Quantification of activation patterns during ventricular fibrillation in open-chest porcine left ventricle and septum

Quantification of activation patterns during ventricular fibrillation in open-chest porcine left ventricle and septum Jian Huang, MD, PhD,a Greg P. Walcott, MD,a Cheryl R. Killingsworth, DVM, PhD,a Sharon B. Melnick, AAS,a Jack M. Rogers, PhD,b Raymond E. Ideker, MD, PhDa,b,c a

From the Department of Medicine, Department of Biomedical Engineering, and c Department of Physiology, Cardiac Rhythm Management Laboratory, University of Alabama at Birmingham, Alabama, Birmingham. b

BACKGROUND A single stationary mother rotor has been hypothesized to be responsible for maintenance of ventricular fibrillation (VF) in the guinea pig. Previous studies have pointed to the ventricular septum as a possible location for a mother rotor in the pig heart. OBJECTIVES The purpose of this study was to test the hypothesis that a mother rotor is located in the septum. METHODS In seven open-chest pigs, we mapped the first 20 seconds of electrically induced VF simultaneously from the posterior left ventricle (LV) and right side of the septum with two electrical arrays. Each array contained 504 electrodes (21 ⫻ 24) spaced 2 mm apart in the LV and 1.5 mm apart in the septum. RESULTS The percentage of VF wavefronts that formed reentrant circuits was significantly lower in the septum (1% ⫾ 1% [mean ⫾ SD]) than in the LV (2% ⫾ 1%). The peak frequency during VF also was significantly smaller in the septum (8.6 Hz ⫾ 3.0 Hz) than in the LV (10.4 Hz ⫾ 3.4 Hz). The mean direction of spread of activation of VF wavefronts was away from the region where the posterior LV free wall intersects the posterior septum in both the LV and septum. CONCLUSIONS The lower incidence of reentry and lower peak frequency in the mapped region of the septum than in the LV indicate that a mother rotor is not present in swine on the RV side of the septum. The mean directions of the VF activation sequences in the LV and septum suggest that if a mother rotor is present during the first 20 seconds of VF, it exists where the posterior LV free wall joins the septum, the region where the posterior papillary muscle inserts. KEYWORDS Arrhythmia; Mapping; Fibrillation (Heart Rhythm 2005;2:720 –728) © 2005 Heart Rhythm Society. All rights reserved.

Introduction A single stationary mother rotor, located in the fastest activating region and giving rise to activation fronts that propagate throughout the remainder of the myocardium, has been hypothesized to be responsible for maintenance of ventricular fibrillation (VF).1,2 Epicardial mapping studies Supported in part by National Institutes of Health Research Grants HL-66256 and HL-28429. Address reprint requests and correspondence: Dr. Jian Huang, Cardiac Rhythm Management Laboratory, Volker Hall B140, 1670 University Boulevard, Birmingham, Alabama 35294-0019. E-mail address: [email protected]. (Received January 21, 2005; accepted March 28, 2005.)

have demonstrated that, during the first 20 seconds of VF, wavefronts in pigs tend to propagate from the posterior basal left ventricle (LV) to the anterior LV and on to the anterior right ventricle (RV),3,4 raising the possibility of a mother rotor in the posterior LV. However, no sustained reentry consistent with a mother rotor was found on the posterior LV epicardium,3,5 even though an intramural mapping study showed that the fastest activating transmural layer was the epicardium.6 Many wavefronts in the posterior LV entered the mapped region from the posterior boundary of the mapping array, adjacent to the posterior descending coronary artery,3 raising the possibility that a mother rotor is located in the RV or septum. Because a previous study showed that the RV activates more slowly than the LV

1547-5271/$ -see front matter © 2005 Heart Rhythm Society. All rights reserved.

doi:10.1016/j.hrthm.2005.03.025

Huang et al

Activation Patterns of VF

Figure 1 Diagrams of the heart indicating the location of the mapping electrodes on the left ventricle (A) and the right side of the septum (B). The black dots represent the individual plaque recording electrodes. The 16 stars in a 4 ⫻ 4 array represent the locations of the plunge needle electrodes. The directions of the x and y components of the conduction velocity vectors of the ventricular fibrillation wavefronts are indicated. PDA ⫽ posterior descending artery.

during VF,4 the more likely site for a mother rotor, if it exists, is the septum. This study tested the hypothesis that a mother rotor is located in the septum.

