© 2001 American Heart Association, Inc.
UltraRapid Communication |
The Distribution of Refractory Periods Influences the Dynamics of Ventricular Fibrillation
From the Department of Cell Biology and Physiology, University of Pittsburgh, School of Medicine, Pittsburgh, Pa.
Correspondence to Guy Salama, PhD, Professor of Cell Biology and Physiology, University of Pittsburgh, School of Medicine, 3500 Terrace St, S314 Biomedical Science Tower, Pittsburgh, PA 15261. E-mail gsalama+{at}pitt.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Abstract—The spatial and dynamic properties of ventricular fibrillation (VF) may be random or related to cellular electrical properties of the normal heart. Local activation intervals (AIs) in VF may depend on the local refractory period (RP), and sustained VF may require a steep action potential (AP) restitution curve. In guinea pig hearts, AP durations (APDs) and RPs on the epicardium are shorter at the apex and progressively longer toward the base, producing gradients of RPs that may influence the spatial organization of VF. In the present study, the influence of APDs on VF dynamics is investigated in perfused guinea pig hearts stained with a voltage-sensitive dye by comparing APD gradients to the dynamics of VF elicited by burst pacing. In VF, AIs had no clear periodicity, but average AIs were shorter at the apex (57.5±8.1 ms) than the base (76.1±1.5 ms, n=6, P<0.05) and had gradients similar to APD gradients (correlation coefficient 0.71±0.04). Analysis of local velocity vectors showed no preferential directions, and fast Fourier transform (FFT) power spectra were broad (10 to 24 Hz) with multiple peaks (n=6). However, the selective inhibition of delayed K+ rectifying currents, IKr (E4031; 0.5 µmol/L, n=3), shifted FFT spectra from complex to a lower dominant frequency (10 Hz) and altered repolarization but retained the correlation between mean AIs and RPs. Thus, VF dynamics are consistent with a multiple wave-make and wave-break mechanism, and the local RP influences VF dynamics by limiting the range of VF frequencies and AIs at each site. The full text of this article is available at http://www.circresaha.org.
Key Words: action potential • refractory period • ventricular fibrillation • action potential duration • fast Fourier transform
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Ventricular fibrillation (VF) had been traditionally thought of as a highly disorganized process of random electrical activity, but more recent studies described a substantial degree of structure and organization to seemingly random events.1 2 3 4 5 Optical mapping of electrical activity using a CCD camera in dog and sheep hearts showed that activation frequencies during VF were spatially organized as a small number of relatively large domains, each with a single dominant frequency.6 7 Each domain with a common frequency in the fast Fourier transform (FFT) spectrum was proposed to represent a mother rotor, occurring in the tissue activating the most rapidly and with the shortest refractory period (RP). Activation fronts in a faster domain were said to undergo Wenckebach block at the boundary with slower domains with some of these activation wavefronts crossing into the slower domain. Even though wandering wavelet reentry could occur in smaller regions where activation could appear disorganized, these regions were thought to be driven by activation fronts arising from neighboring faster domains, which could be traced back to the mother rotor.6 7 Similar results were obtained with smaller rabbit hearts where large domains had a single dominant frequency suggesting that a mother rotor or a 3D scroll was sufficient to maintain VF and activation waves in VF emanate from a single periodic source.7
In contrast, other studies predicted that fibrillation consisted of wandering wavelets with a high level of disorganization, nonrepeatability, resulting in sporadic short-lived, ever-changing reentry.4 8 Still, activation intervals (AIs) in VF were correlated with the RPs of the normal action potential (AP)4 and agents that alter RPs also changed the number of activation fronts per second, demonstrating a correlation between tissue RPs and VF structure.5 Action potential durations (APDs) and RPs are dynamic parameters that change as a function of diastolic interval.9 10 11 The adaptation of the APD to a premature impulse or the restitution kinetics curve was proposed as a predictive cellular property that promotes wavebreaks when the slope of the curve is >1.11 12 According to the wave breakup hypothesis, waves are annihilated and new waves are created continuously to maintain the complex behavior of VF. Plots of AIs versus the preceding AI revealed no clear periodicity, consistent with the continuous wave breakup hypothesis.13 Other studies showed that VF could exhibit multiple wavelet reentry in intact pig hearts, whereas in cryoablated hearts (with a thin layer of surviving epicardium) VF exhibited a single wavelet reentry.14
Both hypotheses regarding the nature of VF implicate spatial heterogeneities of RPs as a major determinant of VF structure and stability. In the continuous wave breakup hypothesis, the correlation between local RP and AI in VF4 would suggest that AIs in VF were spatially heterogeneous, like RPs. In the dominant spiral wave hypothesis, changes in RPs delineate the boundaries separating large domains containing a single spiral wave. Despite the significance of the spatial distribution of RPs, no studies have compared detailed maps of RPs to AIs and frequencies in VF.
Guinea pig hearts are now used to investigate the spatial distribution of VF dynamics and its correlation with gradients of APDs and RPs because the sequence of repolarization and RPs had been extensively characterized in this animal model. In guinea pig hearts, repolarization of the right and left epicardium was shown to start at the apex and to spread systematically toward the base.15 16 17 In the absence of anatomical or ischemic injury, RPs showed no abrupt heterogeneities, changed gradually along the epicardium, and were homogeneous along the endocardium.16 The sequence of repolarization and RPs was independent of cycle length (CL) and of activation sequence because pacing the sinoatrial node or various sites on the ventricles (apex, base, anterior, or posterior) produced markedly different activation patterns yet similar repolarization patterns.15 16 The findings indicated that heterogeneities of intrinsic cellular properties of ventricular myocytes along the epicardium produce gradients of APDs, but the exact distribution of inward (INa and ICa) outward (IK) currents that produce APD gradients is not known. In principle, the same spatial heterogeneities of ionic currents in paced hearts underlie the rapid voltage oscillations and the structure of VF. Whereas the depolarizing inward currents (INa and ICa) can account for rapid depolarizations in VF, the rapid repolarizations are more difficult to explain. Studies on the distribution of Erg (the channel protein underlying IKr) and the kinetics of the IKr current suggest that IKr could account for the gradients of APDs, rapid repolarizations in VF and VF frequencies, making IKr a dominant contributor to VF dynamics (see Discussion).
