Polymorphic Reentrant Ventricular Tachycardia in the Isolated Rabbit Heart Studied by High-Density Mapping
Background— The role of dispersion of refractoriness and reentry for the genesis of polymorphic ventricular tachycardia (VT) has recently become emphasized. We investigated the mechanisms of polymorphic arrhythmias in a 2D preparation confining an area of prolonged refractoriness.
Methods and Results— In 16 Langendorff-perfused rabbit hearts, a sheet of left ventricular epicardium was obtained by a cryoprocedure. Enhanced spatial heterogeneity in a refractory period was created by cooling a central region (diameter=12 mm). This markedly prolonged the refractory period (by 36±14 ms) inside but only slightly prolonged it (by 5±11 ms) outside the cooled area (n=6). During a control procedure, programmed stimulation with up to 3 premature stimuli induced an episode of monomorphic VT in only 1 of 10 hearts. During regional cooling, episodes of polymorphic VT with a maximum duration of 35 seconds could be induced in all hearts. High-resolution mapping (229 electrodes) of epicardial activation revealed that polymorphic VT was caused by a functional reentrant circuit located partially within the region of prolonged refractoriness. The reentrant wavefront was continuously shifting along the border of the cooled region, resulting in beat-to-beat changes in the excitation pattern. Spontaneous termination of polymorphic VT occurred either by a shift of the reentrant circuit outside the cooled region or by a block in the central common pathway during figure-of-8 reentry in the region of prolonged refractoriness.
Conclusions— A shifting functional reentrant circuit was the underlying mechanism of polymorphic VT in a substrate of enhanced spatial heterogeneity of refractoriness.
Received January 21, 2002; revision received April 11, 2002; accepted April 11, 2002.
Polymorphic ventricular tachycardia (VT) has been reported in a variety of structural cardiac disorders and in structurally normal hearts with ion channel abnormalities such as long-QT syndrome or Brugada syndrome.1–3⇓⇓ An enhanced spatial dispersion of refractoriness seems to increase the propensity for polymorphic VT, and this has been associated with the occurrence of triggered activity based on early afterdepolarizations.4 More recently, emphasis has been put on reentry as a possible mechanism underlying polymorphic VT.5–8⇓⇓⇓
In the present study, we investigated the influence of enhancing the spatial dispersion of refractoriness in a well-documented model of 2D anisotropic rabbit myocardium.9–12⇓⇓⇓ In this model, aggressive programmed electrical stimulation can provoke monomorphic VT based on a functional reentry circuit.11,12⇓ We extended the model by creating a central area of prolonged refractoriness in the wall of the left ventricle by local cooling. As a result of this intervention, single or multiple premature beats induced short episodes of rapid polymorphic VT based on a functional reentrant circuit that continuously changed its reentrant pathway. In the present article, we describe the mechanisms underlying the polymorphic nature of VT in this model.
The frozen heart model we used has been extensively described and validated previously.9–12⇓⇓⇓ For the present study, 16 Flemish Giant rabbits of both sexes (weight, 3.2 to 4.0 kg; van der Zwam, Weert, the Netherlands) were used; they were treated in accordance with the guidelines of the American Physiological Society, and treatment was approved by the Animal Investigation Committee of the Maastricht University. After sedation with 0.5 mL/kg Hypnorm (fluanisone 10 mg/mL and fentanyl 0.2 mg/mL) and heparinization (1000 IU IV), the animals were killed by cervical dislocation. The heart was rapidly removed and connected to a Langendorff perfusion system. The coronary arteries were perfused with Tyrode’s solution at 37°C (pH 7.35) and saturated with a mixture of 95% O2 and 5% CO2. The perfusion pressure was 55 mm Hg, which resulted in a flow of 29.9±6.7 mL/min. The complete right ventricle and the endocardial and intramural layers of the left ventricle were cryoablated. The final preparation consisted of a viable sheet of perfused left ventricular epicardium (±1 mm thick) without areas of tissue necrosis. The refractory periods (RP) were stable in time and had a very low spatial variation.9 At the start of each experiment, the absence of conduction defects was verified by pacing at long cycle lengths. In the epicardial layer, there is a characteristic orientation of the fiber axis starting perpendicular to the left anterior descending coronary artery (LAD) and gradually curving toward the apex; this results in an ellipsoid, uniform, anisotropic spread of activation.10 Conduction parallel to this fiber axis is ≈3 times faster than in a transverse direction. In this model, sustained monomorphic VT based on functional reentry may be provoked by aggressive programmed electrical stimulation.11,12⇓
Mapping Electrode and Regional Cooling
For mapping of epicardial activation, we used a spoon-shaped, high-resolution electrode spoon with a specially designed cooling probe that covered the entire free wall of the left ventricle without missing critical areas.9–12⇓⇓⇓ The surface shape of the spoon and the cooling probe matched the curvature of the heart. The spoon electrode contained 192 unipolar electrodes (silver wires measuring 0.25 mm) with an interelectrode distance of 2.5 mm that were evenly distributed around the cooled region. The tip of the cooling probe contained an additional 37 unipolar electrodes (glass-coated silver wires measuring 0.25 mm) evenly spaced at a distance of 1.5 mm. A stainless steel canula in the aorta served as an indifferent electrode.
