Evolution of activation patterns during long-duration ventricular fibrillation in pigs

Abstract

Quantitative analysis has demonstrated five temporal stages of activation during the first 10 min of ventricular fibrillation (VF) in dogs. To determine whether these stages exist in another species, we applied the same analysis to the first 10 min of VF recorded in vivo from two 504-electrode arrays, one each on left anterior and posterior ventricular epicardium in six anesthetized pigs. The following descriptors were continuously quantified: 1) number of wavefronts, 2) wavefront fractionations, 3) wavefront collisions, 4) repeatability, 5) multiplicity index, 6) wavefront conduction velocity, 7) activation rate, 8) mean area activated by the wavefronts, 9) negative peak rate of voltage change, 10) incidence of breakthrough/foci, 11) incidence of block, and 12) incidence of reentry. Cluster analysis of these descriptors divided VF into four stages (stages i-iv). The values of most descriptors increased during stage i (1–22 s after VF induction), changed quickly to values indicating greater organization during stage ii (23–39 s), decreased steadily during stage iii (40–187 s), and remained relatively unchanged during stage iv (188–600 s). The epicardium still activated during stage iv instead of becoming silent as in dogs. In conclusion, during the first 10 min, VF activation can be divided into four stages in pigs instead of five stages as in dogs. Following a 16-s period during the first minute of VF when activation became more organized, all parameters exhibited progressive decreased organization. Further studies are warranted to determine whether these changes, particularly the increased organization of stage ii, have clinical consequences, such as alteration in defibrillation efficacy.

ventricular fibrillation (VF) is the initial recorded rhythm in many cardiac arrest patients (4). Based on the temporal progression of VF, it has recently been proposed that VF can be divided into three phases, each of which should receive a different therapy (4, 41). The first phase, the electrical phase, begins at VF onset and lasts ∼4 min. During this phase, a defibrillation shock should be given immediately (41). The second phase, the circulatory phase, lasts from ∼4 to 10 min after VF onset. Recent evidence (4, 41) suggests that survival is improved during this phase if chest compressions are performed for 1–3 min before giving the defibrillation shock. The third phase, the metabolic phase, begins after ∼10 min of VF. Therapy for this phase is still being developed. The duration of these phases is not known in animals.

Basing treatment on the duration of VF requires development of a reliable method to determine the time elapsed since VF onset. In addition, quantitative knowledge of the temporal evolution of electrophysiological characteristics during VF may provide information that could aid in the development of new therapies as well as to increase understanding of the underlying mechanisms that determine why each stage responds to different treatments.

Wiggers (43) divided VF in dogs into four stages based on mechanical motion. We (14) divided VF in dogs into five stages according to the evolution of quantitative descriptors of activation patterns. The main difference between the two classifications is an electrophysiological stage beginning ∼60 s after VF onset and lasting ∼30 s during which the activation patterns become much more organized before the progressive decrease in organization resumes (14). Huizar et al. (17) recently found a similarly organized stage during VF in vitro in isolated blood-perfused pig hearts.

The nature and progression of VF may not be the same in all species, so results in dogs or pigs should be extrapolated to humans with caution. Because of differences in ion channels, the change in action potential duration with heart rate differs among different species (3). In addition, the Purkinje fiber distribution is different in different species; it is confined to the subendocardium in dogs but extends intramurally almost to the epicardium in pigs (11). We (15) have previously shown that activation patterns are different during the first 20 s of VF in dogs and pigs. We hypothesized that the activation patterns over the first 10 min of VF, i.e., the electrical and circulatory stages, also differ between dogs and pigs. Because the in vitro procedure may alter electrophysiologic characteristics (9), the nature and progression of VF obtained from in vitro preparations should be confirmed from in vivo study. Therefore, the purpose of this study was to observe the evolution of activation patterns during long-duration VF in in vivo pig hearts and compare the results to those previously reported in in vivo dogs (14) and in vitro pigs (17). We also compared the results for the anterior vs. the posterior ventricular walls of in vivo pigs.

METHODS

Animal preparation.

The animal protocol was approved by the University of Alabama at Birmingham Institutional Animal Care and Use Committee. Six pigs (36 ± 8 kg, means ± SD, 4 female and 2 male, and 3.5–4 mo of age) were injected intramuscularly with Telazol (4.4 mg/kg), xylazine (2.2 mg/kg), and atropine (0.04 mg/kg) for anesthesia induction. Anesthesia was maintained with isoflurane in 100% oxygen by inhalation. Core body temperature, arterial blood pressure, arterial blood gases, and serum electrolytes were monitored and maintained within normal ranges throughout the study. Surface ECG lead II was monitored and recorded.

