HRS-Plenary Lecture 2003: From Bedside to Bench: Entrainment & Other Stories (PDF)

Heart Rhythm (2004) 1, 94 –106 www.heartrhythmjournal.com HRS-PLENARY LECTURE 2003 From bedside to bench: Entrainment and other stories Albert L. Waldo Department of Medicine, Di...

Heart Rhythm (2004) 1, 94 –106 www.heartrhythmjournal.com HRS-PLENARY LECTURE 2003 From bedside to bench: Entrainment and other stories Albert L. Waldo Department of Medicine, Division of Cardiology, Case Western Reserve University/University Hospitals of Cleveland, Cleveland, Ohio. KEYWORDS Reentry; The concepts of transient entrainment of reentrant rhythms started with studies of overdrive pacing Entrainment; of atrial flutter (AFL) in patients in the immediate period after open heart surgery. Initial studies Tachycardias; demonstrated the need to achieve a critical pacing rate and a critical duration of pacing at the Rapid pacing critical pacing rate to interrupt AFL. Further pacing studies of AFL, ventricular tachycardia, atrioventricular (AV) reentrant tachycardia, AV nodal reentrant tachycardia, and atrial tachycardia refined the understanding of what occurs during overdrive pacing of reentrant tachycardias, and permitted a mechanistic understanding of transient entrainment as continuous resetting of a reentrant tachycardia to a pacing rate that is faster than the rate of the tachycardia, but which fails to interrupt it. The demonstration of transient entrainment of a tachycardia provides a reliable clinical tool to establish the presence of a reentrant rhythm. Moreover, the principles of entrain- ment have also been applied clinically to assist in effective application of antitachycardia pacing and catheter ablation techniques. © 2004 Heart Rhythm Society. All rights reserved. Introduction or both.2 None of these studies were systematic, and the results were quite variable, i.e., some reported some Entrainment begins with atrial flutter (AFL) and a bed- success, and some reported failure.2 In fact, it was ques- side story. When we started these studies, it wasn’t be- tioned whether one could reliably treat AFL with cardiac cause we knew there was something called entrainment. pacing at all. Also, the then-new concept of triggered So, it is important to provide some background. In the rhythms was evolving1 and played a role in the story. early 1970s, AFL was certainly a long- and well-recog- This was because for a very long time it was thought that nized rhythm, but its mechanism was not very well un- if one could initiate or terminate a tachyarrhythmia with derstood. There were two schools, one that thought it was cardiac pacing, it meant the tachyarrhythmia was due to due to a single focus firing rapidly, and one that thought reentry.1,3 However, when studies began to teach us it was due to a reentrant mechanism.1 And at that time, about triggered rhythms, it became apparent that they there appeared a series of articles in which investigators reported attempts to interrupt AFL using rapid atrial also could be initiated or terminated with cardiac pacing pacing techniques, introduction of premature atrial beats, techniques.1 Thus, we no longer could use pacing initi- ation or termination of a rhythm as the sine qua non for Supported in part by Grants RO1 HL38408 and RO1 HL074189 from identifying whether it was due to reentry. So, this was the the National Institutes of Health, National Heart, Lung, and Blood Institute, perspective when we confronted the management of AFL Bethesda, Maryland. in patients following open heart surgery. Address reprint requests and correspondence to: Albert L. Waldo, One day in the fall of 1972, as Dr. William MacLean, M.D., Division of Cardiology, MS LKS 5038, University Hospitals of Cleveland, 11100 Euclid Avenue, Cleveland, OH 44106. then a postdoctoral fellow in cardiology at The University E-mail address: [email protected]. of Alabama at Birmingham (UAB) Medical Center, and I 1547-5271/$ -see front matter © 2004 Heart Rhythm Society. All rights reserved. doi:10.1016/j.hrthm.2004.02.002 Waldo Entrainment 95 Pacing studies of AFL We soon found that to use overdrive pacing to interrupt AFL, pacing had to be performed at a critically rapid rate. If the overdrive pacing rate was not rapid enough, pacing only increased the AFL to the pacing rate, but with abrupt ter- mination of pacing or slowing of the pacing rate below the intrinsic AFL rate, the AFL was still present (Figure 1).2 When the atria were paced at a sufficiently rapid rate, the AFL was finally interrupted (Figure 2), but we didn’t un- derstand why. Moreover, not only was a critical pacing rate required, but a critical duration of pacing at the critical pacing rate was also required (Figure 2).2 The observations from Figure 2 were seminal, because with this case we recognized that when pacing from a site high in the right atrium, interruption of classical AFL was associated with Figure 1 ECG lead II recorded from a patient with AFL (atrial CL ⫽ 264 ms) (A) and at the end of 30 s of rapid atrial pacing from the appearance of a positive P wave in ECG leads II, III, and a high right atrial site at CLs of 254 ms (B), 242 ms (C), and 232 aVF. This, we recognized, was a marker that the AFL had ms (D). The atrial flutter was transiently entrained at each pacing been interrupted, because a positive atrial complex in those CL. See text for discussion. Time lines are at 1-s intervals. Mod- leads was what was expected when overdriving sinus ified from reference #2. rhythm from a high right atrial pacing site. But the expla- nation for the sudden change in morphology of the atrial were making rounds, Dr. T. Joseph Reeves, Professor and complexes in the ECG (Figure 2) was initially not apparent. Chairman of the Department of Medicine and a superb Moreover, this sudden change in atrial complex morphol- cardiologist, was about to perform DC cardioversion of ogy in the ECG was associated with a decrease in activation AFL in a post– open heart surgical patient. A pair of tem- time from the high right atrial pacing site to the posterior- porary atrial epicardial wire electrodes had been placed at inferior left atrial electrode recording site. The latter was the time of surgery, and Dr. Reeves, aware of recent reports documented by a dramatic shortening (mean 100 ms) of indicating that AFL could be interrupted by rapid atrial activation time from the high right atrial pacing site to the pacing, asked us to try to pace the patient’s atria to convert posterior-inferior left atrial recording site.2 But this, too, the AFL to sinus rhythm. We decided to give it a try. The was not yet understood except as a marker of successful reason we could try was because of the Medtronic SP AFL interruption. 