In this study, fluorescence signals from voltage-sensitive dyes were used to map APDs from 252 sites on paced guinea pig hearts. Next, VF was elicited by burst stimulation, and AIs were correlated to the local APD measured in paced hearts. Local conduction velocity vectors were used to test for the occurrence of Wenckebach-like conduction blocks in VF. AIs and frequency distributions of FFT spectra were correlated with APDs, in the absence and presence of a selective blocker (E4031) of the delayed K+ rectifier current (IKr) to prolong APDs and hence test the link between RPs and VF dynamics. Preliminary reports of these data appeared in abstract form.18
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Preparation
Guinea pigs (female,
400 g) were procured from
Hill Top Lab Animals, Inc (Scottdale, Pa). Experiments were carried out
in accordance with the Animal Welfare Act and the NIH Guide for
the Care and Use of Laboratory Animals. Guinea pigs were
anesthetized (pentobarbital 35 mg/kg) and injected with heparin
(200 U/kg IP). Hearts were perfused in a Langendorff
apparatus with (in mmol/L) NaCl 130,
NaHCO3 25, MgSO4 1.20,
KCl 4.7, dextrose 20, and CaCl2 1.25, at pH 7.4,
and gassed with 95% O2 and 5%
CO2 at 37.0±0.2°C. Perfusion pressure was
adjusted to 70 mm Hg by controlling the flow rate of
perfusate and was continuously monitored. Hearts were placed in
a chamber to reduce movement artifacts without using chemical
uncouplers to arrest
contractions.16 19 20
The voltage-sensitive dye (di-4-ANEPPS, 10 µL of 1 mg/mL dimethyl
sulfoxide) was added to the perfusate while continuously
recording heart rate, electrocardiograms, and
perfusion pressure to ensure that staining did not produce lasting
pharmacological effects.14
Optical Apparatus, Data
Acquisition, and Analysis
An image of the heart was focused on the array by
focusing the image on a graticule (Graticules Ltd), located on a plane
parafocal to the array
(Figure 1A
). Outputs from 256 (16x16) diodes were amplified,
digitized (12-bit) (Microstar Laboratories, Inc) and stored in computer
memory.17 20
Unless stated, all maps were obtained from the anterior surface of the
heart
(Figure 1E).
Figure 1C depicts a map of the array; each box
represents a diode with APs detected by that diode. APDs were
calculated from the activation ([dF/dt]max of
the AP upstroke) and repolarization
([d2F/dt2]max
of the AP downstroke) time
points.19 Maps of APDs
(Figure 1E) differ from maps of activation
(Figure 1D) and demonstrate gradients of APDs from apex to
base.16 17 19
Repolarization time points determined from
(d2F/dt2)max
were shown to be equivalent to 97% recovery to baseline and coincident
with the RP.19
|
VF was induced by burst stimulation (2x threshold, 2-ms
duration, 30- to 50-ms CL for 1 second). After 2 minutes of VF, data
were acquired as a series of scans (4.096 sec/scan) at 2000 frames/sec.
During VF, the local (dF/dt)max was considered
an activation event, if >10% of maximum
(dF/dt)max
(Figure 1
). The time delay between activation events (eg,
AIs) was measured from >50 beats to plot AI
(AIn+1) versus the previous AI
(AIn), as previously
described.13
Local conduction velocities vectors were calculated for each
diode from the differences in activation time points of that diode
(determined from dF/dt)max) and its 8 nearest
neighbors, as previously
described.20 Local velocity
may be overestimated in VF because wavefronts may originate from deeper
layers. The maximum local velocity vector was normally 0.8 m/sec but
exceeded 1.1 m/sec in VF in 18±4% of activation events (n=4 hearts).
Local velocities 
0.95 m/sec (maximum velocity+1 SD) could detect
transmural propagation and were deleted from the
analysis.
FFT analysis used an optimized FFT routine, FFTW (available at http://www.fftw.org) to speed up calculations. Each trace lasting 4.096 seconds (8192 points) was transformed to the power spectrum in the range of 2 to 40 Hz, providing a frequency resolution of 1/4.096=0.244 Hz. FFT analysis >4.096 seconds at the highest frequency resolution of the apparatus provided accurate frequency distributions across the heart. To investigate VF structure in various regions of the ventricles, hearts were rotated around the aortic cannula to record signals from all accessible surfaces (anterior, posterior, and right and left epicardia) (n=4 hearts).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Spatial Distribution of AIs
The initiation of VF is illustrated in Figure 2A
where APs are recorded during 3 basic beats
(CL 300 ms) followed by a burst of electrical stimuli (2 ms in
duration, 50-ms CL for 1 second), which triggered long-lasting VF. Maps
of activation and APDs recorded during the basic beat from the same
heart before the induction of VF are shown in
Figures 1D and 1E. A set of simultaneously
recorded signals during VF shows that signals from the edges of the
heart had lower signal to noise (S/N) ratio
(Figure 2B). Hence, data from the edges were not included in
the analysis of VF dynamics to avoid a bias of the results.
