Ventricular Fibrillation Conduction through an Isthmus
of Preserved Myocardium between Radiofrequency
Lesions
FRANCISCO J. CHORRO, M.D., PH.D.,*,† XAVIER IBAÑEZ-CATALÁ, I.E.E.,††
ISABEL TRAPERO, PH.D.,‡ LUIS SUCH-MIQUEL, PH.D.,§ FRANCISCA PELECHANO, M.D.,
PH.D.,† JOAQUÍN CÁNOVES, M.D., PH.D.,* LUIS MAINAR, M.D., PH.D.,*
ÁLVARO TORMOS, PH.D.,†† JOSE M. CERDÁ, M.D., PH.D.,¶ ANTONIO ALBEROLA, M.D.,
PH.D.,** and LUIS SUCH, M.D., PH.D.**
From the *Service of Cardiology, Valencia University Clinic Hospital Incliva, Valencia, Spain; Departments of
†Medicine, ‡Infirmary, §Physiotherapy, ¶Pathology, **Physiology, Valencia University, Estudi General, Valencia,
Spain; and ††Department of Electronics, Valencia Polytechnic University, Valencia, Spain
Background: Selective local acceleration of myocardial activation during ventricular fibrillation (VF)
contributes information on the interactions between neighboring zones during the arrhythmia. This
study analyzes these interactions, centering the observations on an isthmus of myocardium between
two radiofrequency (RF) lesions.
Methods: In nine isolated rabbit hearts, a gap of preserved myocardium was established between two
RF lesions in the anterolateral left ventricle (LV) wall. Before, during, and after increasing the spatial
heterogeneity of VF by local myocardial stretching, VF epicardial recordings were obtained.
Results: Local stretch in the anterior LV wall decreased the excitable window (17 ± 7 ms vs 26 ± 7 ms; P
< 0.05) and increased the dominant frequency (DFr; 18.9 ± 5.0 Hz vs 15.2 ± 3.6 Hz; P < 0.05) in this zone,
without changes in the non-stretched posterolateral zone (25 ± 4 ms vs 27 ± 6 ms, ns and 14.1 ± 2.7 Hz vs
14.3 ± 3.0 Hz, ns). The DFr ratio at both sides of the gap was inversely correlated to the excitable window
ratio (R = −0.57; P = 0.002). Before (31% vs 26%), during (29% vs 22%), and after stretch suppression
(35% vs 25%), the wavefronts passing through the gap from the posterolateral to the anterior LV wall
were seen to predominate. The number of wavefronts that passed from the anterior to the posterolateral
LV wall was related to the excitable window in this zone (R = 0.41; P = 0.03).
Conclusions: The VF acceleration induced in the stretched zone does not increase the flow of wavefronts
toward the non-stretched zone in the adjacent gap of preserved myocardium. The absence of significant
changes in the electrophysiological parameters of the non-stretched myocardium limits the arrival of
wavefronts in this zone. (PACE 2013; 36:286–298)
Cardiac mapping, mechanical stretch, myocardial activation, radiofrequency, spectral analysis,
ventricular fibrillation
Introduction
Myocardial activation during ventricular
fibrillation (VF) is a complex phenomenon, and
its analysis has produced a variety of data
on the mechanisms that initiate, perpetuate,
Grants: This work was supported by Spanish Ministry of
Science “Instituto de Salud Carlos III” Grants FIS PS09/02417
and RETIC REDINSCOR RD06/0003/0010, and by “Generalitat
Valenciana” Grant PROMETEO 2010/093.
Address for reprints: Francisco J. Chorro, M.D., Ph.D., Servicio
de Cardiologı́a. Hospital Clı́nico Universitario, Avda. Blasco
Ibañez 17, 46010, Valencia, Spain. Fax: 34963862658; e-mail:
[email protected]
Received January 17, 2011; revised October 14, 2012; accepted
October 23, 2012.
doi: 10.1111/pace.12060
modify, or interrupt this arrhythmia.1–9 One of
the aspects that have generated interest among
investigators is the interaction between different
myocardial zones during VF, because there have
been reports of regional differences in myocardial
activation frequency during the arrhythmia.10–21
The magnitude of these differences, the location of
the zones with the greatest activation frequencies,
and the role played by them during VF have been
the subject of a number of studies.10–16,18,19,21 Most
such studies have analyzed the heterogeneities
that occur spontaneously during VF, under conditions where coronary perfusion is both maintained
and suppressed, and both shortly after the start of
VF (short duration VF) and after prolonged periods
of time (long duration VF).
The isolation or suppression of a given myocardial zone, as well as the selective acceleration
©2012, The Authors. Journal compilation ©2012 Wiley Periodicals, Inc.
286
March 2013
PACE, Vol. 36
VF CONDUCTION BETWEEN RADIOFREQUENCY LESIONS
or slowing of local activation, allow us to study
its influence and repercussions upon activation
of the remaining myocardium. Accordingly, areas
of myocardium have been selectively isolated or
modified with the purpose of evaluating how VF
activation is changed by these maneuvers.22–30 The
procedures available for accentuating the regional
differences range from the placement of barriers
or obstacles that complicate conduction between
contiguous zones of myocardium to techniques
that selectively accelerate or slow the activation
of a given ventricular myocardial zone during
the arrhythmia. Regarding the local acceleration
or slowing of VF, local modification of the
characteristics of myocardial activation during VF
has been carried out using different techniques
such as myocardial stretch or localized thermal
variations.27,29,30
Mechanical stretch is an arrhythmogenic
factor. The electrophysiological modifications
produced by myocardial stretch enhance electrophysiological heterogeneity.27,30 Stretch is present
during acute changes in ventricular or atrial
load, as occurs in different clinical scenarios
such as the Valsalva maneuver, pulmonary
thromboembolism, the mechanical complications
of acute myocardial infarction, the onset of tachyarrhythmias, cardiopulmonary bypass during
open-heart surgery, or during regional wall motion
abnormalities, as in the border zone between
normal and ischemic myocardium.