The heart was exposed through a median sternotomy and supported in a pericardial sling. A plaque containing 504 electrodes (24 × 21) with 2-mm spacing between electrodes was sutured to the posterolateral left ventricle (LV), with one edge adjacent to the posterior descending artery. A second identical plaque was sutured to the anterolateral LV with one edge adjacent to the left anterior descending artery (Fig. 1). A catheter (model 6942, Sprint; Medtronic) was inserted with electrodes in the right ventricle (RV) and superior vena cava for defibrillation.

Fig. 1.

Diagrams of the heart indicating the location of the mapping electrodes on the anterior left ventricular (LV) surface (A) and the posterior LV surface (B). Black dots represent the individual plaque recording electrodes. Directions of the x- and y-components of the conduction velocity vectors of the ventricular fibrillation (VF) wavefronts are indicated. PDA, posterior descending artery; LAD, left anterior descending artery; LVA, left ventricular apex.

VF was induced four times with a DC pulse from a 9-V battery applied to the RV. The first three episodes of VF were mapped for 20 s before a rescue shock of 15 J was given. The fourth VF episode was continually mapped for >10 min. The chest was covered with a plastic film after the fourth induction of VF to keep the epicardium warm and moist. The 20-s VF episodes served as control data in another study, whereas the last VF episode was analyzed in this study.

Mapping system and data acquisition.

Two 528-channel mapping systems were synchronized by a common clock. Each mapping system recorded from one of the two plaques. The unipolar electrograms were bandpass filtered with a high-pass filter of 0.05 Hz and a low-pass filter of 500 Hz. Data were sampled at 2 kHz and recorded with 14-bit resolution (44).

Quantification of VF.

Quantitative analysis of VF activation patterns was performed using algorithms discussed in detail elsewhere (31, 33). A five-point digital filter was used to calculate the temporal derivative of each of the 1,008 electrograms. A single temporal sample at a recording site was considered to represent an activation in the underlying tissue when the rate of voltage change (dV/dt) was <−0.3 V/s. The algorithms automatically identified activation wavefronts by grouping activations recorded at neighboring electrodes recorded within 20 ms of each other.

From the automated analysis, the following variables, which are explained in detail elsewhere (13, 31, 33), were analyzed within the mapped area every second during each 10 min VF episode: 1) number of wavefronts; 2) incidence of wavefront fractionation; 3) incidence of wavefront collision; 4) repeatability of the activation patterns; 5) multiplicity of the activation patterns; 6) mean wavefront conduction velocity (32); 7) activation rate, the mean number of activation fronts passing each epicardial site per second (42); 8) mean epicardial area swept out by each wavefront; 9) mean negative peak dV/dt of activations; 10) wavefront breakthrough/focal incidence; and 11) wavefront block incidence. The incidence of reentry was determined as the percentage of wavefront families that contained a reentrant circuit during 5-s intervals of VF. A wavefront family consisted of all wavefronts that interacted with each other through fractionation and collision.

We also calculated the directionality of VF wavefronts from the propagation velocity of each wavefront. The wavefront velocity was estimated by computing the location of the centroid of the wavefront at each time sample of the data stream (32). The velocity of each wavefront centroid was separated into x- and y-vector components. The mean weighted velocity vectors of the wavefronts in the x-direction (positive x from the lateral free wall towards the LAD for the anterior epicardium and from the posterior descending artery towards the lateral wall for the posterior epicardium, Fig. 1) and y-direction (positive y from apex towards base for both the anterior and posterior epicardium; Fig. 1) were calculated for each segment as described elsewhere (23).

Statistical analysis.

Statistical procedures described previously (14) were used to divide VF into different stages. A k-means clustering analysis (26) was employed for partitioning the 600 1-s intervals into subgroups (descriptor 12, incidence of reentry, was excluded from this analysis because it was computed for 5-s intervals). The number of clusters, k, was determined with the cubic clustering criterion (34). The maximum value across the hierarchy levels was used to suggest the optimal number of clusters in the data (34). Differences between the descriptors in each cluster were tested by ANOVA. Differences between descriptors for the anterior and posterior epicardium were tested by paired t-tests, and Bonferroni adjustment was performed for each test at the 0.004 (=0.05/12) level of significance. Differences between each descriptor for the pig anterior epicardium and for the dog anterior epicardium from our previous study (14) were compared by two-way ANOVA. Post hoc tests were followed if a difference was found by two-way ANOVA. The mean normalized rate of change of each parameter during each stage was calculated by subtracting the last value from the first value of the parameter for each stage and dividing this difference by the duration of the stage in seconds. Significance was defined as P < 0.004 for Bonferroni adjustment and <0.05 for other tests.

