Analysis of the Degree of QRS Fusion Necessary for Its Visual Detection
Importance for the Recognition of Transient Entrainment
Background Fixed fusion is the hallmark for the demonstration of transient entrainment. However, the degree of accuracy of its recognition on the surface ECG is unknown. The purpose of the present study was to evaluate the ability to detect fusion in the QRS complex.
Methods and Results While pacing the ventricles at a fixed rate, a model of ventricular fusion was created by introducing late extrastimuli at a second site. In this model, the presence and degree of fusion are known. Pacing sites were the RV apex, outflow tract, and left ventricle in various configurations. We analyzed 433 QRS complexes with different degrees of fusion (or no fusion) in 21 patients. Each QRS was “read” by three investigators blinded to intracardiac recordings but having a reference QRS with no fusion. There was a statistically significant correlation between the degree of fusion and its recognition. Fusion was detected with a sensitivity of 75% and a specificity of 87%. Fusion was accurately detected in all configurations only when >22% of the QRS was fused. In patients with organic left ventricular disease, fusion was better recognized when the driving pacing site was the left ventricle than when it was a right ventricular site. The interobserver agreement was moderate between two pairs of observers and only fair between the remaining pair.
Conclusions Our results suggest that an accurate detection of ventricular fusion can only be accomplished when fusion occurs during a significant proportion of the QRS duration. The potential lack of recognition of minor degrees of fusion may produce underdetection of transient entrainment.
Transient entrainment during pacing has been described in a variety of ventricular and supraventricular tachycardias, and its demonstration supports reentry as the mechanism of the tachycardia.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 The presence of fixed fusion is the hallmark for the demonstration of transient entrainment because it establishes the presence of two simultaneous wave fronts in a cardiac chamber, one generated by the tachycardia circuit and the other by the stimulated impulse. The latter, however, influences the tachycardia circuit because it advances the subsequent tachycardia circuit wave front (therefore, it is accelerated to the pacing rate). This ability of a paced wave front to influence (accelerate) the tachycardia circuit, even if it is generated at a time when a wave front is exiting from the circuit, characterizes the phenomenon of entrainment with fusion and suggests that the tachycardia mechanism is reentrant (rather than focal), with separate entrance and exit to and from the circuit.
Transient entrainment with ventricular fusion in response to ventricular stimulation has been demonstrated in VT,* primarily in patients with coronary artery disease and old myocardial infarction, and in AV reciprocating tachycardias involving an AV bypass pathway.7 11 15 17 21
However, ventricular fusion is not always detectable on the surface ECG during transient entrainment.7 9 10 13 16 22 23 24 There are two potential explanations for this finding: (1) These arrhythmias are not reentrant; in fact, cellular and animal experiments have suggested that rhythms due to automaticity25 and triggered activity26 27 can be transiently entrained under certain circumstances. (2) Fusion exists, but it cannot be recognized. It seems obvious that a critical mass of ventricular myocardium has to be depolarized by the wave front exiting from the tachycardia circuit to produce a change in the QRS morphology so it can be distinguished from the fully paced QRS morphology. On the other hand, if most of the ventricular mass is depolarized as a result of the tachycardia wave front, a critical mass has to be depolarized by the paced wave front to produce a change in the QRS morphology so it can be distinguished from the tachycardia QRS morphology. Thus, the recognition of fusion on the surface ECG is limited by the amount of myocardium that needs to be depolarized in a different fashion and by the extent of the modifications of the ECG waveforms necessary to be recognized by the human eye. Moreover, the ability to recognize fusion has never been tested in a scientific way or compared with a “gold standard.”
To study the ability to recognize ventricular fusion, a “model” of fusion was created by the introduction, during a paced rhythm at one ventricular site (site A), of late premature impulses at a second ventricular site (site B) with a variable degree of prematurity. The aims of the present study were to analyze (1) the degree of fusion necessary to produce changes in the QRS complex detectable by the human eye and (2) other factors that can influence the recognition of fusion on the surface QRS.