Methods Animal preparation The study was approved by the University of Alabama at Birmingham Institutional Animal Care and Use Committee. All animals were managed in accordance with the guidelines established by the American Heart Association on Research Animal use.7 Eight pigs (weight 40 –50 kg, 44 ⫾ 3 [mean ⫾ SD]) were injected intramuscularly with zolazepam-tiletamine 4.4 mg/ kg, xylazine 4.4 mg/kg, and atropine 0.04 mg/kg for anesthetic induction. Anesthesia was maintained with isoflurane in 100% oxygen by inhalation. Core body temperature, arterial blood pressure, arterial blood gases, surface ECG lead II, and serum electrolytes were monitored and maintained within normal ranges throughout the study. The heart was exposed through a median sternotomy and supported in a pericardial sling. A plaque containing 504 electrodes (24 ⫻ 21) with 2-mm space between each electrode was sutured to the posterior lateral LV, with one edge of the plaque adjacent to the posterior descending artery (Figure 1). Two VF episodes were recorded with this LV array. After the pigs were heparinized, two plastic cannulas were passed through the wall of the right atrium into the venae

721 cavae. Cotton tapes were placed around the cavae and their enclosed cannulas. Another cannula was inserted through an incision in the ascending aorta. Under the support of cardiopulmonary bypass, two more VF episodes were recorded with the LV array. An incision was made through the anterior RV wall adjacent to the septum. A second plaque containing 504 electrodes (24 ⫻ 21) with 1.5-mm space between each electrode was placed on the right side of the septum through the incision, which covered about 40% of the septum (Figure 1). Two VF episodes were simultaneously recorded with the LV and septum mapping arrays. The mapping array on the right side of the septum was removed, and 16 plunge needles were inserted in a 4 ⫻ 4 matrix distributed throughout the region previously mapped with the surface array. Each plunge needle contained three electrodes spaced 1 cm apart for recording from the subendocardium of the LV side of the septum, the middle of the septum, and the subendocardium of the RV side of the septum. Two more VF episodes were recorded with the LV array and plunge needles.

Mapping system and data acquisition Two 528-channel mapping systems were synchronized by a common clock for time alignment of the two mapping data streams. One recorded from the LV and the other recorded simultaneously from the septum during the last four VF episodes. 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.8 VF was induced using a DC pulse applied to the RV. VF was allowed to last 20 seconds before defibrillation. VF data were acquired starting from induction of VF. Eight episodes of VF were induced per animal. The first two episodes of VF were induced before any manipulation for cardiopulmonary bypass (before pump group). The second two episodes of VF were induced after instituting cardiopulmonary bypass but before RV incision (before incision group). The third two episodes were induced after RV incision with the plaque on the septum (after incision group). The fourth two episodes were induced with 16 plunge needles in the septum instead of the mapping plaque (plunge needle group). A minimum of 4 minutes was allowed between VF episodes.

Quantification of VF Quantitative analysis of VF activation patterns was performed on a Linux system using algorithms discussed in detail elsewhere.9,10 The algorithms automatically isolate wavefronts by grouping together activations recorded at neighboring electrodes that were recorded within 0.5 ms of each other. A single temporal sample at a recording site was considered to represent an activation in the underlying tissue when dV/dt ⬍ ⫺0.5 V/s.11 A five-point digital filter was used to calculate the temporal derivative of each of the 1,008 electrograms.