RESULTS

The cubic clustering analysis divided porcine VF into four stages for both anterior and posterior recordings. Examples of VF activation sequences during the four stages are shown in Fig. 2 and in animations (online Supplemental Data; Supplemental Material for this article is available online at the Am J Physiol Heart Circ Physiol website). The Supplemental Movies are from the posterior surface of the heart. The detailed data at each individual stage of all parameters are summarized in Table 1. The time course for each stage, while not identical, was similar between the anterior and posterior epicardium (Table 1). The final time course for each stage was the mean time course between the anterior and posterior epicardium from the six animals.

Fig. 2.

Snapshots of activation during stages i-ii (A) and iii-iv (B) of VF in 1 pig. Each colored pixel is an electrode site at which the rate of voltage change (dV/dt) is less than or equal to −0.3V/s sometime during the 15-ms interval represented by each frame. The numbers show the time from the beginning of each VF stage. Different-colored pixels indicate distinct isolated wavefronts. Examples of wavefront fractionation (A: stage i, anterior, 195–210 ms) and collision (A: stage i, anterior, 165–180 ms) are indicated by arrows. Recordings from the same electrode are shown below the activation maps for each stage. Wavefront during stage i on posterior surface propagated in one direction for four cycles immediately after VF induction and then degenerated into chaotic pattern (not shown).

Table 1.

Pig ventricular fibrillation activation pattern descriptors in different stages

	Stage i	Stage ii	Stage iii	Stage iv
Duration, s
Anterior	1–22	23–38	39–195	196–600
Posterior	1–21	22–40	41–179	180–600
Average	1–22	23–39	40–187	188–600
No. of wavefronts
Anterior	61 ± 10	59 ± 4	44 ± 6	23 ± 5
Posterior	79 ± 11*	71 ± 4*	59 ± 7*	50 ± 3*
Fractionation incidence
Anterior	0.23 ± 0.02	0.23 ± 0.01	0.16 ± 0.04	0.03 ± 0.02
Posterior	0.19 ± 0.02*	0.20 ± 0.02	0.18 ± 0.02*	0.13 ± 0.02*
Collision incidence
Anterior	0.20 ± 0.02	0.21 ± 0.01	0.24 ± 0.02	0.08 ± 0.05
Posterior	0.26 ± 0.02*	0.26 ± 0.02*	0.24 ± 0.02	0.17 ± 0.02*
Multiplicity
Anterior	6.9 ± 1.0	5.8 ± 0.6	6.8 ± 0.7	4.5 ± 0.9
Posterior	9.6 ± 1.4*	8.0 ± 0.8*	8.8 ± 0.7*	7.5 ± 1.1*
Repeatability
Anterior	7.8 ± 0.5	8.3 ± 0.3	6.5 ± 1.4	4.4 ± 4.0
Posterior	7.8 ± 0.7	9.0 ± 0.7	6.7 ± 1.4	6.5 ± 0.9*
Velocity, cm/s
Anterior	0.39 ± 0.03	0.35 ± 0.02	0.27 ± 0.04	0.22 ± 0.01
Posterior	0.41 ± 0.03	0.37 ± 0.02	0.27 ± 0.05	0.23 ± 0.01
Activation rate, Hz
Anterior	8.4 ± 0.6	9.2 ± 0.2	8.2 ± 0.6	5.0 ± 1.1
Posterior	9.2 ± 0.2	9.1 ± 0.1	7.9 ± 0.7*	5.3 ± 0.4*
Area swept, mm2
Anterior	394 ± 26	356 ± 7	354 ± 39	249 ± 53
Posterior	437 ± 55	439 ± 17*	346 ± 59	207 ± 21*
dV/dt
Anterior	−0.84 ± 0.04	−0.80 ± 0.03	−0.64 ± 0.05	−0.55 ± 0.01
Posterior	−1.05 ± 0.04*	−1.02 ± 0.05*	−0.73 ± 0.09*	−0.68 ± 0.03*
Breakthrough/focal
Anterior	0.20 ± 0.03	0.17 ± 0.01	0.23 ± 0.06	0.52 ± 0.03
Posterior	0.14 ± 0.02*	0.12 ± 0.02*	0.19 ± 0.04*	0.30 ± 0.03*
Block
Anterior	0.18 ± 0.02	0.17 ± 0.02	0.25 ± 0.10	0.65 ± 0.09
Posterior	0.14 ± 0.02*	0.12 ± 0.02*	0.19 ± 0.07*	0.30 ± 0.03*
Reentry
Anterior	0.03 ± 0.02	0.05 ± 0.01	0.03 ± 0.03	0 ± 0
Posterior	0.05 ± 0.02	0.08 ± 0.02	0.03 ± 0.02	0.01 ± 0.004*

Values are means ± SE; n = 6. dV/dt, rate of voltage change.