The central region of the left ventricular free wall could be temporarily cooled at any temperature by running cooling water through the cooling probe. The probe consisted of a stainless steel pipe (12 mm) that fit exactly in the hole of the mapping electrode. Both the electrodes and the cap of the cooling probe were made of silver to ensure a good and homogeneous transmission of cooling. The exact temperature under the cooling probe was not measured because cooling was only used to create a large gradient of refractoriness. The effect of cooling on the local effective RP (ERP) was demonstrated at a single temperature, whereas the arrhythmogenic effects were determined at 3 temperatures. The cooling effect was never so strong that a part of the tissue became locally inexcitable and prohibited any electrical activation.
In a first set of experiments (6 hearts), we studied the effects of regional cooling on the spatial inhomogeneity of RPs using cooling water at a temperature of 15°C. Electrical stimulation could be performed through any pair of electrodes in the spoon electrode or cooling probe. A constant current stimulator delivered square pulses of 2-ms duration with an amplitude of 4 times diastolic threshold. The ERP was determined with the extrastimulus method, which involved the stepwise lowering of a premature stimulus by 3 ms during regular pacing at 300-ms intervals. The ERP measurements were performed twice for each site (during control and during regional cooling) at multiple positions inside and outside the cooled region. The spatial dispersion of refractoriness was defined as the maximal difference in RP between the sites.
In a second set of experiments (10 hearts), we studied the effects of regional cooling on the inducibility of ventricular arrhythmias at 3 different temperatures of the cooling water (20°C, 10°C, and 0°C). Programmed electrical stimulation was performed at multiple positions inside and around the cooled region and consisted of a train of 10 basic stimuli with a coupling interval of 300 ms followed by 1 to 3 premature stimuli. The coupling interval of the first premature stimulus was shortened in steps of 5 ms until an arrhythmia occurred or capture failed. The coupling interval was then increased by 5 ms and the same sequence was repeated for a second and third premature stimulus. An arrhythmia of <6 beats was defined as repetitive ventricular responses (RVR), whereas episodes of ≥6 beats were designated as polymorphic VT. This distinction was made because the polymorphic nature of a VT may not be apparent for very short arrhythmias and also to be sure that the episode of polymorphic VT was long enough to enable a relevant study of the underlying mechanisms.
Registration and Data Analysis
All signals were individually amplified (bandwidth, 1 to 400 Hz), sampled at 1000 Hz, A/D converted (resolution, 8 bit), and stored on videotape. A computer algorithm was used for automatic detection of the steepest negative deflection in each electrogram marking the moment of local activation. All electrograms were checked and edited manually if necessary. All activation maps have the same anatomic orientation, with the base at the top and the apex at the bottom and the LAD at the left side of the map. The dotted area indicates the central cooled region. Isochrones were drawn every 10 ms, and a solid black line or a double bar indicated block. Conduction block was defined as a local velocity <5 cm/s (50 ms over 2.5 mm or 30 ms over 1.5 mm) associated with a change in direction of the impulse at the other side of the line of block. This was stricter than the criteria used in previous studies with this model11,12⇓ to minimize the inclusion of very slow local conduction or pseudoblock. Similar criteria have been used by other investigators.8,13–15⇓⇓⇓
Average values are given as mean±SD. The paired Student’s t test was used to compare RP measurements during cooling and control. The unpaired Student’s t test was used to compare RP within and around the cooled region during both conditions. Differences were considered to be statistically significant at P<0.05.