722 The algorithms quantitatively calculated the manners in which wavefronts arose, propagated, and terminated. During VF, a wavefront frequently splits into two or more wavefronts. Conversely, two or more wavefronts frequently collide to merge into one or more wavefronts or terminate. The wavefronts before and after fractionation or collision are counted as different wavefronts. From the automated wavefront analysis, the following variables within the mapped area were analyzed during the first 0.5 second of every second during each 20-second VF episode. (1) Number of wavefronts per cm2. This is the number of wavefronts completely or partially within the mapped region normalized with respect to the mapping area. Because the size of the mapping array is different for the septum and LV, we calculated the number of wavefronts per cm2. (2) Mean area swept out by the wavefronts. The area swept out by a wavefront is the number of electrodes that recorded an activation from that wavefront multiplied by the area represented by each electrode (4 mm2 for 2-mm interelectrode spacing and 2.25 mm2 for 1.5-mm interelectrode spacing).10 When reentry was present, electrodes were only counted once. Thus, the maximum area that could be swept out by a wavefront is the area of the plaque. (3) Block incidence. This is the fraction of wavefronts that terminated without fractionating, colliding with other wavefronts, or propagating out of the mapped region.12 (4) Breakthrough/ focal incidence. This is the logical opposite of block. A wavefront that appears in the mapped region without arising from a fractionation or collision event or without propagating in from an edge of the mapped region either has broken through to the epicardium from deeper layers or has arisen de novo.12 (5) Multiplicity of the activation pattern. Multiplicity is a parameter that measures the number of distinct wavefront activation pathways.9 It is a descriptor of the spatiotemporal complexity of the activation sequences. For example, during pacing at a slow rate from a single site, in which the wavefronts sweep over the mapped area in a repeatable activation sequence, the multiplicity is 1. If two wavefronts repeatedly enter the mapping region from opposite sides and collide to form a new wavefront, this pattern has a multiplicity of 3, 1 for each of the 2 original wavefronts and 1 for the new wavefront after the collision. (6) Repeatability of the activation pattern.12 This descriptor expresses the mean number of wavefronts that propagate along each distinct pathway indicated by the multiplicity parameter. For example, if an activation pattern contains 8 wavefronts, 4 of which follow one pathway, and 4 of which follow another, then the multiplicity of the pattern is 2 and the repeatability is 4. If each pattern represents part or all of a reentrant circuit, then repeatability expresses the number of cycles for which reentry persists. (7) Fractionation. This is the fraction of wavefronts that split into two or more wavefronts.10 (8) Collision. This is the fraction of wavefronts that collide and coalesce with one or more other wavefronts.10 (9) Negative peak dV/dt of VF activations. This is the mean rapidity of the peak downslope of activa-

Heart Rhythm, Vol 2, No 7, July 2005 tion, which may be related to the ability of the activation front to excite adjacent tissue. (10) Peak frequency of VF activation. The VF electrograms were analyzed by fast Fourier transform (FFT) analysis. FFT was computed using the Welch method on 2-second intervals of VF, with 1.024second overlapping Hanning windows and 0.5-Hz resolution.13 (11) Conduction velocity. The wavefront velocity was estimated by computing the location of the centroid of the wavefront each 0.5 ms. 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 (posterior to anterior) and y direction (apex to base) were calculated for each segment as described elsewhere.12 (12) Incidence of reentry. This descriptor differs from the previous 11 variables in that it was computed for consecutive 5-second epochs14 rather than 0.5-second epochs. For each epoch, wavefronts were grouped into families consisting of wavefronts connected temporally by fragmentation or collision events. If a family contained a sequence of wavefronts that activated tissue more than once, it was deemed reentrant. The incidence of reentry is the fraction of wavefront families containing a reentrant sequence of wavefronts.

Transmural activation rate in septum The peak frequency in the three unipolar signals recorded from each plunge needle (representing left subendocardium, midmyocardium, and right subendocardium) were computed using the method described for variable no. 10.

Statistical analysis Measured values are expressed as mean ⫾ SD. Repeated measures ANOVA was used to test whether cardiopulmonary bypass and RV incision affected VF activation. Multiple comparisons were made using the Student-NewmanKeuls test of variance. Paired t-tests were used to test for differences of VF activation patterns between the LV and septum (SPSS, Chicago, IL, USA). The two mapping plaques were divided into four equal quadrants. The peak frequencies in these four quadrants were compared by multivariate analysis of variance (MANOVA). P ⬍ .05 was considered significant.

Results Eight pigs were studied. One died during the bypass procedure. Thus, complete data were collected from seven pigs. The seven hearts weighed 183 ⫾ 34 g.

Effect of cardiopulmonary bypass and RV incision procedures on activation patterns Data from all 0.5-second intervals of the 20-second VF episodes are pooled together for the LV and septum sepa-

Huang et al

Activation Patterns of VF

723

Figure 2 Ventricular fibrillation activation pattern descriptors computed before pump, before incision, and after incision. The boxes extend from the 25th to the 75th percentile. The whiskers extend from the 5th to the 95th percentiles. The horizontal lines mark the median, and the black squares indicate the mean. *P ⬍ .01 between before pump group and after incision group for the left ventricle (LV). #P ⬍ .01 after incision between LV and septum.

rately (Figure 2). No significant differences in the LV epicardial VF activation patterns between the before pump and before incision groups or between the before incision and after incision groups were noted (Figure 2). However, cardiopulmonary bypass and RV incision synergistically affected all VF variables measured on the LV epicardium, as shown by a significant difference between the before pump and after incision groups (Figure 2).

lower in the septum (1% ⫾ 1%) than in the LV (2% ⫾ 1%, P ⬍ .01). LV epicardium activated faster than the septum during VF (Figures 2 and 4). Mean peak frequencies were 10.4 ⫾ 3.4 for the LV epicardium and 8.6 ⫾ 3.0 Hz for the septum (P ⬍ .001).