^*P < 0.004, compared with the anterior.

The broad picture of activation is as follows: stage i (1–22 s), large spatial and temporal variation; stage ii (23–39 s), more organized and repeatable activation patterns; stage iii (40–187 s), fragmented and irregular wavefronts; and stage iv, (188–600 s), slower moving and shorter lived wavefronts. The evolution of activation patterns during VF detected from the posterior and anterior epicardium was similar, with the activation on the posterior LV epicardium attenuating slower than activation on the anterior LV epicardium.

Our previous microelectrode study (29) showed that pig action potential amplitude (APA) and diastolic interval (DI) markedly changed during the first 3 min after VF induction, which corresponds to stages i to iii (1–187 s) in this study. We observed in the current study that calculated parameters changed in a similar way as APA and DI in that most parameters changed more during stages i and iii than stage iv (P < 0.05).

Wavefront number and incidence of fractionations and collisions.

In porcine stage i, wavefronts propagated over similar pathways for numerous cycles during the first 10 s after VF induction and then quickly degenerated into numerous smaller wavefronts that propagated in multiple directions. The number of wavefronts decreased briefly during stage ii (Table 1). During stage iii, the number of wavefronts increased again and then continuously declined to reach a plateau during stage iv. Over the entire 10 min of VF, there were 39% more wavefronts on the posterior than the anterior epicardium in pigs (P < 0.01; Fig. 3).

Fig. 3.

Evolution of the number of epicardial wavefronts during the first 600 s of VF for the anterior LV in pig (red), the posterior LV in pig (green), and the anterior LV in dog (black). Data in this and Figs. 4–9 are means from 6 pigs and 6 dogs. Short vertical lines demarcate the times of transition from one stage to the next for the 4 stages in pigs and the 5 stages in dogs. A short stage of increased organization is present (stage ii in pigs and stage iii in dogs) in which the number of wavefronts is briefly decreased. For this and Figs. 4–10, please see the text for additional description.

During the first 300 s of VF, there were 68% more wavefronts on the anterior epicardium in dogs than pigs (P < 0.001); however, after 300 s of VF, there were 27% fewer wavefronts on the anterior epicardium in dog than pigs (P < 0.001; Fig. 3). As with dogs, a stage of increased organization was present in pigs also. However, this more organized stage occurred earlier in the pig as stage ii (23–39 s) than in the dog where it was stage iii (63–86 s).

Examples of wavefront fractionation and collision are shown in Fig. 2. For both the anterior and posterior epicardium in pigs, 70% more wavefronts fractionated and 43% more collided during the first two than during the last two VF stages (P < 0.01; Fig. 4). The incidence of wavefront fractionation was significantly higher on the posterior than the anterior epicardium during stages iii and iv (Table 1). The incidence of wavefront fractionation on the anterior epicardium was 123% higher for dogs than pigs for the 10 min VF (P < 0.01; Fig. 4A). The incidence of wavefront collisions on the pig and dog anterior epicardium was not significantly different during pig stage iii (P > 0.05; Fig. 4B).

Fig. 4.

Evolution of wavefront fractionation (A) and collision (B) during VF.

Multiplicity and repeatability.

During stage i on both the anterior and posterior epicardium in pigs, multiplicity (the number of distinct activation pathways followed by the wavefronts) and repeatability (the number of wavefronts following each activation pathway) quickly increased (Fig. 5), indicating that increased numbers of wavefronts followed similar pathways. During stage ii, multiplicity suddenly decreased and repeatability increased, consistent with the decreased number of wavefronts during this stage. Multiplicity was 53% higher on the posterior than the anterior epicardium in pigs (P < 0.05) throughout the first 10 min of VF (Table 1). Repeatability was 48% higher on the posterior than the anterior epicardium only during stage iv (Table 1). Multiplicity, but not repeatability, was 24% higher for dogs than pigs on the anterior epicardium (P < 0.01).

Fig. 5.

Evolution of multiplicity (A) and repeatability (B) during VF.

Conduction velocity and activation rate.

Conduction velocity was 52% higher during the first two stages compared with the last two stages of VF in pigs (P < 0.01). Conduction velocity decreased during stage iii and reached a plateau during stage iv (Table 1; Fig. 6A). The conduction velocity was not significantly different for the anterior and posterior epicardium during any stage (Table 1). Wavefronts propagated 22% faster on the anterior epicardium in dogs than pigs during the first 120 s of VF (P < 0.01); this duration covers stages i and ii and part of stage iii in pigs, and stage i, ii, and iii and part of stage iv in dogs. However, the wavefronts propagated 20% faster on the anterior epicardium in pigs than dogs 120 s after VF induction (P < 0.01).