1349A triple pulse generator. And this is also a nice story from a different era, but one central to the story of entrain- ment. The SP 1349A was a small, portable, easily hand- carried device that we were able to use at the bedside. This pacemaker, built by Medtronic with independent input from Dr. S. Serge Barold and me, permitted us to pace very rapidly at precise rates with stimuli of up to 50 mA.2 Without this device, it would have been far more difficult to do this work. Using overdrive pacing at rates chosen rather arbitrarily, but faster than the intrinsic rate of the AFL, we tried to terminate it and restore sinus rhythm but couldn’t. Because of the large volume of patients undergoing open heart sur- gery at UAB and because of the significant incidence of postoperative AFL, we subsequently had many opportuni- ties to try again. After rereading the several articles that reported attempts to interrupt AFL with pacing, we realized Figure 2 (A): ECG lead II recorded from the same patient as that no systematic studies had been performed, that some Figure 1 during high right atrial pacing from the same site at a CL of 224 ms. With the 7th beat in this tracing, and after 22 s of atrial reported modest success, and that some reported abject pacing at a constant rate, the atrial complexes suddenly became failure. So, we started to do the atrial pacing systematically positive. (B): ECG lead II recorded from the same patient. With using the SP 1349A pacemaker and the temporary epicar- abrupt termination of pacing, sinus rhythm occurs. (C): Continued dial electrodes placed in the high right atrium near the sinus sinus rhythm. See text for discussion. Asterisk ⫽ same beat. node at the time of the surgery. Modified from reference #2. 96 Heart Rhythm, Vol 1, No 1, May 2004 bedside with the SP1349A temporary pacemaker.6 And what we finally figured out during transient entrainment and interruption of reentrant VT turned out to be true for every single ordered reentrant tachyarrhythmia (AFL, VT, atrio- ventricular reentry tachycardia [AVRT], AV nodal reentry tachycardia, and atrial tachycardia) we have studied.1 More- over, the phenomena observed during transient entrainment and interruption of tachycardias identified the mechanism as reentrant.1 The breakthrough case was one of postoperative VT with 1:1 retrograde atrial activation. That gave us the idea that Figure 3 ECG lead II recorded from a patient with VT at a rate of 150 bpm. Top trace: termination of ventricular pacing (dot) at a perhaps we could pace the atria and interrupt the VT, espe- rate of 160 bpm, with prompt return of VT; middle trace: termi- cially because this VT was relatively slow (141 bpm) and nation of pacing (dot) at a rate of 175 bpm, with prompt return of well tolerated. Figure 4 shows the first of a series of dia- VT; bottom trace: termination of pacing at a rate of 190 bpm, with grams using a figure-of-eight model of reentry with a central successful interruption of the VT. See text for discussion. Modified isthmus to represent the reentrant VT circuit. At the time, from reference #4. we thought that the isthmus represented an area of slow conduction. But as we now know from the work of Wit et al,1 the area of slow conduction in this model is probably We performed a series of atrial overdrive pacing studies where block is thought to occur on either side of the isth- of AFL in patients and reported it.2 In that paper, we coined mus. For purposes of historical accuracy, we shall again use the term “transient entrainment of a tachycardia” and de- the figure-of-eight reentry circuit to illustrate what we fined it as an increase in the rate of a tachycardia to a faster thought was occurring during entrainment and interruption pacing rate, with resumption of the tachycardia either upon of the VT. abrupt cessation of pacing or slowing of the pacing rate below the intrinsic rate of the tachycardia. This was a In Figure 4, the top left panel illustrates the VT wave descriptive definition because we did not yet understand fronts traveling around the reentrant circuit. With the onset entrainment fully. of atrial pacing (Figure 4, top middle panel) at 150 bpm, a rate faster than the VT, the ventricular wave front from the pacing impulse is able to enter the excitable gap of the reentrant circuit, where it travels both antidromically and Studies of ventricular tachycardia (VT) orthodromically. Antidromically, it collides with the ortho- We then started to examine overdrive pacing of other tachy- dromic wave front of the previous beat, creating a fusion arrhythmias. The stories continue with VT. We now dem- beat in the ECG. Orthodromically, it enters the reentrant onstrated that when pacing the ventricles at a rate faster than circuit early, so that it captures the reentrant circuit and the tachycardia rate, a critical pacing rate was required to resets it to the pacing rate. Because this orthodromic wave interrupt the tachycardia.4 What was new was the clear front isn’t blocked, it travels around the reentrant circuit recognition of progressive fusion of the QRS complexes in until it collides with the antidromic wave front of the next the ECG during pacing at two or more constant rates faster paced beat (Figure 4, top right panel). Thus, during pacing, than the rate of the tachycardia, in which each rate failed to there is