Reentrant Wave Fronts in Wiggers’ Stage II Ventricular Fibrillation
Characteristics and Mechanisms of Termination and Spontaneous Regeneration
Abstract The mechanisms of Wiggers’ stage II ventricular fibrillation (VF) are poorly understood. Using computerized mapping techniques, we studied the patterns of activation during Wiggers’ stage II VF in 13 open-chest dogs. In 7 of the 13 dogs, the right ventricular Purkinje fibers and adjacent subendocardial myocytes were ablated with Lugol solution. VF was induced electrically, and 3 to 5 seconds of data were obtained beginning ≈2.5 seconds after the onset of VF. Dynamic displays of the activation patterns and isochronal maps revealed the presence of reentrant wave fronts in 17 of 33 runs of VF in ablated ventricles and in 12 of 45 runs of VF in intact ventricles. The incidence of reentry was not different between the subendocardium-ablated group versus the nonablated group (1.7±1.6 versus 1.2±1.6 rotations per episode of VF, P=.19). There were no differences in the core size (25±19 versus 29±18 mm2), life span (3.4±1.1 versus 3.2±1.2 rotations), or cycle length (111±12 versus 107±8 ms) in ablated ventricles versus intact ventricles, respectively. The core was unstable as it meandered within the mapped area displacing the entire reentrant wave front. In all episodes, the reentrant wave fronts were spontaneously initiated by an interaction between two propagating wave fronts roughly perpendicular to each other. The second wave front met the tail of the first wave front 69±11 ms (range, 40 to 90 ms) after its latest activation, indicating that the interaction occurred during a vulnerable period. The reentrant wave fronts terminated spontaneously (n=7), as the result of interference by an invading wave front (n=19), or meandered off the mapped region (n=3). We conclude the following: (1) Reentrant activities with short life spans and meandering cores are present during Wiggers’ stage II VF in dogs. (2) New reentrant wave fronts are generated when one wave front interacts with another wave front during its vulnerable period. (3) The reentrant wave fronts terminate spontaneously or as the result of interference. (4) Chemical subendocardial ablation does not affect the incidence, life span, cycle length, or core size of the reentrant wave fronts.
In 1930, Wiggers et al1 reported that four stages of electrically induced ventricular fibrillation (VF) were visible by cinematography. The first stage, which lasts for only a few seconds, is the undulatory or tachysystolic stage. The second stage, which lasts for 15 to 40 s, is called the convulsive incoordination stage. This stage is characterized by more frequent waves of contractions, which sweep over smaller sections of the ventricles. The third stage, lasting 2 to 3 minutes, is that of tremulous incoordination. The final stage is the stage of atonic fibrillation, with complete failure of contractility. Among these four stages of VF, Wiggers’ stage II VF is clinically the most significant, because it is during this stage that therapeutic modalities, such as electric countershock, can effectively terminate VF and prevent sudden cardiac death.
By use of computerized mapping techniques, it has been demonstrated that reentrant wave fronts underlie the onset (Wiggers’ stage I) of VF in intact canine ventricles.2 3 4 However, the reentrant wave fronts observed in these studies, which were initiated by the strong electrical stimuli used to induce VF, had an average life span of only 1.36 s before termination.2 Moreover, this time period correlates with a surface electrocardiogram consistent with ventricular tachycardia. The mechanisms of Wiggers’ stage II VF, which correlates with VF on the electrocardiogram, remains poorly understood. One study5 was able to show the presence of a reentrant wave front 20 s after the onset of VF; however, only one episode was demonstrated, and only three cycles were analyzed before defibrillation changed the activation sequence. Other investigators6 7 used unipolar recordings to map VF. Reentrant excitation was either a rare event6 or was reported to be nonexistent.7 These observations raise the possibility that although reentry underlies the onset, or Wiggers’ stage I, of electrically induced VF,2 3 4 the maintenance of Wiggers’ stage II VF is due to entirely different mechanisms. We hypothesize that reentry underlies the mechanism of Wiggers’ stage II VF and that technical limitations were responsible for the poor understanding of these reentrant activities. To test this hypothesis, we constructed a 512-channel computerized mapping system and a bipolar recording plaque electrode array with short interelectrode distances to record activations during VF. Because bipolar recordings are associated with narrower electrograms than are unipolar recordings, succeeding electrograms are less likely to overlap with each other, allowing for more accurate selection of activation times. We also developed methods to automatically select the times of activation and display the activation patterns dynamically on the computer screen. These methods provide better temporal resolution than the conventional isochronal activation maps, which usually cover only short periods of time.
There have been contradicting reports on the importance of the subendocardium and the Purkinje fiber network on the generation and maintenance of VF. Although we4 8 found that subendocardial ablation affected neither the vulnerability to VF nor the life span of reentrant wave fronts during Wiggers’ stage I VF, other authors9 10 have suggested that ablation of the subendocardium and Purkinje fibers may decrease ventricular vulnerability. Therefore, it was of interest to determine if subendocardial and Purkinje fiber ablation affects the characteristics of the reentrant wave fronts during Wiggers’ stage II VF. To test the importance of the subendocardium on the maintenance of VF, we8 developed methods to perform subendocardial ablation without subjecting the dogs to cardiopulmonary bypass. The purposes of the present study were to use computerized mapping techniques to study Wiggers’ stage II VF in normal and in subendocardium-ablated canine ventricles to test the hypothesis that reentrant wave fronts are present during Wiggers’ stage II VF and to determine the mechanisms by which these reentrant wave fronts terminate and regenerate. The effects of subendocardial and Purkinje fiber ablation on the core size, cycle length, life span, and incidence of the reentrant wave fronts were also studied.