Fig. 6.

Mean conduction velocity (A) and mean activation rate (B) during VF.

For both the anterior and posterior epicardium in pigs, the activation rate increased quickly during stage i and reached its highest level during stage ii (Table 1; Fig. 6B). It then decreased during stage iii and remained unchanged during stage iv (Table 1). The activation rate was 18% lower on the anterior epicardium 300 s after VF induction (last stage for both pigs and dogs) than on the posterior epicardium (P < 0.05).

Mean area swept out by wavefronts and mean peak negative dV/dt.

During stage i, the area swept out, i.e., the epicardial area traversed by each wavefront, was larger during the first few cycles at the initiation of VF but then quickly broke down into smaller wavefronts on the anterior and posterior epicardium of the pigs. On both the anterior and posterior epicardium, the area swept out by the wavefronts decreased slowly during stage iii, and reached a fairly constant value during stage iv (Table 1; Fig. 7A). The wavefronts propagated over a mean area that was 78% larger on the anterior epicardium in pigs than dogs for the first 10 min of VF (P < 0.01).

Fig. 7.

Evolution of the area swept out by each wavefront (A) and the peak negative dV/dt (B) during VF.

For both the anterior and posterior epicardium in pigs, the mean negative peak dV/dt was most negative during stages i and ii (Table 1; Fig. 7B). It decreased in absolute value during stage iii, indicating a slower downstroke of activation in the unipolar recordings, and remained constant during stage iv (Table 1; Fig. 7B). The dV/dt was 21% more negative on posterior surface than on anterior surface for the first 10 min of VF in pigs. The peak dV/dt was 55% more negative on the anterior epicardium in dogs than in pigs during the first 300 s of VF (P < 0.01), but was 10% less negative after 300 s (P < 0.05).

Breakthrough/focal and block incidence.

There were 96% fewer wavefronts that appeared de novo on the epicardium during the first two than the last two stages for both the anterior and posterior epicardium in pigs (Fig. 8A; Table 1). This pattern probably represented epicardial breakthrough of wavefronts from the intramural LV but it is possible the wavefronts could have arisen from a true focus. This pattern was 65% more frequent on the anterior than on the posterior epicardium throughout the entire 10 min of VF (P < 0.01). There were 41% fewer wavefronts that exhibited a breakthrough/focal pattern on the anterior epicardium in dogs than pigs for the first 10 min of VF (P < 0.01).

Fig. 8.

Evolution of wavefront breakthrough or foci on the epicardium (A) and block incidence (B) during VF.

For both the anterior and posterior epicardium in pigs, there were 56% fewer wavefronts that totally blocked within the mapped region during the first two than the last two stages (P < 0.01; Table 1; Fig. 8B). There were 70% fewer wavefronts that blocked on the anterior epicardium in dogs than pigs throughout the 10 min of VF (P < 0.01).

Reentrant circuits.

In pigs, only 1.2% of wavefront families formed a reentrant circuit on the epicardium during stage i, while ∼10% formed a reentrant pathway during stage ii and iii (Fig. 9). Almost no reentry was observed during stage iv. The incidence of reentry was significantly higher on the anterior epicardium in dogs than pigs throughout the 10 min of VF (P < 0.01).

Fig. 9.

Incidence of wavefront families that formed reentrant pathways during VF.

Directionality of wavefront propagation during VF.

The x-velocity (parallel to the atrioventricular groove) on both the anterior and posterior epicardium was positive over all stages, which indicates wavefronts propagated mainly clockwise (as viewed from above). The y-velocity (from apex to base) was in opposite directions on the anterior (Fig. 10A) and posterior (Fig. 10B) epicardium during the first and second stages, which indicates wavefronts propagated mainly from the base towards the apex in the anterior epicardium but from the apex towards the base in the posterior epicardium. During the third and fourth stages, the wavefronts propagated mainly from base to apex.

Fig. 10.

Decomposition of the weighted velocity vectors into x- and y-components over the anterior (A) and posterior (B) LV with x-velocity parallel to the atrioventricular groove and y-velocity from apex to base.

DISCUSSION

In this study we report the quantitative changes in electrophysiological properties during the first 10 min of VF in the pig. The principal findings are as follows: 1) a brief distinct increase in organization 23–39 s after VF induction (stage ii); 2) an opposite direction of wavefront propagation along the long axis of the anterior and posterior epicardium during the first and second stages; 3) a large change for most descriptors during stages i and iii compared with stage iv; 4) a much slower decline in the number of wavefronts in the posterior than anterior epicardium; and 5) an increased incidence of epicardial reentry during stages ii and iii.

Evolution of activation patterns during long VF in pigs.

