duringentrainment and termination oftachycardia shown in Figure 2 atpacing cycle length of120 msec. Continuous activation sequenceduring stimulated beats S7, S8, S9, and SI6 is depicted. DuringSq, area below superior vena cava (300-msec isochrone) wasactivated by stimulated orthodromic wave front after reentrantorthodromic wave front of S8 had reached lower right atrial wall(270- and 280-msec isochrones); thus, the difference in activa-tion time between these two isochronal areas was now only 30and 20 msec, respectively. It is, however, safe to assume thatconduction from lower right atrium to area below the superiorvena cava was still blocked, the latter being activated by thestimulated orthodromic wave front of S9. Here, as well as in allsubsequent figures, no isochrones were interpolated between thetwo adjacent areas (note the 300-msec isochrone next to the 270-and 280-msec isochrones). (See textforfurther details and Figure1 for explanation of activation maps.) Bottom panel: Selectedelectrograns during transient entrainment and termination oftachycardia shown in Figure 2 atpacing cycle length of120 msec.The interval between consecutive activations at electrode sitesother than G only decreased after S8. Note progressive change inactivation sequence and electrogram morphology at sites F, E,andD in response to S 7, S8, and S 9. Conduction blockfrom siteI to site J during Sq is indicated by double bar. S10 activated sitesJ, K andA (marked by stars) from a different direction and at ashorter coupling interval, as reflected by change in activationsequence and electrogram morphology.

either impulse. The stimulated orthodromic wave front,however, advanced undisturbed and reset the tachycar-dia by 10 msec. Selected electrograms (Figure 2, bot-tom panel) illustrate the stability of this activationpattern. Different from the spontaneous tachycardia(Figure 1), activation at site G preceded that at site Fduring the pacing period, and the electrogram mor-phology at site F had changed, indicating activationfrom a different direction. Because Sls was the laststimulated beat, the reentrant S15 orthodromic wavefront did not encounter an antidromic wave front withwhich to collide. Instead, activation continued to pro-ceed within the original reentrant pathway. During A1(Figure 2), the spread of activation up to electrode siteD was identical to the activation pattern during pacedbeats. Respective electrodes (A-D) were, therefore,reactivated after 130 msec. As the A1 reentrant wavefront continued within the circuit pathway, electrodesites that had previously been reset by the stimulatedwave front were now activated by this reentrant im-pulse at their original coupling interval of 140 msec.Thus, the A1 activation was entrained to the paced cycleup to the previous site of collision. The couplinginterval returned to 140 msec at all electrode sitesduring the following beat (A2).

Classical Termination of Single-Loop ReentryFurther shortening of the cycle length of stimula-

tion to 120 msec terminated the arrhythmia after atrain of nine stimuli. Epicardial activation maps offour stimulated beats (S7-S1o) and selected atrialelectrograms are shown in Figure 3. Up to S6, only

local capture was achieved, and the original activationsequence was basically undisturbed. Accordingly,there was no change in cycle length other than at siteG (bottom panel). At the time of the S7 stimulation,the reentrant orthodromic wave front of the precedingbeat (S6) had reached site C (top panel). There was an"incomplete" collision in the area below the inferiorvena cava (note the change in the electrogram mor-phology and the shortening of the coupling interval atsite F in the bottom panel). The remaining electrodesites within the circuit were activated at their originalcoupling interval (140 msec). When S8 was introduced(top panel), the S7 orthodromic activation had onlyreached site A. This allowed further advancement ofthe antidromic wave front elicited by S8, until bothimpulses collided in the proximity of electrode site E.Accordingly, the electrogram morphology at site Echanged (bottom panel). The stimulated orthodromicwave front of S8 arrived at the entrance to the slowzone of the reentrant circuit (site H) only 130 msecafter the S7 activation had passed this site, resulting infurther slowing of conduction (illustrated by the in-creasing isoelectric interval between the two compo-nents of electrogram I in the bottom panel). Thus, theorthodromic wave front of S8 had only reached thearea between electrode sites J and K at the time ofinitiation of S (top panel). Again, the activation wavefront induced by S could now advance even further inantidromic direction, before collision occurred nearsite D. In orthodromic direction, site H was reacti-vated after only 120 msec. This interval was probablynot sufficient for the tissue distal to site I to recoverexcitability, resulting in conduction block to sites J, K,and A. Figure 3, bottom panel, documents not onlythe absence of an activation potential at electrodesites J, K, and A during S but also the absence of thesecond deflection in electrogram I. During Slo (toppanel), the complete atrial surface was activated ex-clusively by the activation wave front induced by thisvery stimulus, with a more or less radial spread ofactivation from the site of stimulation and a relativelyshort total atrial activation time (80 msec). Sites A, K,and J (marked by stars in the bottom panel) were nowactivated from a different direction (note the changein the electrogram morphology and the activationsequence in the bottom panel) and at a shortercoupling interval. Activation at site J preceded activa-tion at site I, which is reflected by reversal of thedouble potential in electrogram I. Continuation ofpacing at a cycle length of 120 msec restored theoriginal conduction block, as indicated by the activa-tion sequence and electrogram morphology (bottompanel) during the beat following S1o.Entrainment With a Change in the Reentrant Pathway