The study group consisted of 21 patients (10 men and 11 women ranging in age from 21 to 80 years [mean±SD, 52±18 years]) who underwent an electrophysiological study for a variety of reasons. Nine patients had Wolff-Parkinson-White syndrome, 11 had a history of sustained monomorphic VT, and the remaining patient had syncope of unknown origin. Five patients had prior myocardial infarction, 4 had valvular heart disease, 1 had right ventricular dysplasia, and the remaining 11 had no structural heart disease. The mean left ventricular ejection fraction, calculated by echocardiography or contrast ventriculography, was 0.51±0.14 (range, 0.25 to 0.70). No patient presented with intraventricular conduction defects. Antiarrhythmic drugs were stopped at least five half-lives before the study. All patients gave written informed consent before invasive procedures were performed. Some relevant clinical or electrophysiological variables are depicted in Tables 1⇓ and 2⇓.
Electrical Stimulation Protocol and Recordings
Using standard techniques, we initially placed multipolar catheter electrodes (with 1- to 5-mm interelectrode distance) in the high right atrium, His bundle position, RVA, and, in the patients with Wolff-Parkinson-White syndrome, the coronary sinus. In 10 patients, an additional catheter was placed in the LV, to ablate a left-sided accessory pathway in 7 patients or to perform left ventricular stimulation in 3 patients with documented sustained VT noninducible from the right ventricle. For this particular study, the right atrial or the His bundle catheter was transiently positioned at the RVOT. Electrodes were used selectively to record electrograms and to pace the heart. Data were recorded with a photographic recorder (Honeywell VR 12) at a paper speed of 100 mm/s. Four to six surface ECG leads (I, II, aVF, aVR, V1, and V5) were recorded in all patients along with two to five bipolar intracardiac electrograms. All intracardiac electrograms were filtered at 30 to 500 Hz. We performed stimulation using two programmable stimulators simultaneously (Biotronic UHS 20 and Medtronic 5328). Bipolar cathodal stimulation was performed at the distal pair of each multipolar catheter. Stimuli were rectangular pulses 1 ms in duration at twice the diastolic threshold. To avoid a large separation between the recording and the stimulation sites, catheter electrodes with only 1-mm interelectrode distance were used; the distal pair of electrodes was used to stimulate and the second pair to record electrograms.
Model of Fusion
To create the model of fusion, during continuous pacing from ventricular site A at a constant rate of 400 ms, slightly faster trains (398-ms cycle length) were delivered at ventricular site B. The synchronization between the two stimulators was achieved in the following fashion: stimulation was started at site A in the VVI mode at a paced cycle length of 400 ms. After a few paced beats, synchronized continuous pacing was started at a site B, also in the VVI mode, at a paced cycle length of 398 ms. In this manner, the stimuli generated at site B had progressively greater degrees of prematurity with respect to the stimuli arising at site A, therefore producing QRS complexes with variable degrees of fusion until paced beats at site B captured the electrogram at site A, inhibiting pacing at site A. Stimulation was performed at the RVA, RVOT, and the lateral wall of the LV in two different configurations (RVA-RVOT and RVA-LV). In all but one patient, each site was alternately used as stimulation site A and site B in each configuration (RVA-RVOT, RVOT-RVA, RVA-LV, and LV-RVA). A percentage of fusion for each QRS complex was derived relative to the time interval between the two stimulation artifacts considering the following as explained in Fig 1⇓: (1) When the stimulus artifact at the second ventricular site occurs coincidently with or immediately after the onset of the local electrogram at that site, fusion cannot occur (or is minimal) because the second site is captured by the wave front generated at the first site; the time interval between the two stimuli in this situation (“X” in Fig 1⇓) is considered as a reference because it is the shortest time interval between the two stimulation artifacts without fusion (0% fusion). (2) When the two stimulation artifacts are delivered simultaneously (middle panel in Fig 1⇓), assuming similar conduction time in both directions, fusion occurs during half the time required for ventricular depolarization between the two stimulation sites in the absence of fusion (50% fusion). (3) Between these two situations (right panel in Fig 1⇓), the duration of fusion is derived by subtracting the time interval between the two stimulation artifacts (the certainly unfused part of the QRS duration; “Y” in Fig 1⇓) from “X” (the reference interval over which the second stimulus is “advanced” and thus, the potential time interval over which fusion can occur) and dividing the result by 2 (because conduction operates in both directions; see right panel in Fig 1⇓). The value is finally expressed as a percentage of X, the time interval over which fusion (F) can occur, according to the formula % F=F×100/X.