constant fusion of the QRS complex morphology in interrupt the tachycardia (Figure 3). We put these data the ECG, as the ventricles are activated simultaneously by together and submitted a paper describing transient entrain- wave fronts from two different paced beats, and there is ment of VT but were turned down consecutively by three continuous resetting of the tachycardia to the pacing rate by peer-reviewed journals. On appeal to the third journal, it the orthodromic wave from each pacing impulse. This, in was finally published4 but not without an editorial entitled fact, is the mechanistic definition of transient entrainment: “Entrainment of ventricular tachycardia: all aboard or the Transient entrainment of a tachycardia occurs during pacing end of the line.” The editorial was basically supportive but at a rate faster than the tachycardia rate when each ortho- quite tentative.5 dromic wave front from the pacing impulse resets the tachy- cardia to the pacing rate, and each antidromic wave front from the pacing impulse either collides with the ortho- Understanding entrainment–The breakthrough dromic wave front of the previous beat or is otherwise case and development of the first three blocked. entrainment criteria The actual ECG and ventricular electrogram (VEG) re- cordings (atrial pacing at 150 bpm, with 1:1 capture of the The critical insight into entrainment that led to its under- ventricles) are shown in the bottom panel of Figure 4. Note standing finally resulted from analysis of a single case of that during pacing, the QRS complexes have changed in entrainment and interruption of VT, again performed at the morphology and are narrower (120 ms) compared to those Waldo Entrainment 97 Figure 4 Top panel: Diagrammatic representation of transient entrainment, in this case, of VT during atrial pacing. In this and subsequent diagrams, arrows indicate direction of activation, box represents an isthmus in the reentrant circuit, serpentine line indicates slow conduction [sic], dashed line indicates the excitable gap in the reentry circuit, dot represents a right ventricular electrogram (VEG) recording site, and large arrow (middle and right panels) indicates wave front from the pacing impulse entering the excitable gap of the VT reentry circuit, where it is conducted orthodromically (ortho) and antidromically (anti). Top left panel: Diagram of the reentry circuit (figure-of-eight model) during VT at a rate of 141 bpm. x ⫽ orthodromic wave fronts of the reentrant rhythm. Top middle panel: Introduction of the first pacing impulse (x ⫹ 1) during atrial pacing at a rate of 150 bpm during the VT. The antidromic wave fronts (x ⫹ 1) collide with the orthodromic wave fronts of the previous spontaneous beat (x), resulting in a fusion ventricular beat, which, in effect, interrupts the VT. However, the orthodromic wave front from the pacing impulse continues the tachycardia, resetting it to the pacing rate. Top right panel: Introduction of the second pacing impulse (x ⫹ 2) during atrial pacing at 150 bpm. The antidromic wave fronts (x ⫹ 2) collide with the orthodromic wave fronts (x ⫹ 1) of the previous paced beat, again causing ventricular fusion, which, again, in effect, interrupts the VT. But again, the orthodromic wave front from the pacing impulse (x ⫹ 2) continues the VT, resetting it to the pacing rate. During the period of pacing, the VEG site is always activated by an orthodromic wave front. Bottom left panel: ECG leads I and V1 recorded simultaneously with a bipolar atrial electrogram (AEG) and a unipolar ventricular electrogram (VEG) during VT at a rate of 141 bpm (425 ms CL). In this and Figure 5, the circled number indicates the duration (ms) of each QRS complex. Bottom right panel: ECG leads I and V1, recorded simultaneously with the atrial pacing (A Pace) stimulus (S) artifact (stim) and the unipolar ventricular electrogram (VEG) during atrial pacing at 150 bpom (400 ms CL). Time lines are at 1-second intervals. See text for discussion. Modified after reference #6. during the spontaneous VT, but the VEG morphology was different beats). The orthodromic wave front of the last the same during pacing as during the spontaneous VT, paced beat again resets the tachycardia to the pacing rate, indicating that activation at this recording site was the same but this time, this orthodromic wave front, which travels in both instances. around the reentrant circuit at the pacing cycle length, is The top panel of Figure 5 diagrammatically illustrates unopposed because there is no subsequent pacing impulse. what happened when we stopped pacing. The top left panel Therefore, this last captured beat is entrained, but there is no shows the last paced beat, and, as expected, its antidromic fusion in the ECG. The bottom panel of Figure 5 shows the wave front collides with the orthodromic wave front of the actual ECG and EG recordings. The last pacing impulse previous beat, creating a fusion beat in the ECG (the ven- captured the ventricles at the 400 ms pacing cycle length tricles are activated simultaneously by wave fronts from two (CL), but while the last captured ventricular beat was en- 98 Heart Rhythm, Vol 1, No 1, May 2004 Figure 5 Diagrammatic representation and tracings of the termination of atrial pacing at 150 bpm, illustrating the first entrainment criterion. Top left panel: The wave front from the last pacing impulse (rate 150 bpm) enters the excitable gap of the VT reentry circuit, where it is conducted orthodromically and antidromically. The antidromic wave fronts, xn(a), collide with the orthodromic wave fronts xn-1, causing a fusion beat. The orthodromic wave front xn(o) from the last pacing impulse continues the tachycardia, resetting it to the pacing rate. Top right panel: The orthodromic wave fronts from the last pacing impulse are now unopposed by antidromic wave fronts because there is no subsequent pacing impulse, so that no fusion of ventricular activation occurs despite the presence of transient entrainment. This last entrained beat continues the