Materials and Methods
The research protocol was approved by the institutional animal care and use committee of the Cedars-Sinai Medical Center and followed the guidelines of the American Heart Association. Thirteen adult mongrel dogs were anesthetized with 25 to 35 mg/kg sodium pentobarbital,11 intubated, and ventilated with room air by a respirator (Harvard Apparatus). An arterial line was inserted into the right femoral artery to continuously monitor systemic blood pressure. Blood was drawn to determine the pH, Po2, Pco2, base excess, and bicarbonate concentrations. Normal metabolic status was maintained throughout the study by correcting any abnormal values. A venous line was inserted into the femoral vein to infuse saline and to give supplemental doses of pentobarbital. Rectal temperature was monitored and maintained at 36°C to 37°C by heating the table with warm circulating water. The chest was opened through a medial sternotomy, and the heart was suspended in a pericardial cradle.
The recording plaque electrode array was constructed with stainless steel wires with a diameter of 0.4 mm. These wires were selected because they were durable, therefore decreasing the likelihood of malfunction during experiments. The wires were fully insulated except at the tips, which served as the tissue contact points. The interelectrode distance was 1.6 mm, and the interpolar distance was 0.5 mm, measured from center to center. The bipolar axes of all the bipolar pairs were aligned with the rows of the electrodes. However, because the electrode array was handmade, the alignments may not have been perfect. During the experiments, the wires were hung at the front end, which was usually 20 to 30 cm higher than the heart surface. In addition, an umbilical tape was used to hang the redundant wires in the air so that the full weight of the electrode array was not resting on the heart. In 6 dogs, mapping studies were performed with a plaque of 317 bipolar electrodes containing 21 columns. The same electrode array was later expanded to include 509 bipolar electrodes and was used in the remaining 7 dogs. An additional three channels of recording electrodes were used to register surface electrocardiograms, which were constantly monitored throughout the study.
The recording electrodes were connected to a computerized mapping system (EMAP, Uniservices).12 The electrograms were filtered with a high-pass filter of 0.5 Hz and were acquired at 1000 samples per second. The mapping system has a fixed gain of 10. The sample-and-hold technique was performed sequentially. For 1-kHz sampling rate, the skew between two consecutive channels was 2 μs. The maximal skew (skew between channels 1 and 512) was 1 ms and was not corrected. The analog-to-digital conversions were calibrated with 18-bit resolution. True 16-bit conversions were achieved. The dynamic range of the analog-to-digital convertor was between −4.5 and +4.5 V, or a total of 9 V. Dividing 9 V by 65 536 and then by the gain of 10 resulted in an input resolution of the system of 13.7 μV.
Protocol 1: Computerized Mapping of Intact Ventricles
Five dogs were used in this protocol. A 509-channel recording electrode array was sutured on the epicardial surface of the right ventricular anterior wall, 1 cm below the pulmonary conus. A bipolar pacing electrode was added to the right edge of the recording electrode array to deliver baseline pacing with 5-ms pulse widths at twice diastolic threshold current (Fig 1A⇓). Another pair of stimulation electrodes was added to the area between electrodes 156 and 178 and between electrodes 178 and 158 to deliver the premature stimulus (S2).2 After eight S1 stimuli at a cycle length of 300 ms, a strong S2 stimulus with an average strength of 73±10 mA was delivered to induce VF. The first tested S1-S2 interval was shorter than the effective refractory period of the ventricles, usually <130 ms. If VF was not induced, the S1-S2 interval was increased in 10-ms increments to scan the vulnerable period of the ventricles until VF was induced. If VF was induced, the same S1-S2 interval was used to induce two more VF episodes in that dog so that a total of three episodes of VF were analyzed in each dog.
Protocol 2: Computerized Mapping of Ventricles After Chemical Subendocardial Ablation
Six dogs were used in this protocol. The right ventricular subendocardium was ablated with Lugol solution according to methods previously reported in detail.8 Briefly, umbilical tapes were threaded around the venae cavae and the pulmonary artery in preparation for inflow and outflow occlusion. A catheter was inserted via the right atrial appendage into the right ventricular cavity. The umbilical tapes around the venae cavae and the pulmonary artery were tightened, and the right ventricular cavity was emptied by syringe. Lugol solution (20 to 30 mL) was injected into the right ventricular chamber and maintained for 10 to 20 s. Right bundle branch block always occurred immediately after the injection of the Lugol solution. Warm normal saline was then used to flush the same chamber multiple times to remove the Lugol solution. The occlusions were then released, and the dog was allowed to recover for 30 minutes, or until the blood pressure and the heart rate returned to normal. The total duration of the inflow and the outflow occlusions was ≈2 minutes.
A 317-channel recording electrode array was then sutured on the epicardial surface of the right ventricular anterior wall, 1 cm below the pulmonary conus. Eight pacing wires, 3 mm apart, were sutured to the right edge of the recording plaque. Baseline (S1) unipolar cathodal pacing using a 10-mA 5-ms stimulus was delivered simultaneously from these pacing electrodes, with the chest wall used as the anode, to create planar activation wave fronts (Fig 1B⇑).3 4 To deliver the strong premature stimulus (S2), a patch electrode measuring 3.16 by 0.85 cm was sutured to the upper edge of the plaque. After eight S1 stimuli at a cycle length of 300 ms, a second channel of the programmable stimulator was used to deliver a premature stimulus to a high-voltage stimulator (HVS-02, Ventritex). The S2 was used as an external signal to trigger the immediate delivery from the HVS-02 of a 6-ms 50-V truncated exponential shock to the patch electrode on the edge of the plaque electrode array to induce VF.3 4
Protocol 3: Computerized Mapping of Ventricles Before and After Subendocardial Ablation
Because different electrode arrays were used in the first two protocols, the ability to discern and quantify reentrant wave fronts may be affected; thus, comparing the incidence of reentry between the first two groups may not be accurate. Therefore, two additional dogs were studied. Also, since it is possible that the transient global ischemia inflicted during the ablation procedure may account for any differences seen, one of the two dogs was mapped under control conditions and then again after a sham “ablation” procedure. In each dog, the same 509-channel electrode array was sutured on the right ventricular outflow tract and free wall. Both rectal temperature and epicardial temperature were monitored during the study. In both of these dogs, the temperature differences between these two sites was <1°C. This finding is consistent with those previously reported using the same model.8
VF was induced by rapid ventricular pacing from the left ventricular apex with a cycle length of 100 ms and a duration of 3 to 5 s. Data acquisition started 3 s after the onset of VF for a total of 8 s. After a total of 15 fibrillation/defibrillation episodes, the dogs were prepared for subendocardial ablation as described in protocol 2. In one dog, the subendocardium was ablated with Lugol solution. In the second dog, the same procedure was followed, but normal saline was used instead of the Lugol solution. After the procedure, 15 episodes of VF were mapped in each dog. The dogs were then euthanized, and the mapped tissues were removed for histological examination.
