Vol. 281, Issue 1, H253-H265, July 2001
Patterns of wave break during ventricular fibrillation in
isolated swine right ventricle
Moon-Hyoung
Lee*,1,
Zhilin
Qu*,2,
Gregory A.
Fishbein1,
Scott T.
Lamp2,
Eugene H.
Chang1,
Toshihiko
Ohara1,
Olga
Voroshilovsky1,
Jong R.
Kil2,
Ali R.
Hamzei1,
Nina C.
Wang1,
Shien-Fong
Lin3,
James N.
Weiss2,
Alan
Garfinkel2,
Hrayr S.
Karagueuzian1, and
Peng-Sheng
Chen1
1 Division of Cardiology, Department of Medicine,
Cedars-Sinai Medical Center, and 2 Division of Cardiology,
Departments of Medicine and Physiology and Physiological Science,
University of California School of Medicine, Los Angeles, California
90048; and 3 Department of Physics and Astronomy, Vanderbilt
University, Nashville, Tennessee 37235
 |
ABSTRACT |
Several different patterns of wave
break have been described by mapping of the tissue surface during
fibrillation. However, it is not clear whether these surface patterns
are caused by multiple distinct mechanisms or by a single mechanism. To
determine the mechanism by which wave breaks are generated during
ventricular fibrillation, we conducted optical mapping studies and
single cell transmembrane potential recording in six isolated swine
right ventricles (RV). Among 763 episodes of wave break (0.75 times · s
1 · cm2), optical
maps showed three patterns: 80% due to a wave front encountering the
refractory wave back of another wave, 11.5% due to wave fronts passing
perpendicular to each other, and 8.5% due to a new (target) wave
arising just beyond the refractory tail of a previous wave. Computer
simulations of scroll waves in three-dimensional tissue showed that
these surface patterns could be attributed to two fundamental
mechanisms: head-tail interactions and filament break. We conclude that
during sustained ventricular fibrillation in swine RV, surface patterns
of wave break are produced by two fundamental mechanisms: head-tail
interaction between waves and filament break.
reentry; mapping; electrophysiology; action potentials; restitution
| |
INTRODUCTION |
ON THE BASIS OF
COMPUTER SIMULATIONS, Moe et al. (26) proposed the
multiple wavelet hypothesis of fibrillation, in which they postulated
that cardiac fibrillation is due to the presence of a number of
independent wavelets. The wavelets appear to propagate randomly through
the myocardium and are maintained by seemingly spontaneous wave
splitting (wave break) that constantly regenerates new daughter
wavelets. Subsequent experimental studies demonstrated the presence of
multiple wavelets in both atrial (7, 12, 19) and
ventricular fibrillation (VF) (16, 17, 21). From activation maps of the surface of the heart, two patterns of wave break
have been characterized: wave break when two waves intersect, often
perpendicularly (5, 21), and wave break when a wave front
collides with the refractory tail of a preceding spontaneous (13) or paced (10, 27) activation. A weakness
of electrode mapping studies is that the repolarization patterns of the
wave fronts cannot be determined (5). The wave front-wave
tail interaction can only be deduced based on the patterns of
activation not by direct visualization. Most of the optical
mapping studies of the wave break formation were based on the
interactions between wave fronts induced by pacing (10,
27) and not on the study of spontaneous occurring wave fronts
during Wiggers' stage II VF. Because of these limitations, it is
unclear which of these mechanisms of wave break formation are most
important and, indeed, whether they really are distinct mechanisms
rather than manifestations on the surface of the heart of wave break
occurring deeper in the tissue by a common mechanism. In the present
study, we used high-density optical mapping techniques
(23) and single cell transmembrane potential (TMP)
recordings to study both the activation and repolarization patterns
during VF in isolated swine right ventricles (RV) (17) and
to quantify the frequency of different surface patterns of wave break.
We then performed computer simulations of three-dimensional (3-D)
tissue to determine whether or not the same surface patterns could be
reproduced and to investigate the underlying mechanisms of wave break.
| |
METHODS |
Optical mapping of swine RV.
The research protocol was approved by the institutional animal care and
use committee and followed the guidelines of the American Heart
Association. Six farm pigs (25-32 kg) of either sex were used in
the study. The details of this model have been reported elsewhere
(17). Briefly, the RV was isolated, perfused with oxygenated Tyrode solution, and placed in a tissue bath with either the
epicardium (n = 3) or endocardium (n = 3) facing up for optical mapping. The pseudoelectrocardiogram was
monitored and recorded by a pair of widely spaced bipolar electrodes.
The optical mapping system used in the present study was similar to the
one described previously (23) except that the stimulating light was delivered to the tissue with fiber-optic bundles (model A08550, Fostec; Auburn, NY) through an interference filter (500 ± 30 nm, Omega Optical; Brattleboro, VT) from a stabilized 250-W tungsten
halogen lamp (model 66196, Oriel; Stratford, CT). The RV were stained
for 20 min with 1-2 µM pyrimidine
4-(2-(6-(dibutylamino)-2-naphthalenyl) ethenyl)-1-(3-sulfopropyl)
hydroxide (di-4-ANEPPS; Molecular Probes) added to the Tyrode solution.