Figure 2C (top trace) shows an expanded trace of voltage
changes during VF and the first derivative of the signal (bottom
trace). Movies (not shown) and isochronal maps of activation and
during VF revealed no regular patterns or structural organization
during the first 30 minutes of VF. An example of random activation
patterns is depicted in panels 1 through 6
(Figure 2D), which shows activation maps from 6 sequential VF
beats (labeled 1 through 6 in
Figure 2C). Each VF beat produced a markedly different
activation sequence because the initiation site(s), zones of functional
conduction block, densities of isochronal lines, and directions of
wavefronts varied from one VF beat to the next.
|
VF dynamics were examined by plotting
AIn+1 versus AIn, as
previously described by Garfinkel et
al13
(Figure 3A). The open circles represent a series of
74 AIs detected by a diode recording from the apex, and the
solid triangles represent a series of 53 AIs from a diode
recording from the base. Consistent with electrode
studies,13 AI plots produced
a random distribution, with no discernible clusters of data points,
consistent with a nonperiodic process. However, the
average AIs from sites at the apex and base were statistically
significant different with values of 57.5±8.1 and 76.1±1.5 ms
(P<0.05, n=6), respectively.
The shorter AIs at the apex than the base were found to correlate with
the local APD measured from the same hearts, before VF. AIs were
averaged over 4 seconds from 220 (of 252) diodes and plotted against
APDs, demonstrating a statistically significant correlation
(correlation coefficient 0.70) between the local mean AI in VF and the
local APD in paced hearts
(Figure 3B). A best fit between mean AI and local APD favors
a linear relationship between these two parameters with a
slope of 0.59±0.08 (n=6). Note that the slope is <1, because APDs in
hearts paced every 300 ms are in the range of 160 to 200 ms whereas in
VF, AIs fell in the range of 60 to 120 ms. The correlation of AIs in VF
with local APDs in paced hearts is equivalent to the correlation of AIs
to RPs because APDs measured from the second derivative of the AP
downstroke are coincident with the
RP.19 The correlation
between APDs and AIs also implies that the ionic currents responsible
for APDs in normal conditions influence VF dynamics.
|
Like the restitution kinetics of the AP
amplitude,21 the rate of
depolarization of a VF beat depends on the recovery of inward currents,
which might depend on the previous AI.
Figure 3C plots (dF/dt)max of VF
depolarizations versus the previous AI for a site at the apex and a
site at the base of the heart, showing a linear correlation between the
rate of depolarization of a VF cycle and the duration of the previous
AI. The analysis was repeated for a column of 16 diodes
oriented from the base to the apex; all showed the same tight
correlation between (dF/dt)max and the previous
AI. Best-fit analysis produced lines with similar slopes but
statistically significant differences in x-intercepts in going from the
apex to the base (not shown). In 6 hearts, the correlation coefficients
between (dF/dt)max and previous AI were
0.64±0.09 for sites at the base and 0.63±0.08 for sites at the apex.
These findings indicate that determinants of refractoriness and
recovery from RPs influence VF, resulting in gradients of AIs and
(dF/dt)max in VF.
Local Velocity Vectors in VF
The distribution of the amplitude and orientation of
local velocity vectors were also analyzed as an alternative
approach to detect organized repetitive wavefronts in VF. The
limitation of the approach was that activation wavefronts do not spread
strictly along the surface of the heart but may contain transmural
components of propagation emerging from deeper layers. Thus, conduction
velocities measured from 2D data must be considered as apparent rather
than exact conduction velocities. Errors produced by transmural
activation can be partially overcome by deleting from the
analysis data diodes that record zones of synchronous
activation, which indicate zones of epicardial breakthrough. The
distribution of local conduction velocities during VF was
analyzed from sites at the base
(Figures 4A and 4C) and the apex
(Figures 4B and 4D) of the heart. Local velocity vectors
revealed a complex pattern of angles and amplitudes with no
preferential amplitude or direction, even for a few consecutive cycles
of VF, which is contrary to expectations for repetitive reentry
circuits. However, the distribution of local velocity vectors at the
base
(Figure 4A) showed a slight tendency to cluster near 100
degrees, which might be caused by the neighboring
atrioventricular boundary. The directions of local
vectors were also calculated for 5 adjacent channels at the base and 5
at the apex and are displayed as histograms of angular distribution
(Figures 4C and 4D). Histograms of local velocity vectors
showed no preferential orientation (n=4 hearts). Hence, the
analysis of local velocity vectors did not support the
hypothesis that VF is an organized process.
|
Spatial Distribution of FFT Power
Spectra
Another approach to test for spatial organization in VF
is to analyze FFT spectra from various sites on the heart. As
shown in
Figure 5, FFT spectra recorded from a site on the base
(Figure 5A, top trace) and from the apex
(Figure 5B, bottom trace) were broad, with multiple peaks and
no indication of a single dominant frequency. FFT spectra from the base
had lower frequencies (Figure 5A, 12.05±0.6 Hz) compared with the apex (Figure 5B, 16.3±0.5 Hz). The analysis was then
extended for all sites on the anterior surface of the hearts (n=4
hearts, 220 diodes per heart) to relate the normal gradient of APDs to
the range of FFT frequencies at each site.