Spatial Inhomogeneity in RP by Regional Cooling
In 6 experiments, the effects of regional cooling on the spatial inhomogeneity in ERP were studied. During both control and regional cooling, the ERP was measured at 17±2 sites within the central region and at 21±6 sites around it. The latter sites were at an average distance of 4.7±0.8 mm from the border of the central region. Figure 1 shows an example of the spatial inhomogeneity in ERP in the ventricle during regional cooling at 15°C. Outside the cooled region, the ERP was not altered (146±6 ms versus 147±8 ms) and the variation remained low, demonstrating the stability of the preparation in time. Inside the cooled region, the ERP was significantly prolonged by an average of 36 ms (150±3 versus 186±8 ms) while the ERP variation remained low, indicating a homogeneous and stable effect of cooling over time. Table 1 summarizes the effects of regional cooling on ERP for all 6 hearts. Cooling prolonged the ERP in the central area from 144±6 to 180±15 ms (P<0.001). The spatial dispersion in ERP increased from 4±4 ms during control to 34±7 ms during cooling (range, 27 to 45 ms; P<0.001). In experiments 2, 4, and 6, a small but significant increase in ERP (between 5 and 15 ms) was also found outside the cooled region. In general, local cooling resulted in a sudden steep gradient of refractoriness within 1 to 2 interelectrode distances from the cooling probe.
Inducibility of Polymorphic Arrhythmias
In 10 hearts, the inducibility of arrhythmias was tested by 1 to 3 premature stimuli, both during control and during regional cooling at various temperatures. In Figure 2 an example is given. During control, up to 3 premature stimuli with the shortest possible coupling intervals failed to induce an arrhythmia. During cooling at 20°C, application of 3 premature stimuli induced a polymorphic VT; at 10°C, a single premature stimulus was sufficient to evoke polymorphic VT. During polymorphic VT, beat-to-beat changes in electrogram morphology and marked variations in cycle length occurred. In the example shown in Figure 2, during regional cooling at 20°C, the intervals varied between 121 and 160 ms (mean, 139±10 ms); at 10°C, they varied between 127 and 179 ms (mean, 157±21 ms). Table 2 shows the inducibility of arrhythmias for all hearts during control and regional cooling at different temperatures (20°C, 10°C, and 0°C). During control, premature stimulation induced RVRs in only one heart and sustained monomorphic VT in another heart. In contrast, during regional cooling, episodes of polymorphic arrhythmias were initiated in all hearts. Also, in the heart that had a monomorphic VT induced by 3 premature stimuli during control, only RVRs and polymorphic VT were induced during regional cooling. The inducibility of polymorphic VT tended to increase at lower cooling temperatures. There was no consistent relationship between the number of stimuli needed to induce arrhythmias and the cooling temperature.
Mapping of Initiation of Polymorphic Arrhythmias
Figure 3 shows an example of initiation of polymorphic VT by 3 premature stimuli during regional cooling at 10°C. The polymorphic nature of the tachycardia is obvious from the variations in cycle length (between 124 and 166 ms) and morphology of the electrogram. The upper panels show the last basic beat (S1) and 3 stimulated premature beats (S2, S3, and S4). S1 resulted in a rather uniform epicardial spread of excitation with only a slight slowing of conduction at the cooled region. The wavefront had an ellipsoid shape because conduction was faster parallel than perpendicular to the fiber axis (from upper left to lower right).9–12⇓⇓⇓ The excitation wave elicited by S2 propagated with considerable delay in the cooled region where the tissue was still partially refractory. This resulted in 2 wavefronts propagating around both sides of the cooled region and colliding in the lower left part of the ventricular wall. During S3, conduction was even further depressed, and 2 arcs of functional block now appeared at the cooled region. During S4, the arc of block markedly prolonged at the proximal boundary of the cooled region. Two slowly propagating wavefronts traveled around this line of block and started a figure-of-8 reentry circuit (panel 1).13 However, during the first beats of the resulting VT, the counterclockwise wavefront of the figure-of-8 reentry was blocked at the LAD border (panel 3), and VT continued as a single clockwise reentry circuit (panel 4). As can be seen in panels 1 to 8, the circulating wave never followed its previous pathway but continuously shifted along the border of the cooled region with an ever-changing size and position of the central line of conduction block. Although the circuit was anchored to the cooled region, the exit site of the circulating wave and ventricular activation pattern changed from beat to beat. This explains the continuous variations in intervals and electrogram morphology during polymorphic VT.