In another dog with sustained atrial flutter, theepicardial activation map revealed the activationpattern illustrated in Figure 4A (left). A singlereentrant wave front circulated around a large func-tional/anatomic obstacle that included the inferiorvena cava, the right pulmonary veins, and an arc of


Dow

nloaded from



A) SUSTAINED ATRIAL FLUTTERA

B

c

D

E

F

G

H

J

K

L

A

B) ATRIAL STIMULATION CL 130 msecA AB C

D _

E -

H -

-

J -

K -

L -

A A

A

B -

D

E

F '

H

K-

A-

120

77,C

-^

120

X

A254-

FIGURE 4. Panel A: Epicardial activation sequence (left) and selected electrograms (right) during episode of sustained atrialflutter at cycle length of 140 msec. A-L are electrode sites. Panel B: Epicardial activation sequence and selected atrial electrogramsduring entrainment of tachycardia shown in panelA (sustained atrial flutter) at cycle length of 130 msec. Again, the moment ofstimulation (S) is chosen as the time reference. Shaded area represents tissue activated by reentrant orthodromic wave front ofpreceding beat; dotted area contains site of collision with stimulated antidromic impulse. Panel C: Epicardial activation sequence

and selected electrocardiograms during entrainment of tachycardia shown in panelA (sustained atral flutter) at a cycle length of120 msec. Site of collision has shifted in antidromic direction (electrode site K). Note change inpathway oforthodromic wave fiont.(See Figure 1 for explanation of activation maps.)

S S S S

S S S S S120

¾,7G


Dow

nloaded from



functional conduction block. The slow conductingzone of this reentrant circuit was located in the lowerright atrium and in the isthmus between the inferiorvena cava and the nearby AV ring. As demonstratedby the selected electrograms (Figure 4A, right), thisactivation sequence repeated itself regularly at acycle length of 140 msec.

Atrial stimulation at a cycle length of 130 msecentrained the arrhythmia to the paced rate. Theselected electrograms (Figure 4B, right) illustratethe stability of the activation pattern during thepacing period. Thus, the epicardial activation mapof only one stimulated beat is shown (Figure 4B,left). Stimulation was applied to a site in the leftatrial appendage, outside the reentrant circuit(electrode site G). At the time of stimulation, theorthodromic wave front of the preceding impulsehad constantly reached the area below the inferiorvena cava, and this orthodromic wave front collidedwith the stimulated antidromic wave front in theleft atrium close to electrode site I. The stimulatedwave front also proceeded in orthodromic directionwithin the reentrant pathway and reset the tachy-cardia by 10 msec. Due to the relative prematurityof the stimulated wave front, conduction was nowslower than during the spontaneous tachycardia, sothat electrode sites F and C were separated by sixisochrones (as opposed to three isochrones in Fig-ure 4A). This constant activation pattern was simi-lar to the one in the previous example of transiententrainment (Figure 2) in demonstrating advance-ment of both stimulated wave fronts within theoriginal reentrant pathway and constant collision ofthe stimulated antidromic wave front with the re-entrant orthodromic wave front of the precedingbeat.

Atrial stimulation at a cycle length of 120 msec,however, resulted in a different activation pattern(Figure 4C). The selected electrograms (right) showthat the tachycardia was actually entrained to thepaced cycle of 120 msec and that the activationsequence was constant during the pacing period.Compared with the original activation sequence (Fig-ure 4A), electrode sites G-K were now activated inreverse order, with G recording the earliest activa-tion (note the change in the electrogram morphologyat sites G, H, I, and J). During the precedingentrainment period at 130 msec (Figure 4B, right),these changes had only involved sites G-L. Thus, thetwo pacing cycles resulted in different sites of colli-sion, indicating different degrees of constant fusionat different cycle lengths (i.e., progressive fusion).The epicardial activation map (Figure 4C, left) dem-onstrated that the site of collision had shifted to thearea below the inferior vena cava, further away fromthe site of stimulation. However, the early arrival ofthe stimulated orthodromic wave front at electrodesite E had resulted in an extension of the arc of blockacross the free right atrial wall to the proximity of theAV ring. Therefore, the stimulated orthodromicwave front could no longer proceed within the reen-