tachycardia (dashed lines), which then resumes at its previous spontaneous rate. Bottom panel: ECG leads I and V1 recorded simultaneously with the atrial pacing (A Pace) stimulus (S) artifact (stim) or atrial electrogram (AEG), and the unipolar ventricular electrogram (VEG) at the termination of atrial pacing at a rate of 150 bpm (400 ms CL). In this and subsequent figures, the open circle denotes the last stimulus, the asterisk denotes the last entrained beat, the arrow from the stimulus in the VEG points to the resulting VEG (with the stimulus-to-VEG interval being indicated in ms); the dashed arrow in lead I represents the last antidromic wave fronts, xn(a); and the solid arrow represents the last orthodromic wave fronts, xn(o) from the last pacing impulse. See text for discussion. Modified after reference #6. trained (asterisk), it was not fused. In fact, because the stant rate that is faster than the rate of tachycardia and which orthodromic wave front from the last pacing impulse was fails to interrupt it, there is the demonstration of constant unopposed by a subsequent antidromic pacing wave front, fusion beats in the ECG except for the last captured beat, this beat had the QRS complex morphology of the sponta- which is not fused.1 neous VT. From these observations came the first entrain- Then with atrial pacing at 155 bpm, the same thing ment criterion: during a tachycardia, when pacing at a con- happened. The VT circuit was captured, the morphology of Waldo Entrainment 99 Figure 6 Illustration of the second entrainment criterion. Top panel: Diagrammatic representation of transient entrainment of the VT illustrated in Figure 4 during pacing at 150 bpm, 155 bpm, and 160 bpm. In each example, the antidromic wave fronts from each pacing impulse, xn(a), collide with the orthodromic wave front of the previous beat, xn-1. However, in each case, the orthodromic wave front from each pacing impulse, xn(o), transiently entrains the VT by resetting it to the pacing rate. Because the pacing CL progressively shortens as the pacing rate increases, the antidromic wave fronts from each pacing impulse at each faster rate penetrate the reentry circuit to a greater degree. Therefore, although ventricular fusion occurs in each instance, the degree of fusion is different. This results in progressive fusion in the ECGs (see bottom panel). Bottom panel: ECG leads I and V1 recorded simultaneously during VT at a rate of 141 bpm (A), and during atrial pacing which transiently entrains the VT at 150 bpm (B), 155 bpm (C), and 160 bpm (D). S ⫽ stimulus artifact. See text for discussion. Modified after reference #6. the QRS complexes changed even more, and narrowed to 91 reentrant circuit (Figure 6), explaining the yet again differ- ms in duration (Figure 6). Also, as before, the VEG mor- ent QRS complex morphology of the fusion beats in the phology was unchanged from that during spontaneous VT. ECG leads. The different degree of fusion in the ECG leads And as before, the last pacing impulse captured the ventri- at each pacing rate, which entrains but fails to interrupt the cles at the pacing CL, now 387 ms, but the morphology of tachycardia, is called progressive fusion (Figure 6). From the last captured beat was not fused. Because the pacing CL these observations came the second entrainment criterion: was shorter, the antidromic wave front from each pacing during a tachycardia, when pacing at two or more constant impulse entered the excitable gap earlier, resulting in colli- rates that are faster than the rate of the tachycardia but sion with the orthodromic wave front of the previous beat at which fail to interrupt it, there is the demonstration of a different place (Figure 6). This created constant fusion constant fusion beats in the ECG at each rate, but different QRS complexes in the ECG which were different than seen degrees of constant fusion at each rate.1 when pacing at 150 bpm (Figures 4 and 5). And then when Then with atrial pacing at a rate of 165 bpm (CL of 364 pacing the atria at 160 bpm, there was a still different QRS ms), still with 1:1 conduction to the ventricles (Figure 7, top complex morphology, and the QRS complex was now only panel), initially there is further progressive fusion of the 80 ms in duration (Figure 6), but the ventricular electrode QRS complexes in the ECGs. Then, denoted by the aster- recording site was still activated from the same direction isks, there was an abrupt change in morphology of both the during pacing as during the spontaneous tachycardia, and recorded QRS complexes in the ECG leads and the VEGs. the last captured beat was not fused. Because this pacing CL Immediately before the change in morphology of the QRS was shorter (375 ms), the antidromic wave front of each complexes and of the VEG, the stimulus-to-VEG interval pacing impulse was able to penetrate still further into the was 640 ms. Then, there was localized conduction block to 100 Heart Rhythm, Vol 1, No 1, May 2004 Figure 7 Illustration of the third entrainment criterion. Top panel: ECG leads I and V1 recorded simultaneously with the atrial pacing stimulus (s) artifact (stim), and the unipolar ventricular electrogram (VEG) during atrial pacing at 165 bpm (364 ms CL) which interrupts the VT. The asterisks denote an abrupt change in configuration of the recorded QRS complexes in both ECG leads and the VEG. Before the change in configuration of the QRS complexes and VEG, the stimulus-to-VEG interval is 640 ms. Then, after the localized block to the VEG recording site, this interval becomes 305 ms. Note also that this localized conduction block is associated with a one-cycle increase in the beat-to-beat CL localized to the VEG recording site (from 364 to 425 ms, and then back again to 364 ms). S ⫽ stimulus artifact. The bottom panel is a diagrammatic representation of events recorded in top panel during the interruption of the VT with pacing at 165 bpm. Bottom left panel: The wave front from the pacing impulse enters the excitable gap of the VT reentry circuit, where it is