Figure 6 is a 3D plot, with FFT frequencies plotted along
the z-axis as a function of x-y location on the anterior surface of the
heart. A 2D plot of APDs across the surface of the heart is
superimposed below the 3D plot of FFT frequencies. FFT frequencies
increased gradually from base to apex whereas APDs decreased
progressively from base to apex. Those findings support the hypothesis
that ionic currents responsible for gradients of APDs in normal cardiac
rhythm influence the dynamics of the heart under VF.
|
|
The complex FFT spectra measured at all sites on the
anterior surface of the heart imply that VF was not sustained by one or
several rotors. Still, the possibility that the mother rotor (or zone
with a single dominant frequency) existed in a region beyond the field
of view was investigated by repeating the measurement around the
perimeter of the hearts (n=3).
Figures 7A through 7D illustrate FFT spectra recorded
from the right, left, anterior, and posterior surfaces of a heart,
respectively. Although FFT spectra had somewhat different spectral
characteristics in the different regions of epicardia, all regions
exhibited complex broadband FFT spectra rather than a single dominant
frequency in VT. The narrowest FFT spectra (ie, most organized) were
found in a small region (3x3 mm) at the base of the right
ventricle, which tends to have the longest APDs
(Figure 7A). It is unlikely that the base of the right
ventricle
(Figure 7A) can be the source of a mother rotor that sustains
VF throughout the heart because activation maps showed no periodicity,
dimensions were too small, and the maximum FFT frequency was not faster
than elsewhere in the heart.
|
Ionic Currents Underlying APD Gradients and AIs
in VF
The link between the spatial distribution of APDs and
AIs shown in the present study suggests that the
heterogeneous current distributions that produce APD
gradients likewise influence AIs and frequencies in VF. To investigate
the role of
IKr as a
possible determinant of the spatial distribution of APDs and AIs in VF,
hearts were perfused with the selective
IKr
antagonist E4031. E4031 (0.5 µmol/L) produced the maximum
prolongation of APDs from 172±9 to 241±6 ms (n=3) and marked changes
in the distribution of APDs (compare
Figures 8A and 8B), which indicated that
IKr is
heterogeneously distributed on the epicardium. The effects
of E4031 on VF dynamics were tested by inducing VF, mapping AIs, and
then adding E4031 to the fibrillating hearts and remapping AIs (n=3
hearts). In the absence of E4031, the mean AI maps were similar to
control maps of APDs, with a pattern of short to long AIs from apex to
base
(Figure 8A versus 8C). With E4031, FFT spectra changed from
complex broad energy distributions
(Figure 8E) to narrow-bandwidth lower-frequency
(Figure 8F) spectra, within 10 minutes. With E4031, maps of
mean AIs
(Figure 8D) were again similar to maps of APDs in the
presence of E4031
(Figure 8B).
|
To highlight the link between APDs and AIs in VF, we
selected 2 sites on the heart, one on the right ventricle and another
on the left ventricle, that had similar APDs in control maps but
different APDs after treatment with E4031.
Figures 8G and 8H show Poincaré plots of AIs recorded
from these 2 sites identified on the mean AI maps
(Figures 8C and 8D) by an open circle and an asterisk. In this
example, Poincaré plots show that in the absence of E4031, AIs from
the 2 sites occurred with similar AI values of 45.8±9.7 ms on the left
ventricle (open circle) and 38±9.1 ms on the right ventricle
(asterisk)
(Figure 8G). AI values were consistent with the APD
values measured at the same sites and with the correlation of AIs with
APDs shown in
Figure 3B. With E4031, AIs from these 2 sites separated into
2 clusters with long (asterisks, 137±7.6 ms) and short (open circles,
72.5±22.3 ms) durations
(Figure 8H), consistent with marked dispersion of
long and short APDs recorded at the same sites in the presence of
E4031
(Figure 8B). Maps of mean AIs
(Figures 8C and 8D) showed substantial structure and
organization much like those obtained in maps of APDs with or without
E4031
(Figures 8A and 8B). The data show that the prolongation of
APDs decreased VF frequencies and AIs and the inhibition of
IKr
narrows the bandwidth of FFT spectra most likely as a result of the
substantial increase in RPs throughout the
heart.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The properties of VF and the possible relationship to the electrical properties of the normal myocardium are important concepts that could help us to better understand the mechanisms that sustain VF and improve strategies to interrupt VF. The present study investigates the relationship between VF dynamics and RPs by analyzing their spatial distribution. The main findings are that VF is not organized into zones of reentrant waves with a single dominant frequency but has a broad range of frequencies and that AIs are tightly correlated with the local APD and RP, producing gradients of AIs across the surface of the heart. Hence, heterogeneities of repolarizing K+ currents underlie gradients of APDs and RPs seen in the normal heart and influence VF dynamics.
Dispersion of Refractoriness and VF
Structure
Heterogeneities of excitability and/or conduction
velocities have long been implicated as mechanisms that initiate
arrhythmias.4 22 23 24 25
Enhanced dispersion of refractoriness caused by tissue damage or
pathological condition can produce a functional arc of conduction
block. For instance, a premature impulse originating in an area with a
short RP can capture, propagate, then encounter a zone with longer RPs,
resulting in unidirectional propagation and a reentrant circuit that
progresses to fibrillation. However, once VF reaches a steady state, it
is not known whether properties of the normal myocardium
(ie, gradients of APD and refractoriness) continue to dictate limits on
VF dynamics. Once formed, stable spiral waves can be the source of
activation wavefronts, and their periodicity may in turn depend on the
local excitability and recovery of the tissue where the core is
located. Optical recordings of
atrial26 and
ventricular6
signals using a CCD camera suggested that fibrillation is composed of a
few domains, each with a single dominant frequency separated by
boundaries of slow Wenkebach-like conduction. Chen et
al7 proposed that a single 3D
scroll is sufficient to account for the spatial and temporal features
of VF, and the boundaries between these reentrant waves can be
attributed to zones with different
RPs.6 In contrast, Moe et
al8 proposed a tight
relationship between RPs and AIs in atrial fibrillation, and
Janse4 reported a high degree
of correlation (r=0.95) between
local APDs and AIs of VF in dog hearts. Previous studies did not
directly measure the distribution of RPs and did not have sufficient
spatial resolution to correlate AIs and VF frequencies to RPs at
multiple sites. Hence, the relationship between RPs and VF dynamics and
whether RPs dictate spatial and/or temporal features of VF dynamics
were not firmly established and remained a matter of
conjecture.