Mechanisms Underlying a Shifting Circuit
To demonstrate what forces a functional circuit to shift, in Figure 4 activation maps and a selection of local electrograms are given during 4 consecutive beats of an episode of polymorphic VT. The map of the first beat exhibits a small clockwise reentry circuit half inside and half outside the cooled region. The electrograms recorded at sites a to f follow the circulating wave around a line of functional block in the left part of the cooled region. During the next beat (panel 2), the circulating wave failed to follow its previous pathway because of local conduction block at site c, located inside the area of prolonged refractoriness. The line of block now extended all the way across the cooled region, resulting in delayed activation of sites d to f. During beat 3, local conduction block of the circulating wave at site d again prevented early entry in the region with prolonged refractoriness. Reentry of the cooled region now occurred close to site e, and site d was now actually activated later than sites e and f. Due to this delay in activation inside the area of prolonged refractoriness, the surrounding tissue could recover its excitability and the impulse could exit the cooled region at multiple sites. During beat 4, the position of the reentry circuit shifted to the bottom of the cooled region; sites a, b, and c were now remote bystanders outside the actual reentry circuit. The activation maps and electrograms demonstrate that the shifting position of the reentry circuit was primarily due to the differences in refractoriness between the normal and cooled myocardium. Extension of the functional line of block occurred continuously as the circulating wave tried to reenter the area with prolonged refractoriness from the surrounding tissue with shorter refractoriness.
Termination of Polymorphic VT
All episodes of polymorphic VT terminated spontaneously within 35 seconds. Two different mechanisms for termination were observed: (1) the circulating impulse of a single reentrant circuit was extinguished when it shifted to the anatomic boundary of the preparation, or (2) conduction block occurred in the central common pathway during figure-of-8 reentry located within the region of prolonged refractoriness. A single reentry circuit could transform to figure-of-8 reentry or vice versa during one and the same episode of polymorphic VT (Figure 3). Short lasting RVRs ended more often as figure-of-8 reentry (59%), whereas longer lasting polymorphic VT predominantly (73%) ended by extinction of a single reentry circuit at the boundaries. Figure 5 shows the last 4 activation maps of a polymorphic VT that terminated by a shift of a clockwise circulating wave to the boundary of the preparation. The electrograms at sites a through f around the cooled region show the characteristic beat-to-beat variations in interval and morphology. In panels 1 to 3, the central line of block shifted within the region of prolonged refractoriness. During the second to last beat, a double potential was recorded at site f, signifying local conduction block between sites e and f. Panel 3 shows that this resulted in a slight extension of the central line of block outside the cooled region. At this turning point, the circulating wave propagated slowly, allowing more time for recovery of excitability in the cooled region. During the last beat of VT (panel 4), the circulating wave could therefore propagate at relatively high speed from right to left through the cooled region, and it arrived early again at the previous turning point. At site f, the last interval was very short (103 ms) and the circulating wave failed to activate sites a and a’. The central line of block was forced to extend to the boundary, and the circulating wave was extinguished.