The role played by the zones with the
greatest activation frequencies during fibrillation
in relation to maintenance of the arrhythmia is the
subject of debate. It has been postulated3,10–12 that
VF is the result of zones with increased activation
frequencies that function as driving zones and give
rise to fibrillatory conduction to the rest of the
myocardium—thereby determining maintenance
of the arrhythmia. In this scenario, local interventions designed to eliminate the driving zones,
e.g., radiofrequency (RF) ablation, could interfere
with the maintenance of fibrillation and facilitate
its interruption, either spontaneously or through
pharmacological or electrical procedures.
In an earlier study, we observed that the acceleration of VF produced by myocardial stretching
in the left ventricular anterior free wall did not
give rise to changes in the VF activation patterns
observed at a distance in the left ventricular
posterior free wall,27 and we speculated that local
modifications in electrophysiological properties,
when confined to zones of limited extension,
would induce few changes in the global activation
process during VF. On the other hand, in a similar
experimental preparation, we observed that the
suppression of different zones of ventricular
myocardium, including the anterior part of the left
PACE, Vol. 36
ventricular wall, did not impede the maintenance
of VF.28
This study aims to obtain further information
on the interactions between two myocardial
regions during VF, centering the observations on
an isthmus of preserved myocardium between two
linear RF lesions produced in the anterolateral
wall of the left ventricle. The experimental
model used allows us to create a gradient of
activation frequencies during VF by applying local
mechanical stretch and to analyze the wavefronts
flow between an accelerated zone and a part of the
adjacent myocardium, focusing the analysis in a
narrow zone.
Methods
Experimental Preparation
The procedures employed in this study were
in accordance with the guidelines for the care
and use of animals of the US National Institutes
of Health (NIH) Publication No. 85–23 (revised
1985), and with the Institutional Animal Care and
Use Committee. Nine isolated and perfused rabbit
hearts were used. After anesthesia with ketamine
(25 mg/kg i.m.) and heparinization, the animals
were sacrificed, and the hearts were removed
and immersed in cold (4◦ C) Tyrode solution.
Following isolation, the aorta was connected to a
Langendorff system for perfusing Tyrode solution
at a pressure of 60 mm Hg and a temperature of
37 ± 0.5◦ C. The millimolar composition of the
perfusion fluid was as follows: NaCl 130, NaHCO3
24.2, KCl 4.7, CaCl2 2.2, NaH2 PO4 1.2, MgCl2 0.6,
and glucose 12. Oxygenation was carried out with
a mixture of 95% O2 and 5% CO2 .
A vertical radiofrequency lesion was produced in the lateral wall of the left ventricle, leaving a 5-mm gap in the lesion to allow conduction
between the anterior and posterolateral walls, as
in an earlier study by Pérez et al.,31 who observed
that conduction was preserved with gap sizes of
4.3 ± 0.3 mm. Radiofrequency current (10 W
during 20 seconds) was sequentially delivered
through needle electrodes (length = 12 mm;
diameter = 0.5 mm) inserted into the ventricular
wall perpendicular to the epicardium and ensuring that the inserted portion of the needle was
greater than the thickness of the ventricle wall.28,32
A HAT 10 generator (Osypka AG, RheinfeldenHerten, Germany) was used, and the dispersive
patch was located on the aortic cannula of the
perfusing system. The location of the lesions was
similar in all the experiments. They were always
longitudinal, taking the curvature of the left
ventricular wall into account. From the midzone
of the anterolateral wall, radiofrequency was
applied with the needle electrode, advancing the
March 2013
287
CHORRO, ET AL.
lesion through consecutive punctions toward the
ventricular base, without reaching the zone of
the atrioventricular sulcus, to avoid the trajectory
of the main coronary arteries. Then, leaving a
zone 5 mm in width, successive radiofrequency
applications were made toward the ventricular
apex until reaching a length of 1 cm.
An L-shaped device was introduced through
the left atrium into the left ventricle. The
device consisted of a hollow tube measuring
3.5 mm in diameter, through which a stem
1.5 mm in diameter could be advanced (Fig.
1). The distal end of the stem protruded from
the L-shaped device and consisted of a circular
platform measuring 7.5 mm in diameter, with
which controlled stretching could be induced of
a circumscribed zone of ventricle wall.27,30 To
ensure that stretch was produced by the circular
platform, and not by displacements of the device,
it was proximally and distally affixed to external
supports. The upper part of the device was
affixed with forceps located above the left atrium.
The lower portion of the device was affixed by
mean of a suture thread knotted to the angle of
the L-shaped tube and exteriorized through the
ventricular apex to affix it to an external support to
avoid inappropriate displacements of the device.
The suture thread was exteriorized from the left
ventricle cavity by means of a suture needle. This
zone was at a distance from the gap of preserved
myocardium located in the midzone of the left
ventricular wall.
Left ventricle unipolar electrograms were
recorded using a multiple electrode composed of
240 unipolar stainless steel electrodes (diameter
= 0.125 mm; interelectrode distance = 1 mm)
positioned at the epicardial surface of the left
ventricle free wall encompassing the RF lesion
with the gap in the center of the electrode
and the adjacent zones of the anterior and
posterolateral left ventricle walls. The pressing
location was the anterior portion of the left
ventricular wall, to the left of the isthmus of
preserved myocardium, as shown in Figure 1.
Fixed points marked on the ventricle wall were
used as a reference of the electrode position.
The indifferent electrode was a 4-mm × 6-mm
silver plaque located over the cannulated aorta.
Recordings were obtained with a cardiac electrical
activity mapping system (MAPTECH, Waalre, the
Netherlands). The electrograms were amplified
with a gain of 100–300, broadband (1–400 Hz)
filtered, and multiplexed. The sampling rate of
each channel was 1 kHz. The data obtained
throughout the experiment were stored on digital
storage devices for posterior analysis.
Thirty minutes after positioning the electrodes and the stretching device, the existence
288
Figure 1. Schematic representation of the experimental
preparation. The control (CTRL) stage, where no stretch
was performed, is represented in the left column.