Validity of the Approach and the Animal
Model
In our study, a 16x16 photodiode array was used to
record transmembrane potentials from multiple locations with good
S/N ratio (100:1) and high temporal resolution (2000 frames/sec).
Guinea pig hearts were chosen to study the spatial and temporal
dynamics of VF because the electrophysiology and the initiation and
termination of VF in these hearts have been well characterized and the
gradients of APDs and RPs have been extensively
investigated.17 20
In guinea pig hearts, APDs are short at the apex and are progressively
longer toward the base, resulting in a gradient of APDs on the left,
right, and anterior surface of the
heart.16 17 27
In previous studies, we showed that in normoxic hearts, the
repolarization time point taken as
(d2F/dt2)max
of the AP downstroke was shown to be coincident with the RP such that a
map of APs can be used to generate an instantaneous map of
RPs.19
The size of guinea pig hearts was not a limitation because VF can be readily elicited by burst pacing, was maintained for >30 minutes, and could be terminated and re-induced, in the absence of ischemia or any anatomical block.20 Movement artifacts were reduced effectively by pressing the hearts gently against the glass window of the heart chamber. Numerous precautions were taken to ensure that hearts did not become ischemic when pressed against the glass window. Hearts were perfused at a constant flow rate; the aortic pressure was continuously monitored to check for an increase in coronary resistance, an indication of blocked vessels by the chamber glass. We required that the optical APs had stable plateau phases and long APDs rather than the triangulated shape characteristic of APs in ischemia. These precautions ensured that the heart was not ischemic, particularly within the field of view, and APD recordings were measured several times over 10 minutes before VF to guard against unstable preparations.
Guinea pig ventricular myocytes have been extensively studied, and the ionic channels and currents underlying the generation of the AP are well characterized with notable features such as the lack of Ito currents.28 The present study demonstrates a tight correlation between AIs in VF with local APDs and RPs in paced hearts, indicating that a form of refractoriness persists in VF. Activation waves in VF invade regions of myocardia before the myocytes recover fully from the previous depolarization; yet AIs in VF still influence the next AI, its rate of depolarization, and mimic the distribution of APDs in the normal heart.
Distribution of
IKr May
Underlie APD Gradients and AIs in VF
The distribution of APDs along the epicardium of guinea
pig hearts has been extensively
demonstrated,15 17 27
but the ionic channels that are heterogeneously distributed
on the epicardium to produce APD gradients are unknown. Repolarizing
K+ channels
(Ito,
IKr,
IKs) are
heterogeneously distributed along the
ventricular wall, which could account for the observed
dispersions of repolarization. Several studies are consistent
with the notion that
IKr may
contribute to the gradients of repolarization and the dispersion of AIs
in VF, shown in
Figure 8. The delayed rectifying outward current,
IKr, was
distributed heterogeneously in various species of mammalian
hearts. In ferret hearts, immunohistochemistry of Erg, the channel
protein responsible for the rapid component of the delayed rectifier
current
IKr was
shown to be heterogeneously distributed (decreasing from
apex to base), consistent with the gradients of APDs reported
in the epicardium of guinea pig
heart.29 In rabbit
ventricles, the ratio of
IKr/IKs
currents was found to be greater at the apex than at the base,
consistent with the data on ferret
hearts.29 30
Guinea pig ventricular myocytes do not express the channel
protein underlying the transient outward current,
Ito,
leaving the distribution of
IKr as
the most likely current to contribute to gradients of APD and its
kinetics to account for rapid repolarization during VF beats. In guinea
pig myocytes,
IKr was
known to activate early in the plateau phase of the AP and was
thought to have a negligible contribution to the early phase of
repolarization.31 32
Gintant33 reexamined the
contribution of
IKr to
the early phase of the AP downstroke and proposed that
IKr
contributed a considerable component of the total repolarizing
K+ current. During an AP, most
IKr
channels activate early in the plateau phase, then rapidly
shift to an inactivation state while still in the plateau phase. As the
plateau potential decreases, a threshold voltage for
IKr
channels is reached, resulting in a shift to the open state and a large
K+ repolarizing current before channel
closure.34 Thus,
IKr
could account for APD gradients and rapid voltage repolarizations that
are required to produce voltage oscillations in VF. Indeed,
the IKr
blocker E4031 produced marked changes in VF dynamics, changing a
complex frequency distribution to a major dominant frequency and
considerably longer AIs, particularly at sites with prolonged APDs
(Figures 8G and 8H). Further studies will be needed to
elucidate the relative contribution of
ICa,
IKr, and
IKs in
generating APD gradients, rapid voltage oscillations during
VF, and their roles in wave breakups.
Single or Multiple Wavelets?