One of the investigators (J.A.) reviewed the surface QRS recordings while blinded to intracardiac recordings (intracardiac recordings were removed) and decided whether fusion was present or absent by comparing a QRS complex with 0% of fusion with other QRS complexes with different degrees of fusion or without it. To avoid the additional information that the investigator who blindly reviewed the surface QRS complexes might have derived from the stimulation artifacts buried in the QRS complexes (if the interval between stimulus artifacts was short, fusion was likely), stimulation was repeated in several cases with subthreshold electrical stimulation at site B. In this way, a short interval between stimulus artifacts may not mean fusion. Figs 2⇓ and 3⇓ show examples of QRS complexes with different degrees of fusion obtained with this model in each configuration.
The QRS complexes were divided into six groups based on the different degrees of fusion: group 1, no fusion (No F; QRS complexes with subthreshold electrical stimulation); group 2, <5% fusion; group 3, from 6% to 10% fusion; group 4, from 11% to 15% fusion; group 5, from 16% to 20% fusion; and group 6, >20% fusion. Once the reviewer decided whether fusion was present or not in each QRS complex, the percentage of successfully detected QRS complexes with fusion (or No F in group 1) was calculated in each group and configuration. The conduction time from site A to B in all configurations was measured to evaluate its influence in the detection of fusion. To evaluate the influence of the presence of a left ventricular organic disease, a subanalysis was performed considering patients with and without organic left ventricular disease.
Finally, to study the interobserver variability, blind evaluation of all the QRS complexes was made in the same manner by two additional investigators (J.P.V. and J.D.M.-A.).
Data are presented as mean±SD. Categorical variables were compared by use of the χ2 test. Continuous variables were compared by use of an unpaired two-tailed Student’s t test or a Mann-Whitney U test for nonparametric data. A value of P<.05 was considered statistically significant. We interpreted κ results as follows: values <0.20 are poor, 0.2 to 0.4 is fair, 0.41 to 0.60 represents moderate agreement, 0.61 to 0.80 represents good agreement, and 0.8 to 1.0 represents very good agreement.
Four hundred thirty-three QRS complexes were analyzed: 106 of the RVA-RVOT, 117 of the RVOT-RVA, 97 of the RVA-LV, and 113 of the LV-RVA configurations. Fusion was correctly identified in 273 (75%) of 365 QRS complexes with fusion. Nine (13%) of 68 analyzed QRS complexes without fusion were read as fused QRS complexes, and the remaining 59 were correctly identified as QRS without fusion. Thus, the overall sensitivity for the detection of fusion is 75%, the specificity is 87%, the positive predictive value is 97%, and the negative predictive value is 39%. The ability to detect fusion (or its absence) is considered in Table 3⇓ and Fig 4⇓ for each group and configuration. There was a statistically significant correlation between the degree of fusion and the degree of recognition (see the overall probability value for each configuration in Fig 4⇓). However, also note that in all configurations, there were some QRS complexes without fusion that were misclassified as fused. Although the percentage of QRS read as fused was lowest in the No F group, >5% fusion was necessary to obtain statistically significant differences with the No F group in RVOT-RVA and LV-RVA configurations and >10% in RVA-RVOT and RVA-LV configurations.