The induced fluorescence was collected through a 600-nm long-pass glass
filter (R60, Nikon; Tokyo, Japan) and a 25-mm/f 0.85 video lens
(Fujinon CF25L, Fuji Photo Optical; Omiya City, Japan) with a 12-bit
digital charge-coupled device camera (CA-D1-0256T, Dalsa; Ontario,
Canada) programmed to sample at 3.75-ms intervals. The camera acquired
data from 96 × 96 sites simultaneously over a 35 × 35-mm2 area, resulting in a spatial resolution of 0.36 × 0.36 mm2/pixel. However, due to the need for temporal
and spatial averaging to reduce noise (see below), the effective
resolution was reduced. The face plate of the camera was cooled with an
ethylene-glycol coolant from a refrigerated water bath to 15°C. The
digital images were transferred to a personal computer with a frame
grabber (IC-PCI-DIG16, Imaging Technology; Bedford, MA). During the
experiment, the tissue was immobilized by pinning it to the base of the
tissue chamber. Because the RV did not contract effectively during VF,
we did not use electromechanical uncouplers in this study.
TMPs were recorded from a surface cell using standard techniques
(17). The digitization rate was 5 kHz with 12 bits of accuracy.
All hearts developed VF during excision. VF continued in the excised RV
(17). Continuous TMP recordings were made during VF for
30-60 s. The RV was then stained with di-4-ANEPPS, and the TMP
recordings were repeated. The patterns of activation of VF were
acquired by the charge-coupled device camera.
Signal processing.
The optical signals were temporally filtered and spatially averaged to
reduce noise. For temporal filtering, we applied a five-point time
median filter to each pixel. We took the original first five data
points (frames 1, 2, 3, 4,
and 5), found the median value of those points, and used
that as the new value for point (frame)
3. We then took the next original five points (i.e.,
frames 2, 3, 4, 5, and
6), found the median value of those, and used that as the
new value for point 4. We continued this exercise until the
end of the data. We then took the tracing, inverted the data, and
brought the baseline down to zero, which was defined by the average of
the five lowest fluorescent values recorded by that pixel. Afterward,
we range normalized each pixel. We found the five lowest and five
highest points and took the average of those numbers. We then adjusted
the fluorescent value of each pixel by the same amount so that the
highest pixel value was 255 and the lowest was 0. For each pixel on the
frame, we then averaged the fluorescent values of the pixel and its
eight surrounding pixels. We used this average as the new value for the
pixel. After those averaging procedures were completed, we repeated the
procedure for a second time. At that time, we rezeroed the signal by
bringing the baseline down to zero, defined by the average of the five lowest points of each pixel. We then range normalized the signals again. The maximal signal amplitude was coded white, representing a
fully depolarized state. The minimum signal amplitude was coded black,
representing a fully repolarized state. Each pixel was then assigned a
shade of gray between white and black.
For the computer-assisted automatic detection of wave break, we defined
the occurrence of wave break in a propagating wavelet as the point
where the activation wave front and the repolarization wave back joined
together. The computer first found every adjacent pair of pixels in the
frame that crossed the average value of the data. If the intensity of
the data on which the line coincides is increasing, that edge is
identified as the wave front and colored red. Otherwise, if it is
decreasing, the edge is identified as the wave back and colored blue
(25). The point where the red line meets the blue line is
the wave break. Each of these wave breaks was further analyzed by
visual inspection of consecutive frames. We included in analyses the
wave breaks occurring in large and coherent wave fronts. Excluded from
analyses were the wave fronts that changed from blue to red and vice
versa from frame to frame or the appearance of red and blue dots in
close vicinity, forming a mosaic pattern. We also excluded wave breaks
if there was no continuous propagation of the red line or continued
repolarization of the blue line around the wave break site over at
least two consecutive frames. Because of these selection criteria, the
absolute number of wave breaks reported in this paper might have
underestimated the real quantity of wave breaks in VF. However, the
wave breaks analyzed are likely to be true wave breaks.
We also generated color isochronal maps based on the location of wave
fronts at each frame. First, a start frame and an end frame are
selected by the user. For example, we chose frames
241-259. Second, the amount of frames to be skipped to get a
new color is selected by the user. For example, we chose three. Third,
the program generates on the screen isochronal lines according to the
user's chosen criteria. In this case, the program identified the
location of the activation wave fronts in frames 241,
242, and 243 and labeled those isochronal lines
red. The areas between the red lines were filled with red color. The
program then switched to orange to draw isochronal lines for
frames 244, 245, 246, and so on. At
this point, the areas between the red lines were red, the areas between
the orange lines were orange, and so on. The areas between two colors
were left blank. Therefore, there were gaps of colors on the map. The
program then filled in the gaps between adjacent colors with one color
halfway across and the other color the rest of the way across.
The single cell TMP recorded by glass microelectrode was analyzed to
determine the diastolic interval [DI; the time between 90% of
repolarization of the preceding action potential (AP) to the upstroke
of the current AP] and the AP duration (APD) from upstroke to 90%
repolarization (APD90). Details of the algorithm have been
published elsewhere (17). The DI and APD measured by glass
microelectrode may not be the same as those measured by optical mapping
techniques. Each pixel of the optical map represents fluorescent
changes of hundreds of cells. Because these cells may not activate and
repolarize in phase, the DI is often nonexistent on optical signals.
The APD restitution curve was created by plotting APD90
against the preceding DI. The restitution curve was generated by an exponential fit using ORIGIN 5.0 (Microcal Software; Northampton, MA).
ORIGIN was also used for statistical analyses. Results are expressed as
means ± SD. A P value of <0.05 was considered significant.
Computer simulation.
The partial differential equation for cardiac conduction is as follows
(11)
|
(1)
|
where V is the TMP, Cm the
membrane capacitance, and t is time.
Iion is the total ionic current density of the
membrane. The diffusion tensor (
) = 
/SvCm,
where is the conductivity tensor and
Sv is the surface-to-volume ratio of the cell.