Multiple wavelet reentry has been extensively studied
as a possible mechanism for
VF1 8 where spiral
waves can be subdivided by wave breakups to produce multiple
wavelets.11 35 36
In the present study, FFT spectra had a broad power spectrum with
multiple peaks appearing from a 4-second analysis of VF, and
broad spectra were observed at all sites on the epicardial surface. The
complex power spectra are consistent with the hypothesis that
VF is composed of multiple wavelets undergoing wave breakup. Our data
are inconsistent with a single spiral or 3D scroll because we
failed to observe large areas with a monolithic dominant frequency
(n=7). Several factors could account for the different findings
reported here and previous data from a CCD
camera.3 6 7
The photodiode array provided a significant improvement in sampling
rate and S/N ratio compared with CCD cameras, hence a superior temporal
resolution and accuracy of FFT spectra. On the other hand, CCD cameras
have a greater spatial resolution than photodiode arrays (if spatial
averaging is not required to improve the S/N ratio) so that each FFT
spectrum was derived from smaller regions of tissue. Another difference
is that FFT spectra were determined from 4 seconds of continuous data
compared with 2 seconds in previous studies that used a CCD
camera.7 The longer time of
analysis increases the frequency resolution and improves the
details regarding the frequency distributions in VF signal. However,
the FFT algorithm represents the time-averaged behavior during
the time of analysis and contains no information regarding when
each peak occurred within that time interval. Thus, each peak in the
frequency domain could appear and disappear and hence more peaks are
seen when the time of analysis is increased. To examine the
possibility that the FFT spectra will have a markedly altered energy
distribution, the FFT analysis window was reduced from 4
seconds to 0.5. Windows of 2 seconds (0.5-Hz resolution) did not
produce a single peak in the frequency domain. With a 0.5-second (2-Hz
resolution) window, FFT spectra showed a single peak, but different
regions of the heart had different peak frequencies, which would remain
inconsistent with a highly organized process. Further studies
are required to adjust the window of analysis to encompass the
lifetime of each peak in the frequency domain, which represents
the lifetime of spirals and their breakups.
Species differences may also account for the different findings on the structure of VF. Guinea pig ventricles lack Ito currents, so different ionic currents, with different spatial distributions, are involved in setting limits to VF dynamics. In rabbit heart, Ito currents produce spatial distributions of repolarization that contribute to repolarization in VF, which may alter VF dynamics and VF frequencies. Finally, edema of the ventricular myocardium must be avoided to protect the heart and avert changes of conduction velocities and VF dynamics.
Analysis of Local Conduction
Velocity
The spatial and temporal distribution of local velocity
vectors was used in attempts to quantitatively characterize
Wenckebach-like activation waves in VF that might occur at boundaries
separating rotors of different phases and/or frequencies. Previous
studies had shown that such boundaries were invaded by wavefronts
emanating from the two rotors that could alternatively propagate or be
blocked on encountering functional conduction blocks due to spatial and
temporal changes in refractoriness, resulting in Wenckebach-like
conduction.6 Such periodic
wavefronts were identified on the epicardium of hearts in VF using
space-time plots, an approach that lacks quantitative information
regarding the velocity and orientation of the
waves.6 To quantitatively
assess the occurrence of Wenckebach-like conduction on the epicardium
during VF, the magnitude and direction of local velocity vectors were
analyzed as previously
described20 at multiple
sites on the epicardium for 4-second intervals during VF. These data
illustrate the findings for sites on the apex (5 diodes) and the base
(5 diodes)
(Figure 4) and showed a random distribution of angles, which
is inconsistent with a periodic process or Wenckebach-like
wavefronts.
The restitution kinetics of APDs and conduction velocities are also important to understand the stability of spiral waves.24 36 Restitution curves with steep slopes predict an increase in APD oscillations and increased vulnerability to VF. The role of conduction velocity oscillations is less understood, but simulation studies predict that velocity oscillations produce CL oscillations and conduction blocks that may lead to wave breakups and new spiral waves in VF elicited by burst stimulation.24 The role of conduction velocity oscillations in VF has not been confirmed experimentally because of the technical difficulties of measuring conduction velocities in 3D. Hence, it should be emphasized that local conduction velocities measured in the present study may contain components of transmural propagation even after precautions were taken to eliminate zones of highly synchronous activation. For this reason, the analysis of conduction velocities should be limited to test for the occurrence of Wenkebach-style conduction and should not be overinterpreted to provide evidence for the wave breakup or the mother rotor hypotheses.
Limitations: Ischemia and Effects of 3D
Structure of the Heart
The findings correlating VF dynamics to dispersion of
refractoriness are limited to VF in normoxic hearts with constant
coronary perfusion. In the clinical setting, VF will rapidly
trigger a loss of perfusion pressure, ischemia, and acidosis.
These conditions would compromise the metabolic state of
the heart, depolarize the resting membrane potential, and alter RPs
(ie, postrepolarization refractoriness) and perhaps the dispersion of
refractoriness. Ischemia also alters ionic currents, eg,
ATP-sensitive K+ currents modifying the
myocardial substrate compared with normoxic hearts, thus changing VF
dynamics and the correlation between electrical activity in VF with
those of the normoxic paced heart.