It is commonly thought that most wavefront parameters progressively degrade to zero once VF is induced (39). Tovar et al. (39) continuously recorded monophasic action potentials in isolated rabbit hearts and showed that during the first 5 min of VF progressive action potential shortening occurs accompanied by lengthening of the cycle length (CL) and DI. Tovar and Jones (39) and Robertson et al. (29) also demonstrated with microelectrode recordings that the increased CL in VF is due to an increased DI in rabbits and pigs, respectively. They paced the heart at a CL shorter than the VF CL and found the paced wavefront propagation slowed and quickly blocked (29). The APA and V̇_max decrease secondary to global ischemia in VF may be responsible for this finding. The changes in APD and DI do not progress linearly as VF persists. Instead, the majority of the change in APD and DI occurs during the first 3 min of long-duration VF. After this, the rate of change is much less (29). Most parameters calculated in this study changed in a similar manner as the APD and DI changes, with the largest changes during the first 3 min of VF. One potential cause of the observed changes is the disruption of potassium homeostasis and acid-base equilibrium that occurs in VF (18). ATP-dependent K⁺ (K_ATP) channels are critically dependent on high-energy phosphates and quickly respond to global ischemia.

We (14) previously observed the evolution of activation patterns during VF in dogs using a mapping array that covered ∼20 cm² of the epicardium and divided VF into five stages (see discussion below) More recently, Huizar et al. (17) performed an in vitro experiment using optical mapping to observe the evolution of activation in isolated blood-perfused pig hearts during 10 min of VF and global ischemia. Based on the changes in action potential, DI, activation frequency, wavebreak, and wave direction, they divided pig VF into three stages, i.e., relatively periodic, highly periodic, and aperiodic phases. In the present in vivo study, we (14) applied the electrical mapping and analysis methods similar to those in the dog study, which divided pig VF into four stages. Consistent with Huizar's data, after VF induction, our experimental findings also demonstrated that pig epicardium exhibited electrical activity acceleration and a brief period of increased organization, followed by degeneration of activation. The parameters calculated in our study (number of wavefront, conduction velocity, fractionation, collision, block, and breakthrough incidence) allowed us to separate the degeneration into two stages, i.e., stages iii and iv.

The starting time and duration of each stage are slightly different between the study of Huizar et al. (17) and our studies (14). The different parameters observed and different animal preparations in the two studies might be the major factors for these differences. The number of wavefronts, multiplicity, repeatability, and area swept by each wavefront temporally fluctuated more than other parameters such as conduction velocity, activation rate, and V̇_max.

One of the new findings in this study is that wavefronts propagated in an opposite direction along the long axis of the heart on the anterior (from base to apex) and posterior epicardium (from apex to base) during stages i and ii. In our previous study (16), we mapped the posterior LV epicardium and also found that the wavefronts propagated mainly from the apex towards the base. The finding that the wavefronts mainly propagated clockwise (top view) around the heart is also consistent with our previous findings (16, 23). In this study, we applied mapping arrays on the anterior and posterior epicardium over the papillary muscles which have been shown to serve as anchor points for reentry during VF (24, 25). The orientation of the long axis of the myocytes change markedly at different depths within the ventricular wall (36). In brief, the myocytes are parallel to the long axis of the heart at the outer and inner layers and perpendicular to the long axis in the mid-wall of the LV. The myocyte orientation favors VF wavefronts propagating along the ventricular axis of apex to base.

Comparison of VF activation patterns between pig and dog.

The recording and pattern analysis methods in the current study were similar to those in our previous dog study (14). The common findings between pig and dog are as follows: 1) the first 10 min of VF can be partitioned into stages for both pig and dog according to the evolution of electrophysiological characteristics; 2) the trendline of the evolution for most parameters is similar as VF continues; and 3) a distinct period of increased organization exists both in dogs and pigs. The differences between pigs and dogs for the parameters measured are as follows: 1) immediately after VF induction on pig epicardium, there were significantly fewer wavefronts (Fig. 3) that swept over a larger area (Fig. 7A) with lower multiplicity (Fig. 5A) and higher repeatability (Fig. 5B) in pigs than dogs. However, after 350 s, there were significantly more wavefronts in pigs than dogs; 2) the distinct period of increased organization began 23 s after VF induction in pigs rather than after 63 s as in dogs; and 3) reentrant circuits were less commonly recorded in pigs than dogs, (Fig. 9); the propagation velocity during VF was faster in dogs than pigs during the first 120 s, while the area swept out by the wavefronts was larger in pigs than dogs.