We used no flux boundary condition (11): 
× 
V = 0, where
is the unit vector normal to the boundary. We
assumed the fibers were parallel and uniform in the
x-y plane but rotated along the
z-direction. Therefore, had the following
matrix structure (11)
|
(2)
|
where
|
(3)
|
D|| is the diffusion constant along
the fiber direction, and D
is the transverse
diffusion constant. We used D|| = 0.001 cm2/ms and D = 0.0002 cm2/ms. 
(z) is the angle between the fiber
and the x-axis. We used a uniform fiber rotation angle,
(z) = 
z, where is a constant. Iion in Eq. 1 was taken from the
phase 1 Luo-Rudy (LR1) AP model (24). We varied some
parameters to change the APD and APD restitution. The parameters were
selected so that APD restitution was steep enough to cause spontaneous
wave break in the simulated 3-D tissue when the thickness exceeded 0.4 cm. We used 
Na = 16 mS/cm2, K = 0.0423 mS/cm2, and Si = 0.047 mS/cm2, where is maximum value of the
channel conductance. We also sped up the Ca2+
kinetics, i.e., 
d 
0.5d and
f 0.5f. d and
f are the time constants for the recovery of activation
and inactivation, respectively, of the L-type calcium channel.
We used our advanced numerical method (28) to integrate
Eq. 1 with time steps varying from 0.02 to 0.2 ms. The space
step was 0.015 cm. The tissue size was 4.8 × 4.8 × 0.9 cm3. (In comparison, the swine RV has a thickness of
~5-7 mm.) The total fiber rotation angle was = 120°,
i.e., = 13.3°/mm. The simulation was started with a single
intact scroll wave. Simulations were carried out in the San Diego
Supercomputer Center.
| |
RESULTS |
APD restitution during VF.
To determine the APD restitution curve, DI and APD90 were
calculated from conventional single cell TMP during VF. The average cycle length during VF was 79 ± 17 ms, and the average DI was 12 ± 8 ms. Figure 1A
shows typical TMPs recorded during VF. The APD restitution curve was
created by plotting APD90 against the preceding DI (Fig.
1B). The maximum slope of the APD restitution curves ranged
from 2.4 to 7.6.
View larger version (32K):
[in this window]
[in a new window]
|
Fig. 1.
Action potentials (AP) recorded during ventricular
fibrillation (VF) and AP duration (APD) restitution curves during VF in
6 isolated swine right ventricles (RV). A: typical
transmembrane potential (TMP) recording showing that APD is positively
related to the preceding diastolic interval (DI). B: APD
restitution characteristics in all 6 tissues (1-6). The
maximum slope of the APD restitution curve (APDRmax) is
identified in each of the 6 tissues. CL, cycle length;
APD90, APD from upstroke to 90% repolarization.
|
|
Surface patterns of wave break during VF.
During sustained VF, multiple wavelets coexisted. A total of 763 episodes of wave break were observed in 83 s of VF. Thus new wave
break occurred 0.75 times · s1 · cm2. In 612 episodes (80%), wave break was caused by a wave front encountering the
trailing edge of refractoriness from a previous activation or
neighboring wave. Figure 2 shows an
example. Frame a shows a convex wave front at the bottom of
the mapped tissue moving upward. The central portion of the convex wave
front encounters residual refractoriness left over by a previous
activation (shown by the blue line). Unable to propagate into the
refractory region, the new wave front breaks in its central portion,
splitting into two daughter wavelets (frame b). The two
daughter wavelets turn on themselves, forming a figure-8 (yellow arrows
in frames b-f) around the site of the initial wave
break. The two wavelets then collide and fuse, forming a central common
pathway, which then propagates in both directions (double-headed yellow
arrows, frames f-h). Reentry continues after
frame h to complete two cycles of the figure-8 pattern. Note
that in frame f there is a mosaic pattern in the right upper
quadrant. The junctions of the red and blue dots in that region were
not included for wave break analyses.
View larger version (85K):
[in this window]
[in a new window]
|
Fig. 2.
The wave break and initiation of
reentry by a wave front encountering the trailing edge of
refractoriness from a neighboring wave. Frames a-h show
sequential wave maps recorded by the optical mapping system during VF.
The red line denotes the wave front, and the blue line denotes the wave
tail. Curved yellow arrows indicate the direction of the reentrant wave
fronts. The double-headed yellow arrow shows two waves traveling in
opposite directions. The white vertical bar in frame a = 1 cm.
|
|
Figure 3C shows an isochronal
map of the figure-8 reentrant loop shown in Fig. 2. The color
isochrones indicate the locations of the activation wave fronts at the
frame number shown on the color bar. There are eight colors,
representing activation of 15 frames (56.25 ms) of data. Therefore,
each color represents 7 ms of activation. This isochronal map does not
contain the first reentrant activation at pixels 1 and
2 shown in Fig. 3A. Optical signals in Fig. 3
from selected sites (black and white thin arrows in C) are
plotted in A and B, respectively. In Fig.
3A, optical signals in site 6 (downward orange
arrows) as well as in other sites show large APD oscillations preceding
wave break. The patterns of wave fronts associated with APD
oscillations were similar to each other. The first long red arrow
indicates the smooth uninterrupted propagation of the wave front
immediately before wave break. The short upward red arrow shows the
wave break, which was associated with a progressive shortening of DI
and APD and cessation of propagation in the central portion of the wave
front (black thick arrows), leading to block (double horizontal red
line segments). After wave break occurred, the wave traveled in the
opposite direction (long downward red arrow). Figure 3B
shows that the peripheral portion of the wave front kept propagating
forward without a significant shortening of either DI or APD. Figure
3D shows optical signals from six different sites (numbers
in circles) from one reentrant wave front (long red arrow) as indicated
in C (bottom).
View larger version (69K):
[in this window]
[in a new window]
|
Fig. 3.