trant pathway that had been active during the spon-taneous tachycardia. Instead, entrainment of thetachycardia to the paced rate created a differentorthodromic reentrant path, and only on cessation ofpacing did the reentrant wave front resume theoriginal activation pattern. This latter phenomenonwas probably due to the fact that the reentrant wavefront of the first nonstimulated beat arrived later atelectrode site E than the stimulated orthodromicwave front during the preceding entrainment period.

In Figure 5, top panel, epicardial activation mapsduring another episode of sustained atrial flutter andentrainment of the arrhythmia at stimulated cyclelengths of 115 and 105 msec are shown. The sustainedarrhythmia was due to a single reentrant loop around acombined functional/anatomic obstacle in primarily theleft atrium (Figure 5, top panel, left map). An arc offunctional conduction block extended across the freeleft atrial wall to the proximity of the AV ring, whereconduction was slower than in the remaining part of thecircuit. The reentrant impulse had a revolution time of125 msec and activated electrodes A-E in the free leftatrial wall consecutively (Figure 5, bottom left panel,top electrograms).

Atrial stimulation at a cycle length of 115 msecentrained the arrhythmia to the paced cycle andcreated the expected activation pattern (Figure 5,top panel, middle map). The site of stimulation waslocated within the reentrant circuit, close to elec-trode site E. Collision of the reentrant orthodromicwave front of the preceding impulse with the stim-ulated antidromic wave front occurred after 40msec between sites C and D. Orthodromically, thestimulated wave front entered the original reen-trant pathway, where slowing of conduction re-sulted from the prematurity of the impulse. Se-lected electrograms (Figure 5, bottom left panel,middle electrograms) confirm that site E was nowactivated earlier than site D and that the electro-gram morphology at sites D and E had changed.Sites A-C were still activated consecutively.

Entrainment of the tachycardia to a shorter pacedcycle length (105 msec) resulted in an activation patternsimilar to that during the preceding entrainment period(Figure 5, top panel, right map). The electrogrammorphology at electrode sites A-E did not changesignificantly (Figure 5, bottom left panel, bottom elec-trograms). However, site D was now activated 45 msecafter site E. This was interpreted to indicate conductionblock rather than marked slowing of conduction be-tween the two sites, based on the absence of frag-mented, long-duration electrograms, a hypotheticalconduction velocity of 0.09 m/sec, and our previousfindings on high resolution recordings in this model.9The stimulated wave front proceeded toward the infe-rior pulmonary veins, where it turned medially to enterthe orthodromic reentrant pathway and laterally toreach the free left atrial wall. As a result of this delay inthe spread of the stimulated antidromic wave front, thereentrant orthodromic wave front of the precedingimpulse could activate sites A-C before it coalesced


Dow

nloaded from


1724 Circulation Vol 83, No S May 1991

SUSTAINED ATRIAL FLUTTER ATRIAL STIMULATION CL 115 mssc ATRIAL STIMULATION CL 105 msec

SUSTAINED ATRIAL FLUTTER

A -4

BCD

E


S S S

A r

B rJ

D )J_

115

I

115

C,__

S

4-u~


S S S S S

105 105 105 X105A

B

D j@L > s 11 s

with the stimulated antidromic wave front to activatesite D. Had electrode site D been activated exclusivelyby the reentrant orthodromic wave front, an electro-gram morphology similar to that during the originaltachycardia (Figure 5, bottom left panel, top electro-grams) would have been expected. Collision seemed tooccur in the same area as during stimulation at thelonger (115-msec) cycle. Thus, entrainment at a shorter

FIGURE 5. Top panel: Epicardial activation maps duringepisode of sustained atrial flutter at cycle length of 125 msecand during entrainment of arrhythmia at cycle lengths of 115and 105 msec. During pacing, moment of stimulation ischosen as time reference. Site of collision does not shift inantidromic direction during entrainment at fasterpacing rate.(See text for details and Figure 1 for explanation of activationmaps.) Bottom left panel: Selected atrial electrograms duringspontaneous tachycardia and two episodes of entrainmentshown in top panel. S is the moment ofstimulation. Note thatactivation sequence A-D and electrogram morphology at siteD, though different from spontaneous activation, are similarduring both pacing periods; the difference in activation timebetween sites D and E increases.

cycle length was associated with a change in the path-way of the stimulated antidromic wave front, whichprevented a shift of the site of collision in antidromicdirection (i.e., progressive fusion).