For all three protocols, two patch defibrillation electrodes with an active surface area of 13.5 cm2 (CPI) were sutured to the right and the left ventricular epicardia, distant from the recording electrode array, to deliver rescue shocks within 10 s after the induction of VF. A 4-minute interval was allowed to lapse between each fibrillation/defibrillation episode.
Selection of Activations
The purpose of the present study was to determine the patterns of activation during Wiggers’ stage II VF; therefore, 3 to 5 s of data were analyzed beginning 2.5 s after the onset of VF in protocols 1 and 2, and the initial 4 s of data of each episode were analyzed beginning 3 s after the onset of VF for protocol 3. We began data analysis 2.5 to 3 s after S2 because the reentrant wave fronts initiated by the S2 might persist for 0.15 to 2.75 s.2 At 2.5 to 3 s after S2, we hoped that all reentrant wave fronts directly initiated by the S2 would have died out and that all observed reentry was spontaneously generated during VF.
The time of activation was taken as the time of the fastest slope (dV/dt) of each electrogram (Fig 2⇓).2 13 14 15 The maximal dV/dt in the window of data analysis was first determined by the computer. The investigators then had the option to choose a threshold dV/dt value (a percentage of the maximal dV/dt) and a threshold interval (in milliseconds). In the example shown in Fig 2⇓, the threshold values were 20% and 50 ms, respectively. The computer selected a time as the time of local activation if the dV/dt at that time exceeded the threshold value and if the interval between that time and the time of the previous activation exceeded the threshold interval. Because each channel has a different signal-to-noise ratio, the threshold values selected varied from channel to channel at the investigator’s discretion. The vertical lines in Fig 2B⇓ indicate the times, in milliseconds, of selected activations, with the induction of VF by S2 as time 0.
The computers were not 100% specific and sensitive in selecting activations, as demonstrated in Fig 2B⇑. It is an efficient method to help investigators to detect reentrant wave fronts in VF. However, because of the presence of the noise and artifacts, this method is not suitable for quantitative analysis of the initiation and termination events of the reentrant wave fronts. Therefore, manual editing was always performed for each activation for each episode of VF. The arrows in Fig 2B⇑ point to deflections that will be manually selected as activations. If two deflections (double potentials) were observed, both deflections were selected as activations regardless of the duration of the isoelectric interval (Fig 2B⇑, thick arrow). Therefore, some activations selected may have been electrotonic in origin and did not represent true local activations. The advantage of selecting all deflections as activations is that the investigators did not have to apply artificial criteria to reject or to accept a deflection as a local activation. Dynamic displays of the activation patterns were then visualized on a computer screen in which each electrode site was illuminated when an activation was registered.
The disadvantage of the above method is that it is difficult to accurately convey to a reader the dynamic activation patterns with a limited number of still-frame figures. Therefore, for the purposes of illustration, it was necessary to also construct conventional isochronal activation maps.2 5 The criteria for selecting local activations for conventional isochronal map generation have been reported elsewhere.4 16 Briefly, for the biphasic and the multiphasic waveforms, the maximal slope of the activation complex was selected by the computer to be the time of activation, and only one activation time was assigned to the entire complex if no isoelectric period was present within the complex. If the activation complexes were monophasic with a single maximum or minimum, the time of activation was assigned to be at the peak of the maximal deflection.13 Only one activation was selected in the channel with multiple activations to represent the time of these multiple complexes. The activation that we selected was the one that was the largest and had the steepest slope among all the neighboring activations. This activation time was then used to match the activations on the other channels, thus generating the isochronal map. For channel 368, shown in Fig 2B⇑, the deflection marked by the vertical line 6519 had a steeper slope than the deflection marked by the thick arrow. Therefore, the time 6519 was selected to match the activation time 6482 on channel 369 for the purpose of generating isochronal maps. The isochronal maps were then compared with the dynamic display to ensure that the isochronal maps adequately represented the activation patterns shown in the dynamic display.
After manual editing of the activation times, a dynamic display of the activation patterns of each episode of VF was visualized on a computer screen in which each electrode site is represented by a dot. For each activation registered, a software program directs the corresponding dot on the computer monitor to illuminate. The dot initially illuminates red, then yellow, followed by green, light blue, and finally dark blue. The changing color of each dot acts as an aid in the visual identification of the direction of wave front propagation, particularly when the dynamic display is being observed in fast motion. The total duration of the illumination of one dot by one activation time was manually preset at 50 ms. Therefore, each color persisted for 10 ms. The 50-ms duration was selected because a previous study4 showed that the refractory period of ventricular cells during VF ranged from 48 to 77 ms. Therefore, the 50-ms illumination period approximates the refractory period of the propagating wave front. This period, however, can be set at any value by the investigator. Occasionally, the red color could appear twice within 50 ms of each other on the same channel. This phenomenon was usually due to recording of double potentials on the same channel, which occurred as a result of wave front collision or when the channel was near the core of reentry. The speed of the dynamic display can also be selected by the investigators, usually at 1/15th of the speed as it actually occurred in vivo. The investigator can also advance the dynamic display at fixed time intervals to observe the activation sequence one frame at a time.