While the methods and parameters to observe the evolution of long duration VF are different in different species, we think that k-means clustering in pigs and dogs in vivo and Huizar's in vitro pig studies all somewhat reflect Wiggers' four stages in dog (13). Wiggers used high-speed cinematography to observe ventricular mechanical contraction from which he identified four stages of VF. The first was the undulatory or tachysystolic stage, which consisted of three to six undulatory contractions lasting ∼1 s. The second stage, called the convulsive incoordination stage, lasted for 15–40 s. This stage was characterized by more frequent waves of contractions that sweep over smaller sections of the ventricles. In this stage, synchronous contractions of the ventricles were lost. The third stage, that of tremulous incoordination, lasted 2–3 min, during which the surface of the ventricles appeared to be broken up into progressively smaller independently contracting areas. The final stage was that of atonic fibrillation, with complete failure of contractility.

The current study was based on electrical wavefront activation that partitioned VF into four stages. The digital signals processing and computer algorithms used in our and the studies of Huizar's et al. (17) are more sensitive for fast and small changes during VF. The time for most descriptors to increase to their highest value took 11 s in dog, which is longer than the 6 s it took in pigs that the k-means cluster analysis divided into two stages. Thus stage i in pigs is comparable to stages i and ii in dogs. We (14) and Huizar et al. (17) found a short period of organization during VF that was not found by the cinematography of Wiggers (33). The distinct increase in organization appeared at 23–39 s after VF induction in pigs (stage ii) and 63–86 s in dogs (stage iii). Stages iii and iv in pigs are comparable to stages iv and v in dogs.

The intrinsic substrate differences, such as anatomical or ion channel differences might account for the activation pattern differences. In has been reported that the transient outward K⁺ current (I_to1) is much more prominent in the dog ventricular epicardium than the endocardium (21). I_to1, which is active during phase 1 repolarization, is not expressed in pig myocardium. Instead, a Ca²⁺-activated transient outward Cl current (I_to2) has been shown to contribute to phase 1 repolarization in pigs (18). I_to1 and I_to2 have been demonstrated to play a role in determining membrane excitability and action potential morphology (19). However, I_to1 recovers rapidly and shows little frequency dependence (46). This might partially explain the larger number of wavefronts, faster activation rate, and conduction velocity at the early VF stages in dogs compared with pigs. The prominent I _to1 in dog epicardium but not endocardium could cause the dog epicardium to be more electrically depressed than the endocardium by ischemia. Thus the global ischemia caused by long duration VF could cause electrical activity to be more depressed on dog epicardium than the pig epicardium. The different density of I_k,ATP between epicardium and endocardium provides more evidence to explain the activation difference (9). More recently, Cordeiro at al. (5) showed increasing extracellular K⁺ concentration caused selective depression of epicardial action potentials and transmural conduction slowing and block.

The Purkinje network is a thin layer of tissue that is coupled to the much thicker ventricular myocardium at numerous, discrete sites (28, 40). The Purkinje distribution is limited to the endocardium in dogs but extends almost to the epicardium in pigs (1, 8). A recent study (38) demonstrated that during the first 10 min of VF in isolated dog hearts wavefronts propagate both from the working myocardium into the Purkinje system and from the Purkinje system into the working myocardium (38), suggesting that the Purkinje fiber system contributes to wavefront propagation during VF. If so, then the wavefronts mapped from the epicardium are probably more influenced by the effect of the specialized conduction system in pigs than dogs. Purkinje fibers are more fatigue resistant and continue to function during ischemia longer than working myocardial tissue (2, 10). The extension of the Purkinje network almost to the epicardium in pigs could bridge individual wavefronts into larger wavefronts during VF. The wavefronts syncretized by the Purkinje network should also be less fractionationated compared with those without an almost transmural Purkinje network as in the dog.

Comparison of activation patterns of the anterior and posterior epicardium.

While the activation rate and propagation velocity on the anterior epicardium were similar to that on the posterior epicardium, more wavefronts were present on the posterior epicardium than on the anterior epicardium during VF in pigs. Nanthakumar et al. (23) mapped almost the entire LV epicardium for the first 30 s of VF and found more wavefronts propagated from posterior to anterior than in the opposite direction (23), consistent with the current study that found that most wavefronts propagated clockwise as viewed from the top of the heart. Pak et al. (25) found that the majority of reentrant wavefronts during VF with acute ischemia or during propranolol infusion were located at the posterior papillary muscle. The incidence of reentry has been shown to be less in the septum than in the posterior wall (16). All these findings suggest that there are more wavefronts created in the posterior wall than in other areas (47).