Isochronal map (C)
and optical potentials (A, B, and D)
of wave break. This is the same episode as that shown in Fig. 2.
C, top: an isochronal map with arrows pointing to
the two wave break sites. C, bottom: the same
isochronal map with numbers and letters corresponding to the recordings
shown in A, B, and D. The yellow
segments in A, B, and D indicate the
time window during which the activations in frames a-h
of Fig. 2 were registered.
|
|
Perpendicular wave front intersection.
In 86 episodes (11%), wave break was created by a wave front crossing
roughly perpendicularly to the tail of another wave front, similar to
that reported previously in the canine RV in vivo (21).
Figure 4 shows an example. Frames
a-c show the first wave front propagating from the lower
right to the upper left corner. A second wave front arising at the
bottom left of the mapped region (frame a) spread upward and
to the right (frames b and c). As
this wave front propagated, its perpendicular intersection with the
residual refractoriness left over by the first wave front created
wave breaks (frames c and d) and a
counterclockwise reentrant wave front in the upper left quadrant of the
mapped tissue (frames d-h).
View larger version (82K):
[in this window]
[in a new window]
|
Fig. 4.
The wave break and initiation of
reentry by perpendicular interaction of two wave fronts. Frames
a-h show sequential dynamic wave maps of activation and
repolarization. The red line represents the wave front, and the blue
line shows the refractory wave tail. Curved yellow arrows indicate the
direction of wave front propagation.
|
|
Figure 5 shows the optical signals
(A and C) and an isochronal map of reentry
(B) for the episode shown in Fig. 4. The color isochrones
indicate the locations of the activation wave fronts at the frame
number shown on the color bar. There are seven colors, representing
activation of 19 frames (71.25 ms) of data. Therefore, each color
represents 10.2 ms of activation. Figure 5A shows optical signals from selected sites (black arrows) shown in B
(bottom). The interaction was associated with progressively
smaller potentials (black arrows in A), followed by
cessation of propagation and wave break (double horizontal red line
segments). Potentials just after wave break show that the direction of
propagation (the long downward red arrow) is opposite to that before
wave break. Figure 5C shows reentrant optical signals from
six different sites as indicated in B.
View larger version (64K):
[in this window]
[in a new window]
|
Fig. 5.
Isochronal map (B)
and optical signals (A and C) of wave break
associated with the perpendicular interaction of two waves. This is the
same episode as that shown in Fig. 4. The arrow in B,
top, shows the site of wave break. B,
bottom, shows the same isochronal map as B,
top, with numbers and letters corresponding to the
recordings shown in A and C. The yellow segments
in A and C indicate the time window during which
the activations in frames a-h of Fig. 4 were
registered.
|
|
Spontaneous focal wave fronts arising after the wave back of a
previous wave.
In 65 episodes (8.5%), we observed the generation of new wave breaks
without apparent wave-wave interactions. Figure
6 shows a typical example in which a wave
propagated from right to left (black arrows in frames
a-c). At its wave back, a new target wave suddenly appeared
on the right side (yellow arrows in frames b-h). It
propagated into nonrefractory tissue to the right but was blocked as it
propagated along the direction of the previous wave back. Two new wave
breaks were created (white arrows in frames i-l). Figure 6B shows the optical signals recorded from six
adjacent sites crossing the blue line (A, frame
a). The cells at sites 1-3 were in late phase 2 to
early phase 3 repolarization. In contrast, the cells at sites
4-6 were fully recovered. This is where the new activation
arose to propagate rightward.
View larger version (96K):
[in this window]
[in a new window]
|
Fig. 6.
Wave break occurring without
wave-wave interaction. A: a wave propagating from right to
left (black arrows). From the end of a previous wave back, a wavelet
emerged and propagated in the opposite direction (yellow arrows) with
new wave breaks (white arrows). B: optical signals. The
numbers correspond to the locations shown in A (frame
a). The yellow segment in B indicates the time window
during which the activations in frames a-l in
A were registered.
|
|
TMPs near the site of wave break.
TMP recordings registered by glass microelectrodes were compared with
the patterns of activation registered by the optical mapping system.
Outside of the region of wave break (>2 mm away), AP amplitude, APD,
the maximum change in voltage over time
(dV/dtmax), and DI averaged
54 ± 16 mV, 62 ± 16 ms, 33 ± 19 V/s, and 11 ± 8 ms, respectively. In five episodes, wave break occurred within 2 mm of
the microelectrode recording site, and TMPs showed significant (P < 0.001 for all comparisons) reductions in AP
amplitude (31 ± 15 mV), duration (36 ± 13 ms),
dV/dtmax (11 ± 7 V/s), and DI (5 ± 4 ms). Figure 7 shows the
single cell TMP (left) and the fluorescent changes
(right) at the site of wave breaks. A good correlation
between wave break sites and diminished AP (27) supports
the validity of our automated analysis for detecting wave fronts, wave
backs, and wave break locations (2).
View larger version (66K):
[in this window]
[in a new window]
|
Fig. 7.
Five examples of single cell TMP (left) and
simultaneously registered fluorescent signals (right) at or
near the wave break (middle). Each of the five rows
represents one episode. The horizontal black bar indicates the times
when wave break and reentrant wave fronts were recorded by the optical
mapping system. The yellow arrows point to the site where TMP and
fluorescent signals were recorded in each episode.
|
|
Computer simulations in 3-D cardiac tissue.