A reentrant wave front was defined as a wave front that completed a circular pathway and reentered the area of origin. It was characterized by the juxtaposition of early and late sites on the isochronal map. However, because the core of reentrant wave fronts meandered from beat to beat, the reentrant wave front may not return to the exact same point from which it originated. All other VF wave fronts were classified as either wave fronts that originated from outside the mapped area or as wave fronts that originated from within the mapped area.5 The former wave fronts were characterized by an early site at the edges of the mapped tissue, followed by propagation toward the center. The latter wave fronts had early sites within the mapped region, followed by propagation toward the periphery.
Determining the Core of Reentry
The approximate location of the core of the reentrant wave fronts was first identified by dynamic display. Fig 3⇓ illustrates the method used to determine the core area. The red color indicates the electrodes that were activated within the previous 10 ms. Panels A through I are selected frames from the dynamic display of the activation patterns during VF. The frames are separated by variable intervals. The time, in milliseconds, of each frame is shown in parentheses above the panels; the beginning of data acquisition was time 0. The reentrant wave front rotates in a clockwise direction. Each electrode that was nearest to the core and on the leading edge of the wave front was marked. The reentry was then advanced in 5-ms intervals, and the leading edge of the reentrant wave front was marked each time when an appropriate point showed up. These points may show up as close as 5 ms apart, as shown in panels B and C. However, sometimes >20 ms passed without a red dot showing near the core. These marked electrodes were then connected by lines as shown. Because the core in this example meandered rightward, the loop was not closed in panel D at the same point it began in panel A. The size of the core was estimated from the lines that encircled an area (panel D). This loop represents the perimeter of the core of the first reentrant cycle. Panels E through I show the core identified for the second reentrant cycle. The myocardial fiber orientation is displayed by the arrow at the bottom of the figure. Since the electrical activations were analyzed before any knowledge about the location of the core, this was a blinded method of determining the perimeter. Once the perimeter of the core was identified, the computer counted the number of electrodes encircled by this perimeter. Because the interelectrode distance was 1.6 mm, each electrode represents an area of (1.6 mm)2, or 2.56 mm2. The core size was then calculated by the product of 2.56 mm2 and the number of electrodes encircled.
At the conclusion of the experiments, the dogs were euthanized by an overdose of pentobarbital. The electrode array was removed, and the underlying tissue was excised from the rest of the heart and fixed in 10% buffered formalin solution. A horizontal section was obtained 1 mm from the epicardium to determine the presence, if any, of an anatomic barrier. Transmural sections were also taken to evaluate the effect of Lugol solution on the subendocardial tissue. All tissue samples were processed routinely and embedded in paraffin. Five-micron-thick sections were cut and stained with hematoxylin and eosin for light microscopic evaluation.
All statistical analyses were performed using SYSTAT.17 Results are expressed as the mean±SD. Student’s t tests were used to compare the mean cycle lengths, life spans, core sizes, and incidences of reentry of the reentrant wave fronts before and after ablation of the subendocardium and the Purkinje fibers. The null hypothesis was rejected at a value of P≤.05.
The dogs weighed 24.5±3.8 kg in protocol 1, 22.1±1.0 kg in protocol 2, and 23.6±4.8 kg in protocol 3.
Activation Patterns During Baseline Pacing
Fig 1A⇑ shows the patterns of activation after an S1 stimulus (black arrow) in an intact ventricle. Because subendocardial Purkinje fibers were intact, tissue located distant from the S1 site was activated by transmural wave front propagation, as evidenced by the presence of epicardial breakthrough sites with activation times earlier than that of the surrounding tissue (white arrows). Similar findings have been reported by other investigators.15 18 Fig 1B⇑ shows the pattern of activation after an S1 stimulus in an ablated ventricle. The wave front is planar because the pacing stimulus was delivered simultaneously by eight pacing wires located 3 mm apart along the right edge of the electrode plaque (left edge of panel). The tissue distant from the S1 site did not have recording sites that activated earlier than the surrounding electrodes. This observation is compatible with successful subendocardial ablation.8 There was no evidence of conduction block in the mapped region during S1 pacing in either group of dogs.
Reentrant Wave Fronts in Wiggers’ Stage II VF in Intact Ventricles
A total of 45 episodes of VF were analyzed. Among them, 15 episodes were from protocol 1, and 15 episodes were from each of the two dogs from protocol 3 before subendocardial ablation. In all episodes, the reentry initially induced by the S2 or rapid ventricular pacing terminated because no reentry was observed at the beginning of our data analysis window. Among the 45 runs of VF mapped in dogs with intact ventricles, 12 episodes of reentrant wave fronts were observed in the mapped region. In 5 episodes, the reentry was clockwise, and in 7 episodes, it was counterclockwise. The mean number of rotations (life span) was 3.2±1.2. The mean cycle length was 107±8 ms. Fig 4⇓ shows an example of a reentrant wave front during Wiggers’ stage II VF. A clockwise reentrant wave front was seen at the right edge of the figure, corresponding to the dog’s left side, and lasted three cycles. The two panels in Fig 4⇓ show different isochronal line locations near the core, indicating that the core meanders. Fig 5⇓ shows the actual activations recorded during reentrant excitation. The electrogram recorded on channel 313 in the first cycle (marked time 0) was upright and in the next cycle (marked time 94 ms) was downward. These changes indicate that the wave front might have traveled in different directions in these two consecutive cycles. This could be explained by the meandering nature of the core, which was on one side of electrode 313 in the first cycle and then on the other side of the electrode in the subsequent cycle.
Reentrant Wave Fronts in Wiggers’ Stage II VF in Ablated Ventricles
A total of 33 episodes of VF were mapped. Among them, 18 episodes were from protocol 2, and 15 episodes were from protocol 3. In all episodes, the reentry immediately induced by the S2 terminated because no reentry was observed at the beginning of the data analysis window. In 17 of 33 episodes of VF, reentrant wave fronts were observed in the mapped region. In 7 episodes, the reentry was clockwise, and, in 10 episodes, it was counterclockwise. The mean number of rotations (life span) before termination was 3.4±1.1 (P=.83 when compared with the group with intact ventricles). The mean cycle length was 111±12 ms (P=.12).