Differences in activation patterns between LV and septum

For the fifth and sixth VF episodes, which were recorded simultaneously from the LV and septum, decomposition of the weighted velocity vectors into x and y components over the LV yielded an x velocity (positive direction from posterior descending artery to LAD parallel to the AV groove) of 0.06 ⫾ 0.02 m/s, which was significantly different from 0 (P ⬍ .01, Figure 5). The x velocity over the septum (positive direction from anterior to posterior) was ⫺0.05 ⫾ 0.02 m/s, which also was significantly different from 0 (P ⬍ .01, Figure 5). The y-velocity vectors (positive direction from apex to base perpendicular to the AV grove) were 0.05 ⫾ 0.02 m/s on the LV (P ⬍ .01). However, the y-velocity vectors were 0.02 ⫾ 0.01 m/s on the septum, which was not significantly different from 0. We also compared the x- and y-velocity vectors among the three LV groups and found

Examples of VF activation sequences are shown in Figure 3. There was a significant overall multivariate difference between the septum and LV (P ⬍ .001, Figure 2). There was no significant difference with regard to the number of wavefronts per cm2; however, the incidence of collision and fractionation, mean area swept out by wavefronts, propagation speed, multiplicity, repeatability, and incidence of reentry all were smaller on the right side of the septum than on the LV (P ⬍ .01). The incidence of wavefronts that break through to the surface or block within the mapped region was larger in the septum (P ⬍ .01). The percentage of wavefronts that formed reentrant circuits was significantly

Directionality of VF

724

Heart Rhythm, Vol 2, No 7, July 2005

Figure 3 Simultaneous snapshots of activation during an 85-ms interval of ventricular fibrillation in the left ventricle (A) and septum (B). Frames are spaced every 5 ms. Numbers show time in milliseconds from the first frame. Each colored pixel is an electrode site at which dV/dt ⬍0.5 V/s sometime during the 5-ms interval represented by each frame. Different colored pixels indicate distinct isolated wavefronts. The overall wavefront direction on the LV is from the apex and posterior descending artery to the base and left anterior descending coronary artery (LAD). The overall wavefront direction on the septum is mainly from posterior to anterior. LCX ⫽ left circumflex coronary artery.

Huang et al

Activation Patterns of VF

Figure 4 Simultaneous unipolar recordings from the left ventricle (top trace) and septum (bottom trace) during ventricular fibrillation.

that the mean directions were not altered by bypass and RV incision (Figure 5).

Transmural peak frequency in the septum and peak frequency on the different parts of the mapping plaque The peak frequency measured from the left side, mid, and right side of the septum were 8.6 ⫾ 2.2, 8.7 ⫾ 1.6, and 8.9 ⫾ 1.9 Hz, respectively. The values were not significantly different by multivariate repeated measures. There were no significant differences for the peak frequencies in the four different parts for either the LV or septum plaques.

Discussion The main findings from this study of VF activation are as follows. The incidence of reentry was less on the right side of the septum than on the posterobasal LV epicardium. The septum activated less rapidly than the LV as estimated by the dominant frequency. The dominant frequencies on the LV side and middle of the septum were no higher than the dominant frequency on the mapped portion of RV side of the septum. These findings suggest that a mother rotor is not located in the central portion of the septum that was mapped. In both the LV and septal mapped regions, the mean direction of activation spread was away from the intersection of the posterior LV free wall and the posterior septum. This finding suggests that, if a mother rotor exists, it exists in the region where the posterior LV free wall joins the septum, which includes the posterior papillary muscle insertion.