Stretch stage is represented in the right column. A left
intraventricular device (upper part of the figure) was
used to stretch the left ventricle anterior wall (LVAW).
An epicardial multielectrode (lower part of the figure)
was used to obtain the ventricular fibrillation (VF)
recordings in the stretched zone (SZ) on the LV anterior
wall, in the gap of preserved myocardium between
the RF lesions (RFL), and in the nonstretched zone
on the LV posterolateral wall. Mechanical stretch was
obtained on displacing the stem of the device 6 mm. This
displacement produced an approximate elongation of
12% in the horizontal and vertical axes of the zone
of the ventricle wall located over the circular platform
of the device. A and A´ = distance between two fixed
points located on the horizontal axis of the modified
zone before and during stretching; B and B´ = distance
between two fixed points located on the vertical axis of
the modified zone before and during stretching.
of conduction through the isthmus of preserved
myocardium was checked during constant pacing
at 4 Hz, 7 Hz, and 10 Hz. Then VF was induced
by pacing at increasing frequencies from 4 Hz to
20 Hz, and coronary perfusion was maintained
during the arrhythmia. Ventricular pacing was
March 2013
PACE, Vol. 36
VF CONDUCTION BETWEEN RADIOFREQUENCY LESIONS
performed using bipolar electrodes (diameter =
0.125 mm; interelectrode distance = 1 mm)
located near the multiple electrode. Pacing was
carried out using a GRASS S88 (Grass Instruments
Inc., Quincy, MA, USA) stimulator equipped
with a stimulus isolation unit (SIU5). Stimuli
were rectangular pulses of 2-ms duration and an
intensity twice the diastolic threshold.
Experimental Series
VF recordings were obtained each minute
during a period of 5 minutes in the control phase.
Local myocardial stretch was then applied and
maintained on the anterior left ventricle wall for
10 minutes by displacing the stem of the L-shaped
intraventricular device. In an earlier work,27 we
analyzed the effects of this maneuver on the
vertical and horizontal axes of the left ventricular
wall (tangential to the wall), both in the epicardial
surface over the pressing zone (left anterior wall)
and in the left posterior wall. On displacing the
stem of the L-shaped device 6 mm, we obtained
a mean longitudinal increase in both axes of
around 12% in the anterior left ventricular wall,
without significant modifications in the posterior
left ventricular wall. VF recordings were obtained
each minute during the stretching phase. After
this period, local stretching was discontinued, and
VF was recorded each minute during 5 minutes
after the suppression of stretch. Three zones were
considered for the study: the stretched zone (SZ)
on the anterior wall of the left ventricle, the gap
between the RF lesions, and the nonstretched
zone (NSZ) on the posterolateral wall of the left
ventricle (Fig. 1). No information was obtained
on the effects upon propagation during VF in the
puncture zone of the ventricular apex where the
suture thread knotted to the L-shaped tube was
exteriorized.
Data Analysis
VF Spectral Analysis
Welch’s periodogram was used to obtain the
power spectrum of the signals recorded with
all the unipolar electrodes.33 The periodogram
was calculated for the first 4 s of each signal,
fragmenting the recordings into eight segments
with 50% overlap and using Hamming’s window.
Frequency with a maximum peak of power
spectral density between 5 Hz and 48 Hz was
obtained for each channel (dominant frequency
[DFr]). Moreover, spectral concentration (SpConc)
was calculated for each channel as a percentage
of the total energy contained in the interval
DFr ± 1 Hz.
Both DFr and SpConc were calculated in three
specific stages: control (just before stretching);
PACE, Vol. 36
during stretch (3 minutes after beginning stretch,
given that the maximum effects of stretching are
known to occur at this time)27 ; and poststretch (5
minutes after stretch suppression).
VF Time-Domain Analysis
Activation times in each electrode were determined by identifying the moments of maximum
negative slope of the ventricular electrograms.
The minimal threshold for dV/dt to be judged
as a local deflection was a percentage (20%) of
the maximal negative slope in each channel. If
electrograms exhibited two or more deflections,
then the steepest slope of the activation complex
was assigned as the local activation time. The
intervals between consecutive local activations
during VF (VV intervals), the histogram, and
the median were determined for each channel
of each zone of the multielectrode during time
windows of 2 seconds in the three stages
mentioned above: control (before stretch), during
stretch (3 minutes after beginning stretch), and
poststretch (5 minutes after suppressing stretch).
The functional refractory period (FRP) during VF
was regarded as the average of the five smallest
intervals between two successive activations by
two different wavefronts recorded in an electrode,
as previously described.5 This procedure does
not involve direct measurement of refractoriness
but rather a surrogate measure of the effective
refractory period, which in previous studies has
been seen to behave parallel to the refractoriness
obtained during fixed pacing rates, showing good
correlation with the activation frequency during
VF.5,27
Activation Maps
Activation maps during VF were constructed
every 100 ms in the three 2-second time windows
analyzed immediately before stretch, at the third
minute during the stretch period, and at the
fifth minute of the poststretch period. In each
experiment a total of 20 maps were constructed
in each time window. Isochrones were drawn and
analyzed, determining whether each activation
wavefront swept the area through the gap from left
to right (LR), from right to left (RL), and whether
there was a collision (Col) or a breakthrough
(BrTh) in the gap (Fig. 2) (A and B). A BrTh
was considered when the earliest activation was
located in the gap zone and a centrifugal activation
pattern was seen toward the adjacent zones.
Statistical Analysis
Data are presented as means ± standard deviation. The general lineal model was used to analyze
the differences within each study zone (repetitive
measurements) and between zones. Values of
March 2013
289
CHORRO, ET AL.
Figure 3. Longitudinal sections of two hearts (A and B)
showing the transmural radiofrequency-induced lesions
(arrows). Microscopic views (C = 5x high-power field;
D = 20x high-power field) showing the gap of preserved
myocardium between the radiofrequency lesions and
one of the boundaries of the lesion (D). Coagulative
necrosis of the cells with disruption of myocyte structure
and contracture bands are shown in the damaged
myocardium. Trichromic staining. RF = radiofrequency
lesion.