All QRS complexes with >22% fusion were identified correctly. If each patient is analyzed separately, the percentage of fusion over which fusion was always correctly detected was 22%. When the percentage of fusion was analyzed for each configuration separately, it was always correctly detected when it occurred for>20% of the QRS duration in the RVA-RVOT configuration, >20% in the RVOT-RVA configuration, >22% in the RVA-LV configuration, and >14% in the LV-RVA configuration. No statistically significant differences were obtained in the ability to detect fusion among RVA-RVOT, RVOT-RVA, and RVA-LV configurations. In contrast, fusion was better recognized when ventricular site A was the LV than when it was a right ventricular site (mean percentage of fusion for correct recognition, 7±4% versus 11±5%; P<.05; Fig 5⇓, top). When we separately analyzed patients with and without organic left ventricular disease, fusion was better detected when ventricular site A was the LV than when it was a right ventricular site only in patients with organic left ventricular disease (mean percentage of fusion for correct recognition, 5±3% versus 13±4%; P<.05; Fig 5⇓, bottom left). No significant differences were observed among patients without left ventricular disease (8±4% versus 9±5%, respectively; P=NS; Fig 5⇓, bottom right).
The conduction time was 72±21 ms from RVA to RVOT and 68±18 ms from RVOT to RVA (P=NS). It was 104±17 ms from LV to RVA and 98±19 ms from RVA to LV (P=NS). There was a significantly longer conduction time between LV and RVA than between both right ventricular sites (P<.01). No significant correlation was found between the detection of fusion and the conduction time between the different sites.
The interobserver agreement was moderate between observers 1 and 2 (κ=0.43) and 1 and 3 (κ=0.59). It was only fair between observers 2 and 3 (κ=0.40). There were not enough QRS complexes to establish differences in this issue, depending on the group (degree of fusion) or configuration.
The data obtained in the present study indicate that accurate detection of ventricular fusion (detection in almost all QRS complexes and in all configurations) can only be accomplished when fusion occurs during >22% of the QRS duration. Therefore, minor degrees of QRS fusion can easily be overlooked. Moreover, there is only fair to moderate agreement in the detection of fusion by different observers. When one looks for fusion, there is a small but not negligible percentage of unfused QRS complexes that can be misinterpreted as fused.
Importance of Fusion for the Detection of Transient Entrainment
Three basic criteria were initially recognized for the detection of transient entrainment: (1) acceleration of the tachycardia to the pacing rate with constant fusion on the surface ECG at constant pacing rates except for the last paced complex, which is unfused; (2) progressive fusion on the surface ECG with different degrees of constant fusion at different overdrive pacing rates; and (3) termination of the arrhythmia associated with sudden shortening of the conduction time from the pacing site to a recording site, with reversion of the electrogram morphology to that of pacing from the same pacing site in the absence of an arrhythmia.4 More recently, two other criteria have been recognized: an electrogram equivalent of progressive fusion20 and a constant first postpacing interval in response to pacing trains with a constant pacing cycle length and different number of beats.23 The first two criteria refer to the presence of fusion in the surface ECG. This is the simplest and most conventional way to detect transient entrainment. However, several studies have shown that fusion cannot be recognized in all VTs suspected to be caused by reentrant mechanisms.7 9 10 13 16 22 23 24
Findings of the Fusion Model
To the best of our knowledge, this is the first study that addresses the sensitivity and specificity of the surface QRS for the detection of ventricular fusion in a scientific fashion, comparing the QRS morphology with a “gold standard” given by intracardiac electrograms.
The significant relationship between the degree of fusion and the degree of recognition does suggest a significant accuracy of the surface QRS in the detection of fusion, ie, the more fused the QRS, the more likely it is to be recognized as such.