conducted antidromically, x ⫹ 1(a), and orthodromically, x ⫹ 1(o). The antidromic wave fronts collide with the orthodromic wave fronts of the previous beat (x), causing a fusion beat. Note that this collision is at a site still different than that illustrated in Figure 6. However, this time, the orthodromic wave front is also blocked during the same cycle. Note that sites immediately orthodromically distal to the site of block of x ⫹ 1(o), although activated by the orthodromic wave front (x) of the previous beat, are not activated by x ⫹ 1. Bottom right panel: The next pacing impulse at the same 165 bpm rate (x ⫹ 2) now activates the ventricles. But because the VT was interrupted by the previous (x ⫹ 1) paced beat, the sequence of ventricular activation during x ⫹ 2 is as expected during atrial overdrive pacing of sinus rhythm. Thus the sites which had just demonstrated localized conduction block for one beat, including the VEG site, are now activated from a different direction and with a shorter conduction time. This is because there no longer is a VT reentry circuit direct activation to these latter sites. See text for discussion. Modified from reference #6. the VEG recording site for one beat (i.e., the impulse from This localized conduction block was associated with an the next stimulus never activated this recording site), after increase in the beat-to-beat CL at the VEG recording site which, the impulse from the next stimulus was conducted to from 364 ms to 425 ms and then back to 364 ms. There was that site with a stimulus-to-VEG interval of only 305 ms. no such increase in the QRS complex CLs. In the ECG lead Waldo Entrainment 101 demonstrated localized conduction block gets activated from a different direction and with a shorter conduction time. From this came the third entrainment criterion: during a tachycardia, when pacing at a constant rate that is faster than the rate of tachycardia and which interrupts it, there is the demonstration of localized conduction block to a site or sites for one beat followed by activation of that site or those sites by the next paced beat from a different direction and with a shorter conduction time.1 Now we can understand what happened in Figure 2 in that pivotal case of AFL. Recall that after the 22 s of high right atrial pacing at a constant rate of 373 bpm, suddenly there was Figure 8 ECG lead II also shown in Figure 2, but now shown the appearance of positive P waves in ECG lead II. Now we recorded simultaneously with the stimulus artifact (stim) from know that during entrainment, the antidromic wave front of the high right atrial pacing site. Arrows drawn from each each atrial pacing impulse collided with the orthodromic wave stimulus artifact represent the antidromic (a) and orthodromic front of the previous beat, and each orthodromic wave front (o) wave fronts from each pacing impulse in association with reset the tachycardia to the pacing rate. But when the anti- each atrial complex in the ECG. Note that the appearance of the positive atrial complex in the ECG is associated with block of dromic and orthodromic wave fronts of the same pacing im- both the antidromic and orthodromic wave fronts of the same pulse blocked during the same beat, the tachycardia was inter- pacing impulse during the same cycle. The antidromic wave rupted. Thus, the next pacing impulse captured the atria as if front blocks because it collides with the orthodromic wave front overdriving sinus rhythm. Since pacing was being performed of the previous wave front, as usual. When the orthodromic from the high right atrium, a positive P wave appeared in ECG wave front of the same beat blocks during the same cycle, the lead II (Figure 8). So, we finally understood what happened AFL is interrupted, and continued atrial pacing activates the during entrainment and interruption of classical AFL when atria from the high right atrial pacing site as expected during pacing from the high right atrium. sinus rhythm. Modified after reference #2. I tracing, the dashed arrows represent the antidromic wave fronts, and the associated solid arrows represent the ortho- The fourth entrainment criterion dromic wave fronts from the fifth and sixth pacing impulses shown in the figure. After the block of both the antidromic With these studies, we recognized that it is sometimes hard and orthodromic wave fronts of the sixth pacing impulse in to see or demonstrate some of these entrainment criteria, particularly fusion in the ECG, when pacing the atria during the reentrant circuit of the VT, the ventricles were activated AFL or a reentrant atrial tachycardia. From that evolved the by the seventh pacing impulse as expected during overdrive fourth entrainment criterion: During a tachycardia, when atrial pacing of a sinus rhythm because the VT has been pacing at two constant rates that are faster than the rate of interrupted. This, of course, explains the change in mor- tachycardia but which fail to interrupt it, there is the dem- phology of the QRS complexes and the VEGs. Finally, had onstration of a change in conduction time to and EG mor- the pacing been terminated prior to achieving block of both phology at an electrode recording site.1,7 This is the EG the antidromic and orthodromic wave fronts of the same equivalent of the second criterion. pacing impulse during the same beat, the tachycardia would Figure 9 well illustrates the fourth entrainment criterion not have been interrupted, emphasizing the critical duration in a patient during atrial pacing from the high right atrium of pacing at the critical pacing rate required for interruption (HRA) during an AVRT (CL 339 ms). Shown in both panels of the tachycardia. is ECG lead II recorded simultaneously with EGs from the Termination of the tachycardia is shown diagrammati- proximal pair of electrodes of a catheter placed in the HRA cally (Figure 7, bottom panel). When pacing the atria at 165 and the coronary sinus (CS) during termination of pacing at bpm, the antidromic wave front of the pacing impulse col- CLs