Figure 2. (A) and (B) Four categories of activation maps
were established: (A) activation wavefront sweeping
the gap from left to right (left), and from right to left
(right); (B) collision of wavefronts in the gap (left),
and breakthrough (right). Four electrograms for each
activation map are shown where progression of the
activation wavefront can be observed.
P < 0.05 were considered statistically significant.
Differences between qualitative variables were
analyzed by the χ 2 test.
Results
Characteristics of the RF Lesions
Lesions were transmural, and the mean length
of the gap between them was 5.0 ± 0.2 mm. The
length and the width of the upper part of the RF
lesion were 9.4 ± 0.6 mm and 3.7 ± 0.4 mm,
290
respectively. The lower part of the RF lesion had a
mean length of 10.1 ± 0.7 mm and a mean width of
3.7 ± 0.3 mm. The transmural nature of the lesions
and their limits were determined by macro- and
microscopic analysis of the preparations (Fig. 3).
VF Spectral Analysis
In the control stage, DFr in SZ was significantly higher than in NSZ (Table I). When
stretch was applied, DFr increased significantly
both in SZ and in the gap—significant differences
being observed when comparing SZ with the
gap and with NSZ. After suppressing stretch,
the DFr values returned to the same values of
the control stage—significant differences again being observed between SZ and NSZ. Figure 4 shows
the specific changes in DFr observed during stretch
in the three studied zones. An example of the
spatial distribution of DFr before stretch, during
March 2013
PACE, Vol. 36
VF CONDUCTION BETWEEN RADIOFREQUENCY LESIONS
Table I.
Spectral Analysis Results: Mean ± SD Obtained in Each of the Experimental Stages in the Three Study Zones
SZ
Control
Stretch
Post
GAP
DFr (Hz)
SpConc
15.2 ± 3.6
18.9 ± 5.0*
15.1 ± 1.6
38 ± 11
32 ± 12*
46 ± 12*
DFr (Hz)
15.5 ± 3.6
16.9 ± 4.3*, **
15.4 ± 1.9
NSZ
SpConc
DFr (Hz)
SpConc
30 ± 10**
33 ± 14
32 ± 10**
14.1 ± 2.7**
14.3 ± 3.0**
14.4 ± 2.1**
36 ± 12
37 ± 10**
39 ± 13**
*Significant differences versus control (P < 0.05); **significant differences versus SZ (P < 0.05).
DFr = dominant frequency; NSZ = non-stretched zone; SpConc = spectral concentration; SD = standard deviation; SZ = stretched zone.
Figure 4. Values of the dominant frequency (DFr)
obtained in each experiment in the control phase
(CTRL.) and during stretch (STR.) in the three studied
zones. Abbreviations: SZ = stretched zone; NSZ = nonstretched zone.
in SZ increased, being significantly higher than
the value measured in the gap and in NSZ.
With regard to the FRP, during the control
stage this parameter was lower in SZ than in
NSZ (Table II). During stretch it decreased in SZ.
This reduction was significant when compared
to the gap and NSZ, and also compared to the
control stage. After the suppression of stretch, the
refractory period in SZ was significantly greater
than in the gap and before stretch.
The regression straight line obtained, relating
DFr to the inverse of the mean VV intervals (INVV),
was: DFr = 0.87 INVV + 1.7; R = 0.82; P < 0.0001;
standard error of estimate = 1.6 Hz. The regression
line obtained on relating DFr to FRP was: DFr =
−0.28 FRP + 25.42; R = 0.45; P < 0.0001; standard
error of estimate = 2.6 Hz.
Activation Maps
stretch, and after stretch suppression is shown in
Figure 5.
The same comparisons were made with
SpConc (Table I). Before stretch, SpConc in SZ
was significantly higher than in the gap. During
stretch, SpConc in SZ was significantly reduced
with regard to SpConc in NSZ, and the reduction
in SZ was also significant when compared to the
control stage. After stretch suppression, SpConc in
SZ increased until it became significantly higher
than in the control stage, and also higher than in
the gap and in NSZ.
VF Time-Domain Analysis
Immediately before stretching, VV in the gap
was significantly lower than in SZ (Table II).
During stretch, VV in SZ was significantly reduced
when compared to the gap and NSZ. This
reduction was also significant when compared to
the control values. In this stage, VV in the gap was
also reduced, being significantly lower than in the
control stage. After the suppression of stretch, VV
PACE, Vol. 36
Figure 6 shows percentages of each type of
activation map, classified according to the direction through the gap. No significant differences
were found between the different study stages.
A general tendency toward the predominant pass
of activation wavefronts from right (posterolateral
left ventricle wall) to left (anterior left ventricle
wall) was observed. Both before and during
stretch, and also after stretch suppression, the
percentage of wavefronts that passed through
the gap from the posterolateral wall to the
anterior wall was slightly higher than in the
opposite direction. This slight predominance was
maintained during local VF acceleration in the
anterior left ventricle wall produced during the
stretching phase. A slight increase in the collisions
of wavefronts in the gap was also observed during
stretching.
Mapping of the flow of wavefronts in the gap
zone also revealed that the maximum number of
consecutive left to right transmitted wavefronts
was four in one experiment during the control
stage and four in another experiment during
March 2013
291
CHORRO, ET AL.
Figure 5. Spatial distribution of dominant frequencies (DFr) over the mapped surface obtained
before stretch, during stretch, and after stretch suppression in a particular experiment. Frequency
is gray coded, where dark gray indicates high-frequency values. Radiofrequency lesions are
denoted by dashed zones. For each zone of interest, an electrogram and the corresponding
spectrogram are shown: (1) corresponds to an electrogram from the stretched anterior wall, (2) to
an electrogram from the gap, and (3) to an electrogram from the non-stretch posterolateral wall.
PSD: power spectral density; t(s): time (in seconds), a.u.: arbitrary units. For increased clarity,
only 1-second VF recordings are shown.
the stretch stage. The maximum number of
consecutive right to left transmitted wavefronts
was three in three experiments during the control
stage and four in two experiments during the
stretch stage. In both cases most of the wavefronts
passed the gap only in one direction or the
other isolatedly, and to a lesser degree as two
consecutive wavefronts.