To investigate the mechanisms responsible for the surface patterns of
wave break observed experimentally, we performed computer simulations
of fibrillation in 3-D tissue using the LR1 ventricular AP model. With
steep APD restitution incorporated into the LR1 model, a scroll wave
initiated in simulated tissue broke up to form multiple wavelets
similar to fibrillation. Figures 8-10 illustrate that all three of
the surface patterns of wave break observed experimentally were also
observed on the surface of the simulated 3-D tissue. The first pattern,
that of a wave front running into the trailing edge of refractoriness
of a prior wave, is shown in Fig. 8. The
second pattern, involving collision of perpendicular wave fronts
(21), was also confirmed to be due to collision of a new
wave front with a region of refractoriness left over from the first
wave front (Fig. 9). Finally, the third
pattern, the emergence of a focal-appearing wave front in the trail of a previous wave, was due to development of severe twist in the scroll filament, which broke off intramurally to form a scroll ring
(Fig. 10). The initial target wave
pattern was produced when the arm of the scroll wave arrived at the
upper surface. Subsequently, the filament of the scroll ring drifted
upward and intersected the surface, where it broke to create two new
wave fronts.
View larger version (66K):
[in this window]
[in a new window]
|
Fig. 8.
Three-dimension simulation of
wave break by a wave front running into the trailing edge of
refractoriness. A: surface activation patterns (red = wave front, green = wave back) at the times indicated. The white
arrows indicate the region where this mechanism of wave break occurs.
B: corresponding scroll wave fronts in the tissue (red = rising membrane voltage). C: blowup of the region of wave
break on the upper surface (near the white arrows in A).
Residual refractoriness (green) was left over by a previous wave front,
and when the next wave (red) encountered this refractory region, wave
break occurred, generating two new scroll waves (compare with the
experimental wave break in Fig. 2).
|
|
View larger version (52K):
[in this window]
[in a new window]
|
Fig. 9.
Three-dimensional simulation of
wave break by perpendicular intersection. The surface activation
patterns (A), the scroll wave fronts in the tissue
(B) and the filament (C), and a blowup of the
region of wave break (D) (near the white arrows in
A) are shown. A wave front propagated upward to the right,
leaving an area of residual refractoriness at its wave back. A
different wave front propagated upwards from the left, perpendicular to
the first wave front. It encountered residual refractoriness, causing
wave break and a new scroll filament to form (compare with the
experimental wave break in Fig. 4).
|
|
View larger version (49K):
[in this window]
[in a new window]
|
Fig. 10.
Three-dimensional simulation of
wave break by filament break. A and B: surface
activation patterns (A) and the scroll wave fronts in the
tissue (B). C: scroll filaments in the tissue.
The black filament developed a severe twist that budded off to form a
new scroll ring in C (frame b, blue arrow), marking the
appearance of the target wave in A (frame
b, white arrow) as the arm of the scroll ring emergeed at the top
surface. When the scroll ring filament subsequently arrived at the
surface, it broke in two (C; frame c, blue
arrows), apparent on the surface as two new wave breaks (A;
frame c, white arrows). Further breaks in the black filament
formed new scroll waves (C; frame c), whereas the
green filament disappeared at the lower border (compare with the
experimental wave break in Fig. 6).
|
|
The computer model does not contain papillary muscle or other
endocardial structural inhomogeneities. The only inhomogeneity was the
myocardial fiber orientation. Because the wave break occurs everywhere
in the mapped region, it is not a phenomenon specific to the papillary
muscle (18) or other endocardial structural inhomogeneities.
| |
DISCUSSION |
The main new finding in this study is that different surface
patterns of wave break during fibrillation are all consistent with two
fundamental mechanisms of wave break, namely head-tail interactions
when a wave front encounters residual refractoriness from a previous
wave and filament break. The relation between local residual
refractoriness and the formation of wave break is consistent with that
proposed by Krinsky (20). The filament break, as first
reported by Fenton and Karma (11), can also contribute to
the generation of wave breaks. We arrived at this conclusion because,
using a detailed AP model in simulated 3-D tissue, we could reproduce
all of the surface patterns of wave break observed experimentally by
these two mechanisms. The mechanism of wave break could be
unequivocally identified in the simulation because events below the
tissue surface, as well as the movement of the scroll filaments, could
be directly monitored. Because the simulated 3-D tissue was completely
homogeneous and isotropic, these data also suggest that fixed tissue
anatomic and electrophysiological heterogeneities are not necessary to
account for the observed patterns of wave break during fibrillation.
Spontaneous wave breaks and the perpetuation of VF.
According to the multiple wavelet hypothesis (26),
constant formation of new wavelets occurs through the process of wave splitting (wave break), in which a wave splits into new (daughter) wavelets. The process of wave break depends on the presence of electrophysiological heterogeneity, where regions of refractoriness, arising either dynamically from cardiac restitution properties or
anatomically from obstacles or regional electrophysiological differences, cause localized propagation failure. Because the remaining
portion of the wave fronts continues to propagate, it may result in a
complete reentrant excitation (a spiral wave) (6, 9, 33).
Therefore, the wave break provides new opportunities for reentry,
perpetuating fibrillation. However, because of the complicated patterns
of activation, not all wave breaks lead to the formation of complete
reentry, no spiral wave or reentry is generated, and no phase
singularity is detected. Therefore, the number of wave breaks
determined in this study may not be equal to the number of phase
singularities determined by other investigators (13).
Recently, optical mapping techniques (30) have allowed
direct measurement of repolarization as well as activation during cardiac arrhythmias in animal models (8). Gray et al.
(13) showed an example in which a wave front encountered
refractory tissue, causing wave break. In the present study, we
extended Gray et al.'s (13) observations by
systematically analyzing the patterns of all wave breaks during
sustained VF in perfused swine RV. We found that the mechanism
described by Gray et al. (13) was the most common (80%),
but, in addition, we observed two other surface patterns that occurred
less frequently. In one pattern, the intersection of two perpendicular
wave fronts produced wave break, as has been described previously by
our group using multielectrode mapping (21). We
hypothesized that the wave break occurs due to interaction of the
second perpendicular wave front with the wave back of the first wave.