Mother rotor hypothesis Several electrophysiologic mechanisms have been proposed for maintenance of VF.2,15–18 Most proposed mechanisms postulate the existence of numerous different reentrant wavefronts to maintain the arrhythmia.19 –21 Specifically, VF is thought to be maintained by multiple, disorganized, wandering wavelets that follow constantly changing reen-

725 trant pathways. However, data from small animals suggest that one or two primary wavefronts located in the regions with the shortest refractory periods and fastest activation rates drive the rest of the heart and that these regions, not the entire myocardium, are responsible for maintenance of VF.2,18 In these studies, the investigators used optical recording techniques to map guinea pig hearts during VF. They found that a single, stable rotor was located in the anterior LV, which activates extremely rapidly and drives the activations in the rest of the heart. Wavefronts from the high-frequency area propagate into the surrounding areas where Wenckebach block occurs because of the longer refractory periods in these regions. However, Choi et al19 did not find evidence of a mother rotor in optical recordings from the isolated guinea pig heart. If a mother rotor exits, this finding is significant not only for the mechanism of VF but also for the treatment of VF. If human VF activation is dominated by a few rapidly activating regions, it may be possible to significantly reduce the defibrillation threshold with a lead configuration that generates a high-shock electric field in these regions, or pacing can be performed to capture these regions to stop reentry. A series of studies was performed to search for the region where stable reentry exists in pigs during the first 20 seconds of VF in normal hearts. First, global peak frequency was estimated using sock22 and plunge needle electrodes6 that covered both ventricles. The studies revealed that the epicardial base has the highest peak frequency.6,22 Simultaneous epicardial mapping of RV and LV found that the LV has a higher incidence of reentry with higher peak frequency and more complex activation patterns than the RV.4 High-density mapping of the LV surface revealed that VF wavefronts tend to move from posterior to anterior and that more wavefronts propagate into the mapped region from the posterior edge of the mapping plaque than propagate off this edge of the mapping plaque.3,5 Because the RV activates more slowly than the LV,4 it is not a likely source for these wavefronts entering the posterior LV. All these observations point to the septum as a possible source for the dominant region during VF in pigs. In the present study, we mapped the right side of the septum and did not find evidence of a dominant domain there. Instead, we found that the right side of the septum exhibits less reentry and activates more slowly than the LV. The similar peak frequencies on both sides and the middle of the septum suggest that other portions of the septum do not contain a dominant region either. Therefore, the results of this study provide evidence that a mother rotor is not located in the septum of the pig during the first 20 seconds of VF. These results are consistent with the findings of Valderrábano et al,23 who did not observe a mother rotor on the cut transmural surface of the isolated perfused septum of pig heart. This finding, combined with the findings from earlier studies that mother rotors are not present on the epicardium of the anterior RV free wall or the base of the LV epicardial free wall,3,4 even though the epicardial region is the fastest

726

Figure 5 Velocity vector components. A: The x and y components of the velocity vector for the left ventricle (circles) and the septum (squares). B: The x component of the conduction velocity vector for each animal. The animal numbers are given at the bottom. C: The y component of the conduction velocity vector for each animal. All data are shown as mean ⫾ 95% confidence interval. LAD ⫽ left anterior descending coronary artery; LV ⫽ left ventricle; PDA ⫽ posterior descending artery.

Heart Rhythm, Vol 2, No 7, July 2005 activating region transmurally,6 raises the possibility that a mother rotor does not exist in the pig heart. However, two possible locations for a mother rotor remain untested: the posterior RV free wall and the region where the LV free wall intersects the posterior septum. These two locations are possible candidates for a mother rotor because they are the regions from which wavefronts can originate and propagate into the posterior LV free wall and posterior septum, as observed in this study. Of these two regions, the posterior RV free wall is less likely because its activation rate is slower than that of the LV4,22 and the mother rotor is thought to be located in the fastest activating region.24 Kim et al25 reported that reentry during VF in isolated, perfused pig RVs tends to occur about papillary muscles. Therefore, a possible site for a mother rotor is around the insertion of the posterior papillary muscle, which is at the intersection of the posterior LV free wall and the posterior septum. The fact that the wavefronts in the LV tended to propagate not only anteriorly but also basally in the plaque that covered the posterobasal LV is consistent with a mother rotor around the papillary muscle insertion, which is slightly more than halfway from the base to the apex. However, several considerations are not totally consistent with this idea. In one of the six hearts in the study by Nanthakumar et al3 and in one of the seven hearts in our study, VF wavefronts in the LV propagated toward the posterior interventricular groove, not away from it as in the other hearts. Perhaps in these two hearts, a mother rotor was located at another site, such as the insertion of the anterior papillary muscle. Although we found a significant tendency for the VF wavefronts to pass from the apex toward the base, in two of the hearts, including the heart in which wavefronts propagated toward the posterior interventricular groove (Figure 4B, animal 3), the wavefronts propagated from the base towards the apex. Nanthakumar et al3 did not report a significant direction from apex to base in their study. Although not mentioned in that study, in five of the six hearts the VF mean wavefront direction was from the apex to the base. The one exception was the heart in which activation also propagated toward the posterior interventricular groove (K Nanthakumar, personal communication). Another inconsistency with a mother rotor located at the posterior papillary muscle insertion is that this insertion is primarily endocardial, and Newton et al6 reported that the activation rate was significantly slower toward the endocardium than toward the epicardium in the LV free wall. In addition, they reported that the activation rate was more rapid in the basal half than in the apical half of the LV, whereas the insertion of the posterior papillary muscle is in the upper portion of the apical half of the LV. However, because the posterior papillary muscle partially inserts into the posterior septum so that it is partially hidden beneath the posterior descending coronary artery and the posterior RV free wall, this region has not yet been adequately mapped.