Table II.
Time-Domain Analysis Results: Mean ± SD Obtained in Each of the Experimental Stages in the Three Study Zones
SZ
Control
Stretch
Post
GAP
NSZ
VV (ms)
FRP (ms)
VV (ms)
FRP (ms)
VV (ms)
FRP (ms)
61.5 ± 8.0
48.5 ± 9.8*
65.3 ± 6.4*
34.6 ± 6.2
31.5 ± 4.5*
38.6 ± 7.8*
58.7 ± 7.5**
53.9 ± 9.0*, **
59.8 ± 6.9**
35.5 ± 4.9
34.0 ± 5.4**
35.7 ± 6.7**
63.2 ± 7.0
61.5 ± 7.1**
63.1 ± 8.3**
37.0 ± 6.1**
35.1 ± 6.9**
36.7 ± 7.4
differences versus control (P < 0.05); ** significant differences versus SZ (P < 0.05).
FRP = functional refractory period; NSZ = non-stretched zone; SD = standard deviation; SZ = stretched zone; VV = median of the
ventricular fibrillation intervals.
* Significant
292
March 2013
PACE, Vol. 36
VF CONDUCTION BETWEEN RADIOFREQUENCY LESIONS
Figure 6. Percentages of activation maps, classified according to the direction through the gap,
obtained immediately before stretch (control), during stretch, and after stretch suppression.
Nonsignificant differences were found between the study stages. BrTh = breakthrough pattern.
Number of maps analyzed in each stage = 180.
We identified those wavefronts that were
blocked after crossing the isthmus of preserved
myocardium, and determined the ratio between
the blocked and transmitted wavefronts, considering both the left to right and the opposite direction.
During stretch there was a significant increase in
this parameter on considering the left to right
direction (0.56 vs 0.30; P = 0.02), while the
increase on considering the right to left direction
failed to reach statistical significance (0.39 vs 0.23;
P = 0.11).
Wavefront Propagation through the Gap
and Electrophysiological Parameters
The regression coefficient obtained on relating
the ratio between the number of wavefronts that
passed through the gap from left to right (nLR)
or in the opposite direction (nRL) to the ratio
between DFr at both sides of the gap (DFr Ratio)
was not statistically significant: nLR/nRL = 0.41
(DFr Ratio) + 0.61; R = 0.06; P = 0.75; standard
error of estimate = 1.2 (Fig. 7). Likewise, the
regression coefficient on relating nLR/nRL to the
ratio between FRP in the left and right side of
the gap (FRP Ratio) also failed to reach statistical
significance (Fig. 7): nLR/nRL = −1.77 (FRP Ratio)
+ 2.84; R = 0.25; P = 0.22; standard error of
estimate = 1.2.
The relationships between left and right
directionality of wavefront propagation and the
ratio between the estimated excitable window
at both sides of the gap were also studied.
The difference between the median of the VV
intervals (VVmed) and FRP was considered as
an approximation to the excitable window.34
Figure 8 shows the values of this parameter in each
studied zone. During stretch there was a significant
reduction in SZ. Figure 8 shows the absence of
a significant relationship between nLR/nRL and
the (VVmed-FRP) ratio. However, the regression
PACE, Vol. 36
coefficient between the DFr ratio and the excitable
window (VVmed-FRP) ratio was significant (DFr
ratio = −0.38 (VVmed-FRP) ratio + 1.46; R =
−0.57; P = 0.002; standard error of estimate = 1.5;
Fig. 9).
The relationship between the number of
wavefronts that passed through the gap from
left to right or in the opposite direction and
the electrophysiological parameters determined
in each studied zone (SZ, gap, and NSZ) was
also evaluated. The only significant regression
coefficient was obtained on relating nLR to the
excitable window to the right of the gap (Fig. 9).
On relating nRL to the excitable window to the left
of the gap, the regression coefficient failed to reach
statistical significance (R = 0.18; P = 0.37).
Discussion
The main findings of this study are the
following: (1) the experimental model used shows
a slight predominance of wavefront flow from the
posterolateral wall of the left ventricle toward the
anterior wall, through the isthmus of preserved
myocardium between the RF lesions; (2) during
local VF acceleration induced by myocardial
stretch, the modified zone continues to receive
more wavefronts through the isthmus of preserved
myocardium than the nonmodified zone; (3) VF
acceleration in the left ventricular anterior wall
does not produce acceleration in the posterolateral
wall located on the opposite side of the RF
lesions; and (4) the absence of significant changes
in the electrophysiological parameters in the
posterolateral zone limits the flow of wavefronts
from the accelerated zone toward the NSZ.
Characteristics of the Lesions
The analysis of the flow of wavefronts has
focused on the gap of preserved myocardium
between two linear RF lesions located between
March 2013
293
CHORRO, ET AL.
Figure 7. Scattergrams obtained on relating the ratio
between the number of wavefronts passing through the
isthmus from left to right or in the opposite direction
(nLR/nRL) and the ratio between DFr (DFr Ratio) or the
functional refractory period (FRP Ratio) on the left and
right side of the gap.
the stretched and a NSZ. Conduction through gaps
of preserved myocardium has been previously
analyzed in both atrial and ventricular myocardium. The dimensions of the gap and the
frequency of activation are factors that determine
conduction,31,35,36 because they define the source–
sink relationship and wavefront curvature once
the latter passes through the gap. In atrial
myocardium of rabbit hearts, gaps of 3 mm did not
prevent conduction between contiguous zones.36
In ventricular myocardium of rabbit hearts, Pérez
294
Figure 8. Top: Variations in the excitable window
(VVmed-FRP) in the different stages of the experimental
protocol in the three studied zones. CTRL. = control
phase; STR. = during stretch; POST. = after suppressing
stretch; SZ = stretched zone; NSZ = non-stretched
zone. Bottom: Scattergram obtained on relating the ratio
between the number of wavefronts that passed through
the isthmus from left to right or in the opposite direction
(nLR/nRL) and the ratio between the excitable window
([VVmed-FRP] Ratio) on the left and right side of the
gap.
et al.31 analyzed the effects of a residual isthmus
of surviving tissue upon conduction after linear
ablation. They observed that for gap lengths of 4.3
± 0.3 mm on average, conduction was preserved
at all cycle lengths studied. Cabo et al.,35 in
epicardial myocardium of sheep, found that for the
maximum frequency of stimulation, the critical
isthmus was 2.5 mm. In this study, the width of
the gaps of preserved myocardium was 5.0 ± 0.2
mm, which is greater than the value established in
earlier studies for limiting the flow of wavefronts,
and in all cases studied, bidirectional conduction
through the gap was observed.