The present study confirms that hypothesis. In another pattern, a focal
target wave front appears in the wake of the passing wave, which
develops wave break along the receding wave back of the previous wave
as it expands. In the simulation, this pattern was observed when a
twist developed in the filament of a scroll wave. The severe twist
caused a section of the filament to bud off and form a separate scroll
ring (circular filament), as shown in Fig. 10, which subsequently
migrated to the surface. Initially, the arm of the scroll emerged,
producing the target wave. Subsequently, when the filament arrived at
the surface, two new wave breaks appeared in the target wave. The formation of the new wave breaks was facilitated by preexisting refractoriness due to incomplete recovery of the previous wave front.
In other simulations, we also observed that the twist in the original
filament sometimes arrived at the surface and broke before a scroll
ring had time to form, producing the same pattern observed experimentally.
Recovery of excitability and the formation of wave break.
Given that these were normal hearts, the spatial heterogeneity in
refractoriness was unlikely to be due to fixed anatomic or
electrophysiological heterogeneity. Rather, it is more likely that the
heterogeneity arose from dynamic processes, i.e., APD and conduction
velocity restitution (4). Consistent with this idea,
the slope of the APD restitution curve during VF was found to
be steep (Fig. 1). Although there is a lot of scatter in the data
points on the APD restitution curve, the best fit of an exponential curve to the data in Fig. 1, taking into consideration that APD restitution is not single valued during VF, gave a slope with a maximum
value >1. A maximum slope of >1 is an important criteria for wave
front instability during cardiac arrhythmias (31). A steep
restitution curve contributes to the development of large changes of
APD with small changes in DI (14, 15). When APD is
shortened and small, the safety factor of propagation diminishes, resulting in wave break. Compatible with this hypothesis, extremely short DIs and APDs (black arrows in Fig. 3A) were noted
before the occurrence of wave break in both the experiment and the
simulation. If we deliberately made the slope of APD restitution more
shallow by adjusting model parameters, spiral wave breakup and
fibrillation could have been prevented (29).
Limitations of the study.
Optical mapping is limited to the tissue surface, whereas wave break
and reentry occurred in a 3-D structure. Therefore, we depended on
simulations in 3-D tissue to infer what was happening beneath the
surface in the RV preparations so that we could interpret the surface
patterns of wave break. Thus we cannot exclude the possibility that
other mechanisms of wave break may also be at work. This is
particularly true for the filament break hypothesis because we were not
able to map the filaments under the surface. An alternative hypothesis
to explain the patterns seen in Fig. 6 is reflected reentry
(1), which occurs when recovered tissues and the
recovering or depolarized tissues are adjacent to each other. The
differences of membrane potential results in excitation of the
recovered tissue. Mapping study would show that a wave front (red)
develops from the edge of repolarizing tissue (blue), as shown in Fig.
6. Other possible mechanisms include breakthrough by an entirely
different scroll wave or a focal source that happens to be
active in the area. We also underestimated the frequency of wave breaks
because we could not directly detect wave break beneath the
surface. The inability to map 3-D impulse propagation also prevented
the accurate measurement of conduction velocity, which may be important
in the generation and maintenance of wave break (4).
Another limitation is that we only studied normal excitable tissue
without apparent anatomic obstacles. Therefore, these results may not
be applicable to diseased myocardium.
Wu et al. (34) showed that, in a beating heart, the AP
measurements could be distorted by myocardial movement. Although VF is
not associated with effective and synchronized contractions, some
myocardial movement may be present (32). It was unclear if
these small myocardium movements prevented us from accurately studying
the characteristics of wave breaks and number of wave fronts in VF. We
(22) recently performed an experiment to compare the
effects of electromechanical uncouplers such as diacetyl monoximide (DAM) and cytochalasin D (Cyto D) on VF activation patterns. Our results showed that DAM significantly reduced the slope of APD restitution, the Kolmogorov entropy, and the number of wavelets in VF.
On the other hand, the slope of APD restitution curve, the Kolmogorov
entropy, and the number of wavelets were the same at baseline (no
electromechanical uncoupler) and during Cyto D infusion. These findings
suggest that, for optical mapping of VF, electromechanical uncouplers
are not needed. However, the same conclusion does not apply to the
studies in sinus or paced rhythms, when large movement artifacts may
distort the optical image. In those situations, Cyto D (3,
34) is preferable to DAM as an electromechanical uncoupler.
In addition to the above experimental justification for not using an
electromechanical uncoupler during the study, our 3-D simulation
studies, which are immune to motion artifacts, reproduced all
the mechanisms of wave breaks observed in the optical mapping studies.
These findings further support the major conclusions of the present study.
The optical signals and the TMP recordings were not exactly
the same (Fig. 7). Movement artifacts could partially account for the
differences. In addition, the TMP recordings were made with a standard
glass microelectrode that registers the TMP from a single cell. In
contrast, the optical signals included the activations from multiple
cells. Because the activation patterns in VF are complex, neighboring
cells may be activated by different wave fronts. Therefore, recordings
from a single cell and from a group of cells may look different.
In conclusion, during VF, wave break occurs when wave fronts encounter
residual refractoriness left over from a previous wave or when a scroll
wave filament arrives at the surface and breaks. These two fundamental
mechanisms are sufficient to explain all of the surface patterns of
wave break observed experimentally during sustained VF in normal swine RV.
| |
ACKNOWLEDGEMENTS |
We thank Ling-Tao Fan, Avile McCullen, Meiling Yuan, and Lucas
Huang for technical assistance and Elaine Lebowitz for administrative assistance.
| |
FOOTNOTES |
*
M.-H. Lee and Z. Qu contributed equally to this study.