An example of figure-of-8 type termination is given in Figure 6. Panel 1 shows that VT was due to a single clockwise reentry circuit along a small line of block within the cooled region, with a large offspring wavelet propagating toward the apex. During the next activation (panel 2), the main circuit shifted a little to the right, the heart of the cooled region was momentarily inexcitable (long interval in electrograms B and C), and a long arc of conduction block toward the apex appeared. The offspring clockwise wavelet of the previous beat collided with a counterclockwise offspring of the circulating wave left of the cooled region. During the next beat (panel 3), the cooled region had recovered its excitability, and the clockwise circulating wave exited so early that it was blocked toward both the LAD and the base (local interval about 100 ms). However, VT did not terminate because an offspring wavelet from the apex reentered the cooled region from site a to b to c, divided into a clockwise and counterclockwise impulse, and started a new figure-of-8 reentry circuit around the cooled region. However, neither of the 2 opposed waves succeeded in finding an entrance point into the cooled region because they only had half of the perimeter of the area of prolonged refractoriness at their disposal. The figure-of-8 reentry waves finally fused at site i; thus, they were blocked at the central common pathway that was still refractory, and the polymorphic VT terminated.
Arrhythmogenicity of Spatial Dispersion of Refractoriness
In a variety of early experimental studies, it was established that dispersion of refractoriness predisposes to the development of ventricular arrhythmias.16 In canine left ventricle, Kuo et al17 used a combination of global cooling and regional warming to create a critical dispersion of refractoriness of 111±16 ms. Programmed electrical stimulation from the region with short refractoriness resulted in inducibility of ventricular fibrillation. In the postinfarction dog heart, inducibility of sustained monomorphic VT was also related to a steep gradient of refractoriness between adjacent areas in the surviving tissue.14 There seem to be inherent differences in repolarization characteristics in the ventricles that are due to ion channel differences within the right an left ventricular wall and from endocardium to epicardium.18 Studies in long-QT models show that certain interventions may enhance this inherent dispersion in refractoriness and set the stage for the occurrence of polymorphic VT.2,4⇓ The frozen heart model created in our laboratory is characterized by rather uniform refractoriness, and aggressive electrical stimulation is required to induce sustained monomorphic VT.11,12⇓ In the present study, we applied regional cooling of an area of ≈1 cm2 to create a steep gradient of refractoriness, and this made the ventricle vulnerable to polymorphic arrhythmias. Similarly, in a guinea pig frozen heart model, Pastore and Rosenbaum19 needed high pacing rates to induce a discordant alternans in repolarization between adjacent regions that facilitated the induction of ventricular arrhythmias during control. When a structural anatomic barrier leading to regional uncoupling was introduced, a critical repolarization gradient and arrhythmia induction occurred at much lower pacing rates.
A Shifting Reentry Circuit Causing Polymorphic VT
Polymorphic VT has been explained by mechanisms such as competition of multiple ventricular sources, wandering of a single source, or a single source with multiple exit sites.6 In experimental long-QT models, the initiation of polymorphic VT coincided with triggered activity due to early afterdepolarizations.2,4⇓ However, there are mapping studies providing evidence that reentry may be the underlying mechanism of polymorphic VT. In the postinfarction canine model of El-Sherif,13 polymorphic VT was occasionally observed when figure-of-8 reentry degenerated into multiple asynchronous reentry circuits. Schmitt et al15 observed polymorphic VT with a torsades des pointes morphology that was due to small changes in the size of functional reentrant circuits with a variable exit point to the ventricle. In 2D computer models and with voltage-dye mapping of superfused sheets of ovine myocardium or isolated perfused intact hearts, polymorphic VT and ventricular fibrillation were caused by drifting single or figure-of-8 scroll waves.5–7⇓⇓ The drifting was attributed to nonuniformities in electrophysiological properties inherent to the normal tissue, although these were not systematically studied. Of interest, many of the arrhythmias only occurred after the temperature was lowered to 30°C to 35°C.