On the other hand, minor degrees of fusion may not deform the surface QRS, or alternatively, minor deformations in the QRS complex morphology may be overlooked because the observer is aware that spurious minor changes may be due to factors such as overlapping of the P waves, rate-related changes in QRS morphology, different degrees of superimposition of T waves from the previous beat as pacing rate changes, and finally, extracardiac signals or noise. For these same reasons, some QRS complexes without fusion (QRS with subthreshold stimulation from site B) can be erroneously detected as fused.
An additional finding was that the recognition of fusion was dependent, to some extent, on the pacing configuration. In patients with organic left ventricular disease, fusion was easier to detect when the initial forces of the QRS were generated at the LV. The reason for this finding may be that the spread of activation is slower when it arises in fibrotic or ischemic areas, and therefore the wave front proceeding from an opposite (healthy) site activates more ventricular mass at a given percentage of fusion (which represents time during which fusion is taking place), leading to an easier detection. In the study by Kay et al19 in which most patients had ischemic heart disease, transient entrainment with surface QRS fusion was detected in 10 (83%) of 12 cases of VT with left bundle-branch block morphology and left ventricular stimulation versus in 9 (69%) of 13 cases of VT with right bundle-branch block morphology and right ventricular stimulation. Aizawa et al,24 in a group of patients with VT unassociated with coronary artery disease, found transient entrainment with surface QRS fusion in 22 (95.7%) of 23 cases of VT with left bundle-branch block morphology originating from the LV and in 21 (84%) of 25 cases of VT with right bundle-branch block morphology originating from the right ventricle. Although the type of bundle-branch block morphology only provides limited information about the origin of a VT, these two studies would suggest that ventricular fusion could be easier to detect in VT of right ventricular or interventricular septal origin, with left ventricular stimulation. This would be in agreement with the findings of the present model of fusion, with an endocardial origin of the hypothetical VT. It would suggest that the difference found in these two studies relates to a difference in the ability to recognize fusion rather than to a difference in the mechanism of VT.
Reasons for Failure to Recognize Fusion During Transient Entrainment
In a previous study,13 a possible cause of the failure to recognize fusion during transient entrainment was demonstrated. If the exit site and presystolic electrograms from the VT circuit are retrogradely (antidromically) captured, fusion will not occur because the entire ventricle except part of the VT circuit is activated from the paced impulse.
In the present study, a second cause for this phenomenon is recognized: lack of sensitivity of the surface ECG to detect ventricular fusion if it occurs over a small percentage of the QRS duration.
We did not record all 12 surface ECG leads. Thus, it is conceivable that with more leads, the sensitivity for the detection of fusion could have been increased. However, most studies on transient entrainment of VT have used a similar number of leads.6 8 9 10 13 23 Stimulation was performed at the endocardial surface of the ventricle. The most frequent anatomic location of VT circuits is the subendocardial region, at least in patients with previous myocardial infarction. However, substrates involving subepicardial and deep septal layers have also been described.28 The extent to which this can influence the recognition of ventricular fusion is unknown. In the present study, a limited number of pacing configurations were tested. It is possible that the specific site of VT origin may influence the ability to observe fusion. Of particular importance may be the fact that pacing sites were far from each other. Closely located pacing sites, simulating pacing close to the exit of a reentrant circuit and even reentrant wavelets of different dimensions, could influence the results and the degree of “observed” fusion. We used the distal pair of electrodes of the catheters to stimulate the heart and the second pair to record electrograms. Despite the use of catheter electrodes with only 1-mm interelectrode distance, we cannot completely exclude a minimal degree of local capture of a pacing stimulus at a site at which the local electrogram of the adjacent pair was captured by the paced wave front of the other pacing site. Such an occurrence could have only happened in “marginal conditions” (ie, instances considered as 0% fusion) and are unlikely to influence the overall results.