Another possibility explaining the difference in the number of wavefronts between the anterior and posterior epicardial surface could be the epicardial temperature difference between them. While we covered the chest with a plastic film and the thick mapping plaque served as an insulator to protect the anterior surface during the long duration VF, the edges of the mapping plaque did not seal to the epicardial surface. We did not monitor the anterior and posterior epicardial surface temperature in the study; thus we are not certain how well those procedures prevented the anterior surface temperature from decreasing during the long duration VF. The temperature of the anterior epicardial surface could still have been lower than the posterior surface. A previous transmural mapping study from our laboratory demonstrated that the endocardial-epicardial gradient of electrical activity that developed during the first 10 min of VF is caused by ischemia rather than temperature (45). In that study, one group of animals was maintained with the heart surface at room temperature, and in another group, the heart surface was temperature controlled at 38°C (45). While that study demonstrated that a decrease of heart surface temperature close to room temperature did not dramatically change long-duration ventricular fibrillation activation patterns, for this study we cannot exclude the possibility that the number of wavefronts and fractionation difference between the anterior and posterior surface was influenced by a temperature difference.

Clinical relevance.

This study demonstrates that there are rapid dynamic changes in activation patterns during VF. There is also a very brief stage in which activation is more organized than in other VF stages in pigs. Combined with our previous dog study (14) and Huizar's pig study (13), the phenomena observed from pig and dog long duration VF are helpful for us to understand the probability of defibrillation success and the timing of defibrillation shocks to improve the success rate, because it has been shown that defibrillation shocks synchronized to a relatively less chaotic interval during VF are more efficient than shocks during more chaotic periods (7, 12). Studies are needed to investigate the relationship of the surface ECG to myocardial electrophysiological changes during long VF to estimate the duration of VF from the ECG. It is important to study the electrophysiological evolution of human VF, since it is important for resuscitation success during prolonged VF. The spectral characterization of ECG during VF has been reported to predict the success of defibrillation (6, 20, 27). However, it is not known how the body surface spectral characteristics, such as amplitude, median, and dominant and peak frequency, relate to the characteristics of VF wavefront propagation. Stationarity and periodicity are fundamental assumptions of the Fourier transform, the most frequently used computational method for obtaining spectral characteristics of VF signals (35). In addition, the body surface ECG is affected by the electrical activity of the whole heart. Therefore, the ECG can exhibit a typical VF pattern even when LV endocardial activation is highly organized and synchronous, because the transmural LV and the RV during this period still contain fibrillatory wavefronts (30).

The current pig study and the previous dog study (14) add electrophysiological information in addition to the already known mechanical information about the circulatory phase of VF in humans ( 4–10 min of VF; Refs. 4, 41) The circulatory phase in humans corresponds to stages iv and v in the dog model and stage iv in the pig model. All the electrophysiological characteristics degenerated to low levels during this period. Activation rate and conduction velocity were very slow. Activation after a successful shock in this phase may fail to propagate due to prolonged ischemia. This would not only inhibit cardiac contraction but could cause a reentrant circuit or triggered activation that might reinitiate VF. It is possible that chest compressions for 1–3 min before giving the defibrillation shock during the circulatory phase of VF could cause blood flow to the ventricles that improves cardiac function not only by a direct effect on cardiac contractility but also by improving conduction of the cardiac impulse.

Limitations.

The limitations for extracellular electrical mapping studies from epicardial arrays described in our previous study (13) are also present in this study. We mapped only 40% of the epicardial area and performed no transmural mapping so that complete activation pathways could not be determined and all reentrant circuits were probably not detected. The evolution of VF described in this study used anesthetized pigs with normal hearts, which may differ from clinical VF, which most often occurs in cardiac ischemia or heart failure patients without anesthesia. The evolution of VF in diseased hearts may differ from that in normal hearts.

Isoflurane anesthesia was used in the study, which blocks IKs (37). IKs is a “slow” K channel that is active during repolarization. Selective blockade of IKs in canine models of malignant ischemia showed a modest prolongation of the ventricular effective refractory period and an antirrhythmic effect (22). A prolonged ventricular effective refractory period may affect the evolution of VF.

GRANTS

This study was supported in part by the National Heart, Lung, and Blood Institute Grants HL-85370, HL-66256, HL-91138, and HL-28429 and an American Heart Association Science Development Grant.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: K.-A.C. and J.H. performed experiments; K.-A.C., D.J.D., L.L., J.M.R., and J.H. analyzed data; K.-A.C., D.J.D., L.L., R.E.I., and J.H. interpreted results of experiments; K.-A.C., D.J.D., L.L., and J.H. prepared figures; K.-A.C. and J.H. drafted manuscript; K.-A.C., D.J.D., L.L., J.M.R., R.E.I., and J.H. approved final version of manuscript; D.J.D., J.M.R., R.E.I., and J.H. edited and revised manuscript; R.E.I. and J.H. conception and design of research.