Another limitation is that this study does not address the role of 3D structure of the heart and how this modifies VF dynamics. Cryoablation and chemical ablation of the specialized conductile system4 and modeling studies show that the 3D structure of hearts can contribute to complex wavefronts in VF. In hearts with thick walls, the propagation of the activation wave in the deeper layers produces more complex pathways, and the emergence of these waves to the surface results in more complex epicardial wavefronts. In the left ventricle of guinea pig hearts, the apex has a thicker wall than the base and one should consider the possibility that the apparent correlation between AIs and APDs may be due to gradients of tissue thickness across the heart. However, this study examined VF dynamics around the perimeter of the heart and failed to find zones with periodic spiral waves. If wall thickness is the important parameter to correlate VF dynamics, then according to this hypothesis, the middle region between the right and left ventricles should show the shortest AIs. Our data reveal no clear relationship between AIs and anatomical landmarks and support the gradient of RPs as the important factor limiting AIs. The influence of wall thickness was examined by cryoablation of the endocardium to produce a thin layer of surviving tissue on the epicardium as described by Allessie et al.2 However, in agreement with previous reports, VF was very difficult to induce by burst stimulation in cryoablated hearts (n=6 hearts), possibly as a result of the reduction of total tissue mass and/or the ablation of the Purkinje system, which may also play a pivotal role in sustaining VF.4 14 37 38 An elucidation of the role of Purkinje fibers to promote persistent VF and as a determinant of VF dynamics will hopefully be achieved with the development of 3D mapping techniques.
|
Acknowledgments |
|---|
This work was supported by grant awards from the National Institutes of Health, R01 HL57929 and HL59614 (to G.S.), and a predoctoral fellowship from the Western Pennsylvania Affiliate of the American Heart Association (to B.-R.C). Thanks are due to the staff of our Departmental Machine Shop: Scott J. McPherson and William B. Hughes for the construction of optical components and heart chamber.
|
Footnotes |
|---|
Original received August 10, 2000; resubmission received February 1, 2001; revised resubmission received February 21, 2001; accepted February 21, 2001.
Preliminary report of these data appeared in abstract form (Pacing Clin Electrophysiol. 2000;23:609).
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- Abildskov JA. Additions to the wavelet hypothesis of cardiac fibrillation. J Cardiovasc Electrophysiol. 1994;5:553–559.[Medline] [Order article via Infotrieve]
- Allessie MA, Schalij MJ, Kirchhof CJ, Boersma L, Huybers M, Hollen J. Experimental electrophysiology and arrhythmogenicity: anisotropy and ventricular tachycardia. Eur Heart J. 1989;10(suppl E):2–8.
- Gray RA, Pertsov AM, Jalife J. Spatial and temporal organization during cardiac fibrillation [published erratum appears in Nature. 1998;393:191]. Nature. 1998;392:75–78.
- Janse MJ. Vulnerability to ventricular fibrillation. Chaos. 1998;8:149–156.[Medline] [Order article via Infotrieve]
-
Rogers JM, Huang J,
Smith WM, Ideker RE. Incidence, evolution, and spatial distribution of
functional reentry during ventricular fibrillation in pigs.
Circ Res. 1999;84:945–954.
[Abstract/Free Full Text] -
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.
[Abstract/Free Full Text] -
Chen J, Mandapati
R, Berenfeld O, Skanes AC, Jalife J. High-frequency periodic sources
underlie ventricular fibrillation in the isolated rabbit
heart. Circ Res. 2000;86:86—93.
[Abstract/Free Full Text] - Moe G, Rheinbolt WC, Abildskov JA. A computer model of atrial fibrillation. Am Heart J. 1964;67:200–220.[Medline] [Order article via Infotrieve]
-
Boyett MR, Jewell
BR. A study of the factors responsible for rate-dependent shortening of
the action potential in mammalian ventricular muscle.
J Physiol. 1978;285:359–380.
[Abstract/Free Full Text] - Chialvo DR, Gilmour RF, Jalife J. Low dimensional chaos in cardiac tissue. Nature. 1990;343:653–657.[Medline] [Order article via Infotrieve]
- Gilmour RF, Chialvo DR. Electrical restitution, critical mass, and the riddle of fibrillation. J Cardiovasc Electrophysiol. 1999;10:1087–1089.[Medline] [Order article via Infotrieve]
- Karma A. Electrical alternans and spiral wave breakup in cardiac tissue. Chaos. 1994;4:461–472.[Medline] [Order article via Infotrieve]
- Garfinkel A, Chen PS, Walter DO, Karagueuzian HS, Kogan B, Evans SJ, Karpoukhin M, Hwang C, Uchida T, Gotoh M, Nwasokwa O, Sager P, Weiss JN. Quasiperiodicity and chaos in cardiac fibrillation. J Clin Invest. 1997;99:305–314.[Medline] [Order article via Infotrieve]
- Janse MJ, Wilms-Schopman FJ, Coronel R. Ventricular fibrillation is not always due to multiple wavelet reentry. J Cardiovasc Electrophysiol. 1995;6:512–521.[Medline] [Order article via Infotrieve]
- Efimov IR, Ermentrout B, Huang DT, Salama G. Activation and repolarization patterns are governed by different structural characteristics of ventricular myocardium: experimental study with voltage-sensitive dyes and numerical simulations. J Cardiovasc Electrophysiol. 1996;7:512–530.[Medline] [Order article via Infotrieve]
-
Kanai A, Salama
G. Optical mapping reveals that repolarization spreads anisotropically
and is guided by fiber orientation in guinea pig hearts.
Circ Res. 1995;77:784–802.
[Abstract/Free Full Text] -
Salama G,
Lombardi R, Elson J. Maps of optical action potentials and NADH
fluorescence in intact working hearts.
Am J Physiol. 1987;252:H384–H394.
[Abstract/Free Full Text] - Choi BR, Salama G. Spatial distribution of activation intervals and conduction velocities (CV) during ventricular fibrillation (VF). Pacing Clin Electrophysiol. 2000;23:609. Abstract.
-
Efimov IR, Huang
DT, Rendt JM, Salama G. Optical mapping of repolarization and
refractoriness from intact hearts.
Circulation. 1994;90:1469–1480.
[Abstract/Free Full Text] -
Salama G, Kanai
A, Efimov IR. Subthreshold stimulation of Purkinje fibers interrupts
ventricular tachycardia in intact hearts:
experimental study with voltage-sensitive dyes and imaging techniques.