of 308 ms (Figure 9, panel A) and 292 ms (Figure 9, lided, as usual, with the orthodromic wave front of the panel B). In the recordings from the CS, the arrows from previous beat. But now the orthodromic wave front of that each stimulus artifact (S) indicate the AEGs (a) that resulted same pacing impulse also blocked, so that an area of the from that stimulus. Note that the AEGs at the CS site in ventricles was never activated by this paced beat. This is panel A were identical to that during the spontaneous localized conduction block. With block of both the anti- rhythm, and that conduction time during pacing was long dromic and orthodromic wave fronts of the atrial pacing (390 ms). However, with pacing at a shorter CL (292 ms) in impulse during the same beat, the tachycardia was inter- panel B, the AEG morphology changed at the CS site during rupted, and the functionally determined components of the entrainment, indicating that this site was now activated from reentrant circuit disappeared. Thus, with the next atrial a different direction and a shorter conduction time (90 ms). pacing impulse at the same 165-bpm rate, the site that just Note also that the tachycardia was not interrupted by pacing 102 Heart Rhythm, Vol 1, No 1, May 2004 Figure 9 Illustration of the fourth criterion of transient entrainment during atrial pacing from the high right atrium (HRA) during an atrioventricular reentrant tachycardia (AVRT), CL 339 ms. In the upper part of both panels A and B, ECG lead II is recorded simultaneously with bipolar EGs from the proximal pair (p) of catheter electrodes in the HRA and the coronary sinus (CS) during termination of pacing at a CL of 308 ms (panel A) and 292 ms (panel B). In the CS recordings, the arrows from each stimulus (s) indicate the atrial complexes (a) that result from that stimulus. V ⫽ ventricular complex. The last arrow in each panel goes to the last transiently entrained beat. S ⫽ stimulus artifact. The bottom diagrams in each panel represent the recorded events during pacing at each CL. The left panel of each diagram indicates the pacing impulse (x ⫹ 1) from the HRA entering into the reentry circuit, whereupon it is conducted antidromically and orthodromically. The antidromic wave front collides in the atria with the orthodromic wave front (x) of the previous beat in panel A but in the AV bypass pathway in panel B. However, in each case, the orthodromic wave front, which is not blocked during that beat, resets the tachycardia to the pacing rate, continuing the tachycardia. Thus, with the last paced beat, the unopposed orthodromic wave front from the last paced beat continues the tachycardia. In panel B, note that with resumption of the AVRT, the CS will again be activated orthodromically (it had been activated antidromically during pacing at 292 ms CL). Serpentine line represents slow conduction in the AV node. See text for discussion. Modified after reference #7. Figure 10 Diagrammatic representation of one form of concealed entrainment of VT by pacing from an isthmus in a reentry circuit. Left panel shows the last paced beat (xn) during concealed entrainment. The square wave indicates the pacing site. xn-1 ⫽ orthodromic wave front of the next-to-last paced beat. The antidromic wave front from the last pacing impulse blocks (unidirectional block) within or very close to the isthmus. The orthodromic wave front from the last pacing impulse activates the ventricles virtually identically as during the spontaneous VT, so that there is no fusion QRS complex morphology in the ECG. Middle and right panels: Following the last paced beat (x n), the VT resumes at its previous rate. See text for discussion. Waldo Entrainment 103 Figure 11 Diagrammatic representation of another form of concealed entrainment during ventricular pacing at rates of 150, 155, 160, and 165 bpm from a site immediately distal to an area of slow conduction in the reentry circuit. Top left panel: Diagrammatic representation of the same VT seen in Figures 4 –7. Dashed lines represent the excitable gap. Xn-1 ⫽ the orthodromic wave fronts of the next-to-last pacing impulse. X ⫽ circulating reentrant wave fronts of the VT. Top middle panel: The large arrow represents the wave front from the last pacing impulse entering the reentrant circuit, where it is conducted orthodromically, Xn(o), and antidromically, Xn(a). The antidromic wave front blocks in an area of slow conduction (unidirectional block). The orthodromic wave fronts reset the tachycardia to the pacing rate. Top right panel: The orthodromic wave fronts from the last pacing impulse continue around the reentrant circuit, so that the VT resumes at its previous rate. Bottom left 2 panels: Diagrammatic representations of concealed entrainment during ventricular pacing at rates of 155 bpm and 160 bpm from the same site shown in the top panel. Note the absence of progressive fusion. X ⫹ 1(o) and X ⫹ 1(a) indicate the orthodromic and antidromic wave fronts, respectively, from the pacing impulse. Bottom right 2 panels: Diagrammatic representation of the interruption of VT pacing at 165 bpm from the same pacing site as shown in the upper and 2 left panels. The wave front from the last pacing impulse at a rate of 165 bpm enters into the VT reentrant circuit. The antidromic wave front of the pacing impulse is again blocked (unidirectionally) in the same place as during pacing at 150, 155, and 160 bpm. The orthodromic wave front from the pacing impulse travels around the reentrant circuit as before, but now it is also blocked, presumably in the area of slow conduction. However, this cannot be appreciated either from the ECG or from standard electrode catheter recording sites. With the termination of pacing, the orthodromic wave front from the last pacing impulse blocks as just described. Only then, with no further pacing, can it be appreciated that the VT has been interrupted. See text for discussion. Modified after reference #10. at either rate. This fulfilled the fourth criterion for demon- Concealed entrainment stration of