Fig 6⇓ shows consecutive isochronal maps of one reentrant wave front. The wave front rotated in the counterclockwise direction and lasted four cycles. Fig 7⇓ shows electrograms along the circular pathway of the reentrant wave front shown in Fig 6⇓. The arrows point to the direction of wave front propagation of the four reentrant cycles.
Incidence of Reentry
The incidence of reentry, defined as the mean number of rotations per episode (4 s) of VF analyzed, was 1.2±1.6 under baseline conditions and 1.7±1.6 after chemical ablation (P=.19). There was also no difference in the incidence of reentry between baseline conditions and after the sham ablation procedure with normal saline (0.5±1.1 versus 0.7±1.2, P=.55).
In Fig 3⇑, note that the core has its long axis parallel to the fiber orientation, a phenomenon that was observed in 19 episodes of reentry. In the remaining 10 episodes, the shape of the core did not manifest an apparent long axis. The mean core size of reentry was 29±18 mm2 in intact ventricles and 25±19 mm2 in ablated ventricles (P=.47).
The Core Meanders
To demonstrate that the core meanders, we studied the path of the core by connecting the inner electrodes on the leading edge of consecutive reentrant excitations. Fig 8A⇓ shows the path of the leading inner point of a reentrant wave front for three consecutive reentrant excitations. The numbers indicate the sequences of activations. This figure demonstrates that the core of reentry is not stationary and fixed but instead meanders from one cycle to the next. This phenomenon was observed in each episode of reentry. Fig 8B⇓ shows electrograms in or near the core. The electrograms from top to bottom correspond to the white dots from left to right, respectively, in Fig 8A⇓. Note that electrograms 138 to 141 were always outside of the core, hence registering large and single bipolar electrograms. Electrodes 142 and 143 were initially inside the core and thus registered low-amplitude slowly rising bipolar electrograms. However, the core meandered toward the dog’s left side (right edge of panel) in the third reentrant excitation. As a result, electrodes 142 and 143 were no longer within the core and thus registered large and single bipolar electrograms (arrows). These findings demonstrate that at a given electrode site, the core meanders in and out over short periods of time.
Spontaneous Initiation of Reentrant Wave Fronts
Analysis of the dynamic display of the activation patterns during the initiation or generation of new reentrant wave fronts revealed a consistent pattern. In each case, a reentrant wave front was initiated by a wave front crossing roughly perpendicular to the tail of another wave front. Fig 9⇓ shows selected frames from a dynamic display that illustrates this phenomenon. Panel A shows a wave front that propagates from the top to the bottom (filled arrow). The unfilled arrow points to a second wave front, which is propagating from the left edge of the mapped tissue rightward. The perpendicular intersection of these two wave fronts created a wave break (asterisk, panel B), followed by reentry (panels C through F). Fig 10⇓ shows the actual activations registered. The lower four electrodes registered the wave front at the bottom of Fig 9A⇓. Electrodes 9, 34, 80, and 102 registered the wave front at the top of Fig 9A⇓. The activations recorded in electrode 102 show that the top wave front propagated to this area 50 ms after the lower wave front activated the same area. The interval between these two deflections (50 ms) was the “intersection interval.” Fig 10⇓ also shows that after the first activation, the activation sequences have changed. For example, the sequence of activation was 141-181-220 in the first activation. This sequence was changed to 220-181-141 in the third activation. These changes are compatible with the creation of a reentrant wave front shown in Fig 9B⇓ through 9F.
To determine whether or not the intersection interval is randomly distributed throughout the VF cycle, we plotted the intersection intervals, which are shown in Fig 11⇓. Twenty-three episodes (9 in intact ventricles and 14 in ablated ventricles) were included. Six episodes could not be determined because the initiation sequences occurred near the edge of the plaque. This intersection interval did not randomly distribute throughout the VF cycle length. Rather, there was a clustering of the interval between 40 and 90 ms, with a mean of 69±11 ms. This clustering indicates that a vulnerable period is present in the VF wave front. Interaction of two wave fronts during this vulnerable period results in the initiation of reentrant excitation.
Termination of Reentrant Wave Fronts
Analysis of the dynamic display data revealed that in 19 episodes, the reentrant wave fronts terminated as the result of interference by an invading wave front, and in 7 episodes, it terminated spontaneously. In 3 episodes, the mode of termination is unknown because the reentrant wave front meandered off the mapped region before it terminated. These findings are similar to those found for the reentrant wave fronts in Wiggers’ stage I VF.4 Both modes of termination were observed in intact and ablated ventricles.
Fig 12⇓ is an example of spontaneous termination. It shows selected frames from a dynamic display of a reentrant wave front. The arrows are immediately in front of the leading edge of the wave front and point in the direction of propagation. Panels A through H demonstrate clockwise reentry. Panels I and J show the wave front terminating spontaneously. This termination is quickly followed (panels K and L) by a wave front from outside the mapped area propagating across the mapped region.
Fig 13⇓ is an example of reentry terminated by interference. It shows selected frames from the dynamic display. Panels A through C show clockwise reentry. The closed arrows are immediately in front of the wave front and point in the direction of propagation. Panels B and C show another wave front (open arrows) propagating from the top of the mapped region downward. In Panel D, this wave front invades the reentrant wave front, terminating the reentry. Panels E and F show that reentrant activity has ceased.