El-Sherif et al8 used an anthopleurin infusion in the intact canine heart to prolong refractoriness, especially at the midmyocardial level, facilitating nonsustained polymorphic arrhythmias. Tridimensional mapping with plunge electrodes showed that the arrhythmias were initiated by focal endocardial activity but seemed to be maintained mostly by one or more circulating waves that traveled around both ventricles, with their central line of functional block at the midmyocardial level. In our model, we used local cooling to create an area with prolonged refractoriness, facilitating the induction of polymorphic VT based on a continuously shifting functional reentry circuit. The reentrant wavefront continuously traveled in and out of the cooled region but was always forced to extend its pathway before it could reenter the area with prolonged refractoriness, as observed previously by Fast and Pertsov.20 The line of block was shortcut at the tail when the impulse traveled toward the normal tissue again, resulting in a constantly changing exit point to the ventricle with a beat-to-beat variation in activation pattern. The circulating wave was confined to the area with prolonged refractoriness, but there was no evidence for fixed anatomic structures, such as a small vessel, or local tissue damage to anchor the reentrant pathway in a fixed position or transform polymorphic VT or ventricular fibrillation to monomorphic VT, as observed in other studies.5,6,19⇓⇓
Instability and Termination of Polymorphic Reentry
In our experiments, all episodes of polymorphic VT were nonsustained and terminated within <1 minute. This propensity for self-termination of polymorphic VT was also observed in the experimental studies of Pertsov et al,5 Gray et al,7 and El-Sherif et al,8 although in the intact heart, VT could also degenerate into ventricular fibrillation. In our study, 2 mechanisms for termination of polymorphic VT were observed. Usually, the single circulating wave was extinguished when the central line of conduction block extended from the cooled region to the boundary of the ventricle. Boundary effects of nonstationary reentry in a finite medium were also observed by Davidenko6 and Fast and Pertsov.20 In their studies, the circulating wave was not confined to a specific region and simply drifted to the boundary of the preparation. In contrast, in our model, the circulating wave was usually confined to the central cooled region, preventing free drift. Termination was preceded by a marked shift of the exit site from the central cooled region toward an area that had not yet recovered excitability, extending the line of block to an anatomic boundary. Ikeda et al21 also showed that a shortcut of the central line of block of a functional atrial reentry circuit forced the circulating wave to annihilation at a boundary. Consistent with our findings, termination of nonsustained reentrant VT in humans was preceded by acceleration and a marked change in direction of the circulating wave.22
The second mechanism of termination we observed was block in the central common pathway of figure-of-8 reentry. El-Sherif13 described how instability and shortcut of the 2 central lines of conduction block terminated the circulating waves of monomorphic figure-of-8 reentry at the central common pathway. In our model of polymorphic VT, figure-of-8 reentry was never stable and either terminated or transformed into a single circuit (Figure 3). Due to the inhomogeneous refractoriness, varying length and location of the lines of block, and varying conduction velocities depending on relation to fiber orientation, the revolution time of 2 circulating waves was rarely equal, and usually 1 of the 2 waves would gain dominance to form a single reentrant circuit. However, when the 2 circulating waves simultaneously filled the entire ventricle in opposite directions, they became trapped by the anatomic boundaries. This forced them to reenter the common pathway in the cooled region, which could not recover its excitability in time, resulting in termination of the polymorphic VT. The instability of polymorphic VT in our model is explained by the fact that the area of prolonged refractoriness constantly forces the circulating wave(s) to find new pathways until it runs into a dead end that is either created by its own wake of refractoriness or is an anatomic boundary.
Relevance of the Study
Evidence is gathering that polymorphic VT may be due to reentrant activation. This is difficult to study in humans, and only a few experimental studies have been performed. In some, inhomogeneities in electrophysiological properties were implied as the driving force for drifting reentry, but this was not always systematically defined or demonstrated.5–7⇓⇓ Some limitations of the mapping techniques in time and space exist when studying the heart with epicardial optical voltage-dye mapping5–7⇓⇓ or tridimensional plunge electrodes.8 In our simplified 2D model, the regional prolongation of refractoriness was well defined in space, and we were able to perform detailed mapping of the complete activation patterns during the induction, shifting, and self-termination of reentrant wavefronts. Our results provide additional evidence that polymorphic VT may be based on a shifting functional reentry circuit.
The cooling water was kept constant at each temperature, but the exact tissue temperature under the cold probe was not determined. Differences in tissue temperature may have existed between experiments. In 3 experiments, we observed an increase in ERP outside the central area; this could have been due to local cold transmission. The measurement of ERP at multiple sites was very time-consuming and could only be performed at a single cooling temperature in each experiment. We were unable to combine it with the study of inducibility of arrhythmias in the same heart. The effect of local cooling on conduction velocity was not studied. The anisotropic conduction properties inherent to our model9–12⇓⇓⇓ probably influenced ventricular activation, but their contribution to arrhythmogenicity was not studied. We did not obtain a surface electrocardiographic recording of the induced arrhythmias to correlate this to clinical types of arrhythmias.