transient entrainment. As diagrammatically illus- trated in the bottom half of Figure 9, panels A and B, the CS A tachycardia may be transiently entrained and even inter- site was activated orthodromically during pacing at a CL of rupted without being able to demonstrate any of the entrain- 308 ms but antidromically during pacing at a CL of 292 ms. ment criteria, i.e., there may be concealed entrainment.1,8 –11 Actually, in Figure 9, panel A, the disappearance of a When this occurs, to prove the presence of entrainment, positive P wave in the ECG with the last captured beat that pacing must be performed from a site demonstrating man- was entrained but not fused, is apparent, also fulfilling the ifest entrainment. There are two types of concealed entrain- first entrainment criterion. Finally, note also in Figure 9, ment described. panel B that with the last paced beat, the CS site was first This next diagram (Figure 10) illustrates and explains activated antidromically and then orthodromically by the one form of concealed entrainment in a figure-of-eight re- same beat because the orthodromic wave front from the last entrant VT circuit. During pacing from an isthmus in the paced beat is not blocked. circuit, with pacing at a rate a little faster than the sponta- 104 Heart Rhythm, Vol 1, No 1, May 2004 Figure 12 Demonstration of concealed entrainment of AFL. ECG lead III is recorded simultaneously with bipolar AEGs from the sulcus terminalis (ST) and Bachmann’s bundle (BB) at the termination of 30 s of rapid atrial pacing (352 bpm–left panel; 365 bpm–right panel) from the postero-inferior left atrium in a patient with AFL (rate of 320 bpm). Comparison of events during rapid atrial pacing and its termination at each rate fails to show any evident transient entrainment criteria, even despite interruption of the AFL by pacing at 365 bpm. In fact, only when pacing was stopped was it realized that AFL had been interrupted. S ⫽ stimulus artifact. Asterisk ⫽ last captured beat during atrial pacing. Time lines are at 1-s intervals. See text for discussion. From: Waldo AL, Carlson MD, Hethorn RW. Atrial flutter: Transient entrainment and related phenomena. In: Zipes DP, Jalife J, eds. Cardiac Electrophysiology: From Cell to Bedside. Philadelphia: WB Saunders Co; 1990. p. 530 –537. neous rate, the orthodromic wave front of the pacing im- There is another form of concealed entrainment that we pulse captures the circuit and resets it, but the antidromic described,8,10 but which is not widely appreciated. It occurs wave front blocks it. Some suggest the latter blocks because when pacing from a site orthodromically distal to an area of it collides with the previous orthodromic wave front coming slow conduction. Using as illustration the same VT shown around the reentrant circuit. We suggest another possibility, in Figures 4 through 7, if the ventricles are paced at rates of that the antidromic wave front blocks in the isthmus, per- 150, 155, and 160 bpm from a site immediately distal to an haps in an area of slow conduction, so that the antidromic area of slow conduction in the reentry circuit (Figure 11), wave front never really exits. In any event, at best, there is the tachycardia will be entrained because the pacing im- minimal antidromic penetration. Therefore, a fusion mor- pulse will enter the excitable gap and reset the tachycardia phology is not appreciated in the ECG because virtually all to the pacing rate. However, despite the fact that the tachy- sites are activated the same way during pacing as during the cardia is entrained and subsequently interrupted when pac- tachycardia except for a very small area activated anti- ing at 165 bpm, none of the criteria of transient entrainment dromically. Because the tachycardia is entrained but there is will be demonstrable. When pacing orthodromically distal no evidence of that in ECG, the entrainment is concealed. to an area of slow conduction at 150 bpm (Figure 11, top Even with interruption of the tachycardia, there is no pos- panel), each pacing impulse enters the excitable gap of the sible demonstration of any of the entrainment criteria. What reentry circuit. The antidromic wave front encounters block happens with the last paced beat is that the orthodromic (either unidirectional or, less likely we suggest, due to wave front from the pacing impulse goes around the reen- collision with the orthodromic wave front of the previous trant circuit, and the tachycardia resumes (Figure 10). Un- beat) in the area of slow conduction. The orthodromic wave derstanding this form of concealed entrainment has become front of each pacing impulse resets the tachycardia to the very useful, particularly during ablation of the so-called pacing rate. When pacing is terminated, the orthodromic AFL isthmus and during ablation of VTs with an isthmus, in wave front of the last paced beat continues the tachycardia. concert with criteria of Stevenson et al,12 which use the Since the last captured beat has the same morphology as that postpacing interval to demonstrate the participation of sites throughout the period of pacing,13 the first criterion of in the reentrant circuit. transient entrainment cannot be met. Waldo Entrainment 105 With pacing at 155 or 160 bpm (Figure 11, bottom remains the way we can clinically identify reentrant panel), because the pacing CL is shorter than pacing at 150 rhythms. And it is now used regularly as a clinical tool, not bpm, the wave front from each pacing impulse enters the only to identify the mechanism, but also as an essential aid reentry circuit earlier, but block of the antidromic wave in antitachycardia pacing and catheter ablation procedures. front still occurs at the same place. The orthodromic wave front continues to activate the ventricles as during pacing at 150 bpm. Thus, the QRS complex during pacing at these faster rates does not change, and progressive fusion, the Acknowledgments second entrainment criterion, will not be present. Therefore, the fourth criterion also will not be present. Finally, when