Fig 14⇓ shows the actual activations shown in Fig 13⇑. Two wave fronts are seen. The lower five channels registered the activations of the reentrant excitation. The upper three channels registered outside interference. Initially, both wave fronts arrive at channel 229 almost simultaneously. The upper wave front (channels 124, 164, and 226) was not able to preexcite the cells near channel 229, allowing reentry to continue. However, the seventh activation of the upper wave front arrived early and preexcited channels 229 (arrow a) and 293 (arrow b), resulting in abrupt termination of reentry. The subsequent activations are much more synchronized, indicating the absence of reentry. These findings indicate that an excitable gap is present in the reentrant wave fronts during Wiggers’ stage II VF and that the presence of this excitable gap contributes to the termination of reentry.
Nonreentrant Wave Fronts
Because reentrant excitation was an uncommon occurrence, most wave fronts on the epicardium did not appear to participate in reentry. By and large these nonreentrant wave fronts were observed to enter the mapped region from outside the mapped area (Fig 15A⇓). In both intact and subendocardium-ablated ventricular epicardium, the wave fronts arrived at the mapped region from all directions. There was no apparent preference of the wave fronts to travel either along or across the fiber orientation. A significant number of wave fronts had the earliest activation site within the mapped region, most likely representing the epicardial breakthrough of transmural activation (Fig 15B⇓). The incidence of wave fronts with early sites within the mapped area was 0.63±0.19 times per second of data analyzed per area of epicardium mapped in intact ventricles. The incidence of wave fronts originating within the mapped area during VF was significantly lower in ablated ventricles (0.25±0.17 times per second per square centimeter, P=.001).
No anatomic barriers were present in any of the tissue specimens. In transmural sections of subendocardium-ablated ventricles, the Purkinje fibers and the adjacent subendocardial contractile myofibers were necrotic. The layer of necrotic subendocardial myocardial cells approximated a zone of up to six or seven myocardial cells, or roughly a depth of 0.5 mm (Fig 16⇓). These histological findings are essentially the same as those reported previously.8
Reentrant Wave Fronts in Wiggers’ Stage II VF
Using computerized mapping techniques, we tested the hypothesis that reentrant wave fronts are present during Wiggers’ stage II VF. The results of the present study demonstrate the presence of reentrant wave fronts during stage II VF in both intact and ablated canine ventricles. Baseline pacing showed no evidence of conduction block in the areas studied, and histological examination revealed no evidence of anatomic barriers. Therefore, the reentrant wave fronts observed in the present study are instances of functional reentry. These functional reentrant wave fronts, or rotors,19 provide the constant source of activation required to maintain VF.
Mechanism of Spontaneous Regeneration of Reentrant Wave Fronts
The present study also showed that these reentrant wave fronts had a very limited life span, while VF persisted. Therefore, it was important to determine how the reentrant wave fronts spontaneously regenerate during the course of VF. We demonstrated that during stage II VF, new reentrant wave fronts are spontaneously initiated by a wave front crossing roughly perpendicularly to the tail of a preceding wave front. It appears that both the relative direction and the relative timing of the two intersecting wave fronts are important in the initiation of reentry. The direction is important because all initiating events were associated with two wave fronts that were roughly perpendicular to each other. The timing is important because the two wave fronts must intersect at a critical interval. Fig 17⇓ schematically illustrates this concept. Panel A shows two wave fronts intersecting at roughly a 90° angle to each other. Panel B shows that the wave break occurs when the lower part of the wave front meets refractory tissue, while the upper part of the same wave front continues to propagate in its original direction. Panel C shows that reentry occurs when the tissue that was refractory in panel B is repolarized. These findings, that a wave break sets the stage for the induction of reentry, are consistent with those shown by Pertsov et al,20 who used in vitro preparations. Because the wave break occurred only when the intersection interval was within a narrow range of 40 to 90 ms, we propose that a vulnerable period is present in the wave front, during which an electrical stimulus provided by a second wave front, invading at a right angle, can induce reentry.
The vulnerable period estimated in the present study was determined by analyzing the wave break induced by the interaction of two propagating wave fronts. However, previous studies of the vulnerable period during regular rhythm were usually performed with premature electrical stimulations.21 22 23 24 25 Future studies using a timed premature stimulus of known current strength will be helpful in more accurately determining the vulnerable period and current strength that are needed to induce reentrant wave fronts during VF.
Strength of the Stimulus and Induction of Reentry
It has been demonstrated that a vulnerable period is present during sinus or paced rhythm, during which an electrical stimulus can induce functional reentry and VF.2 3 21 23 Computer simulation studies20 26 and experiments using in vitro preparations20 have also shown that an electrical stimulus can induce functional reentry. A major difference between the in vivo and the in vitro experiments was the strength of the stimulus required to induce reentry. Whereas the electrical stimulus used to induce reentry in thin in vitro preparations was of a low strength (two to five times threshold),20 the induction of reentry in the intact ventricles requires a very high current strength, often exceeding 100 times the pacing threshold.24 27 The inability of a small electrical stimulus to induce reentry in vivo, in contrast to in vitro, raises doubts about the relevance of the computer simulation studies19 and the in vitro studies20 to the induction of reentry in intact hearts. Therefore, the result of the present study, which indicates that reentry during Wiggers’ stage II VF in vivo can be initiated by a stimulus no stronger than that provided by a propagating wave front, is significant. Because many wave fronts are present at any given time during VF, it is highly likely for one wave front to interact during its vulnerable phase with another wave front at a right angle, thereby inducing reentry.
We do not know why a relatively small electrical stimulus provided by the propagating wave front could induce reentry during VF but not during sinus or paced rhythm. One possible explanation is that the action potential duration during VF is much shorter than that during sinus rhythm.28 Because a shortening of the action potential duration promotes reentry in vitro,29 the same mechanism may facilitate the induction of reentry during Wiggers’ stage II VF.
Mechanisms of Termination of Reentrant Wave Fronts
The present study found that the termination of reentrant wave fronts occurred either spontaneously or as the result of invasion of the reentrant circuit by an outside wave front. These findings are similar to those found for Wiggers’ stage I VF.2 4 These findings indicate that an excitable gap is present in the reentrant wave fronts during Wiggers’ stage II VF and that the presence of this excitable gap contributes to the termination of reentry. The multiple modes of termination may explain, in part, why the reentrant wave fronts have notably short life spans during Wiggers’ stage II VF.