Termination at Two Different SitesThe epicardial activation maps shown in Figure 6

were obtained from an experiment in which sus-

115

1


Dow

nloaded from



2

FIGURE 6. Epicardial activation maps during ternination of spontaneous tachycardia at cycle length of 150 msec by fourstimulated beats at cycle length of 115 msec. Last tachycardia beat before initiation ofpacing (An), four stimulated beats (S,-S4),and first spontaneous beat after cessation ofpacing (NI) are shown. * Indicates site of stimulation. (See Figure 1 for explanationof activation maps.)

tained atrial flutter at a cycle length of 150 msec wasinterrupted by a train of four stimulated beats at acoupling interval of 115 msec. During the sustainedarrhythmia (An), the reentrant activation proceededas a single loop around a large functional/anatomicobstacle in the right atrium. A long arc of functionalconduction block extended from the circumferenceof the superior vena cava to the lower right atrium.There it was paralleled by a shorter arc that reachedthe AV ring. Conduction in the area bounded bythese two arcs, as well as in the isthmus between thesuperior vena cava and the nearby AV ring, wasconsiderably slower than in the remaining part of thecircuit. Slow conduction and conduction block wasalso evident in the lateral right atrial wall, althoughthis area was not part of the reentrant pathway.Stimulation was applied to a site near the tip of theright atrial appendage. Introduction of a train of fourpaced beats at a cycle length of 115 msec resulted inprogressively deeper penetration of the stimulatedwave front in antidromic direction and progressivelyslower conduction in orthodromic direction. Thisprogressive conduction delay was due to furtherslowing of conduction primarily within the area be-tween the superior vena cava and the nearby AVring, which acted like a "buffer" for the second slowzone of the reentrant circuit in the free right atrial

wall. Thus, despite the coupling interval of 115 msec,the stimulated orthodromic wave front of S2 and S3arrived in the lower right atrium 140 msec after thepreceding activation. During S4, the stimulatedorthodromic wave front finally encountered an arc ofconduction block between the superior vena cava andthe AV ring, while the stimulated antidromic wavefront was extinguished after collision with the reen-trant orthodromic wave front of S3. After a pause of230 msec, sinus rhythm resumed (N1).

Termination of the tachycardia was also attemptedwith seven stimulated beats at a cycle length of 115msec (Figure 7). The reinduced arrhythmia had thesame cycle length and activation sequence as in theprevious example (A, of Figures 6 and 7). Onceagain, with each stimulated beat the site of collisionshifted further in antidromic direction, while thespread of the stimulated orthodromic wave front wasprogressively delayed. However, different from theprevious example (Figure 6), the conduction delayaround the superior vena cava was less pronouncedand did not compensate for the prematurity of thestimulated beats. As a result, the stimulated ortho-dromic wave front of S3 and S4 reactivated the lowerright atrium with a coupling interval of 120 msec.Whereas S3 was further delayed in this second slowzone of the reentrant circuit, conduction failed on


Dow

nloaded from



FIGURE 7. Epicardial activa-tion maps showing introductionofseven stimulated beats at cyclelength of 115 msec during sametachycardia as shown in Figure 6with termination of onrgnal (S4)and induction of different tachy-cardia (S7). Last tachycardiabeat before initiation of pacing(An), 7 stimulated beats (S1-S7),and the following spontaneousbeats are shown. (See text fordetails and Figure 1 for explana-tion of activation maps.)

arrival of the stimulated orthodromic wave front ofS4. This conduction block in the lower right atriumterminated the underlying arrhythmia, as indicatedby the collision of the stimulated antidromic with thestimulated orthodromic wave front of S5 during thefollowing cycle.

Table 1 summarizes data from multiple attempts atpacing termination obtained during the experiment

shown in Figures 6 and 7. The data underline thesignificance of the cycle length of stimulation and thenumber of stimulated beats. Atrial stimulation atcycle lengths of 135 and 145 msec entrained thearrhythmia. Thus, by definition, termination couldnot be achieved, irrespective of the number of stim-ulated beats. Pacing cycles of 105 msec and shorter,on the other hand, no longer resulted in regular atrial

S?# Th


Dow

nloaded from



TABLE 1. Minimal Number of Stimulated Beats Required forTermination of Tachycardia at Various Pacing Cycle Lengths

Cycle length (msec)n 105 115 120 125 135 145

109876 + _ _

5 + - _ _4 + - _ _ _

3 - - _ _ _

2 - - _ _ _

n, Number of stimulated beats; +, termination; -, failure.C7ycle length of sustained arrhythmia was 150 msec (Figures 6

and 7).

capture. Within these margins, the minimum numberof stimulated beats required for tachycardia termina-tion was directly related to the cycle length of stim-ulation, so that at least six beats at a cycle length of125 msec or four beats at a cycle length of 115 msechad to be applied. Figure 6 illustrates why less thanfour beats at 115 msec were unable to interrupt thearrhythmia. Only repeated activation of the ortho-dromic pathway at the pacing rate produced theprogressive slowing of conduction that culminated inconduction block during S4.