Because our study was based on analysis of QRS morphology and not on intracardiac mapping, we cannot derive measurements of fusion in terms of ventricular mass or even percentage of endocardial surface but only in terms of the proportion of the QRS duration over which the phenomenon of fusion takes place with certainty. It is obviously possible that fusion might have extended beyond that part of the QRS that is considered to be fused; if that were the case, fusion would have been occurring in a greater proportion of the QRS than was estimated. However, this would have only decreased the calculated sensitivity. Thus, our data represent the maximum possible sensitivity for the detection of fusion.
When overstimulating a reentrant tachycardia, constant fusion and therefore transient entrainment do not develop immediately because the pacing wave front has to “peel back” the amount of myocardium that is influenced from the pacing site. Our model of fusion, by pacing at two different rates, did not enable us to produce a stable degree of prematurity to obtain constant fusion for at least several beats. Therefore, we cannot know if the analysis of a higher number of beats with the same morphology and with the same degree of fusion (or no fusion) could influence the recognition of the fusion (or its absence) by the observers compared with the analysis of only single QRS complexes with the possible effect of other conditions (respiratory movements, P wave, etc) that could influence the QRS morphology.
All available studies on transient entrainment of VT failed to demonstrate the phenomenon in certain individual cases. The information derived from the present study showing that minor (and even moderate) degrees of ventricular fusion can be overlooked would suggest that fusion could be present even in those cases in which it cannot be demonstrated. Furthermore, the significant interobserver variability for the detection of fusion would favor a cautious interpretation of the presence of fusion, ideally involving the coincidence of more than one interpretation. This precludes any suggestion in favor of other mechanisms rather than reentry simply because the demonstration of entrainment with fusion is lacking. In such cases, the demonstration of criteria for entrainment that do not involve changes in QRS morphology may be particularly relevant.
The presence of entrainment without fusion (so-called concealed entrainment) is currently being used as a criterion for localization of adequate target sites for VT ablation.29 30 31 32 However, it is recognized that the success rate for VT ablation, at least in coronary artery disease, is far from ideal at the present time. The lack of sensitivity of the surface QRS for the recognition of fusion observed in the present study could suggest that some of the observed failures may be due to the failure to recognize fusion, ie, instances in which fusion was present might have been considered as concealed entrainment.
Selected Abbreviations and Acronyms
|No F||=||no fusion; QRS complexes with subthreshold electrical stimulation|
|RVA||=||right ventricular apex|
|RVOT||=||right ventricular outflow tract|
Reprint requests to Jesús Almendral, MD, Cardiología (planta 5), Hospital General Gregorio Marañón, Doctor Esquerdo, 46, 28007 Madrid, Spain.
1 References 2, 6, 9, 10, 13, 14, 16, 19, 20, 22-24.
- Received February 6, 1997.
- Revision received June 19, 1997.
- Accepted July 3, 1997.
- Copyright © 1997 by American Heart Association
Waldo AL, MacLean WAH, Karp RB, Kouchoukos NT, James TN. Entrainment and interruption of atrial flutter with atrial pacing: studies in man following open heart surgery. Circulation. 1977;56:737-745.
Henthorn RW, Plumb VJ, Arciniegas JG, Waldo AL. Entrainment of ‘ectopic atrial tachycardia’: evidence for reentry. Am J Cardiol. 1982;39:920. Abstract.
Waldo AL, Plumb VJ, Arciniegas JG, MacLean WA, Cooper TB, Priest MF, James TN. Transient entrainment and interruption of the atrioventricular bypass pathway type of paroxysmal atrial tachycardia: a model for understanding and identifying reentrant arrhythmias. Circulation. 1983;67:73-83.
Portillo B, Mejías J, Leon-Portillo N, Zaman L, Myerburg RJ, Castellanos A. Entrainment of atrioventricular nodal reentrant tachycardias during overdrive pacing from high right atrium and coronary sinus, with special reference to atrioventricular dissociation and 2:1 retrograde block during tachycardias. Am J Cardiol. 1984;53:1570-1576.
Okumura K, Henthorn RW, Epstein AE, Plumb VJ, Waldo AL. Further observations on transient entrainment: importance of pacing site and properties of the components of the reentry circuit. Circulation. 1985;72:1293-1307.