Circ Res. 1994;74:604–619.
[Abstract/Free Full Text] -
Gettes LS, Reuter
H. Slow recovery from inactivation of inward currents in mammalian
myocardial fibres. J Physiol
(Lond). 1974;240:703–724.
[Abstract/Free Full Text] -
Baker LC, London
B, Choi BR, Koren G, Salama G. Enhanced dispersion of repolarization
and refractoriness in transgenic mouse hearts promotes reentrant
ventricular tachycardia.
Circ Res. 2000;86:396–407.
[Abstract/Free Full Text] -
Cao JM, Qu Z, Kim
YH, Wu TJ, Garfinkel A, Weiss JN, Karagueuzian HS, Chen PS.
Spatiotemporal heterogeneity in the induction of
ventricular fibrillation by rapid pacing: importance of
cardiac restitution properties. Circ
Res. 1999;84:1318–1331.
[Abstract/Free Full Text] -
Han J, Moe G.
Nonuniform recovery of excitability in ventricular muscle.
Circ Res. 1964;14:44–60.
[Abstract/Free Full Text] -
Kuo CS, Munakata
K, Reddy CP, Surawicz B. Characteristics and possible mechanism of
ventricular arrhythmia dependent on the dispersion
of action potential durations.
Circulation. 1983;67:1356–1367.
[Abstract/Free Full Text] -
Skanes AC,
Mandapati R, Berenfeld O, Davidenko JM, Jalife J. Spatiotemporal
periodicity during atrial fibrillation in the isolated sheep heart.
Circulation. 1998;98:1236–1248.
[Abstract/Free Full Text] -
Laurita KR,
Girouard SD, Rosenbaum DS. Modulation of ventricular
repolarization by a premature stimulus: role of epicardial dispersion
of repolarization kinetics demonstrated by optical mapping of the
intact guinea pig heart. Circ
Res. 1996;79:493–503.
[Abstract/Free Full Text] - Hume JR, Uehara A, Hadley RW, Harvey RD. Comparison of K+ channels in mammalian atrial and ventricular myocytes. Prog Clin Biol Res. 1990;334:17–41.[Medline] [Order article via Infotrieve]
- Brahmajothi MV, Morales MJ, Reimer KA, Strauss HC. Regional localization of ERG, the channel protein responsible for the rapid component of the delayed rectifier, K+ current in the ferret heart [published erratum appears in Circ Res. 1997;81:455]. Circ Res. 1997;81:128–135.
-
Cheng J, Kamiya
K, Liu W, Tsuji Y, Toyama J, Kodama I. Heterogeneous
distribution of the two components of delayed rectifier
K+ current: a potential mechanism of the
proarrhythmic effects of methanesulfonanilide class III agents.
Cardiovasc Res.. 1999;43:135–147.
[Abstract/Free Full Text] -
Sanguinetti MC,
Jurkiewicz NK. Two components of cardiac delayed rectifier
K+ current: differential sensitivity to
block by class III antiarrhythmic agents.
J Gen Physiol. 1990;96:195–215.
[Abstract/Free Full Text] -
Zeng J, Laurita
KR, Rosenbaum DS, Rudy Y. Two components of the delayed rectifier
K+ current in ventricular
myocytes of the guinea pig type: theoretical formulation and their role
in repolarization. Circ Res. 1995;77:140–152.
[Abstract/Free Full Text] -
Gintant GA.
Characterization and functional consequences of delayed rectifier
current transient in ventricular repolarization.
Am J Physiol Heart Circ
Physiol. 2000;278:H806–H817.
[Abstract/Free Full Text] -
Zhang S, Zhou Z,
Gong Q, Makielski JC, January CT. Mechanism of block and identification
of the verapamil binding domain to HERG potassium channels.
Circ Res. 1999;84:989–998.
[Abstract/Free Full Text] - Chen PS, Garfinkel A, Weiss JN, Karagueuzian HS. Spirals, chaos, and new mechanisms of wave propagation. Pacing Clin Electrophysiol. 1997;20:414–421.[Medline] [Order article via Infotrieve]
-
Qu Z, Weiss JN,
Garfinkel A. Cardiac electrical restitution properties and stability of
reentrant spiral waves: a simulation study.
Am J Physiol. 1999;276:H269–H283.
[Abstract/Free Full Text] -
Damiano RJ, Smith
PK, Tripp HF, Asano T, Small KW, Lowe JE, Ideker RE, Cox JL. The effect
of chemical ablation of the endocardium on ventricular
fibrillation threshold.
Circulation. 1986;74:645–652.
[Abstract/Free Full Text] - Janse MJ, Kleber AG, Capucci A, Coronel R, Wilms-Schopman F. Electrophysiological basis for arrhythmias caused by acute ischemia: role of the subendocardium. J Mol Cell Cardiol. 1986;18:339–355. [Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  T.-J. Wu, S.-F. Lin, A. Baher, Z. Qu, A. Garfinkel, J. N. Weiss, C.-T. Ting, and P.-S. Chen Mother Rotors and the Mechanisms of D600-Induced Type 2 Ventricular Fibrillation Circulation, October 12, 2004; 110(15): 2110 - 2118. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. J. Chorro, J. Guerrero, A. Ferrero, A. Tormos, L. Mainar, J. Millet, J. Canoves, J. C. Porres, J. Sanchis, V. Lopez-Merino, et al. Effects of acute reduction of temperature on ventricular fibrillation activation patterns Am J Physiol Heart Circ Physiol, December 1, 2002; 283(6): H2331 - H2340. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||