Termination Followed by Induction of a DifferentReentrant Circuit

Failure of overdrive pacing to terminate atrialflutter was not only the result of entrainment of thearrhythmia but could also be due to induction of adifferent reentrant circuit or reinitiation of the orig-inal tachycardia. As shown in Figure 6, atrial pacingat a cycle length of 115 msec terminated the under-lying arrhythmia after a train of four beats. During atrain of seven beats at the same cycle length (Figure7), it was again the stimulated orthodromic wavefront of 84 that encountered conduction block, al-though this time at a different site. During Ss and 86,the stimulated orthodromic and antidromic wavefronts simply proceeded around both ends of thecentral obstacle and collided with each other on theopposite side of the arc. Both stimulated wave frontswere blocked during 87. Thus, the original reentrantcircuit could not have been active on cessation ofpacing. However, the pacing train was followed by aself-terminating run of three reentrant beats. Therewas an area on the base of the right atrial appendage,outside the actual reentrant pathway, that showedvarious degrees of conduction block during the pac-ing period. Although this area was not activatedduring 84 and Ss (marked by stars in Figure 7), it wasfinally reached by 86. 87 reinduced. an arc of block inthis area, parallel to and cranially contiguous with theAV ring. The S7-stimulated wave front could, there-

fore, only proceed around the free end of this arc ofblock, where it coalesced with the reentrant antidro-mic wave front of S6. A common wave front thenproceeded cranially, parallel to the AV ring. Becauseconduction in the isthmus between the arc of blockand the AV ring was very slow and because nofurther stimulation was applied, the tissue on theproximal side of the arc of block was able to recoverexcitability. Activation continued as a single reen-trant wave front around a functional arc of blockparallel to the AV ring (A1). Although the initialrevolution time was 160 msec, conduction across thefree right atrial wall improved during the followingbeat. Thus, this cycle was completed after 140 msec(A2). During A3, conduction failed within the slowzone of the reentrant circuit. The following beatshowed a normal sinus activation.

Termination Followed by Reinitiation of the OriginalReentrant CircuitThe activation sequence shown in Figure 8 was

recorded in a dog in which the reentrant wave frontcirculated either clockwise (Figure 8) or counter-clockwise (Figure 5) around a similar, yet not iden-tical, functional/anatomic obstacle in the left atrium.The right atrial tissue seemed to be severely compro-mised and exhibited various degrees of conductionblock. Clockwise reentrant activation (An) had arevolution time of 115 msec, with slowing of conduc-tion in the isthmus between the arc of block and theAV ring. The arrhythmia could be entrained to acycle length of 110 msec and was terminated by atrain of six stimulated beats at a cycle length of 105msec, with S6 being the critical beat (data not shown).Figure 8 illustrates an attempt with six stimulatedbeats at a cycle length of 100 msec. Stimulation wasapplied to a site between the pulmonary veins.Up to S3 (Figure 8), the stimulated wave front

progressively advanced in antidromic direction witheach beat. Conduction block near the site of stimu-lation forced a change in the pathway of the stimu-lated orthodromic wave front during S2. The length ofthis new pathway compensated for the prematurity ofthe stimulus, so that the free left atrial wall was stillactivated at a coupling interval of 110 msec. Thisinterval shortened to 100 msec during S3. As a result,the stimulated orthodromic wave front of S3 wasblocked at the entrance to the slow zone along theAV ring. During S4 and S5, the impulse proceeded inboth directions around the remainder of the centralobstacle before the orthodromic wave front collidedwith the antidromic wave front of the same stimulusin the free left atrial wall. Block of the stimulatedantidromic wave front and delayed conduction of thestimulated orthodromic wave front of S6 allowed thelatter impulse to reexcite the tissue in the upper leftatrium. Thus, on cessation of pacing, the originalreentrant tachycardia resumed (A,+1).