Huang et al

Activation Patterns of VF

Although difficult, such a mapping study is needed to test the hypothesis that a mother rotor is located at the insertion of the posterior papillary muscle.

Comparison of activation patterns in the septum and LV The number of wavefronts per cm2 is not significantly different between the septum and LV, but the lower multiplicity, fractionation, and incidence of collision in the septum indicate that VF activation in this area is less complex by these measures than in the LV. However, the higher incidence of block and breakthrough and the smaller mean area activated by each wavefront in the septum than in the LV indicate VF is less organized in the septum by other measures. These findings support the results of previous studies that VF organization is not a single measure but consists of several variables that do not always change in concert.26,27 The decreased conduction velocity in the septum compared with the LV is consistent with a lower dV/dt in the septum and probably is the reason for the lower peak frequency in the septum. Whereas the epicardial surface of the LV is smooth, the interventricular septum is corrugated with bundles of muscular fibers in the trabeculae carneae.28 The smaller area swept out by wavefronts may be related to the increased structural discontinuities and nonuniformities of tissue in the septum compared to the LV, which prevent the wavefronts from propagating smoothly. The high incidence of block and breakthrough also support this assumption. Our results differ from those of Ikeda et al,28 who reported more wavefronts in the septum than in the LV of isolated, perfused canine heart. They investigated the properties of the ventricles by isolating the ventricular free walls and septum individually, hence eliminating all interactions of the wavefronts between each structure. Other possible reasons for the different results are that activation patterns in the in vivo heart differ from those in the ex vivo heart,29 and the increased sympathetic tone caused by hypotension and ischemia resulting from VF cannot occur in the ex vivo heart.30

Study limitations The incision between the anterior RV and LV free walls probably altered the spread of VF wavefronts in the region. However, the insertion of the posterior RV free wall into the LV near the recording plaque on the posterior LV still was intact. Cardiopulmonary bypass has several deleterious effects on the heart.31 Although cardiopulmonary bypass and the RV incision procedures did not individually significantly alter the electrophysiologic measures in the posterior LV, they did alter the measures in a cumulative manner. Thus, the spread of activation away from the intersection of the posterior LV wall and the septum after bypass and RV incision may not have been present in the intact heart.

727 However, the main direction of spread of VF wavefronts in the LV was not altered by the procedures (Figure 5), providing supporting evidence that activation arose in this region in the intact heart as well. One of the detrimental effects of cardiopulmonary bypass is a degree of ischemia that causes the heart to be less well oxygenated at the time of VF induction. Thus, the ischemia caused by cardiopulmonary bypass before induction of VF may have been responsible for the observed quantitative changes in VF, causing the heart during the first 20 seconds of VF to be equivalent to normal hearts after at least 2 minutes of VF.32 The absence of a mother rotor in the mapped regions during the first 20 seconds of VF in normal pig hearts does not necessarily mean that stable reentry is not present during later periods of VF or in the presence of drugs or cardiac disease. Others have reported two types of VF: type I, which consists of constantly changing multiple wandering wavelets, and type II, which consists of more regularly repeating wavefronts and in which mothers rotors can sometimes be identified.33,34 Type II VF has been observed after administration of the drug D600, in acutely ischemic regions, and after longer duration of VF.20,34

Acknowledgments We thank Frank L. Vance, Melody A. Kinzalow, Reuben L. Collins, and Tracy L. Gamblin for assistance with the experimental preparation. We acknowledge Kate Sreenan for assistance with manuscript preparation.