- ↵Nguyen PT, Scheinman MM, Seger J. Polymorphous VT: clinical characterization, therapy, and the QT interval. Circulation. 1986; 74: 340–349.
- ↵Brugada J, Brugada R, Brugada P. Right bundle-branch block and ST-segment elevation in leads V1 through V3: a marker for sudden death in patients without demonstrable structural heart disease. Circulation. 1997; 95: 457–460.
- ↵Pertsov A, Davidenko G, Salomonsz R, et al. Spiral waves of activation underlie reentrant activity in isolated cardiac muscle. Circ Res. 1993; 72: 631–650.
- ↵Gray RA, Jalife J, Panfilov A, et al. Nonstationary vortexlike reentrant activity as a mechanism of polymorphic VT in the isolated rabbit heart. Circulation. 1995; 91: 2454–2469.
- ↵El-Sherif N, Caref E, Yin H, et al. The electrophysiologic mechanism of ventricular arrhythmias in the long QT syndrome: tridimensional mapping of activation and recovery patterns. Circ Res. 1996; 79: 474–492.
- ↵Schalij MJ, Lammers WJ, Rensma PL, et al. Anisotropic conduction and reentry in perfused epicardium of rabbit left ventricle. Am J Physiol. 1992; 263: H1466–H1478.
- ↵Brugada J, Boersma L, Kirchhof CJH, et al. Reentrant excitation around a fixed obstacle in uniform anisotropic ventricular myocardium. Circulation. 1991; 84: 1296–1306.
- ↵Boersma L, Brugada J, Kirchhof C, et al. Mapping of reset of anatomic and functional reentry in anisotropic rabbit ventricular myocardium. Circulation. 1994; 89: 852–862.
- ↵Schalij M, Boersma L, Huijberts M, et al. Anisotropic reentry in a perfused 2-dimensional layer of rabbit ventricular myocardium. Circulation. 2000; 102: 2650–2658.
- ↵El-Sherif N. The figure of 8 model of reentrant excitation in the canine post-infarction heart.In: Zipes D, Jalife J, eds. Cardiac Electrophysiology and Arrhythmias. Orlando: Grune and Stratton; 1985: 363–378.
- ↵Gough W, Mehra R, Restivo M, et al. Reentrant ventricular arrhythmias in the late myocardial infarction period in the dog, 13: correlation of activation and refractory maps. Circ Res. 1985; 57: 432–444.
- ↵Schmitt H, Cabo C, Costeas C, et al. Mechanisms for spontaneous changes in QRS morphology sometimes resembling torsades de pointes during reentrant ventricular tachycardia in a canine infarct model. J Cardovasc Electrophysiol. 2001; 12: 686–694.
- ↵Surawicz B. Dispersion of refractoriness in ventricular arrhythmias.In: Zipes D, Jalife J, eds. Cardiac Electrophysiology: From Cell to Bedside. Philadelphia: Saunders; 1990: 377–385.
- ↵Kuo CS, Munakata K, Reddy P, et al. Characteristics and possible mechanism of ventricular arrhythmia dependent on the dispersion of action potential durations. Circulation. 1983; 67: 1356–1367.
- ↵Antzelevitch C, Sicouri S, Litovsky S, et al. Heterogeneity within the ventricular wall: electrophysiology and pharmacology of epicardial, endocardial, and M-cells. Circ Res. 1991; 69: 1427–1449.
- ↵Pastore J, Rosenbaum D. Role of structural barriers in the mechanism of alternans-induced reentry. Circ Res. 2000; 87: 1157–1163.
- ↵Ikeda T, Uchida T, Hough T, et al. Mechanism of spontaneous termination of functional reentry in isolated canine right atrium: evidence for the presence of an excitable but nonexcited core. Circulation. 1996; 94: 1962–1973.
- ↵Pogwizd S, Chung M, Cain M. Termination of VT in the human heart: insights form three-dimensional mapping of nonsustained and sustained VTs. Circulation. 1997; 95: 2528–2540.