Dow

nloaded from



DiscussionThe present study demonstrates epicardial activa-

tion patterns during entrainment and termination ofcircus movement atrial flutter in the canine sterilepericarditis model. It is shown that entrainment ofthe arrhythmia does not necessarily result in a stan-dard activation pattern. Variations of the classicalactivation pattern during pacing-induced terminationof circus movement atrial flutter are also presented.

Comparison With Previous StudiesTransient entrainment of a tachycardia was first

described during rapid pacing of atrial flutter' andthen during ventricular tachycardia,14 but the mech-anism was not initially understood. Subsequent stud-ies on transient entrainment and interruption of theAV bypass type of paroxysmal atrial tachycardia ledto a mechanistic concept, based on deductive analysisof the surface electrocardiogram and a limited num-ber of intracardiac recordings.2 Mapping studies dur-ing entrainment of reentrant atrial arrhythmias haveused the isolated, perfused canine heart.8 In thismodel, a surgically induced lesion forced the reen-trant activation to proceed as a single wave front inthe supravalvular anulus of the tricuspid ring.'516Because the reentrant pathway, and thus the pathwayof any stimulated wave front, is anatomically prede-termined and because the propagating impulse en-counters only normal atrial tissue, the demonstratedactivation sequence during entrainment of this ar-rhythmia8 is not likely to reflect the clinical situation.Atrial flutter in the canine sterile pericarditis model,on the other hand, seems to closely resemble itsclinical counterpart.913 The reentrant pathway is, atleast in part, determined by the functional propertiesof the atrial tissue, which is inhomogeneously af-fected by the underlying inflammation.9 Epicardialmapping in the in situ canine heart, as opposed to theendocardial approach in an isolated heart prepara-tion, is not limited by possible mechanical alterationsof the atria, hemodynamic changes, and the absenceof humoral and autonomic influences. On the otherhand, the in vivo mapping method is unable to mapatrial septal activation. It seems, however, that atleast the major portion and the crucial componentsof the reentrant circuits in the canine sterile peri-carditis model are located in the free atrial wall.9 Theatrial septal tissue, which is not involved in theinflammatory reaction, is not expected to exhibit slowconduction or functional conduction block. Thus,even though our epicardial activation maps might notalways reflect the actual activation path for the totallength of the reentrant pathway, the possible involve-ment of the interatrial septum is not likely to producean essentially different activation pattern.Although the isochronal maps were constructed

manually, it is very unlikely that the changes in thereentrant pathway were due to errors in the measure-ment of activation times. Not only were the observed

changes in the activation sequence profound but sowere the changes in local electrogram configurations.

Entrainment ofAtrial Flutter: EpicardialActivation Sequence

Our study has demonstrated classical activationpatterns for both entrainment and termination ofatrial flutter and has lent credence to previouslyproposed electrocardiographic criteria.24'17 The pres-ence of fusion beats in the surface electrocardiogramhas been explained as the result of a collision betweenthe reentrant and the stimulated wave fronts, withconstant fusion being represented by a constant site ofcollision for each paced beat.2,8 We have shown clas-sical entrainment to be associated with antidromic andorthodromic propagation of the stimulated wave frontwithin the original reentrant pathway, where theorthodromic activation sequence was similar to thatduring the spontaneous tachycardia. The last stimu-lated beat was entrained up to the previous site ofcollision, while its reentrant orthodromic wave frontdid no longer encounter an antidromic wave front withwhich to collide. During entrainment at faster pacingrates, the site of collision could shift in an antidromicdirection. Termination of the arrhythmia was associ-ated with progressive slowing of the orthodromic wavefront and, thus, a progressive shift of the site ofcollision in an antidromic direction. When the ortho-dromic impulse was finally blocked, the tissue betweenthe site of this block and the site of collision was notactivated. This area was subsequently activated di-rectly by the stimulated wave front of the followingbeat after a relatively short conduction time. Conduc-tion block to a site in the reentrant pathway wasfollowed by activation of that site from a differentdirection and with a shorter coupling interval.Two variations of the classical activation pattern

were observed: 1) The orthodromic wave front nolonger proceeded within the original reentrant path-way (Figure 4), although the electrocardiographiccriteria of progressive fusion suggested classical en-trainment. This observation has several implications.First, it emphasizes the limitations of a small numberof recording sites when interpreting a complex acti-vation pattern. Second, it raises the question whetherthe term entrainment, as it has been defined,2-4'14 isapplicable to a situation in which a different reen-trant pathway is used during the pacing period. Andthird, it suggests a limited specificity of the criteriaproposed for identification of classical entrainment,if the term "entrainment" is inappropriate for thissituation. 2) Conduction block in antidromic direc-tion forced a change in the pathway of the antidromicwave front so that entrainment of the arrhythmia to ahigher pacing rate did not result in a shift of the siteof collision. It has been suggested that the ability orinability to demonstrate the entrainment criteria atdifferent pacing sites can be used to locate the slowzone of a reentrant circuit.4 However, as shown inFigure 5, the location of the pacing site relative to theslow zone of a reentrant circuit is not necessarily the