References 1. Gray RA, Pertsov AM, Jalife J. Spatial and temporal organization during cardiac fibrillation. Nature 1998;392:675– 678. 2. Zaitsev AV, Berenfeld O, Mironov SF, Jalife J, Pertsov AM. Distribution of excitation frequencies on the epicardial and endocardial surfaces of fibrillating ventricular wall of the sheep heart. Circ Res 2000;86:408 – 417. 3. Nanthakumar K, Huang J, Rogers JM, Johnson PL, Newton JC, Walcott GP, Justice RK, Rollins DL, Smith WM, Ideker RE. Regional differences in ventricular fibrillation in the open-chest porcine left ventricle. Circ Res 2002;91:733–740. 4. Rogers JM, Huang J, Pedoto RW, Walker RG, Smith WM, Ideker RE. Fibrillation is more complex in the left ventricle than in the right ventricle. J Cardiovasc Electrophysiol 2000;11:1364 –1371. 5. Rogers JM, Huang J, Melnick SB, Ideker RE. Sustained reentry in the left ventricle of fibrillating pig hearts. Circ Res 2003;92:539 – 45. 6. Newton JC, Ideker RE. Estimated global transmural distribution of activation rate and conduction block during porcine and canine ventricular fibrillation. Circ Res 2004;94:836 – 842. 7. AHA Special Report. Position of the American Heart Association on Research Animal Use. Circulation 1985;71:849A– 850A. 8. Wolf PD, Rollins DL, Simpson EV, Smith WM, Ideker RE. A 528 channel system for the acquisition and display of defibrillation and electrocardiographic potentials. In Murray A, Arzbaecher R, editors: Proceedings Computers in Cardiology. Los Alamitos, CA: IEEE Computer Society Press, 1993:125–128.

728 9. Rogers J, Usui M, KenKnight B, Ideker R, Smith W. Recurrent wavefront morphologies: a method for quantifying the complexity of epicardial activation patterns. Ann Biomed Eng 1997;25:761– 768. 10. Rogers JM, Usui M, KenKnight BH, Ideker RE, Smith WM. A quantitative framework for analyzing epicardial activation patterns during ventricular fibrillation. Ann Biomed Eng 1997;25:749 –760. 11. Bollacker KD, Simpson EV, Hillsley RE, Blanchard SM, Gerstle RJ, Walcott GP, Callihan RC, King MC, Smith WM, Ideker RE. An automated technique for identification and analysis of activation fronts in a 2-D electrogram array. Comput Biomed Res 1994;27: 229 –244. 12. Huang J, Rogers JM, KenKnight BH, Rollins DL, Smith WM, Ideker RE. Evolution of the organization of epicardial activation patterns during ventricular fibrillation. J Cardiovasc Electrophysiol 1998;9: 1291–1304. 13. Welch PD. The use of Fast Fourier Transform for the estimation of power spectra: a method based on time averaging over short, modified periodograms. IEEE Trans Audio Electroacoust 1967; AU-15:70 –73. 14. Rogers JM, Huang J, Smith WM, Ideker RE. Incidence, evolution and spatial distribution of functional reentry during ventricular fibrillation in pigs. Circ Res 1999;84:945–954. 15. Moe GK. On the multiple wavelet hypothesis of atrial fibrillation. Arch Int Pharmacodyn 1962;140:183–187. 16. Weiss JN, Garfinkel A, Karagueuzian HS, Qu Z, Chen PS. Chaos and the transition to ventricular fibrillation: a new approach to antiarrhythmic drug evaluation. Circulation 1999;99:2819 –2826. 17. Gray RA, Jalife J, Panfilov AV, Baxter WT, Cabo C, Davidenko JM, Pertsov AM. Mechanisms of cardiac fibrillation: drifting rotors as a mechanism of cardiac fibrillation. Science 1995;270:1222–1225. 18. Samie FH, Berenfeld O, Anumonwo J, Mironov SF, Udassi S, Beaumont J, Taffet S, Pertsov AM, Jalife J. Rectification of the background potassium current: a determinant of rotor dynamics in ventricular fibrillation. Circ Res 2001;89:1216 –1223. 19. Choi BR, Liu T, Salama G. The distribution of refractory periods influences the dynamics of ventricular fibrillation. Circ Res 2001;88: e49 – e58. 20. Wu TJ, Lin S-F, Baber A, Qu Z, Garfinkel A, Weiss JN, Ting CT, Chen P-S. Mother rotors and the mechanisms of D600-induced type 2 ventricular fibrillation. Circulation 2004;110:2110 –2118. 21. Chen PS, Wu TJ, Ting CT, Karagueuzian HS, Garfinkel A, Lin SF, Weiss JN. A tale of two fibrillations. Circulation 2003;108:2298 – 2303. 22. Newton JC, Johnson PL, Justice RK, Smith WM, Ideker RE. Estimated global epicardial distribution of activation rate and conduction block during porcine ventricular fibrillation. J Cardiovasc Electrophysiol 2002;13:1035–1041. 23. Valderrábano M, Lee M-H, Ohara T, Lai AC, Fishbein MC, Lin S-F, Karagueuzian HS, Chen P-S. Dynamics of intramural and transmural