Effect of Subendocardial Ablation
Although we4 8 found that subendocardial ablation affected neither the vulnerability to VF nor the life span of reentrant wave fronts during Wiggers’ stage I VF, other authors9 10 have suggested that ablation of the subendocardium and Purkinje fibers may decrease ventricular vulnerability. Therefore, it was of interest to determine if subendocardial and Purkinje fiber ablation affects the incidence or characteristics of the reentrant wave fronts during Wiggers’ stage II VF. The findings of the present study, however, do not support this hypothesis. Our results showed no significant change in the incidence, life span, cycle length, or the mean core size of the reentrant wave fronts after subendocardial ablation. These data indicate that the generation and the maintenance of reentry do not depend on an intact subendocardium.
Spiral waves of excitation have been observed in disparate dynamic systems, including physical,30 chemical,31 and biological32 33 systems. By assuming that the heart behaves like a generic nonlinear excitable medium, theoretical biologists have predicted spiral waves (scroll waves in three dimensions) of excitation as the mechanism of functional reentry during VF in the heart.34 In vitro experimental mapping studies using normal two-dimensional isolated canine and sheep epicardial muscle20 35 and diseased human ventricular tissue29 have revealed that reentrant spiral waves of excitation can be induced by critically timed premature stimuli.
In the present study, we demonstrated that many of the characteristics of reentrant activities during Wiggers’ stage II VF dovetail with the theory of spiral waves. The generation of new reentrant wave fronts by critically timed wave fronts that intersect at roughly right angles to each other is compatible with the theory of spiral waves.20 26 The termination of the reentrant wave fronts by invading wave fronts demonstrated in the present study provides evidence for the presence of an excitable gap, which is also consistent with spiral wave theory.20 36 Furthermore, we have demonstrated that the cores of the reentrant wave fronts meander. A meandering core is considered a characteristic feature of rotors37 whose two-dimensional manifestation are spiral waves. Although these features are consistent with spiral wave theory, the reentrant wave fronts shown in Figs 4⇑ and 6⇑ do not demonstrate wave fronts spiraling outward from a central core. Rather, they appear to travel around a line of block, not inconsistent with the leading circle morphology first reported by Allessie et al.38 The mechanism by which reentrant wave fronts display a true spiral morphology in in vitro ventricular tissue preparation20 35 but not during VF in intact ventricles is unknown. This discrepancy leaves open the question about the fundamental nature of reentry in Wiggers’ stage II VF.
Limitations of the Study
There are some important limitations of the present study. Because transmural or three-dimensional patterns of activation were not mapped, the incidence of reentrant wave fronts reported in the present study is valid only for the right ventricular epicardium. The incidence of reentry may be different if transmural mapping is performed or if the left ventricle is mapped. Indeed, it is likely that the incidence of reentry is underestimated because transmural activation patterns were not mapped. For example, some of the nonreentrant wave fronts observed on the epicardium could be part of a transmural reentrant circuit.
The demonstration of reentrant activities does not necessarily invalidate the multiple wavelet hypothesis of fibrillation.39 Although the reentrant activities may serve as the source of activation during VF, the generation of these reentrant wave fronts depends on the critical interaction of the wavelets. Therefore, it is reasonable to postulate that both reentrant wave fronts and the presence of multiple wavelets are important in the maintenance of Wiggers’ stage II VF.
The present study demonstrates that functionally defined reentrant wave fronts are present during Wiggers’ stage II VF and that these rotors, which have short life spans and have meandering cores, may provide the constant source of activation that maintains VF. It also demonstrates that the spontaneous initiation of these reentrant wave fronts depends on the critical interaction between two wave fronts and that their termination occurs either spontaneously or as the result of interference. Finally, the present study demonstrates that subendocardial and Purkinje fiber ablation does not affect the incidence, life span, cycle length, or the core size of the reentrant wave fronts.
This study was performed during the tenure of an ACC/Merck Fellowship (Dr Hwang) and an AHA/Wyeth-Ayerst Established Investigatorship Award (Dr Chen) and was supported in part by a Specialized Center of Research (SCOR) Grant for Sudden Death (HL-52319) and a FIRST Award (HL-50259) from the National Institutes of Health, by an American Heart Association National Center Grant-in-Aid (92009820), by the Electrocardiographic Heartbeat Organization, and by the Ralph M. Parsons Foundation. The authors wish to thank Drs James S. Forrester and Prediman K. Shah for their support, Avile McCullen and Meiling Yuan for technical assistance, and Elaine Lebowitz for secretarial assistance.
- Received May 22, 1995.
- Accepted December 18, 1995.
- © 1996 American Heart Association, Inc.
Wiggers CJ, Bell JR, Paine M. Studies of ventricular fibrillation caused by electric shock, II: cinematographic and electrocardiographic observation of the natural process in the dog’s heart: its inhibition by potassium and the revival of coordinated beats by calcium. Am Heart J. 1930;5:351-365.
Chen P-S, Wolf P, Dixon EG, Danieley ND, Frazier DW, Smith WM, Ideker RE. Mechanism of ventricular vulnerability to single premature stimuli in open chest dogs. Circ Res. 1988;62:1191-1209.
Frazier DW, Wolf PD, Wharton JM, Tang ASL, Smith WM, Ideker RE. Stimulus-induced critical point: mechanism for electrical initiation of reentry in normal canine myocardium. J Clin Invest. 1989;83:1039-1052.
Cha Y-M, Birgersdotter-Green U, Wolf PL, Peters BB, Chen P-S. The mechanism of termination of reentrant activity in ventricular fibrillation. Circ Res. 1994;74:495-506.
Chen P-S, Wolf PD, Melnick SD, Danieley ND, Smith WM, Ideker RE. Comparison of activation during ventricular fibrillation and following unsuccessful defibrillation shocks in open-chest dogs. Circ Res. 1990;66:1544-1560.