This study was done during the tenure of a Fellowship Grant from the
College of Medicine, Yonsei University, Seoul, Korea (to M.-H. Lee).
This study was supported in part by a Myung Sun Kim Memorial Foundation
Grant (to M.-H. Lee), a Cedars-Sinai Electrocardiographic Heartbeat
Organization Foundation Award (to H. S. Karagueuzian), the Kawata
and Laubisch Endowments (to J. N. Weiss), a Pauline and Harold
Price Endowment (to P.-S. Chen), National Heart, Lung, and Blood
Institute Grants P50-HL-52319, R01-HL-58533, and R01-HL-66389, American
Heart Association National Center Grants-in-Aid 9750623N and 9950464N,
a University of California-Tobacco Related Diseases Research Program
6RT-0020, and the Ralph M. Parsons Foundation, Los Angeles, CA.
Address for reprint requests and other correspondence: P.-S. Chen, Rm.
5342, CSMC, 8700 Beverly Blvd., Los Angeles, CA 90048 (E-mail: chenp{at}cshs.org).
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 16 March 2000; accepted in final form 22 February 2001.
| |
REFERENCES |
1.
Antzelevitch, C,
Jalife J,
and
Moe GK.
Characteristics of reflection as a mechanism of reentrant arrhythmias and its relationship to parasystole.
Circulation
61:
182-191,
1980[Abstract/Free Full Text].
2.
Athill, CA,
Ikeda T,
Kim YH,
Wu T-J,
Fishbein MC,
Karagueuzian HS,
and
Chen P-S.
Transmembrane potential properties at the core of functional reentrant wavefronts in isolated canine right atria.
Circulation
98:
1556-1567,
1998[Abstract/Free Full Text].
3.
Biermann, M,
Rubart M,
Moreno A,
Wu J,
Josiah-Durant A,
and
Zipes DP.
Differential effects of cytochalasin D and 2,3-butanedione monoxime on isometric twitch force and transmembrane action potential in isolated ventricular muscle: implications for optical measurements of cardiac repolarization.
J Cardiovasc Electrophysiol
9:
1348-1357,
1998[ISI][Medline].
4.
Cao, J-M,
Qu Z,
Kim Y-H,
Wu T-J,
Garfinkel A,
Weiss JN,
Karagueuzian HS,
and
Chen P-S.
Spatiotemporal heterogeneity in the induction of ventricular fibrillation by rapid pacing: importance of cardiac restitution properties.
Circ Res
84:
1318-1331,
1999[Abstract/Free Full Text].
5.
Cha, Y-M,
Birgersdotter-Green U,
Wolf PL,
Peters BB,
and
Chen P-S.
The mechanisms of termination of reentrant activity in ventricular fibrillation.
Circ Res
74:
495-506,
1994[Abstract/Free Full Text].
6.
Chen, P-S,
Wolf PD,
Dixon EG,
Danieley ND,
Frazier DW,
Smith WM,
and
Ideker RE.
Mechanism of ventricular vulnerability to single premature stimuli in open-chest dogs.
Circ Res
62:
1191-1209,
1988[Abstract/Free Full Text].
7.
Cox, JL,
Canavan TE,
Schuessler RB,
Cain ME,
Lindsay BD,
Stone C,
Smith PK,
Corr PB,
and
Boineau JB.
The surgical treatment of atrial fibrillation. II. Intraoperative electrophysiologic mapping and description of the electrophysiologic basis of atrial flutter and atrial fibrillation.
J Thorac Cardiovasc Surg
101:
406-426,
1991[Abstract].
8.
Davidenko, JM,
Persow AV,
Salomonza R,
Baxter W,
and
Jalife J.
Sustained vortex-like waves in normal isolated ventricular muscle.
Proc Natl Acad Sci USA
355:
349-351,
1990.
9.
Davidenko, JM,
Pertsov AM,
Salomonsz R,
Baxter W,
and
Jalife J.
Stationary and drifting spiral waves of excitation in isolated cardiac tissue.
Nature
355:
349-351,
1992[Medline].
10.
Davidenko, JM,
Salomonsz R,
Pertsov AM,
Baxter WT,
and
Jalife J.
Effects of pacing on stationary reentrant activity. Theoretical and experimental study.
Circ Res
77:
1166-1179,
1995[Abstract/Free Full Text].
11.
Fenton, F,
and
Karma A.
Vortex dynamics in three-dimensional continuous myocardium with fiber rotation: filament instability and fibrillation.
Chaos
8:
20-47,
1998[ISI][Medline].
12.
Gray, RA,
Pertsov AM,
and
Jalife J.
Incomplete reentry and epicardial breakthrough patterns during atrial fibrillation in the sheep heart.
Circulation
94:
2649-2661,
1996[Abstract/Free Full Text].
13.
Gray, RA,
Pertsov AM,
and
Jalife J.
Spatial and temporal organization during cardiac fibrillation.
Nature
392:
75-78,
1998[Medline].
14.
Karagueuzian, HS,
Khan SS,
Hong K,
Kobayashi Y,
Denton T,
Mandel WJ,
and
Diamond GA.
Action potential alternans and irregular dynamics in quinidine-intoxicated ventricular muscle cells. Implications for ventricular proarrhythmia.
Circulation
87:
1661-1672,
1993[Abstract/Free Full Text].
15.
Karma, A.
Electrical alternans and spiral wave breakup in cardiac tissue.