Dow

nloaded from



FIGURE 8. Epicardial activa-tion maps showing introductionof six stimulated beats at cyclelength of 100 msec during atrialflutter at cycle length of 115 msecwith tennination (S3) and rein-duction (S6) of original tachycar-dia. Last tachycardiac beat be-fore initiation ofpacing (An), sixstimulated beats (S-S6), andfirstspontaneous beat after cessationofpacing (An+j) are shown. (SeeFigure 1 for explanation of acti-vation maps.)

critical factor for the inability to demonstrate pro-gressive fusion. Previous investigators8 have alsonoted that the extent of the antidromic shift duringsubsequent entrainment episodes is not proportionalto the decrease in the pacing cycle length.8 Theclassical entrainment concept was derived from elec-

trocardiographic findings in patients with the AVbypass type of paroxysmal atrial tachycardia,2 and thesuggested activation pattern was confirmed previ-ously in a model of reentry with anatomically prede-termined pathways.8 It is likely that the variations inthe classical entrainment pattern that we measured


Dow

nloaded from



are characteristic of a functional model of reentryand not a fixed reentrant circuit.

Termination and Reinitiation of Circus MovementAtrial Flutter

The mechanisms of termination of circus movementatrial flutter by programmed electrical stimulation weresimilar to those demonstrated for figure-eight reentranttachycardias in the ventricle.7 With every stimulatedbeat, the increasing conduction delay in the slow zone

shortened the interval at which consecutive stimulatedorthodromic wave fronts arrived at this critical area. Itcan be assumed that the tissue with the longest refrac-toriness within the slow zone was finally reached by thestimulated wave front before its refractoriness expired,resulting in conduction block. However, different fromthe figure-eight reentrant circuit, there could be more

than one critical site for termination within the singleloop (Figures 6 and 7).

Continuation of pacing beyond tachycardia termina-tion could result in reinitiation of the same circuit or

induction of a different, potentially faster reentrantrhythm. The latter phenomenon is one of the limita-tions of antitachycardia pacing in the treatment ofreentrant arrhythmias.18-20 In our study, this was alwaysdue to termination of the underlying arrhythmia withthe first few stimulated beats, followed by initiation of adifferent circuit by subsequent stimuli. Similar observa-tions have been previously reported during terminationof figure-eight reentrant tachycardias by programmedelectrical stimulation7 and during the initiation of re-

entrant ventricular arrhythmias by burst pacing.2'

Clinical ImplicationsFailure of rapid pacing to terminate atrial flutter

despite atrial capture may involve the followingmechanisms: 1) entrainment of the arrhythmia, 2)stimulation with less than the critical number ofbeats, and 3) stimulation with more than the criticalnumber of beats, resulting in either reinitiation of thesame or induction of a different reentrant circuit.The fact that single reentrant loops in the atria mayinvolve more than one critical site for terminationmight have implications for ablative interventions,because it offers the possibility of choosing the more

suitable site for respective procedures.

AcknowledgmentThe authors wish to thank Gerald Cohen and his

staff in the animal laboratory for their technicalassistance.

References1. Waldo AL, MacLean WAH, Karp RB, Kouchoukos NT,

James TN: Entrainment and interruption of atrial flutter withatrial pacing: Studies in man following open heart surgery.Circulation 1977;56:737-745

2. Waldo AL, Plumb VJ, Arciniegas JG, MacLean WAH, Coo-per TB, Priest MF, James TN: Transient entrainment andinterruption of the atrioventricular bypass pathway type of

paroxysmal atrial tachycardia: A model for understanding andidentifying reentrant arrhythmias. Circulation 1983;67:73-83

3. Waldo AL, Henthorn RW, Plumb VJ, MacLean WAH: Dem-onstration of the mechanism of transient entrainment andinterruption of ventricular tachycardia with rapid atrial pac-ing. JAm Col] Cardiol 1984;3:422-430

4. Okumura K, Henthorn RW, Epstein AE, Plumb VJ, WaldoAL: Further observations on transient entrainment: Impor-tance of pacing site and properties of the components of thereentry circuit. Circulation 1985;72:1293-1307