Heart Rhythm, Vol 2, No 7, July 2005

24.

25.

26.

27.

28.

29.

30.

31. 32.

33.

34.

reentry during ventricular fibrillation in isolated swine ventricles. Circ Res 2001;88:839 – 848. Samie FH, Mandapati R, Gray RA, Watanabe Y, Zuur C, Beaumont J, Jalife J. A mechanism of transition from ventricular fibrillation to tachycardia: effect of calcium channel blockade on the dynamics of rotating waves. Circ Res 2000;86:684 – 691. Kim YH, Xie F, Yashima M, Wu TJ, Valderrabano M, Lee MH, Ohara T, Voroshilovsky O, Doshi RN, Fishbein MC, Qu Z, Garfinkel A, Weiss JN, Karagueuzian HS, Chen PS. Role of papillary muscle in the generation and maintenance of reentry during ventricular tachycardia and fibrillation in isolated swine right ventricle. Circulation 1999;100: 1450 –1459. Huang J, Rogers JM, Killingsworth CR, Walcott GP, KenKnight BH, Smith WM, Ideker RE. Improvement of defibrillation efficacy and quantification of activation patterns during ventricular fibrillation in a canine heart failure model. Circulation 2001;103:1473– 1478. Huang J, Skinner JL, Rogers JM, Smith WM, Holman WL, Ideker RE. The effects of acute and chronic amiodarone on activation patterns and defibrillation threshold during ventricular fibrillation in dogs. J Am Coll Cardiol 2002;40:375–383. Ikeda T, Kawase A, Nakazawa K, Ashihara TT, Namba TT, Takahashi TK, Sugi K, Yamaguchi T. Role of structural complexities of septal tissue in maintaining ventricular fibrillation in isolated, perfused canine ventricle. J Cardiovasc Electrophysiol 2001;12:66 –75. Qin H, Kay MW, Chattipakorn N, Redden DR, Ideker RE, Rogers JM. Effects of heart isolation, voltage-sensitive dye, and electromechanical uncoupling agents on ventricular fibrillation. Am J Physiol Heart Circ Physiol 2003;284:H1818 –H1826. Opthof T, Ramdat Misier AR, Coronel R, Vermeulen JT, Verberne HJ, Frank RGJ, Moulijn AC, van Capelle FJL, Janse MJ. Dispersion of refractoriness in canine ventricular myocardium: effects of sympathetic stimulation. Circ Res 1991;68:1204 –1215. Murphy GJ, Angelini GD. Side effects of cardiopulmonary bypass: what is the reality? J Card Surg 2004;19:481– 8. Huang J, Rogers JM, Killingsworth CR, Singh KP, Smith WM, Ideker RE. Evolution of activation patterns during long-duration ventricular fibrillation in dogs. Am J Physiol Heart Circ Physiol 2004;286:H1193– H1200. Wu TJ, Lin SF, Weiss JN, Ting CT, Chen PS. Two types of ventricular fibrillation in isolated rabbit hearts: importance of excitability and action potential duration restitution. Circulation 2002; 106:1859 –1866. Liu Y-B, Pak H-N, Lamp ST, Okuyama Y, Hayashi H, Wu T-J, Weiss JN, Chen P-S, Lin S-F. Co-existence of two types of ventricular fibrillation during acute regional ischemia in rabbit ventricles. J Cardiovasc Electrophysiol 2004:1433–1440.