Chaos
4:
461-472,
1994[ISI][Medline].
16.
KenKnight, BH,
Bayly PV,
Gerstle RJ,
Rollins DL,
Wolf PD,
Smith WM,
and
Ideker RE.
Regional capture of fibrillating ventricular myocardium: evidence of an excitable gap.
Circ Res
77:
849-855,
1995[Abstract/Free Full Text].
17.
Kim, Y-H,
Garfinkel A,
Ikeda T,
Wu T-J,
Athill CA,
Weiss JN,
Karagueuzian HS,
and
Chen P-S.
Spatiotemporal complexity of ventricular fibrillation revealed by tissue mass reduction in isolated swine right ventricle. Further evidence for the quasiperiodic route to chaos hypothesis.
J Clin Invest
100:
2486-2500,
1997[ISI][Medline].
18.
Kim, Y-H,
Xie F,
Yashima M,
Wu T-J,
Valderrábano M,
Lee M-H,
Ohara T,
Voroshilovsky O,
Doshi RN,
Fishbein MC,
Qu Z,
Garfinkel A,
Weiss JN,
Karagueuzian HS,
and
Chen P-S.
Role of papillary muscle in the generation and maintenance of reentry during ventricular tachycardia and fibrillation in isolated swine right ventricle.
Circulation
100:
1450-1459,
1999[Abstract/Free Full Text].
19.
Konings, KTS,
Kirchhof CJHJ,
Smeets JLRM,
Wellens HJJ,
Penn OC,
and
Allessie MA.
High-density mapping of electrically induced atrial fibrillation in humans.
Circulation
89:
1665-1680,
1994[Abstract/Free Full Text].
20.
Krinsky, VI.
Spread of excitation in an inhomogeneous medium.
Biophys J
11:
776-784,
1966.
21.
Lee, JJ,
Kamjoo K,
Hough D,
Hwang C,
Fan W,
Fishbein MC,
Bonometti C,
Ikeda T,
Karagueuzian HS,
and
Chen P-S.
Reentrant wave fronts in Wiggers' stage II ventricular fibrillation: characteristics, and mechanisms of termination and spontaneous regeneration.
Circ Res
78:
660-675,
1996[Abstract/Free Full Text].
22.
Lee, M-H,
Lin S-F,
Ohara T,
Omichi C,
Chudin E,
Garfinkel A,
Weiss JN,
Karagueuzian HS,
and
Chen P-S.
The effects of diacetyl monoxime and cytochalasin D on action potential restitution and the dynamics of ventricular fibrillation in isolated swine right ventricles.
Am J Physiol Heart Circ Physiol
280:
H2689-H2696,
2001[Abstract/Free Full Text].
23.
Lin, S-F,
Roth BJ,
and
Wikswo JP, Jr.
Quatrefoil reentry in myocardium: an optical imaging study of the induction mechanism.
J Cardiovasc Electrophysiol
10:
574-586,
1999[ISI][Medline].
24.
Luo, CH,
and
Rudy Y.
A model of the ventricular cardiac action potential. Depolarization, repolarization, and their interaction.
Circ Res
68:
1501-1526,
1991[Abstract/Free Full Text].
25.
Mandapati, R,
Asano Y,
Baxter WT,
Gray R,
Davidenko J,
and
Jalife J.
Quantification of effects of global ischemia on dynamics of ventricular fibrillation in isolated rabbit heart.
Circulation
98:
1688-1696,
1998[Abstract/Free Full Text].
26.
Moe, GK,
Rheinboldt WL,
and
Abildskov JA.
A computer model of atrial fibrillation.
Am Heart J
64:
200-220,
1964.
27.
Pertsov, AM,
Davidenko JM,
Salomonsz R,
Baxter WT,
and
Jalife J.
Spiral waves of excitation underlie reentrant activity in isolated cardiac muscle.
Circ Res
72:
631-650,
1993[Abstract/Free Full Text].
28.
Qu, Z,
and
Garfinkel A.
An advanced algorithm for solving partial differential equation in cardiac conduction.
IEEE Trans Biomed Eng
46:
1166-1168,
1999[ISI][Medline].
29.
Qu, Z,
Weiss JN,
and
Garfinkel A.
Cardiac electrical restitution properties and stability of reentrant spiral waves: a simulation study.
Am J Physiol Heart Circ Physiol
276:
H269-H283,
1999[Abstract/Free Full Text].
30.
Salzberg, BM,
Davila HV,
and
Cohen LB.
Optical recording of impulses in individual neurones of an invertebrate central nervous system.
Nature
246:
508-509,
1973[Medline].
31.
Weiss, JN,
Garfinkel A,
Karagueuzian HS,
Qu Z,
and
Chen P-S.
Chaos and the transition to ventricular fibrillation: a new approach to antiarrhythmic drug evaluation.
Circulation
99:
2819-2826,
1999[Abstract/Free Full Text].
32.
Wiggers, CJ.
The mechanism and nature of ventricular fibrillation.
Am Heart J
20:
399-412,
1940[ISI].
33.
Winfree, AT.
Spiral waves of chemical activity.
Science
175:
634-636,
1972[Abstract/Free Full Text].
34.
Wu, J,
Biermann M,
Rubart M,
and
Zipes DP.
Cytochalasin D as excitation-contraction uncoupler for optically mapping action potentials in wedges of ventricular myocardium.
J Cardiovasc Electrophysiol
9:
1336-1347,
1998[ISI][Medline].
Am J Physiol Heart Circ Physiol 281(1):H253-H265
0363-6135/01 $5.00
Copyright © 2001 the American Physiological Society
TOP
ABSTRACT
INTRODUCTION