March 2013
PACE, Vol. 36
VF CONDUCTION BETWEEN RADIOFREQUENCY LESIONS
Figure 9. Top: Relationship between the number of
wavefronts that passed through the gap from left to right
(nLR) and the excitable window to the right of the gap
(VVmed- FRP)(R). Bottom: Relationship between the DFr
ratio at both sides of the gap and the excitable window
(VVmed-FRP) ratio.
Effects of Stretch upon VF Activation Frequency
Before the stretching phase, there was a small
difference in DFr between the anterior and posterolateral zones of the left ventricle wall, in the
same way as in an earlier study.19 The magnitude
of the difference is smaller than that reported by
other investigators using guinea pig hearts.10 In
swine, slightly higher activation frequencies have
been recorded in the posterobasal zone of the left
ventricle both in perfused and in non-perfused
hearts (short duration VF).16 In earlier studies27,30
we observed that acute local myocardial stretch
PACE, Vol. 36
accelerates VF. In these studies the increase in
activation frequency during the arrhythmia was
associated with a reduction in refractoriness, and
was accompanied by an increase in complexity of
the VF activation patterns. In the SZ, the epicardial
activation maps were more complex, though
without an increment of complete reentrant
activation patterns.27 These changes did not
produce significant VF modifications at a distance
from the SZ. In this study, VF was also accelerated
in the SZ, and the changes produced in the left
ventricle anterior wall were not accompanied by
acceleration in the posterolateral zone of the LV.
In addition to the increased DFr in the SZ, spectral
analysis revealed a reduction in SpConc indicating
more heterogeneous and irregular activation. This
latter parameter was likewise smaller in the
gap zone before the stretching phase and after
stretch suppression—probably because this is a
zone of interaction of the wavefronts coming
from both sides of the lesion. Analysis in the
time domain in turn has shown shortening of
the VV intervals during local stretch applied to
the anterior wall, with no significant variations
in the posterolateral wall—in coincidence with
the data supplied by the spectral analysis.
Before stretch application, the differences in the
median of the VV intervals between the anterior
and posterolateral wall did not reach statistical
significance, while after stretch suppression a
significant increase was observed in this parameter
in the anterior left ventricle wall. Thus, in this
study the information afforded by the frequencyand time-domain methods is discordant in relation
to the statistical significance of the differences
between the two zones. Usually both timedomain and frequency-domain analytical methods
yield analogous information in the analysis of
the activation frequency during VF,5,27 though
certain conditions such as the presence of double
potentials can give rise to discrepancies between
the two methods.14,27 In this study, as in earlier
studies,5,27 the regression straight line obtained
on relating DFr to the inverse of the mean VV
interval was statistically significant, and the slope
was close to 1.
Wavefront flow through the Gap
In guinea pig hearts with greater frequency
gradients than in this study,10 a predominant
wavefront direction was observed from the fastest
toward the least fast zones. In swine hearts, where
the observed frequency gradients were smaller,
Nanthakumar et al.16 conducted epicardial mapping studies and demonstrated that in shortduration VF, the wavefronts tended to propagate
from the posterior basal left ventricle to the
anterior left ventricle, and to the anterior right
March 2013
295
CHORRO, ET AL.
ventricle. Rogers et al.,17 also in swine hearts,
analyzed the flow of wavefronts and the relative
activation frequencies of two epicardial arrays
situated in the left ventricle near the anterior
septum and in a more lateral position. They
observed that wavefronts from the septal array
had a greater influence upon the lateral array
than the other way around, and when the data
obtained before the use of a multineedle array
were included in the analysis, the septal array
was slightly faster than the lateral array. In this
study, before the stretch phase, the frequency
gradient was small and of a magnitude similar
to that reported in earlier studies in the same
experimental model,19 and the flow of wavefronts
through the gap of preserved myocardium showed
a slight predominance from the posterolateral
zone toward the anterior zone. During myocardial
stretch, activation in the anterior wall was
accelerated, resulting in a marked increase in
frequency gradient between the anterior wall and
the posterolateral wall. However, this increase was
not accompanied by an increased wavefront flow
toward the posterior wall, and the slight pass
predominance from the posterior wall toward the
anterior wall persisted, with a tendency toward
increased collisions and breakthrough patterns in
the gap zone.
During the stretch phase, significant shortening of myocardial refractoriness was observed
in the modified zone, as in a previous study,27
without significant changes on the other side of
the RF lesions. This factor is directly related
to the frequency of VF,5,12,13,27 and is probably
one of the main limiting factors of wavefront
arrival to the nonmodified zone. The fact that
VF acceleration was not accompanied by a larger
proportion of wavefronts passing through the gap
of preserved myocardium from the SZ raises
the question of whether the accelerated zone is
a source of wavefronts toward the rest of the
myocardium or whether this zone mainly modifies
its capacity to house more wavefronts, including
those arriving from adjacent zones, because of
its lesser refractoriness. In the former case we
would expect a greater flow of wavefronts toward
the non-modified zone, but this behavior has not
been observed. Instead, we recorded a tendency
toward an increased number of collisions and
breakthrough patterns. The first finding would be
related to the interaction of a larger number of
wavefronts, while the latter could be because of an
increased arrival in this zone of wavefronts from
the subendocardium or midmyocardium. This has
been seen in the stretch zone in earlier studies.27,30
In this study, we observed no significant
relationship between the ratio of wavefronts
passing to one side or the other of the gap of
296
preserved myocardium and the ratio of DFr, FRP,
or the excitable window at both sides of the gap.