My sincere thanks to the NASPE Annual Scientific Ses- pacing at a sufficiently rapid rate (165 bpm), interruption of sion Program Committee for inviting me to give this dis- the VT occurs because now the orthodromic wave front tinguished lecture. It is an honor and a special privilege for blocks independently, presumably in the area of slow con- me for several reasons, not only because I have NASPE in duction (Figure 11, bottom right panel), along with block of my blood, but also because the theme of this 24th Annual the antidromic wave front of the same paced beat. However, Scientific Session, “From bench to bedside and back,” is the only way to know that interruption of the tachycardia something I’ve tried to do throughout my career. has occurred is to stop pacing because the third criterion of We wish to acknowledge so many colleagues who have transient entrainment is not demonstrable. In sum, pacing helped us develop the concepts and data on entrainment. Men- from a site orthodromically distal to an area of slow con- tioned here are those with whom we have published studies: duction may not permit the demonstration of manifest en- William A.H. MacLean, Ken Okumura, Richard W. Henthorn, trainment. Brian Olshansky, Terry B. Cooper, Vance J. Plumb, Joaquin Figure 12 is a representative example of this type of G. Arciniegas, Thomas N. James, Nicholas T. Kouchoukos, concealed entrainment in a case of AFL during pacing from Robert B. Karp, Paul G. Hess, Marshall F. Priest, Andrew E. the CS. The left panel demonstrates the termination of Epstein, Pedro Brugada, Hein J.J. Wellens, Akihiko Shimizu, pacing at a rate (352 bpm) faster than the rate of the AFL. Akira Nozaki, Yoram Rudy, Dalmo Moreira. Although the atria are clearly captured at the pacing rate, note the absence of anything to distinguish the last captured atrial beat from the other captured atrial beats or even from References the atrial complexes seen during AFL. With cessation of pacing, the AFL resumes at its previous rate. The right panel 1. Waldo AL, Wit AL. Mechanism of cardiac arrhythmias and conduc- demonstrates pacing at 365 bpm. Despite pacing at a faster tion disturbances. In: Fuster V, Alexander RW, O’Rourke RA, Roberts rate, there is again no change in the morphology of the atrial R, King SB, Wellens HJJ, editors. Hurst’s The Heart. 10th ed. New complexes compared with that during pacing at the slower York: McGraw-Hill Companies, 2001:751–796. 2. Waldo AL, MacLean WAH, Karp RB, Kouchoukas NT, James TN. rate. However, this time, with cessation of pacing, the AFL Entrainment and interruption of atrial flutter with atrial pacing: Studies has been interrupted. There is no evidence during pacing in man following open heart surgery. Circulation 1977;56:737–745. that the AFL had been interrupted. In fact, the only way to 3. Wellens HJJ. Value and limitations of programmed electrical stimu- know that the AFL has been interrupted is to stop pacing. lation of the heart in the study and treatment of tachycardias. Circu- lation 1978;57:845– 853. 4. MacLean WAH, Plumb VJ, Waldo AL. Transient entrainment and interruption of ventricular tachycardia. Pace Card Electrophysiol 1981; Conclusions 4:358 –366. 5. Fisher JD. Entrainment of ventricular tachycardia: All aboard or the Clearly, we have learned quite a lot about entrainment over end of the line? Pace Card Electrophysiol 1981;4:467– 468. the years. We have learned that transient entrainment of a 6. Waldo AL, Henthorn RW, Plumb VJ, MacLean WAH. Demonstration of the mechanism of transient entrainment and interruption of ventric- reentrant tachycardia occurs during pacing at a rate faster ular tachycardia with rapid atrial pacing. J Am Coll Cardiol 1984;3: than the tachycardia rate when each orthodromic wave front 422– 430. from the pacing impulse resets the tachycardia to the pacing 7. Henthorn RW, Okumura K, Olshansky B, Waldo AL. A fourth crite- rate and each antidromic wave front from the pacing im- rion for transient entrainment: The electrogram equivalent of progres- pulse either collides with the orthodromic wave front of the sive fusion. Circulation 1988;77:1003–1012. 8. Okumura K, Henthorn RW, Epstein AE, Plumb VJ, Waldo AL. Fur- previous beat or is otherwise blocked. That holds for every ther observations on transient entrainment: Importance of pacing site reentrant rhythm studied, including atrial tachycardia, AFL, and properties of the components of the reentry circuit. Circulation AV nodal reentrant tachycardia, AVRT, and VT.1 Sinus 1985;72:1293–1307. node reentry has not been studied. Unfortunately, one can- 9. Frank R, Tonet JL, Kounde S, Farenq G, Fontaine G. Localization of the not always demonstrate the entrainment criteria despite area of slow conduction during ventricular tachycardia. In: Brugada P, Wellens HJJ, editors. Cardiac Arrhythmias: Where To Go From Here? overdrive pacing of a reentrant rhythm. When that happens, Mt. Kisco, NY: Futura Publishing Co., Inc., 1987:191–208. concealed entrainment occurs. 10. Waldo AL, Henthorn RW. Use of transient entrainment during ven- The ideas and concepts of entrainment, I’m glad to say, tricular tachycardia to localize a critical area in the reentrant circuit for have been carried further by many others. Entrainment still ablation. Pace Card Electrophysiol 1989;12:231–244. 106 Heart Rhythm, Vol 1, No 1, May 2004 11. Feld GK, Fleck P, Chen P-S, et al. Radiofrequency catheter ablation ular tachycardia late after myocardial infarction. Circulation 1993;88: for the treatment of human Type I atrial flutter: Identification of a 1647–1670. critical zone in the reentrant circuit by endocardial mapping tech- 13. Okumura K, Olshansky B, Henthorn RW, Epstein AE, Plumb VJ, niques. Circulation 1992;86:1233–1240. Waldo AL. Demonstration of the presence of slow conduction during 12. Stevenson WG, Khan H, Sager P, et al. Identification of reentry circuit sustained ventricular tachycardia in man. Circulation 1987;75:369 – sites during catheter mapping and radiofrequency ablation of ventric- 378.