5. Waldo AL, Henthorn RW, Plumb VJ: Relevance of electro-grams and transient entrainment for antitachycardia devices.PACE 1984;7:588-600

6. Touboul P, Saoudi N, Atallah G, Kirkorian G: Electrophysi-ologic basis of catheter ablation in atrial flutter. Am J Cardiol1989;64:79J-82J

7. El-Sherif N, Gough WB, Restivo M: Reentrant ventriculararrhythmias in the late myocardial infarction period: 14.Mechanisms of resetting, entrainment, acceleration, or termi-nation of reentrant tachycardia by programmed electricalstimulation. PACE 1987;10:341-371

8. Boyden PA, Frame LH, Hoffman BF: Activation mapping ofreentry around an anatomic barrier in the canine atrium:Observations during entrainment and termination. Circulation1989;79:406-416

9. Schoels W, Gough WB, Restivo M, El-Sherif N: Circusmovement atrial flutter in the canine sterile pericarditis mod-el: Activation patterns during initiation, termination and sus-tained reentry in vivo. Circ Res 1990;67:35-50

10. El-Sherif N, Smith RA, Evans KA: Canine ventricular arrhyth-mias in the late myocardial infarction period: 8. Epicardialmapping of reentrant circuits. Circ Res 1981;49:255-265

11. Restivo M, Gough WB, El-Sherif N: Ventricular arrhythmiasin the subacute myocardial infarction period: High-resolutionactivation and refractory patterns of reentrant rhythms. CircRes 1990;66:1310-1327

12. Mehra R, Zeiler RH, Gough WB, El-Sherif N: Reentrantventricular arrhythmias in the late myocardial infarction peri-od: 9. Electrophysiologic-anatomic correlation of reentrantcircuits. Circulation 1983;67:11-24

13. Page PL, Plumb VJ, Okumura K, Waldo AL: A new animalmodel of atrial flutter. JAm Coll Cardiol 1986;8:872-879

14. MacLean WAH, Plumb VJ, Waldo AL: Transient entrain-ment and interruption of ventricular tachycardia. PACE 1981;4:358-366

15. Frame LH, Page RL, Hoffman BF: Atrial reentry around ananatomic barrier with a partially refractory excitable gap. CircRes 1986;58:495-511

16. Frame LH, Page RL, Boyden PA, Fenoglio JJ, Hoffman BF:Circus movement in the canine atrium around the tricuspidring during experimental atrial flutter and during reentry invitro. Circulation 1987;76:1155-1175

17. Portillo B, Mejias J, Leon-Portillo N, Zaman L, Myerburg RJ,Castellanos A: Entrainment of atrioventricular nodal reen-trant tachycardias during overdrive pacing from high rightatrium and coronary sinus. Am J Cardiol 1984;53:1570-1576

18. Fisher JD, Mehra R, Furman S, Ostrow E: Comparative effec-tiveness of pacing techniques for termination of well toleratedsustained ventricular tachycardia. PACE 1983;6:915-922

19. Naccarelli GV, Zipes DP, Rahilly TG, Heger JJ, PrystowskyEN: Influence of tachycardia cycle length and antiarrhythmicdrugs on pacing terminations and accelerations of ventriculartachycardia. Am Heart J 1983;105:1-5

20. Roy D, Waxman HL, Buxton AE, Marchlinski FE, Cain ME,Gardner MJ, Josephson ME: Termination of ventriculartachycardias: Role of tachycardia cycle length. Am J Cardiol1982;50:1346-1350

21. El-Sherif N, Mehra R, Gough WB, Zeiler R: Reentrantventricular arrhythmias in the late myocardial infarction peri-od: II. Burst pacing versus multiple premature stimulation inthe induction of reentry. JAm Coll Cardiol 1984;4:295-304

KEY WORDS * programmed electrical stimulation * reentryarrhythmias


Dow

nloaded from


W Schoels, M Restivo, E B Caref, W B Gough and N el-Sherifduring entrainment and termination of single-loop reentry in vivo.

Circus movement atrial flutter in canine sterile pericarditis model. Activation patterns

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1991 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.83.5.1716

1991;83:1716-1730Circulation.

http://circ.ahajournals.org/content/83/5/1716the World Wide Web at:

The online version of this article, along with updated information and services, is located on

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer information about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. FurtherEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculationpublished in Requests for permissions to reproduce figures, tables, or portions of articles originallyPermissions:


Dow

nloaded from

http://circ.ahajournals.org/content/83/5/1716

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


circus movementatrial flutter in canine sterile...

Documents