On isolatedly analyzing the number of wavefronts
passing to the left or the right, the only electrophysiological parameter yielding a significant
correlation was the excitable window to the right
of the gap on considering wavefront flow from left
to right. During stretch, this parameter was not
significantly modified in NSZ—a fact that could
explain the absence of significant increments in
the flow of wavefronts from left to right. On
considering the flow of wavefronts from right
to left, the relationship between the number of
wavefronts and the excitable window to the left of
the gap failed to reach statistical significance. If the
relationship observed on analyzing wavefront flow
in the opposite direction had been maintained, a
decrease in nRL could have been expected during
stretch, because the excitable window decreased
in SZ—a situation which we failed to observe.
The DFr increment in SZ during stretch did not
depend upon the flow of wavefronts but on the
electrophysiological changes produced by stretch.
The observed significant relationship between the
DFr ratio and the excitable window (VVmed-FRP)
ratio shows that the accelerating effect of stretch in
the zone in which it is applied is associated with
a decrease in the excitable window, and therefore
in the capacity to receive wavefronts from other
zones.
Limitations
Radiofrequency lesions may have modified
the VF activation patterns, though this does not
preclude analysis of the changes in wavefront
flow through the gap of preserved myocardium
following acceleration of the VF in the anterior
wall of the left ventricle. Analysis is facilitated by
focusing the study on this zone located between
the stretched left ventricle anterior wall and the
nonstretched posterolateral wall. However, on
circumscribing the analysis to the gap zone and
adjacent myocardium, no information has been
obtained on the flow of wavefronts toward the
septum and anterior wall of the right ventricle,
because epicardial mapping was carried out
positioning the center of the multiple electrode
in the zone of the isthmus. The mapped surface
corresponds to this zone and to the adjacent zones
of the anterior and posterior walls of the left
ventricle. No information was obtained on the
basal or apical zone. For this reason we have no
direct information on the existence of reentries
around the RF lesions. Nevertheless, analysis
of the flow of wavefronts through the isthmus
revealed no uniform patterns compatible with
persistent reentry around the obstacles formed
by the radiofrequency lesions. Another limitation
March 2013
PACE, Vol. 36
VF CONDUCTION BETWEEN RADIOFREQUENCY LESIONS
is derived from the epicardial analysis, which
precludes the generation of precise data on the
characteristics of intramural activation. On the
other hand, although the method used did not
give rise to modifications in global coronary flow
during the stretching phase in an earlier study,27
this does not rule out the possible existence of
local modifications in flow that could influence
the variations in myocardial activation during the
arrhythmia.
Clinical Implications
In this study, we have seen that the accelerated site located in the SZ does not increase
the flow of wavefronts in the adjacent gap of
preserved myocardium. The results indicate that
the accelerated site exerts no greater influence
upon activation in the posterolateral wall, and
does not appear to act as a driver site for activity
in this zone—though the study has the limitation
of circumscribing the analysis of the directioning
of excitation propagation to the zone of the gap
of preserved myocardium—no information being
obtained regarding the propagation toward other
regions. We speculate that local interventions
designed to reduce or abolish the activity of the
fast zones during VF would have little influence
upon the activity of the rest of myocardium,
if not accompanied by global actions. Such
interventions are different from those destined
to control the triggering of VF; for example, the
RF ablation of selected areas, which have been
shown to be effective in reducing the number of
arrhythmic episodes.
Conclusion
In the experimental model used, a slight
predominance is observed in wavefront flow from
the posterolateral wall of the left ventricle toward
the anterior wall, through the isthmus of preserved
myocardium located between the RF lesions.
During the local VF acceleration induced by
myocardial stretch, the modified zone continues
to receive more wavefronts through the isthmus
of preserved myocardium than the NSZ, and no
significant changes in activation frequency are
seen in this latter zone. The absence of significant
changes in the electrophysiological parameters of
the nonstretched myocardium limits the arrival of
wavefronts from the SZ.
References
1. 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.
2. Ideker RE, Huang J. Our search for the porcine mother rotor. Ann
Noninvasive Electrocardiol 2005; 10(Suppl 1):7–15.
3. Jalife J. Ventricular fibrillation: Mechanisms of initiation and
maintenance. Annu Rev Physiol 2000; 62:25–50.
4. Weiss JN. Ventricular fibrillation. New insights into mechanisms.
Ann NY Acad Sci 2004; 1015:122–132.
5. Chorro FJ, Cánoves J, Guerrero J, Mainar L, Sanchis J, Such L,
Lopez-Merino V. Alteration of ventricular fibrillation by flecainide,
verapamil, and sotalol. An experimental study. Circulation 2000;
101:1606–1615.
6. Chorro FJ, Guerrero J, Ferrero A, Tormos A, Mainar L, Millet
J, Canoves J, et al. Effects of acute reduction of temperature on
ventricular fibrillation activation patterns. Am J Physiol Heart Circ
Physiol 2002; 283:H2331–H2340.
7. Nash MP, Mourad A, Clayton RH, Sutton PM, Bradley CP,
Hayward M, Paterson DJ, et al. Evidence for multiple mechanisms
in human ventricular fibrillation. Circulation 2006; 114:536–
542.
8. Tabereaux PB, Dosdall DJ, Ideker RE. Mechanisms of VF
maintenance: Wandering wavelets, mother rotors, or foci. Heart
Rhtyhm 2009; 6:405–415.
9. Ten Tusscher KHW, Mourad A, Nash MP, Clayton RH, Bradley CP,
Paterson DJ, Hren R, et al. Organization of ventricular fibrillation
in the human heart: Experiments and models. Exp Physiol 2009;
94:553–562.