Johnson EE, Rollins DL, Wolf PD, Smith WM, Ideker RE. Mechanism of ventricular fibrillation as mapped with 524 closely spaced simultaneously recorded epicardial electrodes. Circulation. 1992;86(suppl I):I-820. Abstract.
Chen P-S, Wolf PL, Cha Y-M, Peters BB, Topham SL. Effects of subendocardial ablation on anodal supernormal excitation and ventricular vulnerability in open-chest dogs. Circulation. 1993;87:216-229.
Damiano RJ, Smith PK, Tripp HF, Asano T, Small KW, Lowe JE, Ideker RE, Cox JL. The effect of chemical ablation of the endocardium on ventricular fibrillation threshold. Circulation. 1986;74:645-652.
Janse MJ, Wilms-Schopman F, Wilensky RJ, Tranum-Jensen J. Role of the subendocardium in arrhythmogenesis during acute ischemia. In: Zipes DP, Jalife J, eds. Cardiac Electrophysiology and Arrhythmias. Orlando, Fla: Grune & Stratton; 1994:353-362.
Bonometti C, Hwang C, Hough D, Lee JJ, Fishbein MC, Karagueuzian HS, Chen P-S. Interaction between strong electrical stimulation and reentrant wavefronts in canine ventricular fibrillation. Circ Res. 1995;77:407-416.
Funada T, Iwase T, Iwa T. Method for computerized display of epicardial maps. Biol Eng Comput. 1983;21:418-423.
Chen P-S, Shibata N, Wolf P, Dixon EG, Danieley ND, Sweeney MB, Smith WM, Ideker RE. Activation during ventricular defibrillation in open-chest dogs: evidence of complete cessation and regeneration of ventricular fibrillation after unsuccessful shocks. J Clin Invest. 1986;77:810-823.
Frazier DW, Krassowska W, Chen P-S, Wolf PD, Danieley ND, Smith WM, Ideker RE. Transmural activations and potentials on three-dimensional anisotropic canine myocardium. Circ Res. 1988;63:135-146.
Chen P-S, Cha Y-M, Peters BB, Chen LS. Effects of myocardial fiber orientation on the electrical induction of ventricular fibrillation. Am J Physiol. 1993;264:H1760-H1773.
Wilkinson L. SYSTAT: The System for Statistics. Evanston, Ill: SYSTAT Inc; 1989.
Rawling DA, Joyner RW, Overholt ED. Variations in the functional electrical coupling between the subendocardial Purkinje and ventricular layers of canine left ventricle. Circ Res. 1985;57:252-261.
Winfree AT. Rotors, fibrillation, and dimensionality. In: Holden AV, Panfilov AV, Hogeweg P, eds. Whole Heart Modeling. Manchester, England: Manchester University Press; 1994:1-41.
Pertsov AM, Davidenko JM, Salomonsz R, Baxter WT, Jalife J. Spiral waves of excitation underlie reentrant activity in isolated cardiac muscle. Circ Res. 1993;72:631-650.
King BG. The Effect of Electric Shock on Heart Action With Special Reference to Varying Susceptibility in Different Parts of the Cardiac Cycle. New York, NY: Aberdeen Press, Columbia University; 1934. Thesis.
Han J, Moe GK. Nonuniform recovery of excitability in ventricular muscle. Circ Res. 1964;14:44-60.
Wiggers CJ, Wegria R. Ventricular fibrillation due to single, localized induction and condenser shocks applied during the vulnerable phase of ventricular systole. Am J Physiol. 1940;128:500-505.
Hoffman BF, Gorin EF, Wax FS, Siebens AA, Brooks CM. Vulnerability to fibrillation and the ventricular-excitability curve. Am J Physiol. 1951;167:88-94.
Shumway NE, Johnson JA, Stish RJ. The study of ventricular fibrillation by threshold determinations. J Thorac Surg. 1957;34:643-653.
Cha Y-M, Peters BB, Birgersdotter-Green U, Chen P-S. A reappraisal of ventricular fibrillation threshold testing. Am J Physiol. 1993;264:H1005-H1010.
Zhou X, Wolf PD, Rollins DL, Afework Y, Smith WM, Ideker RE. Effects of monophasic and biphasic shocks on action potentials during ventricular fibrillation in dogs. Circ Res. 1993;73:325-334.
Hwang C, Karagueuzian HS, Fishbein MD, Czer L, Trento A, Weiss JN, Chen P-S. Demonstration of spiral waves of excitation in human ventricular tissue. Circulation. 1994;90(suppl I):I-466. Abstract.
Winfree AT. Spiral waves of chemical activity. Science. 1972;175:634-636.
Lechleiter J, Girard S, Peralta E, Clapham D. Spiral calcium wave propagation and annihilation in Xenopus laevis oocytes. Science. 1991;252:123-126.
Winfree AT. Rotors in normal ventricular myocardium. Proc Kon Ned Akad V Wetensch. 1990;93:513-536.
Davidenko JM, Pertsov AM, Salomonsz R, Baxter W, Jalife J. Stationary and drifting spiral waves of excitation in isolated cardiac tissue. Nature. 1991;355:349-351.
Winfree AT. How does ventricular tachycardia decay into ventricular fibrillation? In: Shenasa M, Borggrefe M, Breithardt G, eds. Cardiac Mapping. Mt Kisco, NY: Futura Publishing Co; 1993:655-682.
Allessie MA, Bonke FI, Schopman FJ. Circus movement in rabbit atrial muscle as a mechanism of tachycardia, III: the ‘leading circle’ concept: a new model of circus movement in cardiac tissue without the involvement of an anatomical obstacle. Circ Res. 1977;41:9-18.
Moe GK, Rheinboldt WL, Abildskov JA. A computer model of atrial fibrillation. Am Heart J. 1964;64:200-220.