Okumura K, Olshansky B, Henthorn RW, Epstein AE, Plumb VJ, Waldo AL. Demonstration of the presence of slow conduction during sustained ventricular tachycardia in man: use of entrainment of the tachycardia. Circulation. 1987;75:369-378.
Almendral JM, Gottlieb CD, Rosenthal ME, Stamato NJ, Buxton AE, Marchlinski FE, Miller JM, Josephson ME. Entrainment of ventricular tachycardia: explanation for surface electrocardiographic phenomena by analysis of electrograms recorded within the tachycardia circuit. Circulation. 1988;77:569-580.
Arenal A, Almendral J, Albertos J, Delcán JL. Determinants of ventricular fusion during entrainment of orthodromic tachycardia. Eur Heart J. 1990;11:304. Abstract.
Suyama K, Ohe T, Kurita T, Maruyama T, Takaki H, Aihara N, Kamakura S, Shimizu W, Matsuhisa M, Shimomura K. Significance of ventricular pacing in manifest entrainment during orthodromic atrioventricular reentrant tachycardia with left-sided accessory pathway. Pacing Clin Electrophysiol. 1992;15:1114-1121.
Henthorn RW, Okumura K, Olshansky B, Plumb VJ, Hess PG, Waldo AL. A fourth criterion for transient entrainment: the electrogram equivalent of progressive fusion. Circulation. 1988;77:1003-1012.
Ormaetxe JM, Almendral J, Arenal A, Martínez-Alday JD, Pastor A, Villacastín J, Delcán JL. Ventricular fusion during resetting and entrainment of orthodromic supraventricular tachycardia involving septal accessory pathways: implications for the differential diagnosis with AV nodal reentry. Circulation. 1993;88:2623-2631.
Aizawa Y, Niwano S, Chinushi M, Tamura M, Kusano Y, Miyajima T, Kitazaba H, Shibata A. Incidence and mechanism of interruption of reentrant ventricular tachycardia with rapid ventricular pacing. Circulation. 1992;85:589-595.
Martínez-Alday JD, Almendral J, Pastor A, Arenal A, Villacastín JP, Peinado R, Medina O. Evaluation of the usefulness of a variety of criteria for the detection of transient entrainment during ventricular tachycardia. Circulation. 1993;88(suppl I):I-459. Abstract.
Daugman KH, Hoffman BF. The effects of single premature stimuli on automatic and triggered rhythms in isolated canine Purkinje fibers. Circulation. 1985;71:813-822.
Damiano BP, Rosen MR. Effects of pacing on the triggered activity induced by early afterdepolarizations. Circulation. 1984;69:1013-1025.
Moak JP, Rosen MR. Induction and termination of triggered activity pacing in isolated canine Purkinje fibers. Circulation. 1984;69:149-162.
Cranefield PF, Aronson RS. Initiation of sustained rhythmic activity by single propagated action potentials in canine cardiac Purkinje fibers exposed to sodium-free solution or to ouabain. Circ Res. 1974;34:477-481.
Stevenson WG, Khan H, Sager P, Saxon LA, Middlekauff HR, Natterson PD, Wiener I. Identification of reentry circuit sites during catheter mapping and radiofrequency ablation of ventricular tachycardia late after myocardial infarction. Circulation. 1993;88:1647-1670.
Morady F, Harvey M, Kalbfleisch SJ, El-Atassi R, Calkins H, Langberg JJ. Radiofrequency catheter ablation of ventricular tachycardia in patients with coronary artery disease. Circulation. 1993;87:363-372.
Kim YH, Sosa-Suárez G, Trouton TG, O’Nunain SS, Osswald S, McGovern BA, Ruskin JN, Garan H. Treatment of ventricular tachycardia by transcatheter radiofrequency ablation in patients with ischemic heart disease. Circulation. 1994;89:1094-1102.