10. Samie FH, Berenfeld O, Anumonwo J, Mironow SF, Udassi S,
Beaumont J, Taffet S, et al. Rectification of the background
potassium current. A determinant of rotor dynamics in ventricular
fibrillation. Circ Res 2001; 89:1216–1223.
11. Chen J, Mandapati R, Berenfeld O, Skanes AC, Jalife J. Highfrequency periodic sources underlie ventricular fibrillation in the
isolated rabbit heart. Circ Res 2000; 86:86–93.
12. 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.
PACE, Vol. 36
13. Choi BR, Liu T, Salama G. The distribution of refractory periods
influences the dynamics of ventricular fibrillation. Circ Res 2001;
88:e49–e58.
14. Valderrábano M, Yang J, Omichi C, Kil J, Lamp ST, Qu Z, Lin SF, et
al. Frequency analysis of ventricular fibrillation in swine ventricles.
Circ Res 2002; 90:213–222.
15. 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; 13:1042–1043.
16. Nanthakumar K, Huang J, Rogers JM, Johnson PL, Newton JC,
Walcott GP, Justice RK, et al. Regional differences in ventricular
fibrillation in the open-chest porcine left ventricle. Circ Res 2002;
91:733–740.
17. Rogers JM, Huang J, Melnick SB, Ideker RE. Sustained reentry in
the left ventricle of fibrillating pig hearts. Circ Res 2003; 92:539–
545.
18. Newton JC, Smith WM, Ideker RE. Estimated global transmural
distribution of activation rate and conduction block during porcine
and canine ventricular fibrillation. Circ Res 2004; 94:836–842.
19. Chorro FJ, Guerrero J, Trapero I, Such-Miquel L, Mainar L, Cánoves
J, Blasco E, et al. Time-frequency analysis of ventricular fibrillation.
An experimental study. Rev Esp Cardiol 2006; 59:869–878.
20. Rogers JM, Walcott GP, Gladden JD, Melnick SB, Ideker RE, Kay
MW. Epicardial wavefronts arise from widely distributed transient
sources during ventricular fibrillation in the isolated swine heart.
New J Phys 2008; 10:015004.
21. Umapathy K, Masse S, Sevaptsidis E, Asta J, Ross H, Thavandiran N,
Nair K, et al. Regional frequency variation during human ventricular
fibrillation. Med Eng Phys 2009; 31:964–970.
22. Garrey WE. The nature of fibrillary contraction of the heart. Is
relation to tissue mass and form. Am J Physiol 1914; 33:397–414.
23. Damiano RJ, Asano T, Smith PK, Cox JL. Effect of the right
ventricular isolation procedure on ventricular vulnerability to
fibrillation. J Am Coll Cardiol 1990; 15:730–736.
24. Kim YH, Garfinkel A, Ikeda T, Wu TJ, Athill CA, Weiss JN,
Karagueuzian HS, et al. 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 1997; 100:2486–2500.
March 2013
297
CHORRO, ET AL.
25. Valderrábano M, Kim YH, Yashima M, Wu TJ, Karagueuzian HS,
Chen PS. Obstacle-induced transition from ventricular fibrillation
to tachycardia in isolated swine right ventricles: Insights into the
transition dynamics and implications for the critical mass. J Am
Coll Cardiol 2000; 36:2000–2008.
26. Pak HN, Oh YS, Liu YB, Wu TJ, Karagueuzian HS, Lin SF,
Chen PS. Catheter ablation of ventricular fibrillation in rabbit
ventricles treated with beta-blockers. Circulation 2003; 108:3149–
3156.
27. Chorro FJ, Trapero I, Guerrero J, Such LM, Canoves J, Mainar L,
Ferrero A, et al. Modification of ventricular fibrillation activation
patterns induced by local stretching. J Cardiovasc Electrophysiol
2005; 16:1087–1096.
28. Chorro FJ, Blasco E, Trapero I, Canoves J, Ferrero A, Mainar L,
Such-Miquel L, et al. Selective myocardial isolation and ventricular
fibrillation. Pacing Clin Electrophysiol 2007; 30:359–370.
29. Tormos A, Chorro FJ, Millet J, Such L, Cánoves J, Mainar L, Trapero
I, et al. Analyzing the electrophysiological effects of local epicardial
temperature in experimental studies with isolated hearts. Physiol
Meas 2008; 29:711–728.
30. Chorro FJ, Trapero I, Such-Miquel L, Pelechano F, Mainar L,
Cánoves J, Tormos A, et al. Pharmacological modifications of the
298
31.
32.
33.
34.
35.
36.
stretch-induced effects on ventricular fibrillation in perfused rabbit
hearts. Am J Physiol Heart Circ Physiol 2009; 297:H1860–H1869.
Pérez FJ, Wood MA, Schubert CM. Effects of gap geometry
on conduction through discontinuous radiofrequency lesions.
Circulation 2006; 113:1723–1729.
Chorro FJ, Guerrero J, Cánoves J, Trapero I, Mainar L, Pelechano F,
Blasco E, et al. Changes in the spectral characteristics of ventricular
fibrillation in lesions produced by radiofrequency ablation. An
experimental study. Rev Esp Cardiol 2008; 61:394–403.
Oppenheim A, Schafer R. Digital Signal Processing. Engelwood
Cliffs, Prentice Hall, 1975.
Wijffels MC, Dorland R, Mast F, Allessie MA. Widening of
the excitable gap during pharmacological cardioversion of atrial
fibrillation in the goat: Effects of cibenzoline, hydroquinidine,
flecainide, and d-sotalol. Circulation 2000; 102:260–267.
Cabo C, Pertsov AM, Baxter WT, Davidenko JM, Gray RA, Jalife J.
Wave-front curvature as a cause of slow conduction and block in
isolated cardiac muscle. Circ Res 1994; 75:1014–1028.
Chorro FJ, Mainar L, Sanchis J, Canoves J, Llavador E, Such LM,
Cerda M, et al. Mapping of atrial activation patterns after inducing
contiguous radiofrequency lesions: An experimental study. Pacing
Clin Electrophysiol 2001; 24:147–156.
March 2013
PACE, Vol. 36